云杉（Picea asperata）天然林可持续经营理论与技术
|
摘要
云杉分布区是长江上游重要的水源涵养林地区，在我国生态环境保护和国民经济建设中有着举足轻重的地位。近年来由于“天然林保护工程”和“退耕还林工程”的实施，云杉已成为该流域主要树种。因此，对长江上游水源涵养地区的最主要建群种云杉天然林和人工林经营理论与技术进行超前系统研究已迫在眉睫。自20世纪90年代以来，在四川西北部高山林区和甘肃白龙江、洮河流域林区，对云杉的分布、林分结构、林分生产力及碳汇功能、生境质量评价、种群增长与调节等方面的内容作了大量深入系统的调查研究工作。共设标准样地328个，解析木400株，生物量样地202个，土壤理化分析样品200份。应用数理统计和概率分布的理论进行林分结构研究，探索林分直径、树高与立木株数之间的基本分布规律；采用生态系统研究方法，揭示林分生产力及其与林分结构和生境的关系；运用数量化理论及多元统计分析方法和系统分析原理，建立生境质量评价综合评判系统；通过对云杉种群增长的Logistic模型拟合，建立不同生境条件下云杉种群的Logistic模型并对云杉种群增长与生境因子的关系进行数量分析，建立相关数学模型，揭示云杉种群的拥挤效应与密度制约的变化规律，并对云杉种群中单株增长与生境因子关系进行系统研究，取得以下研究结果。
1)在云杉生境质量评价的研究中，首先采用相关分析方法，研究生境因子之间的相关程度，找出生境因子间的关系。通过多元统计分析方法，筛选对生产力影响较大的主导因子，然后利用主导因子评价生境条件质量和划分类型。最后，编制出云杉地位指数表等数表。研究结果表明，生境因子中相关程度较高的是地形因子中的地貌、坡向与土壤因子中的腐殖质层厚度、土壤湿度，其次是植被类型。通过多种统计分析方法筛选影响云杉林分生产力主导因子，即地貌类型、坡向、腐殖质层厚度、植被类型、土壤湿度、土壤母质堆积形式和土壤平均容重等为主导因子。并对多元线性回归、逐步回归，数量化理论Ⅰ和主成分分析等统计方法在筛选主导因子方面的作用作了评价。
按照筛选出的主导因子，建立云杉生境质量评价和分类系统。即产区——生境区(地貌类型)——生境组(植被类型)——生境型(坡向和土壤母质形式)——生境类型级(土壤湿度和A层土壤厚度)，对划分出的类型和生产力的特征进行论证与描述，并且应用系统聚类、主分量排序方法对分类结果进行定量检验，取得比较一致的结果。另外，为使生境质量评价直观、简捷，便于在生产实践中应用，研制出了云杉(天然林)地位(生境)指数表、多因素数量化生境质量评定表、生境质量等级评定表和云杉生境类型反应表，并对这些表进行检验，阐述它们的应用方法。
为了深入认识森林生态系统的立地指数及其时空动态变化，利用卫星遥感为研究手段，在岷江上游的四川西北部松潘镇江关流域研究云杉森林生态系统立地指数的空间分布特点，探索有关遥感反演模型的建立，并通过精度评估，分析这种高技术应用价值和潜力。研究结果表明，遥感植被指数NDVI和TNDVI与野外实测云杉立地指数(SI)基本为线性相关。通过对模型拟合结果和实际测定结果的比较研究，发现在1：1比例的分析图中，NDVI和TNDVI的遥感反演模型都有很好的拟合效果与较高的精度，说明通过遥感植被指数的方法测定森林立地指数具有较高的实用价值。
2)在云杉林分结构的研究中，利用五种分布函数对328个样地的资料进行了直径——株数(D—N)和树高——株数(H—N)分布的拟合，共计算了约1600个分布函数的拟合与X~2检验，从而证明了在正态分布、对数正态分布、Γ分布、β分布和weibull分布中，weibull分布对云杉林分的D—N分布和H—N分布拟合的效果最好，X~2检验的接受率为64％。通过对weibull分布密度函数中的三个参数：a——分布的位置参数；b——分布的尺度参数；c——分布的形状参数的多元统计分析，揭示了a、b、c三个参数与林分年龄、林分密度和优势木高等林分因子之间的关系，位置参数a与林分密度的关系密切；尺度参数b与优势木高的关系密切；形状参数c与林分密度的关系密切。分析了这些统计特征参数变化受林分因子制约的规律，所建立的逐步回归和多元线性回归方程，可用于测报，即用林分年龄、林分密度和优势木高可以预测出该林分weibull分布的特征参数，从而能详细的表现林分中各径阶和各高度级中株数分布的状况，以进行产量的预测预报，为编制林分收获量表和经营设计方案服务。
同时，分别研究了在多世代云杉纯林中和云冷杉天然混交林中D—N分布和H—N分布的规律。通过研究发现：在云杉多世代纯林中，整个林分的D—N分布与H—N分布有非连续的间断，不论是weibull分布、Γ分布还是其它的分布函数拟合效果都较差，但如果分世代用weibull函数进行拟合，则可获得满意的效果。云冷杉混交林中，全林分的D—N分布与H—N分布拟合效果不佳，分树种后，用weibull函数进行拟合，则效果有显著改善，特别是对居于中下层的桦木，效果最明显。从群落演替，林分发育和世代更替的角度对这一现象进行了分析。
对各林分的D—N分布与H—N分布的特征参数与林分生产力的相关性研究发现，分布的形状、尺度均对生产力有影响，可以利用分布的特征参数预测林分生产力的高低。
3)在林分生产力的研究中，详细地研究了天然云杉林分的生物量和生产力。生物量的调查样地作了202个，共伐树605株，掘取根系54株，并调查了幼树下木生物量样方960个；草本地被物生物量样方上千个。通过获取的大量资料，研究了云杉天然林分的生物量、生产力及其分配。乔木层平均生物量是212．77×10~3·kg hm~(-2)；幼树和下木层生物量是11．395×10~3kg·hm~(-2)；草本地被层生物量分别为2．71×10~3kg·hm~(-2)和1．38×10~3kg·hm~(-2)。林分平均生物量为230．37×10~3kg·hm~(-2)(含枯枝落叶层)。
