【作者】 李振华；

【导师】 王彦辉；

【作者基本信息】 中国林业科学研究院 ， 生态学， 2014， 博士

【摘要】 在我国西北干旱半干旱地区，植被覆盖率低和生态环境恶劣严重限制着区域经济和社会发展。尽快恢复植被覆盖是改善生态环境的关键举措。然而在以往植被恢复过程中存在着植被种类选择不当、在不适宜立地造林和造林密度过大等问题。事实上，一个地区的水资源总量及其空间分配的异质性在很大程度上决定着所能承载的植被类型、格局、数量和结构。为给合理恢复植被提供依据，除认识植被蒸散耗水特性外，还需掌握特定地点的地形、土壤厚度及含水量等因子的空间变化及其对水文循环、水资源形成及植被生长影响的尺度效应。为此，本文在六盘山北侧半干旱区叠叠沟小流域内选取了一系列典型坡面和典型植被样地，研究植被蒸散及组分的时空变化，同时分析立地因子、植被结构等水文要素的坡面变化及空间尺度效应，主要结论如下：1）华北落叶松液流速率存在优势度差异不同优势度华北落叶松的液流速率有明显差异。优势度大的树木获取光照和水分的能力更强，其液流活动在一天内启动早、结束晚、到达峰值快，峰值更大；其瞬时液流速率对太阳辐射和饱和水汽压差变化的响应更敏感，但对土壤含水量的响应敏感性较弱。这导致其日均液流速率明显高于优势度小的树木。但不同优势度树木间的液流速率相对差异较稳定。相关分析表明，液流速率与优势度（相对树高）、树高呈极显著正相关关系（P＜0.01），与冠长和胸径显著正相关（P＜0.05），而与冠幅和边材面积正相关但不显著（P＞0.05）。利用建立的液流速率（Js）与优势度（h）之间的线性关系（Js=0.0051h+0.0494，R2=0.95）计算了样地内所有单株的液流速率及其平均值。该值比不考虑优势度影响的常用算法的结果低16%。因此，建议今后在利用样树液流速率测定结果经尺度扩展估计林分液流速率时，增加考虑优势度等树形因子的影响。2）华北落叶松林蒸腾响应潜在蒸散和土壤可用水分的关系华北落叶松林的日蒸腾量随根区土壤可用水分的变化遵循逻辑斯蒂增长曲线，即：随土壤水分增多，日蒸腾量先快速近线性增大；当土壤水分达到某个阈值后，日蒸腾量趋于平稳，基本不再变化。不同潜在蒸散条件下，日蒸腾量对土壤水分变化的响应模式非常相似，但决定响应曲线的土壤水分阈值及日蒸腾量最大值存在差异，两者均随潜在蒸散升高而增大（P＜0.01）。基于日蒸腾量对土壤水分和潜在蒸散的响应关系，建立了同时考虑土壤供水能力和大气蒸发潜力的复合模型（T=(-0.05PET2+0.65PET)×(1-EXP(-5.39REW))），能很好地（R2=0.92）解释华北落叶松林的日蒸腾量变化。3）典型植被蒸散及水量平衡特征冠层截留率（量）存在明显的植被类型差异。以2012年7~10月份（降水量为366mm）的总截留率来评价，最大的是白桦林，为15%（56mm）；其次是沙棘灌丛、华北落叶松林、山桃林和杂灌丛，变化在11~12%（40~44mm）；最低的是虎榛子灌丛，为9%（31mm）。林下蒸散的植被类型差异在生长季初较小，月蒸散量变化在38.64~49.84mm，在7、8月份的峰值期间差异较大，月蒸散量变化在60.59~104.93mm。变化明显的是阳坡草地、山桃林、虎榛子灌丛和半阳坡草地，不明显的是华北落叶松林和沙棘灌丛。对林下蒸散影响最大的气象因子是气温，其次是潜在蒸散、太阳辐射强度和空气湿度。华北落叶松林在2011、2012年的冠层截留分别占总蒸散的18%、12%，林木蒸腾占25%、22%，林下蒸散占57%、65%。生长季内，各组分占蒸散的比例随时间变化的规律不同：冠层截留比例变化在0~32%（2011年）、6~24%（2012年），但整体趋势平稳；林木蒸腾比例（6~51%、7~33%）呈逐渐降低趋势，林下蒸散比例（39~69%、53~83%）则呈逐渐升高趋势。2011、2012年生长季降水量分别为424、526mm，华北落叶松林蒸散占降水的比例分别为106%、103%，产流量16.4、165.3mm，消耗外界输入的水分分别为126.4、12.2mm。阳坡草地蒸散占降水的73%，产流量分别为111.4、238.2mm；半阳坡草地蒸散占降水的70%、68%，产流量分别为139.0、259.9mm。4）华北落叶松林分结构的坡面差异及尺度效应华北落叶松冠层叶面积指数（LAI）和地上生物量存在坡向、坡位差异。冠层LAI、地上生物量均表现为阴坡高于半阴坡，并且整体上随坡位下降均呈增加趋势。土壤厚度、坡面汇流和土壤水分差异能较好地解释林分结构的坡位差异。冠层LAI和地上生物量的坡面变化表现出明显的尺度效应。冠层LAI的坡面平均值在阴坡上随坡长每增加100m，对应升高0.22，高于半阴坡的0.16，可见其尺度效应在阴坡强于半阴坡；地上生物量的坡面平均值在阴坡上随坡长每增加100m，对应增加4.92t·hm-2，低于半阴坡的6.28t·hm-2，可见其尺度效应在阴坡弱于半阴坡。林分结构在不同坡位处的观测值对整个坡面平均值的代表性有很大差异。本文拟合了不同坡位处冠层LAI、地上生物量的观测值与整个坡面平均值的比值随离开坡顶水平距离变化的关系，可据此将特定坡位样地的观测值换算为整个坡面的平均值，从而减少野外调查中的样地数量，提高坡面参数的估算精度。5）华北落叶松林蒸散与产流的坡面差异及尺度效应华北落叶松林蒸散和产流具有明显的坡向、坡位差异。蒸散表现为阴坡高于半阴坡，产流则相反；随坡位下降蒸散呈增加趋势，产流则减小。蒸散在阴坡上的坡面尺度效应（坡长每增加100m，蒸散的坡面平均值对应增加28.3mm）强于半阴坡（19.5mm/100m）；与之相反，产流在半阴坡上的坡面尺度效应（34.7mm/100m）强于阴坡（16.8mm/100m）。拟合了不同坡位处林分蒸散及其组分、产流与整个坡面平均值的比值随离开坡顶水平距离变化的关系，可据此将特定坡位样地的相应特征值换算为整个坡面的平均值。更多还原

【Abstract】 Low coverage of vegetation and harsh environment severely limit the economic and socialdevelopment of the regionsin northwest China. Especially in the arid and semi-arid areas,it isthe key to rebuild the vegetation for the purpose of eco-enviroment improvement. In the pastwork of vegetation restoration, there existedmany mistakes, such as selectingunfitvegetation,afforestation in unsuitable area or with astand densitylager than the carryingcapacity of the soil water. In fact, limited water resources and its strong heterogeneityallocating in space determinein a great degreethe vegetation type, pattern and structure. Toprovide a reasonable basis for vegetation restoration, it is necessary to understand the variationand scale effect of topography, soil thickness and water content and other site factors in space,and the current distribution and variation of vegetation structure and scale effect, as well as therelationship between vegetation and site factors. To this end, we selected a series of slopes andplots in a semi-arid region northwest of the Liupan Mountains, to study the temporal andspatial variation of evapotranspiration and its components in typical vegetation, and to quantifythe slope variation and spatial scale effect of the hydrological