腾格里沙漠东南缘半固定沙丘土壤水分影响因素

doi:10.7522/j.issn.1000-694X.2025.00227

摘要/Abstract

摘要：

土壤水分是干旱沙区固沙植被格局和过程的驱动力，在沙区不同微地貌上呈现出较大的差异。以腾格里沙漠东南缘半固定沙丘为研究对象，分析了土壤水分在迎风坡、背风坡、丘顶和丘底的分布特征，利用基于偏最小二乘法的结构方程模型（PLS-SEM）确定了地形、灌木和草本因子对土壤水分的影响路径。同时，采用基于可解释机器学习算法（SHAP）模型筛选了地形和植被影响土壤水分的重要因子并阐明了其影响机制。结果表明：丘底的表层土壤含水量最高，为1.21%，而背风坡上中层和深层土壤含水量最高分别为2.25%和2.43%。随着草本盖度的增加和坡向的变化，不同深度土壤含水量呈现先降低后升高的趋势。表层土壤水分以3 m的高差为阈值界限随着高差的变化呈现先降低后升高的趋势；中深层土壤水分随生物量和灌木盖度的增加也呈现先增加后减少趋势，二者分别以30 g和40%为阈值界限。

关键词: 腾格里沙漠, 半固定沙丘, 土壤水分, 地形-植被因子, 结构方程模型, SHAP模型

Abstract:

Soil moisture is a key driver of the sand fixation vegetation patterns and processes in arid sandy regions， exhibiting significant variability across different microtopography types in the sandy area. This study focuses on soil moisture in the semi-fixed dunes on the southeastern edge of the Tengger Desert， analyzing its distribution characteristics across four microtopographic types： windward slope， leeward slope， the top of dune， and the bottom of dune. A partial least squares structural equation model （PLS-SEM） was employed to determine the influence paths of topography， shrub， and herbaceous factors on soil moisture. Additionally， an interpretable machine learning model （SHAP） was used to identify the key factors influencing soil moisture among topographic and vegetation factors and elucidate their mechanisms. The results indicate that the surface soil moisture at the bottom of dune is highest， at 1.21%， while the mid- and deep-layer soil moisture on the leeward slope reaches 2.25% and 2.43%， respectively. As herbaceous coverage increases and slope direction changes， soil moisture at different depths first decreases and then increases. Surface soil moisture shows a trend of decrease followed by increase as height difference changes， with a threshold of 3 meters in elevation. The soil moisture in the middle and deep layers also exhibits a trend of increasing and then decreasing with the rise in biomass and shrub cover， with thresholds of 30 g and 40%， respectively.

Key words: Tengger Desert, semi-fixed dune, soil moisture, topography-vegetation factor, structural equation model, SHAP model

中图分类号:

杨述睿, 杨甜, 张璐, 张定海, 戚海迪. 腾格里沙漠东南缘半固定沙丘土壤水分影响因素[J]. 中国沙漠, 2025, 45(5): 328-337.

Shurui Yang, Tian Yang, Lu Zhang, Dinghai Zhang, Haidi Qi. Moisture influencing factors on semi-fixed sand dunes in southeast edge of Tengger Desert[J]. Journal of Desert Research, 2025, 45(5): 328-337.

图/表 9

图1 试验样地示意图

Fig.1 Schematic diagram of test sample plots

图2 4种微地貌上表层、中层和深层土壤水分的分布注：**表示P<0.01，极显著，****表示P<0.001，极其显著

Fig.2 Distribution of soil moisture in surface， middle and deep layers on four microtopography types

表1 4种微地貌上表层、中层和深层土壤平均水分含量及变异系数

Table 1 Average soil moisture content and coefficient of variation in surface， middle and deep layers of four microtopography types