林分净生产量乔木层为4676kg·hm~(-2)·a~(-1)，折算其光能利用率为1．35％；林分为6838．5 kg·hm~(-2)·a~(-1)，折算其光能利用率为1．63％，林分的总生产力大致为18．85Kcal·m~2·d~(-1)。
为了充分揭示云杉天然林分生产力与生境条件之间的关系，分别将定性与定量的方法相结合，单因子分析与多因子分析相结合，研究了云杉林分生产力与自然分布、林型、海拔、坡向、坡度、坡位、土壤厚度、腐殖质层厚度、土壤质地、土壤含水率、土壤容重等因子之间的相互关系，找出了影响林分生产力的主导因子，并用生态学的观点对其成因进行了分析。
4)在云杉天然林生态系统碳汇功能的研究中，对云杉天然林生态系统中乔木层各器官、林下植被及土壤中的碳含量，以及碳贮量及其空间分布和碳年净固定量进行了系统测定。结果表明：在调查区域，云杉天然林分平均生物量为230．37×10~3·kg hm~(-2)。云杉天然林生态系统各组分的平均碳含量为树干0．5785gc·g~(-1)，树皮0．4712gc·g~(-1)，树枝0．5122gc·g~(-1)，树叶0．4827gc·g~(-1)，树根0．5239gc·g~(-1)，灌木层平均碳含量0．4991gc·g~(-1)，草本层平均碳含量0．4634gc·g~(-1)，地被层平均碳含量0．4321gc·g~(-1)，枯落物层平均碳含量0．3944gc·g~(-1)，土壤碳含量平均值为0．0141gc·g~(-1)，随土层深度增加各层次土壤碳含量逐渐减少。云杉林分生态系统平均总碳贮量为273．79×10~3 kg·hm~(-2)，其中乔木层109．30×10~3kg·hm~(-2)，占云杉林生态系统总碳贮量的39．92％，灌木层5．69×10~3kg·hm~(-2)，占2．08％，草本层1．26×10~3kg·hm~(-2)，占0．46％，地被物层0．60×10~3kg·hm~(-2)，占0．22％，枯落物层0．83×10~3kg·hm~(-2)，占0．30％，林内土壤(0～100cm)碳贮量为156．11×10~3kg·hm~(-2)，占57．01％．云杉林的碳库分布序列为土壤(0～100 cm)>乔木层>灌木层>草本层>枯落物层>地被物层。云杉天然林分平均净生产总量为6838．5kg·hm~(-2)·a~(-1)，碳年总净固量平均为3584．98kg·hm~(-2)·a~(-1)，其中乔木层净生产量为4676 kg·hm~(-2)·a~(-1)，占林分总量的68．38％，碳年平均固定量2552．99 kg·hm~(-2)·a~(-1)，占林分总量的71．21％。研究还表明：云杉天然林分在高山峡谷区的生产力大于台原区；在四种云杉林型中，偏干灌木——云杉林分平均净生产量和碳年净固定量最高，分别为7515．65kg·hm~(-2)·a~(-1)和3779．54kg·hm~(-2)·a~(-1)，其次是中生箭竹——云杉林，分别为7125．75kg·hm~(-2)·g~(-1)和3591．71 kg．hm~(-2)·a~(-1)，再是干性草类——云杉林，分别为6517．0 kg·hm~(-2)·a~(-1)和3272．37kg·hm~(-2)·a~(-1)，湿润苔藓——云杉林平均净生产量和碳年净固量最低，分别只有3672．75kg·hm~(-2)·a~(-1)和1829．46 kg·hm~(-2)·a~(-1)。另外，通过对林分生物量和蓄积量进行多因素立地因子和林分结构因子数量化分析结果：影响林分生物量变化的前10个主导因子是：侧枝夹角、侧枝粗度、针叶密度、土壤平均含水率、林分年龄、叶面积指数、郁闭度、林分密度、海拔梯度和坡度。通过分析认为，在云杉森林生态系统中，林分结构往往是制约林分生产力的主要方面，生境因子相对于结构其位置稍显次要。模型具有较高的复相关系指数，说明有较好的测报作用。
5)在云杉种群增长的研究中，将调查的240个标准地，按照年龄组的序列，分别计算出各年龄组的种群平均生物量，然后按20个年龄组的种群年龄与种群生物量拟合了云杉种群增长的Logistic方程，分别用林分密度等级、海拔梯度、不同云杉林型建立了云杉种群增长的Logistic模型，并运用数量化理论(Ⅰ)方法分别用林分生物量、林分蓄积量构建了不同生境条件云杉种群增长综合评判模型。结果表明：云杉种群生物量增长的基本模式，为“S”型增长，用Logistic方程进行拟合的效果很好。但在不同生境条件下，云杉种群的增长速度和环境容纳量会发生一些有规律的变化。生境条件好，则环境容纳量高，增长速度较快；反之，则低，则慢。影响种群增长的主导因子通过多变量统计分析其序列是树冠冠体结构>土壤水分条件>种群结构>地形因子。
6)在云杉种群调节机理的研究中，分析了云杉种群密度调节的过程和机制，以及个体在不同密度的生境条件下的变化规律，还研究了云杉种群的拥挤效应与云杉密度制约现象，云杉种群中单株增长与密度的关系及其变化规律，另外，还对云杉种群中单株增长与生境因子关系进行了系统研究。结果表明：(1)云杉种群的调节主要是通过密度制约方式进行，在种群呈现自疏时，可以用-3／2幂定律进行描述；(2)生境条件的优劣限制着云杉种群的调控能力，优良的生境往往对密度制约的种群调节起缓冲作用，较差的生境则产生促进的效果，而非密度制约因子(生境条件)对调控能力和调控强度起一种促进或缓冲的作用，这种作用可以由密度竞争效应的乘幂式W=KN~(-α)中的参数α的绝对值来表征，α值偏大，产生促进作用，α值偏小，则起缓冲作用；(3)影响种群对单株增长调节的主导因子是种群空间的光合营养结构(叶面积指数、侧枝夹角、种群密度)>地形因子>土壤营养条件>种群年龄>土壤水分；(4)云杉种群的增长与调节是一个问题的两个方面，两者紧密相关，种群的增长不断受到调节的反馈制约，不断地调整由于增长产生的种群内部的矛盾。对于云杉种群自我调控的方式、涉及到的因素便有了一个比较系统、清楚的认识。这一结果，可为其他木本植物种群自我调控的研究借鉴。