factors, such as the siteconditions and forest structure,etc. The main conclusions are as follows:1. Difference in sap flow density among Larix principis-rupprechtii Mayr trees ofdifferent degree of dominanceThe sap flow density (Js) in trees of higher degree of dominance started earlier in themorning but ended later at night than that in trees of lower degree of dominance. In addition,the maximum of Jsduring the daytime appeared earlier and was higher in higher trees. Thus thedaily average of Jsin higher trees was apparently higher as well. Furthermore, the Jsin highertrees was more sensitive to the solar radiation and the vapor pressure deficit on a5-min timescale, while on daily scale to the soil drought, higher trees showed less sensitive, implying astronger ability for obtaining light and water. In general, however, no significant difference was found in the pattern of Jsresponse to environment conditions among sample trees, and therelative difference in Jswas relatively stable. Correlation analysis indicated that the mostimportant factors positively affecting the Jswere the degree of dominance (relative height) andthe tree height (P＜0.01), then were the canopy length and diameter at breast height (P＜0.05),and then the canopy width and sapwood area (P＞0.05). In an improved approach, the Jsforthe forest stand was taken as the average of Jsfor all individual trees in the stand which wascalculated by using the linear relation between Jsand degree of dominance. It was16%lessthan the value calculated by the average of Jsfor the five sample trees in the widely used basicapproach. Therefore it is proposed that the degree of dominance should be taken into accountin the up-scaling approach for the stand Jsor transpiration estimate in future.2.Coupling effects of soil moisture and evaporative demandon the larch transpirationDaily transpiration(T)of the larch plantation varied from0.08to2.18mm·day-1for thewholegrowing season.In any given soil watercondition, the T displayed scattered, andmuchmore scattered in better conditions, showing the coupling effects of evaporative demand(PET) and soil water (REW). A logistic relationship(T=Tmax·(1-exp(k·REW)) was derived todescribe the T varying with the rising REW. Additionally, theTmax(the maximum of T under agiven PET condition) increasedwith the PET (Tmax=a·PET2+b·PET). After combining these twofunctions together, a more general model covering the whole variation range of PET and REWwas determined as T=(-0.05·PET2+0.65·PET)·(1-exp(-5.39·REW). This model fitted themeasured data well and could explain92%of the T variation in the larch plantation, and thuscan be used to describe the joint influence of REW and PET on T.3. Characteristics of evapotranspiration and water balance in several kinds of vegetationCanopy interception were different among various kinds of typical vegetation. During theperiod of July to October,2012,with a cumulative precpitation of366mm, the interception washighest in the birch forest (15%of the precipitation),then wasthe sea buckthorn shrubs, larchplantation, mountain peach plantation and miscellaneous shrubs(11-12%,40-44mm), thelowest9%(31mm)was the ostryopsis shrubs. The difference of the evapotranspiration under the canopy (38.64-49.84mm per month)among vegetation was lower in the beginning of the growing season, but higher in the peakperiod,60.59-104.93mm per month. Obvious change of the evapotranspiration was found inthe sunny slope grassland, mountain peach plantation,ostryopsis shrubsand semi-sunny slopegrassland, whereas not in the larch plantation or the sea buckthorn shrubs. Temperature was themost influential factor, then was the potential evapotranspiration, solar radiation and humidity.The canopy