【作者】 曾辰；

【导师】 王全九；

【作者基本信息】 中国科学院研究生院（教育部水土保持与生态环境研究中心） ， 土壤学， 2010， 博士

【摘要】 吐鲁番点为代表的传统农耕粘壤土,保水保肥性较好,导水性和透气性较差。该类土导水率相对较低而扩散率相对较高,灌溉应采取“少次多量”的模式。灌水器应采用较小的滴头流量和较大的滴头间距。鄯善点为代表的戈壁改良含石砾砂壤土,透气性较好,保水保肥性较差,易产生深层渗漏。该种土导水率相对较高而扩散率相对较低,灌溉应采取“多次少量”的模式。灌水器应采用较大的滴头流量和较小的滴头间距。成龄葡萄的蔓长变化较小,在试验期间内变化不大。枝条数目、枝条长度、枝条节间长度、葡萄叶片数目、叶片平均中脉长、平均单叶面积以及地上净增生物量的增长主要集中在新梢生长期,在开花期后,增长速度逐渐趋于平缓,在果粒膨大期后各性状变化幅度不大。枝条直径的增长速度在新梢生长期较小,在花期及果粒膨大期前期逐渐加快,从果粒膨大期后期开始枝条增粗速度逐渐减慢。葡萄叶面积指数在展叶期迅速增大,在开花期后,葡萄叶面积指数增长速度开始减慢,在果粒成熟期至葡萄采收时达到一年的最大值。此后叶面积指数迅速下降,直至埋墩。成龄葡萄叶片中叶绿素含量为叶绿素a>叶绿素b。高水、中水、对照沟灌三个处理间相同叶绿素类型含量无显著差异。低水处理较低且差异显著。以不同位置叶片叶绿素含量比较,迎光处成熟叶片>背光处成熟叶片>迎光处幼叶片。葡萄果粒的纵径、横径、果粒单粒体积进入膨大期后迅速增大,进入果粒成熟期时速度减缓,但在采收前再次迅速增大,呈现出快-慢-快的增长趋势。葡萄的果型指数进入膨大期后迅速减小,进入果粒成熟期后开始逐渐趋于稳定。采收时葡萄果型指数呈正态分布,不同水分处理果型指数在1.25-1.41之间。葡萄果粒糖度进入膨大期后开始增大,进入果粒成熟期时速度增快,在果粒成熟期后期至采收前阶段速度逐渐减缓,呈现出慢-快-慢的变化趋势。同一串葡萄的糖度表现为串上部>串中部>串下部。同一条蔓葡萄串的平均糖度表现为蔓顶部>蔓底部>蔓中部。以鄯善为代表的戈壁改良砂壤土地区,葡萄根系分布范围相对较广。其最大垂直分布深度可超过1.4 m,根系最大密度在40-60 cm附近,其水平分布范围几乎覆盖整个沟垄范围,集中在沟道及垄上沿蔓方向1.8 m范围内。以吐鲁番为代表的传统农耕粘壤土区,葡萄根系分布范围相对较为集中。其最大垂直分布深度一般不超过1 m,根系最大密度在20-40 cm附近,其水平分布只在沟道及垄沿蔓方向2 m范围内但集中分布在1.2 m范围内。在沟灌改滴灌后,垄面及沟平面下0-40 cm有效吸收根系密度有显著增加,在60 cm以下根系密度变化不大。成龄葡萄叶片长、叶片宽、叶片中脉长三者之间存在着较好的线性关系;叶面积与叶长平方、叶宽平方以及叶长宽乘积之间均呈较好的线性关系;叶片面积与叶片长、叶片宽、叶片中脉长之间也存在着良好的幂函数关系;葡萄的枝条长、枝条第三节间长、枝条上的平均叶片数、叶片平均中脉长相互间呈指数函数关系;成龄葡萄的果粒重与果粒体积间呈良好的线性关系,与果粒纵径间呈良好的指数函数关系;地上净增干物质量与枝条长呈良好的幂函数关系;地上净增干物质量与枝条叶片数呈良好的“S”形关系。在葡萄采收时,葡萄含糖量与单粒重之间线性关系。生育期内葡萄叶面积变化可用Logistic模型很好地表达。Logistic模型模拟叶面积变化过程效果较好。吐鲁番点和鄯善点成龄葡萄根系二维分布根长相对密度可用e指数形式来表达,拟合效果均较好模拟结果较为准确。田间用来测定葡萄耗水方法主要有三种,水量平衡法、微气象学法、植物蒸腾和棵间土壤蒸发结合法。三种测定方法计算结果相近,变化规律一致。水量平衡法计算所得数据为多日耗水量平均值,难以获得瞬时变化规律;茎流计结合土面蒸发法与微气象法均能够连续获得葡萄耗水量,但是由于葡萄树个体差异,在尺度扩展上存在一定问题,同时该方法需要对土面蒸发进行准确测量,尤其针对棵间土面不均匀的地方。气象法由于对气象要素敏感,因此测量结果波动性较大。Penman-Monteith方法与其它13种计算方法有很好的线性相关关系。相关性最好的两种方法分别是ASCE Penman-Monteith、Kimberly Penman(1982)方法。相关性差的方法是FAO-24 Blaney-Criddle与Hargreaves。实际作物系数与耗水量密切相关。不同生育期作物系数差异较大,总体呈下降趋势。在充分供水情况下,吐鲁番地区成龄葡萄作物系数均远远高于FAO推荐值。其中初始阶段平均高出约60 %,中期和后期平均高出推荐值100 %左右。在晴天,葡萄蒸腾日变化呈双峰曲线。阴天,葡萄蒸腾速率随气温的波动而波动。阴天的葡萄最大蒸腾速率和平均蒸腾速率较晴天低,但在夜间,两种条件下葡萄蒸腾速率均维持在0.08 mm/hr左右。葡萄日均耗水强度从萌芽期开始逐渐增大,到膨大期达到最大,在果粒成熟期及枝条成熟期逐渐降低。对于滴灌而言,葡萄耗水强度随着灌水量的增加而增加。不同水分处理的葡萄耗水模数均呈双峰型,即萌芽期耗水比例小,新梢生长期出现耗水模数的第一个峰值,花期耗水模数最小,果实膨大期出现耗水模数的第二个峰值,该峰值较第一个峰值大,果浆成熟期和枝蔓成熟期耗水模数逐渐减小。滴灌灌溉水几乎百分之九十以上被葡萄所消耗,灌水效率较高;地面沟灌水分效率较低,鄯善含砾石砂壤土灌水利用率为31 %,吐鲁番粘壤土为70 %。采用滴灌措施,实现的节水并不是真实的节水,仅是提高了水资源的利用效率。土面蒸发主要发生在滴灌直接湿润的种植沟及沟垄交界面处。在葡萄生长初期直至采收前,葡萄耗水量以蒸腾为主,而在采收后,土面蒸发量比例增大。对于耗水总量模拟,使用实际作物系数计算得到的各生育期耗水总量模拟值与实测值计算结果相近;对于土面蒸发模拟,使用双作物系数法中的土壤蒸发系数通过反复迭代能够较好的模拟土面蒸发变化过程;对于根系吸水的葡萄蒸腾模拟,通过Hydrus2D软件,采用本研究中所的到的各项参数以及根系分布,能够较好的模拟鄯善砂壤土和吐鲁番粘壤土两种条件根系吸水影响下的土壤水分运动。葡萄叶片水势日变化呈现出先增大-后减小单峰曲线的变化趋势。葡萄光合作用和蒸腾作用日变化趋势一致,均呈现出增大-减小-增大-减小变化,在一日之内存在两个高峰。葡萄光合作用和蒸腾作用在生育期内变化趋势一致,在展叶期后逐渐增大,在花期、果粒膨大期和果粒转熟期葡萄光合作用和蒸腾作用均达到最大,转熟期后期葡萄光合作用和蒸腾作用逐渐降低。从影响葡萄的环境因子来看:影响葡萄光合速率蒸腾速率水分利用效率的主要因素包括太阳净辐射、大气温度、相对湿度以及土壤含水量四个影响因子;在各影响因素之中,太阳净辐射是最主要的环境因子,其次为气温和土壤含水量;监测葡萄干旱的指标包括土壤水分,在果粒膨大期,土壤水分的下限为田间持水量的74-83 %;土壤水势,400-420 cm左右;冠层温度差以及叶水势,临界值为-13 bar。葡萄产量与全生育期耗水量之间呈现较好的二次抛物线关系;葡萄水分生产效率与全生育期耗水量之间也呈现较好的二次抛物线关系。在保证最高产量和最大水分利用效率的前提下,吐鲁番试验点最优葡萄耗水量为628-663方/亩,鄯善试验点为767-933方/亩。Aquacrop模型模拟葡萄生产力效果较好,大部分处理误差在10 %以内。更多还原