The distribution of picea, stand structure, standing forest productivity and carbon sequestration functions, evaluation of site quality, population growth and adjustment as well were investigated in the high mountain forest region in the northwest of Sichuan province and river area forest regions of Bai Longjiang and Yaojiang in Gansu Province since the 90's with the objective of filling gaps in this field in this paper. At same time, stand structure were investigated based on probability distribution and mathematical statistic methods to seek after the relations between stand diameter, tree height and the number of living tree; the relations between stand productivity, stand structure and habitats were revealed by ecological methods; in addition, we found the judgment system about evaluation of site quality by quantification methods, multi-statistical method and system analysis. The related mathematic models were founded by the analysis of relation between picea population growth and habitats factors based on gaining the Logistic models of various habitats after fitted Logistic models of picea population growth; and the model can reveal the variation of picea population crowding effect and density dependence. It describes four achievements after the systematic studying the relationship of individual growth and habitats factors in this paper
1. The relations among habitats factors were found after studying the degree of correlation of habitats factors by correlation analysis (including simple correlation analysis and canonical correlation analysis) in the part of evaluation of picea site quality. Whereafter, we sought after the dominant factors, which limit greatly the productivity, to evaluate site condition and form type classifications by the multi-statistical method. Finally, some tables including site index list of picea were drew up to describe and evaluate picea site quality quantificationally.