interceptionwas18%,12%of the total evapotranspiration in the growingseason of2011and2012, respectively in the larch plantation. The tree transpirationtookaproportion of25%,22%; meanwhile the evapotranspiration,57%,65%, respectively. Theseasonal change of the ratio of each component to the total evapotranspiration was different.The ratio of interception changed from0to32%,6%to24%, in2011and2012, respectively,with a comparatively smooth tendency; the ratio of transpiration changed from6%to51%,7%to33%, with a gradual decrease; however the ratio of theevapotranspiration under the canopychanged from39%to69%,53%to8%, with a gradual increase during the growing season.The cumulative amount of precipitation was424,526mm in2011and2012, respectively.The evapotranspiration of the larch plantationaccount for106%,103%of the precipitation; thewater yield was16.4,165.3mm and external inputs of water for use was126.4,12.2mmin2011and2012, respectively. The evapotranspiration of sunny slope grasslandaccounted for73%of the precipitation;the water yield was111.4,238.2mm. The evapotranspiration ofsemi-sunny slope grasslandaccounted for70%of the precipitation in2011,68%in2012andthe water yield was139.0,259.9mm, respectively.4.Slope variation and scale effect of the forest structureObvious distinctions were found in the canopy leaf area index (LAI) and abovegroundbiomass among slope directions and positions. The averagesof LAI and aboveground bimasswere higher on the shady slopethan on the semi-shady slope.And along the slope, both of themperformed an increase with the position decreasing,whichcould bewell related tothe soil depthor moisture conditions in different slope positions. Scale effect occurred in the slope variation of the forest stucture. Theslope average ofcanopy LAI increased0.22per100mincrease in the slope length on the shady slope, whichwas higher than0.16on thesemi-shady slope. Therefore, the scale effect of the LAI wasstronger on the shady slope. The slope average of aboveground biomass increased4.92t·hm-2per100m increase in the slope length on the shady slope, which was lower than6.28t·hm-2on the semi-shady slope, implying a stronger scale effect on the semi-shady slope.Furthermore, obvious differencewas obeseverd in the values measured in different slopepositions for representatingthe entire slope. Relationships between the ratio of plot value toslope average value of the canopy LAI or aboveground biomass with the horizontal distancefrom the slope top was fitted.Based on this, values of vegetation structure obtained at a specificposition of a slope can be translated into an average for the entire slope.It is very helpful fortheefficiency and accuracy of the slope parameter estimates.5. Slope variation and scale effect of the forest evapotranspiration and water yieldObvious distinctions were found in the forest evapotranspiration (ET) and water yield (Y)among slope directions and positions. The ET was higher on the shady slope than on thesemi-shady slope, whereas the Y was lower on the shady slope. With thepositon decreasingalong the slope, the ET showed a gradual increase, whereas the Yperformed in a converse way.The scale effect of the ET was stronger onthe shady slope (28.3mm/100m, i.e. the slopeaverage of forest ET increased28.3mm per100m increase in the slope length) than on thesemi-shady slope (19.5mm/100m). In contrast, the scale effect of the Y was weaker on theshady slope (16.8mm/100m) than on the semi-shady slope (34.7mm/100m).Relationshipsbetween the ratio of plot value to slope average value of the ET or the Y with the horizontaldistance from the slope top was fitted.Based on it, the slope average can be estimated upscalingfrom the value measured on any one specific plot on the slope.更多还原