滩地人工林幼林不同时间尺度CH4通量变化特征--基于涡度相关闭路系统的研究

Dynamics and regulation of CH_4 fluxes in a poplar plantation on a floodplain

长江滩地是甲烷(CH_4)排放的潜在热点区域,然而目前其CH_4通量的变化特征及控制因子尚未被揭晓。基于涡度相关闭路系统进行为期2年多的连续观测,旨在揭示长江滩地杨树(Populus deltoides)人工林幼林CH_4通量在不同时间尺度上的变化特征及其调控机理。结果显示,全年和部分未淹水月份表现出白天排放强而夜间排放弱的平均日变化特征,且淹水前、淹水期间和退水后分别表现出日间双峰型(7:00和10:00)、日间与夜间各一峰的双峰型(10:00和23:00),以及典型的日间单峰型(10:00)。淹水年份(2012年)在夏季(6—8月)排放最强,在春末(5月)和秋末冬初(11—12月)排放最弱,而未淹水年份(2013年)在初夏(6月)排放最强,在盛夏(7月)和秋末(11月)转变为较弱的吸收。淹水年份的年排放量((128.0±42.4)mmol/m^2)是未淹水年份((51.5±29.1)mmol/m^2)的2.5倍。滩地人工林幼林CH_4通量的日变化和季节变化最可能受到摩擦风速、水位和土壤温度的调节,而年际间的巨大差异主要由淹水状况决定。

Atmospheric methane （CH4） concentration has been rising in recent years after a short cease, resulting in an increased global warming potential 28 times that of CO2 in 2011. This phenomenon has caused extensive concern about the atmospheric CH4 sink and source as well as the dynamics of CH4 flux at different temporal scales. In fact, changes in CH4 flux are more complicated than those of CO2, owing to the complex interactions between biotic and abiotic regulations. In the past decade, the eddy covariance method has been widely used to assess changes in CH4 flux and enhanced the understanding of the dynamics and regulations of the flux. The floodplain of the Yangtze River is a hotspot for CH4 emission because of frequent water inundation, although limited evidence has been reported. In this study, CH4 flux was continuously measured with a close-path eddy covariance system using a Fast Methane Analyzer （FMA DLT-100, Los Gatos Research, Inc. USA） for more than two years （January 2012-February 2014） in a young Populus deltoides plantation on a floodplain of the Yangtze River. Throughout the experiment, several environmental factors were observed in order to reveal CH4 flux dynamics and the environmental controls at different temporal scales. This plantation was re-established in late January 2012 after clear-cutting of the former mature one, and the field was inundated with water in 2012 for approximately 40 d （July 11th-August 20th）, but not in 2013. We found that CH4 flux was highly variable at several temporal scales. At the half-hourly to daily scale during certain months of non-inundation and over both entire years, the mean diurnal exhibited similar patterns, i.e., the emissions were strong and reached their maximum during the day while they were weak at night. However, this pattern was not apparent in most non-inundation months. In the months of inundation （July-August 2012）, the mean diurnal variations showed a double-peak pattern before and during inundation, while a single-peak pattern was observed after inundation. The two peaks before inundation appeared at 7：00 and 10：00, while those during inundation appeared at 10：00 and 23：00. The peak observed after inundation appeared at approximately 10：00. At the monthly to yearly scale, CH4 fluxes in the two years exhibited patterns in seasonal variation. During the year of inundation （2012）, the strongest emission appeared in summer （June-August）, whereas the weakest emission appeared in late spring （May） and late autumn through early winter （November-December）, at which point the system behaved as a CH4 source. In the year without inundation （2013）, the strongest emission appeared in early summer （June） while the weakest emissions were found in mid-summer （July） or late autumn （November）. At the yearly scale, CH4 emission in the inundation year （（128.0±42.4） mmol/m^2） was 2.5 times of that in the non-inundation year （（51.5±29.1） mmol/m^2）, where the emission in the inundation months was responsible for 83.8% of the annual total. We also found that the CH4 flux at the half-hourly and daily scales was mainly modulated by the friction velocity, water table, and soil temperature, whereas the large interannual difference was mainly determined by inundation frequency and magnitude.