The results showed that: the habitats factors with higher degree of correlation were terrain factors (including landform and exposure) and soil factors (including humus depth and soil moisture), and landform factors and vegetation types have high degree of correlation. the dominant factors that influence greatly the productivity were landform type, exposure, humus depth, vegetation type, soil moisture, cumulus-form of soil parents material and mean soil bulk density. Besides, multiple regression, stepwise regression, quantification method and principal component analysis etc., the statistical methods applied to search for the dominant factors, were evaluated.
The classification system of picea habitats quality evaluation, according to the principle of classification factors with huge capacity and simple-measuring, was established based on the dominant factors from the studying results. And the classification system was divided into five levels: producing area, habitation region (landform type), habitats group (vegetation type), and habitat form (exposure and soil parent form), and site type (soil moisture and soil depth in A layer). At same time, this paper gives the describing and argumentation to the divisory types and the characteristics of productivity. The classification results were similar to the results of quantity test by system cluster and principal component sequencing.
Further, picea (natural forest) site (habitation) index list, multi-factors quantificational habitats quality judge table, habitats quality levels judge table and picea habitats type response table were drew to make the evaluation of site quality strong visual and efficient then easy for applying in the production practice. And we test those tables and lists, explain how to use them to expend the evaluation system in the production practice.
In order to cognize the site index of forest ecosystem and its space-temporal variation, the characteristics of space distribution about Picea forest ecosystem was studied by satellite remote sensing, seeking after the related inverse model foundation on remote sensing, then analyzing the value and potentiality of this high-science & technology in application by related accuracy judgement. The results show that remote sensing vegetation indexes NDVI and TNDVI and site index from field measuring have linear relation on the whole. The inverse model of NDVI and TNDVI show high-fitting and larger accuracy, so it accounts for the high value of remote sensing in application of determining forest site index.