【Abstract】 The Turpan study site represented traditional farming clay loam soil. This soil has a better retain of water and nutrients and poor water conductivity. The soil has a relatively low hydraulic conductivity and relatively high diffusion rate. The irrigation model should take as "less time large amounts", and the emitter should take a small discharge rate and large emitter spacing. The Shanshan study site represented Gobi desert improved gravel sandy loam soil. This soil and a poor water retain and fertilizer, and easy to produce deep leakage. This kind of soil has a relatively high hydraulic conductivity and relatively low diffusion rate. The irrigation model should take as "several time small amounts", and the emitter should take a large discharge rate and small emitter spacing.The variation of grapevine length was small and can be ignored during the study periods. The number of branches, branch length, branch internode length, number of grape leaves, average leaf midrib length, and average leaf area and ground net biomass growth was increased rapidly in the shoot growing period. After flowering period, the growth rate gradually becoming slowly and the variation become slightly after the fruit enlargement period. The branch diameter growth rate was smaller in the shoot growth period, the growth rate accelerated in flowering and fruit enlargement period, and the speed gradually slowed down in the late fruit enlargement period. Grape leaf area index increases rapidly in the leaf expansion period, grape leaf area index began to slow down after flowering period and reached maximum value in grapes harvest period of one year. Leaf area index decreased rapidly thereafter, until the buried pier. The leaf chlorophyll content of mature grape showed chlorophyll a> chlorophyll b. The chlorophyll content showed no significant differences among high, middle water treatments and furrow irrigation treatment, but the chlorophyll content was significantly smaller for the low water treatment. Chlorophyll content in different position showed mature leaves for welcome light > backlight for mature leaves> welcome light for young leaves.The longitudinal diameter, transverse diameter and single berry volume of grape fruit increases rapidly after fruit-development stage, and then slow down after fruit mature stage, but increased rapidly again before the harvest, showed a fast-slow-fast growth trend. The grape fruit sharp index decreased rapidly after fruit-development stage, and became stable after fruit mature stage. Grape sharp index were between 1.25-1.41 under different water treatments. Grape sugar increased after fruit-development stage, and increased rapidly after fruit mature stage, then slow down before harvest, showed a slow - fast - slow the trend. The grape sugar content expressed upper string> string middle> lower strings in same cluster and expressed as top> bottom > middle in same vine.The distribution of grape roots for Shanshan improved sandy loam area showed relatively wide range. The maximum depth of vertical distribution was more than 1.4 m, and the maximum density presented at 40-60 cm. The horizontal root distribution was covering almost the whole ridge and furrow range, and concentrated in the