微地貌	0~40 cm		40~200 cm		200~300 cm
微地貌	平均土壤水分/%	变异系数	平均土壤水分/%	变异系数	平均土壤水分/%	变异系数
背风坡	0.71	0.62	2.25	0.52	2.43	0.36
丘底	1.21	0.41	2.23	0.25	1.98	0.23
丘顶	0.81	0.87	1.57	0.46	2.01	0.27
迎风坡	0.33	0.57	0.86	0.41	1.09	0.51

表2 外部模型有效性检验

Table 2 External model validity test

潜变量	克伦巴赫α系数	DG. rho系数	第一特征根	第二特征根
地形	0.89	0.92	2.58	0.53
灌木植被	0.61	0.84	1.43	0.56
草本植被	0.66	0.80	2.09	0.99
土壤水分	0.79	0.88	2.14	0.70

图3 基于PLS-SEM模型的地形和植被因子对土壤水分的影响路径注：*表示P<0.05，**表示P<0.01，***表示P<0.001

Fig.3 Effects of topography and vegetation factors on soil moisture based on PLS-SEM Model

表3 两种机器学习模型对表层、中层和深层土壤水分的模型评价结果

Table 3 Model evaluation results of two machine learning models for surface， middle and deep soil moisture

评价指标	算法	表层	中层	深层
$R 2$	RF	0.84	0.70	0.83
$R 2$	XGBoost	0.68	0.61	0.59
MSE	RF	0.19	0.52	0.30
MSE	XGBoost	0.31	0.45	0.36
MAE	RF	0.34	0.50	0.35
MAE	XGBoost	0.28	0.58	0.41
MAPE/%	RF	5.25	13.5	5.47
MAPE/%	XGBoost	7.11	13.7	11.59

表3 两种机器学习模型对表层、中层和深层土壤水分的模型评价结果

Table 3 Model evaluation results of two machine learning models for surface， middle and deep soil moisture

评价指标	算法	表层	中层	深层
$R 2$	RF	0.84	0.70	0.83
$R 2$	XGBoost	0.68	0.61	0.59
MSE	RF	0.19	0.52	0.30
MSE	XGBoost	0.31	0.45	0.36
MAE	RF	0.34	0.50	0.35
MAE	XGBoost	0.28	0.58	0.41
MAPE/%	RF	5.25	13.5	5.47
MAPE/%	XGBoost	7.11	13.7	11.59

图4 基于随机森林模型的表层、中层和深层土壤水分SHAP解释图

Fig.4 SHAP interpretation map of surface， middle and deep soil moisture based on random forest model

图5 基于随机森林模型的影响不同深度土壤水分变量的重要性排序

Fig.5 Ranking of the importance of variables affecting soil moisture at different depths based on random forest models