2. The distribution of Diameter-number (D-N) and tree height-number (H-N) were fitted by five distribution functions using the data of 328- plot in the study on picea stand structure, that is ,we fitted 1600 distribution functions and made X~2 tests which proved D-N and H-N were fitted best by weibull distribution with the acceptance rate of 64%in X~2 tests among five distribution functions--normal distribution, lognormaldistribution, gamma distribution, beta distribution and weibull distribution--becauseweibull distribution has many advantages such as strong-flexibility, easy-solution and so on.Multi-statistical analysis to three parameters of weibull distribution densityfunction--location parameter (a), scale parameter (b) and figure parameter (c)--revealed the relations between parameters (a, b, c) and stand age, density, height of dominant tree and other factors. The analysis showed location parameter and figure parameter were intimately related to the stand density, and scale parameter was closely connected with the height of dominant tree. The rule of those statistical characteristicparameters decided by stand factors--stepwise regression function and multi-linearregression function--can be uses to predict the stand yield, through forecasting thecharacteristic parameters of weibull distribution function, which indicated the detail distributions of medium diameter class and the number of plants in each height grades, based on stand age, density and height of dominant tree. So the rule may serve for building the table on stand yield number and managing design scheme.
The distribution rules of D-N and H-N about mickle generation picea forest and natural spruce-fir mixed forest were studied. The data showed that: The distribution of D-N and H-N of mickle generation picea forest present some discrete discontinuities, and they were fitted badly by the five distribution functions, however the fitted results would be improved after separating generation by weibull function. There were not so good fitted results to the distribution of D-N and H-N of spruce-fir mixed forest indicated too, while it can be improved evidently esp. to birch in the middle and lower layer by separating tree species using weibull function. We explain the phenomenon from succession of community, stand growth and generation alternation.
The correlation between characteristic parameter of D-N and H-N and stand productivity were analyzed, and the results show that figure and scale parameter would affect on the yield, so the characteristic parameters of distribution functions can predict the stand productivity.
From the studying, this paper reveal the rules of distribution of D-N and H-N in the natural picea stand structure, and it also gives some new point of view in the choosing of distribution model, multi-statistical analysis of distribution characteristic parameters and stand factors, the analysis of relations between distribution characteristic and stand generation alternation and succession of community, and the correlation between distribution figure and stand productivity; and those views can promote the study of stand structure.
3. The biomass and productivity of picea natural forest were studied in detail in the study of stand productivity. We selected 202 investigational plots, cut down 605 trees, and digged 54 tree-roots to survey biomass. The biomass under young tree was investigated in 960 plots, and the biomass of greensward and ground cover in thousands of plots. The biomass and productivity and distribution of picea natural forest will be gained from huge amounts of material. The results showed: the mean biomass in arbor layer is 212.77×10~3kg·hm~(-2); biomass of young tree and undergrowth is 11.395×10~3kg·hm~(-2), the biomass of greensward and ground cover are 2.71×10~3kg·hm~(-2) and 1.38×10~3kg·hm~(-2) respectively. The mean biomass of forest (including forest floor) is 230.37×10~3kg·hm~(-2).