furrow and 1.8 m toward ridge direction. The distribution of grape roots for Turpan traditional farming clay loam area was relatively concentrated. The maximum depth of vertical distribution was less than 1 m, and the maximum density presented at 20-40 cm. The horizontal root distribution was less than 2 m but concentrated in the 1.2 m range toward ridge direction. After using drip irrigation, the absorb root increased significant in the ridge and 0-40 cm under surface of furrow. The root density has little change Changes under 60 cm.There existed a linear relationship among leaf length, leaf width, leaf midrib length of mature grape; there existed a linear relationship among leaf area, leaf length square, leaf length by leaf width; there existed a good power function relationship among leaf area and leaf length, leaf width, leaf midrib length; there existed a exponential function relationship among grape branch length, branch length of last third quarter, average number of leaves on branches, average leaves length; there existed a good linear relationship between grape berry weight and berry size; there existed a exponential function relationship between grape berry weight and longitudinal diameter; there existed a power function relationship between ground dry weight of shoot and branch length; there showed "S"-shaped relationship between ground dry weight of shoot and leaf number. In the grape harvest period, grape sugar content showed linear relationship to berry weight.Changes of leaf area index growth process can be well expressed by Logistic model. The two-dimensional relative length density distribution of mature grape roots can be well simulated with exponential form for Shanshan and Turpan study sites.Methods to determine and measure grape water consumption in field include water balance method, micro-meteorology method, and plant transpiration and soil evaporation method. The results of three methods are similar with the same variation. Water balance method to calculate the data for the average water consumption for several days, it is difficult to obtain instantaneous variation; Sap Flow with soil surface evaporation and micrometeorological methods were able to obtain continuous glucose water consumption, but due to individual differences in vine in the scaling there are certain problems, but the method requires accurate measurement of soil surface evaporation. Penman-Monteith method has a good linear relationship to the other 13 kinds of calculation methods. The best correlations between two methods were the ASCE Penman-Monteith, Kimberly Penman (1982) method. The poor correlations were FAO-24 Blaney-Criddle and Hargreaves method. The actual crop coefficient and water consumption are closely related. The crop coefficients were quite different at different growth stages and showed downward trend. The mature grape crop coefficient values of Turpan area were much higher than the FAO recommended under sufficient water supply condition. The initial stage was about 60% and mid-and late stages were about 100% higher than the recommended values.The grape transpiration showed diurnal bimodal curve in sunny