【作者】 王宇；

【导师】 陈丽华；

【作者基本信息】 北京林业大学 ， 生态环境工程， 2010， 博士

【摘要】 本文以北京生态涵养带常见树种的苗木、成林单株林木和林分为研究对象,采用小型蒸渗仪法和热扩散技术两种方法对7种盆栽苗木的耗水规律进行对比研究,同时采用热扩散技术对5种成林单株林木的树干液流规律进行研究,揭示了盆栽苗木和成林单株林木的耗水规律,并对成林单株林木树干液流与各微气象因子的关系分别进行拟合,建立了树干液流与各微气象因子关系的拟合模型。此外,为精确模拟林木树干的液流规律,建立了基于灰色关联分析与RBF人工神经网络的林木单株耗水预测模型。在此基础上,通过树干边材面积与胸径的关系研究,建立了基于树干液流的林分尺度耗水模型,为北京生态涵养带的森林建设、经营及改造提供了准确的基础数据和科学的理论支撑。研究结果表明：(1)小型蒸渗仪法具有操作简单并且便于应用的特点,包裹式茎流计法精度较高,但操作相对复杂且仪器较为贵重。采用小型蒸渗仪法和包裹式茎流计法均能揭示盆栽苗木的耗水规律,通过对两种方法求得的耗水量值进行两配对样本的T检验,发现结果并无显著差异。综合比较两种试验的研究结果,阔叶乔木苗木树种蒸腾速率大小为：盐肤木>元宝枫>刺槐>槲树；灌木苗木树种蒸腾速率大小为：荆条>黄栌；针叶乔木苗木树种采用小型蒸渗仪法测定蒸腾速率大小为：油松>侧柏。(2)通过包裹式茎流计对苗木的昼夜耗水特征研究发现,灌木树种的夜间耗水量占全天耗水量的比重最大,针叶乔木树种夜间耗水量占全天耗水量的比重次之,阔叶乔木树种夜间耗水量占全天耗水量的比重最小。(3)在对单株成林树干液流的研究中发现：针叶乔木中,油松和侧柏的液流速率在观测当日均呈现单峰型,且侧柏的树干液流速率在5个月中典型观测日中的表现均高于油松；阔叶乔木中,在典型晴朗天气条件下的日平均树干液流速率大小依次为：刺槐>槲树>栓皮栎。通过比较5个树种的耗水量发现,不仅是树干液流速率和蒸腾时间影响耗水总量,边材面积的大小也会直接对耗水的总量产生影响,油松的树干液流速率低于侧柏,但是因其边材面积较大,表现在耗水总量上,则是油松>侧柏。(4)在典型晴朗天气条件下,林木本身的生理结构和生态特性决定了其与各微气象因子的关系并不相同,但将林木的树干液流速率分别与空气温度、空气相对湿度、太阳辐射和土壤温度进行拟合,发现拟合方程相关系数很高,均在0.9以上,但与平均风速的关系均不显著。(5)选择在2008年8月份典型晴朗天气条件下,对油松的树干液流速率建立基于灰色关联分析与RBF人工神经网络的林木单株耗水预测模型,通过实测数据检验,该径向基函数网络模型预测的绝对误差最大为0.0577 cm／h,最小为0.0009 cm／h,平均为0.0111 cm／h；相对误差最大为1.6613%,最小为0.0557%,平均为0.6134%,显然预测效果较好。(6)本文从林木边材面积和胸径关系入手,建立了5个树种的边材面积和胸径关系的拟合方程,发现各树种的二者相关性很高。以油松为例,通过对不同胸径油松的树干液流速率进行研究,发现液流速率有随着胸径增大而增加的趋势,不同胸径油松的液流速率日变化趋势基本一致。此外,基于Richards生长模型,建立了不同胸径油松日累计耗水量的预测模型,并以此为基础,通过统计不同径级林木的株数,扩展到林分尺度耗水量的研究。通过对试验地的林分耗水量进行计算得出：油松林在典型晴朗天气条件下的日耗水量为17769.8 kg／hm2。更多还原