Stand net primary productivity were gained by calculating according to formula, the data showed: net primary productivity in arbor layer is 4676kg·hm~(-2)·a~(-1) converting 1.35% utilization ratio of light energy; that of stand is 6838.5kg·hm~(-2)·a~(-1) converting 1.63% utilization ratio of light energy; and the whole productivity of forest is about 18.85Kcal·m~2·d~(-1)
We combined qualitative and quantificational methods and bound single-factor analysis with multi-factor analysis in order to reveal fully the relation between stand productivity of picea natural forest and habitats. The correlation among picea standproductivity and habitats factors--natural distribution, forest type, altitude, exposure,slope, slope location, soil depth, humus depth, soil texture, soil moisture, and soil bulk density--were studied, then we found out the dominant factors influencing on the stand productivity, finally, we analyzed the reason by ecology. This study is infrequency in the domestic researches for large number of biomass data, and using together of variousmethods--single-variance regression, multiple variance regression, stepwise regression,and quantification method--in analyzing the relation between productivity and habitatsconditions
The carbon storage, space distribution, annual net amount of fixed carbon, and the carbon density of organs from arbor layer, undergrowth vegetation and soil in the picea natural forest ecosystem were measured systematically in the study of carbon sequestration functions. The results show that the mean biomass of picea natural forest is 230.37×10~3kg·hm~(-2); the components of mean carbon density in the picea natural forest ecosystem indicated that the stem is 0.5785gc·g~(-1), the bark is 0.47 12gc·g~(-1), the branch is 0.5122gc·g~(-1), the leaf is 0.4827gc·g~(-1), and the root is0.5239gc·g~(-1). And the mean carbon densities of bush layer, ground cover, forest floor and soil are 0.4991gc·g~(-1), 0.4634gc·g~(-1), 0.4321gc·g~(-1), 0.3944gc·g~(-1) and 0.0141gc·g~(-1) respectively. Besides, the soil carbon density will decrease with soil height increase gradually. The whole carbon storage of picea forest ecosystem is 273.79×10~3kg·hm~(-2), and the carbon storage in arbor layer, brush layer, herb layer, ground cover, forest floor and forest soil in this ecosystem occupy 39.92%, 2.08%, 0.46%, 0.22%, 0.30% and 57.01% of the whole storage respectively, they are 109.30×10~3 kg·hm~(-2), 5.69×10~3 kg·hm~2, 1.26×10~ 3kg·hm~2, 0.60×10~3kg·hm~(-2), 0.83×10~3kg·hm~(-2), 156.11×10~3 kg·hm~(-2) in order. The Caron stock distribution order of picea forest from large to small is soil(0-100 cm) >arbor layer>brush layer>herb layer>forest floor>ground cover. The mean net productive amount of picea natural forest is 6838.5kg·hm~(-2)·a~(-1) accounting for 63.38%; and annual net amount of fixed carbon is 2552.99 kg·hm~(-2)·a(-1), accounting for 71.21% of the whole amount. The results also indicate that the productivity of picea stand in alpin-canyon is higher than that in the hilly-plateau, and the order of the mean net amount of productivity(MNAP) and annual net amount of fixed carbon(ANAFC) from large to small to the four picea forest types is leaning dry shrub-picea forest with 7515.65kg·hm~(-2)·a~(-1) MNAP and 3779.54kg·hm~(-2)·a~(-1) ANAFC, mid-generation arrow bamboo-picea forest with 7125.75kg·hm~(-2)·a~(-1) MNAP and 3591.71kg·hm~(-2)·a~(-1) ANAFC, drying grass-picea forest with 6517.0kg·hm·a~(-1)MNAP and 3272.37kg·hm~(-2)·a~(-1) ANAFC, and moist-bryophyte-picea forest with 3672.75kg·hm~(-2)·a~(-1) MNAP and 1829.46kg·hm~(-2)·a~(-1) ANAFC. Besides, the quantitative analysis of the multi-factor to site and stand structure through stand biomass and stock showed: the former 10 dominant factors influenced the changing of stand biomass are included angle of branch stem, branch stem thickness, needle density, mean soil moisture, stand age, leaf area index, canopy density, stand density, altitude gratitude, and slope. The results indicate that, on the whole, stand structure is the main aspect of limiting the stand productivity, while habitats factor is slightly unimportant than stand structure. The model has high multiple correlation indexes accounting for the better predicted function.