days. The grape transpiration rate fluctuated with air temperature fluctuations in cloudy days. The maximum transpiration rates of grapes in cloudy days were lower than sunny days. At night, the transpiration rates under two conditions were maintained at 0.08 mm / hr. Daily consumption of grape were gradually increased from bearing shoot stage, then reached max value at fruit-development stage, and gradually decreased at fruit mature stage, fruit mature stage. For drip irrigation, the grape consumption rate was increase with irrigation quote. The grape water consumption modulus showed two peaks under different water treatments. The first water modulus peak appeared at shoot growth period, the second water modulus peak appeared at fruit-development stage. The water modulus showed low values at flowering period and fruit mature stage, fruit mature stage. Ninety percent of irrigated water was consumed by grape with high irrigation efficiency. The furrow irrigation method showed low irrigation efficiency. Furrow irrigation efficiency was 31% for Shanshan gravel sandy loam soil and was 70% for Turpan clay loam soil. Soil surface evaporation occurs mainly in drip irrigation ditch and the interface of ridge and furrow. Water mainly consumed by transpiration from early growth to harvest but reach a larger proportion of soil surface evaporation after harvest.The water consumption amount calculated by actual crop coefficient were similar than measured results. The soil surface evaporation process can well simulated by dual crop coefficient method. The soil water movement influenced by root water uptake of the two study sites can be well monitored by Hydrus 2D software and parameters in the front.Water potential of grape leaves showed increase-decrease, single peak trend. Photosynthesis and transpiration of grapes both showed increases-decreases- increasing-decreasing changes in the day. Grape photosynthesis and transpiration showed same trend in the growth period, that increased after leaf developing stage, and reached the maximum values during flowering period, fruit-development stage and fruit mature stage, then decreased after late fruit mature stage.Photosynthetic rate,transpiration rate and water use efficiency have significantly correlated with solar radiation, air temperature, relative humidity and soil moisture content. Among these factors, the net radiation was the most important factors, following by air temperature and soil moisture. Grape drought monitoring indicators, including soil moisture in the fruit enlargement period, 74-83% of FC; soil water potential, 400-420 cm; deficit between canopy temperature and leaf water potential, the critical value of -13 bar.Grape yield presented good parabolic relationships between water consumption in the whole growth period; grape water use efficiency showed good parabolic relationship between water consumption in the growth period. In ensuring the maximum yield and maximum water use efficiency, the best grape water consumption amount was 628-663 m~3 /mu for Trupan and 767-933 m~3 /mu for Shanshan study sites. Aquacrop model can well simulate water productivity of grapes.更多还原