图6 基于随机森林模型的SHAP特征依赖

Fig.6 SHAP feature dependence graph based on random forest

参考文献 43

[1]	Wang C， Fu B， Zhang L，et al.Soil moisture-plant interactions：an ecohydrological review［J］.Journal of Soils and Sediments，2019，19（1）：1-9.
[2]	Asbjornsen H， Goldsmith G R， Alvarado-Barrientos M S，et al. Ecohydrological advances and applications in plant-water relations research：a review［J］. Journal of Plant Ecology，2011，4（1）：3-22.
[3]	D'Odorico P， Laio F， Porporato A，et al.Ecohydrology of terrestrial ecosystems［J］.BioScience，2010，60（11）：898-907.
[4]	Western A W， Zhou S L， Grayson R B，et al.Spatial correlation of soil moisture in small catchments and its relationship to dominant spatial hydrological processes［J］.Journal of Hydrology，2004，286（1）：113-134.
[5]	Liu Y， Cui Z， Huang Z，et al.Influence of soil moisture and plant roots on the soil infiltration capacity at different stages in arid grasslands of China［J］.Catena，2019，182：104147.
[6]	Anderson C M， Kustas P W， Norman M J.Upscaling and Downscaling：a regional view of the soil-plant-atmosphere continuum［J］.Agronomy Journal，2003，95（6）：1408-1423.
[7]	凤紫棋，孙文义，穆兴民，等.黄土高塬沟壑区植被恢复方式对小流域土壤水分的影响［J］.中国水土保持科学，2023，21（4）：1-10.
[8]	郭艳菊，马晓静，许爱云，等.宁夏东部风沙区沙化草地土壤水分和植被的空间特征［J］.生态学报，2022，42（4）：1571-1581.
[9]	Asbjornsen H， Ashton M S， Vogt D J，et al.Effects of habitat fragmentation on the buffering capacity of edge environments in a seasonally dry tropical oak forest ecosystem in Oaxaca，Mexico［J］.Agriculture，Ecosystems & Environment，2004，103（3）：481-495.
[10]	Potts D L， Scott R L， Bayram S，et al.Woody plants modulate the temporal dynamics of soil moisture in a semi‐arid mesquite savanna ［J］.Ecohydrology，2010，3（1）：20-27.
[11]	Li X Y， Yang Z P， Li Y T，et al.Connecting ecohydrology and hydropedology in desert shrubs：stemflow as a source of preferential flow in soils ［J］.Hydrology and Earth System Sciences，2009，13（7）：1133-1144.
[12]	Moeslund J E， Arge L， Bøcher P K，et al.Topography as a driver of local terrestrial vascular plant diversity patterns［J］.Nordic Journal of Botany，2013，31（2）：129-144.
[13]	Qiu Y， Fu B， Wang J，et al.Soil moisture variation in relation to topography and land use in a hillslope catchment of the Loess Plateau，China［J］.Journal of Hydrology，2001，240（3）：243-263.
[14]	Perring F.Topographical gradients of chalk grassland［J］.The Journal of Ecology，1959，88：447-481.
[15]	Biederman L A， Whisenant S G.Using mounds to create microtopography alters plant community development early in restoration［J］.Restoration Ecology，2011，19（101）：53-61.
[16]	张世才，王慧娟，张定海，等.腾格里沙漠东南缘3种沙丘4种微地貌上土壤水分与地形-植被因子之间的关系［J］.甘肃农业大学学报，2023，58（3）：160-168.
[17]	孙琰蕙，张定海，张志山.腾格里沙漠不同类型沙丘土壤水分含量与地形-植被因子关系研究［J］.干旱区地理，2022，45（5）：1570-1578.
[18]	李新荣，张志山，谭会娟，等.我国北方风沙危害区生态重建与恢复：腾格里沙漠土壤水分与植被承载力的探讨［J］.中国科学生命科学，2014，44（3）：1-10.
[19]	Hair J F， Matthews L M， Matthews R L，et al.PLS-SEM or CB-SEM：updated guidelines on which method to use［J］.International Journal of Multivariate Data Analysis，2017，1（2）：107-123.
[20]	Edeh E， Lo W J， Khojasteh J.Review of partial least squares structural equation modeling （PLS-SEM） using R：a workbook［J］.Structural Equation Modeling：A Multidisciplinary Journal，2023，30（1）：165-167.
[21]	Tavakol M， Dennick R.Making sense of Cronbach's alpha［J］.International Journal of Medical Education，2011，2：53-55.
[22]	Mehmetoglu Mehmet.Partial least squares approach to structural equation modeling for tourism research［M］//Chen J S.Advances in Hospitality and Leisure.Leeds，UK：Emerald Group Publishing Limited，2012：43-61.
[23]	Dijkstra T K， Henseler J R.Linear indices in nonlinear structural equation models：best fitting proper indices and other composites［J］.Quality & Quantity，2011，45（6）：1505-1518.