4. 240 sample plots were indagated according to the age-group order, and then the mean population biomass in each age-group was worked out respectively in the study of picea population growth. The population biomass calculated and the 20 age-groups were fitted by the Logistic function to present picea population growth, and the Logistic models were established based on the data of stand density class, altitude gratitude and various picea stand type. In addition, the comprehensive judge model of picea population growth in different habitats condition was built based on the data of stand biomass and stock by the quantitation method (I) The results show: 1 the essential model of picea population biomass growth show "s" , fitted well by the Logistic function, while, population growth speed and carrying capacity of circumstance would vary on the rule-the carrying capacity of circumstance is high with high speed growth in the superior habitats condition, vice versa. 2 Cui-Lowson model did not apply in the study of picea population for the complex calculation; and the accuracy of fitting population growth by grey-system Verhulst model and GM (1.1) model did not get the satisfied results. 3 the statistical analysis of multi-variation showed that the order of dominant factors of limiting the population growth is tree canopy bulk, soil water, population structure, and landform, however, the order of factors will be changed as to population biomass and stock.
In the study on Accommodation mechanism, the process and mechanism of accommodation of picea population density were analyzed, and crowding effect and density dependence of picea population were investigated besides the alteration of the individual in the habitats condition of different density were investigated because organs constitute one tree in population, at same time, the variation and the relation between individual and component growth and density in the picea population were studied, and we also investigated systematically the relation of organ biomass and population density and the relation between individual growth and habitats factors. The results show: 1 density dependence accommodate the population growth, and it can be descried by the -3/2 power law in the self-scanting process. 2 habitats conditions limit the accommodation capacity of picea population. Non-density dependence factors will prompt or buffer the capacity and strength the accommodation, which can be characterized by the absolute value of a in the power function of density competition effect W=KN~(-a), in which, if a leans great, the function is prompting, otherwise, is buffering. On the whole, the superior habitats will buffer the density dependence of population accommodation, while inferior habitats will accelerate it. We discussed the accommodation of picea population in the individual and component levels. The results indicated: individual accommodation was the comprehensive behavior of component accommodation, and habitats conditions would prompt or buffer the accommodation of two levels. The order of sensibility of responding to accommodation is flower, fruit, branch, leaf, root, stem, and bark. 4 the dominant factors of influence on individual accommodation were structure of mitsukazu nutrient in the population space (leaf area index, branch stem included angle, and population density), landform, soil nutrient, population age, and soil moisture. 5 the growth and accommodation of pices population are the two sides of one thing, and they are intimately correlate. The population growth will be continuously responded and influenced by the accommodation, and the accommodation will adjust the inner conflict from growth.
The work was divided into four parts which had close relation. The evolution of habitats factor and site quality provided the background of research about picea stand structure, population growth and accommodation, and population productivity, which reveal the essential characteristic of potential productivity and the dominant environmental factors that limited the picea stand productivity; while, the research on the variation of picea stand structure in various habitats type, increase-decrease of productivity and the mechanic of population growth and accommodation will further indicate the form and strength of affect on the stand growth and development.
Obviously, habitats condition effect on the stand structure, and stand structure and habitats condition together limited the picea population growth and accommodated the stand productivity. In our work, the relation between habitats conditions, stand structure and productivity has been analyzed qualitative and quantificationally. In addition, the population biomass in the age-group order were fitted by the Logistic function, and we established the Logistic models of picea population growth, various stand density levels, altitude gratitude, and forest type, then built different habitats conditions. In the comprehensive judge model about picea population growth, the bulk dependence and crowding effect were explained; the variation and the relation between individual, component growth and density were investigated systematically based on the theory of population accommodation. The results of this paper can be used in the practice of picea management, directing artificial reforestation, selection of afforestation density and form, and tending and intermediate falling and other practice to actualize the maximum increasing of stand productivity under bringing out the superior stand structure and good management form condition.
