Queensland University of Technology, Institute for Future Environments, Brisbane, Queensland, Australia.
Queensland University of Technology, Institute for Future Environments, Brisbane, Queensland, Australia.
Sci Total Environ. 2015 Apr 15;512-513:8-18. doi: 10.1016/j.scitotenv.2015.01.011. Epub 2015 Jan 19.
Variations in interannual rainfall totals can lead to large uncertainties in annual N2O emission budget estimates from short term field studies. The interannual variation in nitrous oxide (N2O) emissions from a subtropical pasture in Queensland, Australia, was examined using continuous measurements of automated chambers over 2 consecutive years. Nitrous oxide emissions were highest during the summer months and were highly episodic, related more to the size and distribution of rain events than soil water content. Over 48% of the total N2O emitted was lost in just 16% of measurement days. Interannual variation in annual N2O estimates was high, with cumulative emissions increasing with decreasing rainfall. Cumulative emissions averaged 1826.7±199.9 g N2O-N ha(-1) yr(-1) over the two year period, though emissions from 2008 (2148±273 g N2O-N ha(-1) yr(-1)) were 42% higher than 2007 (1504±126 g N2O-N ha(-1) yr(-1)). This increase in annual emissions coincided with almost half of the summer precipitation from 2007 to 2008. Emissions dynamics were chiefly driven by the distribution and size of rain events which varied on a seasonal and annual basis. Sampling frequency effects on cumulative N2O flux estimation were assessed using a jackknife technique to inform future manual sampling campaigns. Test subsets of the daily measured data were generated for the pasture and two adjacent land-uses (rainforest and lychee orchard) by selecting measured flux values at regular time intervals ranging from 1 to 30 days. Errors associated with weekly sampling were up to 34% of the sub-daily mean and were highly biased towards overestimation if strategically sampled following rain events. Sampling time of day also played a critical role. Morning sampling best represented the 24 hour mean in the pasture, whereas sampling at noon proved the most accurate in the shaded rainforest and lychee orchard.
年际降雨量的变化会导致短期田间研究中对年一氧化二氮排放预算估计产生很大的不确定性。本研究利用连续两年的自动气室进行的连续测量，研究了澳大利亚昆士兰州亚热带牧场中一氧化二氮（N2O）排放的年际变化。N2O 排放最高值出现在夏季，并且具有高度的突发性，与降雨事件的大小和分布的关系比土壤含水量更为密切。在总排放量中，超过 48%的 N2O 仅在测量日的 16%中损失。年际 N2O 估算值的变化很大，随着降雨量的减少，累积排放量增加。两年期间，平均累积排放量为 1826.7±199.9 g N2O-N ha(-1) yr(-1)，但 2008 年（2148±273 g N2O-N ha(-1) yr(-1)）的排放量比 2007 年（1504±126 g N2O-N ha(-1) yr(-1)）高 42%。这种年排放量的增加与 2007 年至 2008 年夏季降水量的近一半相一致。排放动态主要受雨事件的分布和大小的驱动，雨事件的季节性和年际变化。使用 Jackknife 技术评估了采样频率对累积 N2O 通量估算的影响，为未来的手动采样活动提供信息。通过从 1 到 30 天的定期时间间隔选择测量通量值，为牧场和两个相邻土地利用（雨林和荔枝果园）生成了每日测量数据的测试子集。如果在降雨事件后策略性采样，则与每周采样相关的误差高达亚日平均值的 34%，并且存在高度的高估偏差。采样时间也起着至关重要的作用。在牧场中，早晨采样最能代表 24 小时的平均值，而在阴凉的雨林和荔枝果园中，中午采样最准确。
