晋西黄土区2种典型森林树种夜间液流特征及对环境因子的响应  

Nocturnal sap flow characteristics of two typical forest tree species and responses to environmental factors in the loess region of West Shanxi,China

作　　者：付照琦 胡旭 田沁瑞 葛艳灵 周红娟 吴小云 陈立欣[1,2] FU Zhao-Qi;HU Xu;TIAN Qin-Rui;GE Yan-Ling;ZHOU Hong-Juan;WU Xiao-Yun;CHEN Li-Xin(College of Soil and Water Conservation,Beijing Forestry University,Beijing 100083,China;National Station for Forest Ecosystem Research in Jixian County,Jixian,Shanxi 042200,China)

机构地区：[1]北京林业大学水土保持学院,北京100083 [2]山西吉县森林生态系统国家野外科学观测研究站,山西吉县042200

出　　处：《植物生态学报》2024年第9期1128-1142,共15页Chinese Journal of Plant Ecology

基　　金：国家重点研发计划(2022YFF1302501)。

摘　　要：了解晋西黄土区典型森林树种夜间水分利用对准确评价流域水循环和碳循环至关重要。但目前对于晋西黄土区典型森林树种夜间液流的变化特征及对夜间环境因子的响应尚不明晰。为了探究该地区典型树种的夜间液流分配特征及对环境因子的响应,于2021年6-10月,采用热扩散探针(TDP)对山西省吉县蔡家川小流域刺槐(Robinia pseudoacacia)人工林和蒙古栎(Quercus mongolica)天然林中树木树干液流、土壤水分和小气候进行连续监测,利用时间序列模型(ARMAX model)区分夜间液流中夜间蒸腾(Tn)和茎干补水(Re)。利用机器学习模型XGboost和SHAP解释框架分析夜间液流量(Q_(Nighttime))与夜间环境因子的关系。结果表明:生长季刺槐和蒙古栎夜间蒸腾量占比(Tn/Q_(Nighttime))分别为73.97%和30.12%,茎干补水占比(Re/Q_(Nighttime))分别为26.03%和69.88%。机器学习XGboost模型结果表明夜间饱和水汽压差(VPD)和夜间土壤水分含量(SWC)是分别影响刺槐和蒙古栎夜间液流的主导因子。VPD、夜间气温(T_(a))、夜间相对湿度(RH)和整日液流量(Q_(Daily))分别达到0.30 kPa、19.84℃、85.3%和339.11 kg·h^(-1)后对刺槐夜间液流起促进作用。VPD达到1.5 kPa后对蒙古栎夜间液流起抑制作用,T_(a)和夜间风速(WS)分别达到23.11℃和0.58 m·s^(-1)后促进蒙古栎夜间液流。SWC对刺槐和蒙古栎夜间液流的影响具有一致性,即SWC在0-12.9%范围内会促进夜间液流。因此构建夜间液流模型时应考虑环境因子的影响。刺槐夜间液流主要用于夜间蒸腾作用,而蒙古栎夜间液流主要用于茎干补水。夜间VPD和SWC分别是影响刺槐和蒙古栎夜间液流的主导因子。该研究揭示了晋西黄土区2种典型树种夜间液流分配特征及与环境因子的响应,这对于准确理解气候变化下黄土区典型森林树种的稳定性具有重要意义。Aims Understanding the nighttime water use of typical forest tree species in the west Loess Plateau is crucial for accurately evaluating the watershed-scale water and carbon cycles.However,the nighttime sap flow dynamics and its response to environmental factors are currently unclear for typical forest species in the west Loess Plateau.Methods We monitored the sap flow with thermal dissipation probes(TDP)and concurrent environmental factors of Robinia pseudoacaciaplantation and Quercus mongolica natural forest in the Caijiachuan watershed from June to September 2021.The time series model(ARMAX model)was used to distinguish between nighttime transpiration(T_(n))and stem water replenishment(Re)in nocturnal sap flow(Q_(Nighttime)).We used the machine learning model XGboost and Shapley additive interpretation framework to analyze and explain the relationship between Q_(Nighttime) and nighttime environmental factors.Important findings The proportion of nighttime transpiration(Tn/Q_(Nighttime))was 73.97%and 30.12%for the R.pseudoacacia and the Q.mongolica,respectively,during the growing season.The proportion of stem water filling(Re/Q_(Nighttime))was 26.03%and 69.88%for the two species,respectively.The XGboost model showed that nighttime vapor pressure deficit(VPD)and nighttime soil water content(SWC)were the main factors driving QNighttime in R.pseudoacacia and Q.mongolica,respectively.VPD,nighttime air temperature(T_(a)),nighttime relative humidity(RH),and daily sap flow(Q_(Daily))can promote the QNighttime when they reach 0.30 kPa,19.84℃,85.3%,and 339.11 kg·h^(-1).For Q.mongolica,VPD inhibited QNighttime when it reached 1.5 kPa,while Ta and nighttime wind speed(WS)promoted QNighttime when it reached 23.11℃and 0.58 m·s^(-1),respectively.The effects of SWC on QNighttime in R.pseudoacacia and Q.mongolica were similar.The SWC within 0-12.9%range would promote the occurrence of QNighttime.Therefore,our study highlights the importance of considering the magnitude of all environmental factors in establishing the noct

关 键 词：夜间液流 环境因子 夜间蒸腾 黄土地区 机器学习模型 

分 类 号：S718.5[农业科学—林学]