[24]	Wetzels M， Odekerken-Schröder G， Van Oppen C.Using PLS path modeling for assessing hierarchical construct models：guidelines and empirical illustration［J］.MIS Quarterly，2009：177-195.
[25]	Zipper S C， Soylu M E， Booth E G，et al.Untangling the effects of shallow groundwater and soil texture as drivers of subfield-scale yield variability［J］.Water Resources Research，2015，51（8）：6338-6358.
[26]	李新荣，张志山，黄磊，等.我国沙区人工植被系统生态-水文过程和互馈机理研究评述［J］.科学通报，2013，58（5）：387-410.
[27]	Song G， Li X R， Hui R.Biological soil crusts increase stability and invasion resistance of desert revegetation communities in northern China［J］.Ecosphere，2020，11（2）：1
[28]	Chamizo S， Cantón Y， Lázaro R，et al.The role of biological soil crusts in soil moisture dynamics in two semiarid ecosystems with contrasting soil textures［J］.Journal of Hydrology，2013，489（6）：74-84.
[29]	Zhong J M， Xiao H D， Yong G，et al.Interactive effects of wind speed，vegetation coverage and soil moisture in controlling wind erosion in a temperate desert steppe，Inner Mongolia of China［J］.Journal of Arid Land，2018，10（2）：534-547.
[30]	Srivastava A， Saco P M， Rodriguez J F，et al.The role of landscape morphology on soil moisture variability in semi‐arid ecosystems［J］.Hydrological Processes，2021，35（1）：e13990.
[31]	John A， Olden J D， Oldfather M F，et al.Topography influences diurnal and seasonal microclimate fluctuations in hilly terrain environments of coastal California［J］.PLoS ONE，2024，19（3）：1-19.
[32]	Singh S.Understanding the role of slope aspect in shaping the vegetation attributes and soil properties in Montane ecosystems［J］.Tropical Ecology，2018，59（3）：417-430.
[33]	Wunsch M， Betzler C， Eberli G P，et al.Sedimentary dynamics and high-frequency sequence stratigraphy of the southwestern slope of Great Bahama Bank［J］.Sedimentary Geology，2018，363：96-117.
[34]	Gong X， Zhang H， Ren C，et al.Optimization allocation of irrigation water resources based on crop water requirement under considering effective precipitation and uncertainty［J］.Agricultural Water Management，2020，239：106-264.
[35]	Zheng L， Wang X， Li D，et al.Spatial heterogeneity of vegetation extent and the response to water level fluctuations and micro-topography in Poyang Lake，China［J］.Ecological Indicators，2021，124：107420.
[36]	Liu Y， Du J， Xu X，et al.Microtopography-induced ecohydrological effects alter plant community structure［J］.Geoderma，2020，362：114-119.
[37]	Mata-González R， Averett J P， Abdallah M A，et al.Variations in groundwater level and microtopography influence desert plant communities in shallow aquifer areas［J］.Environmental Management，2022，69（1）：45-60.
[38]	Rita A， Bonanomi G， Allevato E，et al.Topography modulates near-ground microclimate in the Mediterranean Fagus sylvatica treeline［J］.Scientific reports，2021，11（1）：8122.
[39]	Zhao L， Fang Q， Yang Y，et al.Stemflow contributions to soil erosion around the stem base under simulated maize-planted and rainfall conditions［J］.Agricultural and Forest Meteorology，2020，281：107814.
[40]	Li B B， Li P P， Zhang W T，et al.Deep soil moisture limits the sustainable vegetation restoration in arid and semi-arid Loess Plateau［J］.Geoderma，2021，399：115122.
[41]	Yang L， Wei W， Chen L，et al.Spatial variations of shallow and deep soil moisture in the semi-arid Loess Plateau，China［J］.Hydrology and Earth System Sciences，2012，16（9）：3199-3217.
[42]	Thomas A， Yadav B K， Šimůnek J.Root water uptake under heterogeneous soil moisture conditions：an experimental study for unraveling compensatory root water uptake and hydraulic redistribution［J］.Plant and Soil，2020，457：421-435.
[43]	D'Odorico P， Caylor K， Okin G S，et al.On soil moisture-vegetation feedbacks and their possible effects on the dynamics of dryland ecosystems［J］.Journal of Geophysical Research：Biogeosciences，2007，112（G4）：1-11.