【Abstract】 Seedlings and individual plants of common trees species and common stands in Eco-conservation Belt of Beijing were selected as the research objects in this article. A compared research that rules of water consumption of seedlings of seven tree species (Pinus tabulaeformis, Robinia pseudoacacia, Quercus. dentata, Acer truncatum, Rhus chinensis, Vitex negundo Var.heterophylla, cotinus coggygria) in pot experiment measured with micro-lysimeter and thermal dissipation technology was done and the rules of sap flow of individual plants five tree species(Pinus tabulaeformis, Platycladus orientalis, Q. variabilis, Robinia pseudoacacia, Acer truncatum) were studied to reveal the characters of transportation. The relationship between the sap flow of individual plants and per micro-meteorological factor was studied, and the fitting models between which were established. In addition, a prediction model for water consumption of individual plant based on grey relational analysis and RBF artificial neural network was founded to simulate the rules of sap flow accurately. Subsequently, the stand scale extended model based on sap flow was set up according to the relationship between sapwood area and DBH. The results above could provide exactly basic data and scientific theory support for the forest construction, management and reconstruction in Eco-conservation Belt of Beijing. The results showed that:(1) The character of micro-lysimeter method is simple to operate and convenient for application, on the contrary, the packaged stem sap flow gauge method has higher precision and complex operation relatively, and the instrument is more expensive. Both the micro-lysimeter method and the packaged stem sap flow gauge method can reveal the rules of transportation of seedlings, however, there is no significant differences between the two results according to the paired sample T test on the water consumption measured by two methods. The two experiment results were comprehensive comparative analyzed, which showed that the transportation ordination of broadleaf arbor seedlings was:Rhus chinensi> Acer truncatum> Robinia pseudoacacia> Quercus. dentata; the transportation ordination of shrub seedlings was:Vitex negundo Var.heterophylla> cotinus coggygria; the transportation ordination of coniferous arbor seedlings measured micro-lysimeter with was: Pinus tabulaeformis> Platycladus orientalis.(2) According to the characters of diurnal water consumption measured by the packaged stem sap flow gauge, it is revealed that the proportion of night water consumption of shrub seedlings is the biggest, the proportion of coniferous arbor seedlings is lower than that of shrub seedlings, and the proportion of broadleaf arbor seedlings is the lowest.(3) During the research on the sap flow of individual plant, it is founded that both the diurnal change curves of the sap flow velocity of Pinus tabulaeformis and Platycladus orientalis were single peak type, and the sap flow velocity of Platycladus orientalis is higher than that of Pinus tabulaeformis in typical sunny day during the five months; the diurnal mean sap flow velocity ordination in typical sunny day is:Robinia pseudoacacia> Quercus. dentata> Q. variabilis. Compared with the transportation water consumption of five tree species, not only the sap flow velocity and the transportation time had influence on the water consumption, but also the sapwood area did. Although the sap flow velocity of Pinus tabulaeformis is lower than Platycladus orientalis, but the sapwood area of Pinus tabulaeformis is larger, so the total water consumption of Pinus tabulaeformis is higher than that of Platycladus orientalis.(4) In typical sunny days, the physiological structure and ecological characteristics of the individual plant leaded to the different relationship between the individual plant and each micro-meteorological factor. However, after fitting the sap flow velocity to air temperature, Air Relative Humidity, solar radiation, soil temperature, it is showed that the correlation coefficient of the fitting equation is more than 0.9, but the correlation coefficient of the fitting equation between the sap flow velocity and the average wind speed was not significant.(5) The water consumption prediction model of individual plant based on grey relational analysis and RBF artificial neural network was established by the measured data in typical sunny days in Aug.2008. According to the test by measured data, the biggest absolute error of this prediction model is 0.0577 cm/h, the minimum absolute error is 0. 0009 cm/h, and the average value is 0.0111 cm/h. And the biggest relative error of this prediction model is 1.6613%, the minimum absolute error is 0.0557%, and the average value is 0.6134%. Obviously the model has better prediction effect.(6) The fitting equations, the correlation coefficient of which is good, between sapwood area and DBH of five tree species (Pinus tabulaeformis, Platycladus orientalis, Q. variabilis, Robinia pseudoacacia, Acer truncatum) were established. Taking Pinus tabulaeformis for example, the sap flow velocity increases with DBH changing bigger, and the diurnal variation trend of the sap flow velocity of Pinus tabulaeformis was basically identical. In addition, the prediction model for daily cumulative water consumption of Pinus tabulaeformis with different DBH was established based on Richards growth model. Subsequently, this model was extended to predict the water consumption of the stand scale, according to the number of different diameter class trees. The results of the research on stand scale water consumption in the experimental field showed that:the water consumption of Pinus tabulaeformis stand is 17769.8 kg/hm2in typical sunny days.更多还原