1)在云杉生境质量评价的研究中，首先采用相关分析方法，研究生境因子之间的相关程度，找出生境因子间的关系。通过多元统计分析方法，筛选对生产力影响较大的主导因子，然后利用主导因子评价生境条件质量和划分类型。最后，编制出云杉地位指数表等数表。研究结果表明，生境因子中相关程度较高的是地形因子中的地貌、坡向与土壤因子中的腐殖质层厚度、土壤湿度，其次是植被类型。通过多种统计分析方法筛选影响云杉林分生产力主导因子，即地貌类型、坡向、腐殖质层厚度、植被类型、土壤湿度、土壤母质堆积形式和土壤平均容重等为主导因子。并对多元线性回归、逐步回归，数量化理论Ⅰ和主成分分析等统计方法在筛选主导因子方面的作用作了评价。
按照筛选出的主导因子，建立云杉生境质量评价和分类系统。即产区——生境区(地貌类型)——生境组(植被类型)——生境型(坡向和土壤母质形式)——生境类型级(土壤湿度和A层土壤厚度)，对划分出的类型和生产力的特征进行论证与描述，并且应用系统聚类、主分量排序方法对分类结果进行定量检验，取得比较一致的结果。另外，为使生境质量评价直观、简捷，便于在生产实践中应用，研制出了云杉(天然林)地位(生境)指数表、多因素数量化生境质量评定表、生境质量等级评定表和云杉生境类型反应表，并对这些表进行检验，阐述它们的应用方法。
为了深入认识森林生态系统的立地指数及其时空动态变化，利用卫星遥感为研究手段，在岷江上游的四川西北部松潘镇江关流域研究云杉森林生态系统立地指数的空间分布特点，探索有关遥感反演模型的建立，并通过精度评估，分析这种高技术应用价值和潜力。研究结果表明，遥感植被指数NDVI和TNDVI与野外实测云杉立地指数(SI)基本为线性相关。通过对模型拟合结果和实际测定结果的比较研究，发现在1：1比例的分析图中，NDVI和TNDVI的遥感反演模型都有很好的拟合效果与较高的精度，说明通过遥感植被指数的方法测定森林立地指数具有较高的实用价值。
2)在云杉林分结构的研究中，利用五种分布函数对328个样地的资料进行了直径——株数(D—N)和树高——株数(H—N)分布的拟合，共计算了约1600个分布函数的拟合与X~2检验，从而证明了在正态分布、对数正态分布、Γ分布、β分布和weibull分布中，weibull分布对云杉林分的D—N分布和H—N分布拟合的效果最好，X~2检验的接受率为64％。通过对weibull分布密度函数中的三个参数：a——分布的位置参数；b——分布的尺度参数；c——分布的形状参数的多元统计分析，揭示了a、b、c三个参数与林分年龄、林分密度和优势木高等林分因子之间的关系，位置参数a与林分密度的关系密切；尺度参数b与优势木高的关系密切；形状参数c与林分密度的关系密切。分析了这些统计特征参数变化受林分因子制约的规律，所建立的逐步回归和多元线性回归方程，可用于测报，即用林分年龄、林分密度和优势木高可以预测出该林分weibull分布的特征参数，从而能详细的表现林分中各径阶和各高度级中株数分布的状况，以进行产量的预测预报，为编制林分收获量表和经营设计方案服务。
同时，分别研究了在多世代云杉纯林中和云冷杉天然混交林中D—N分布和H—N分布的规律。通过研究发现：在云杉多世代纯林中，整个林分的D—N分布与H—N分布有非连续的间断，不论是weibull分布、Γ分布还是其它的分布函数拟合效果都较差，但如果分世代用weibull函数进行拟合，则可获得满意的效果。云冷杉混交林中，全林分的D—N分布与H—N分布拟合效果不佳，分树种后，用weibull函数进行拟合，则效果有显著改善，特别是对居于中下层的桦木，效果最明显。从群落演替，林分发育和世代更替的角度对这一现象进行了分析。
对各林分的D—N分布与H—N分布的特征参数与林分生产力的相关性研究发现，分布的形状、尺度均对生产力有影响，可以利用分布的特征参数预测林分生产力的高低。
3)在林分生产力的研究中，详细地研究了天然云杉林分的生物量和生产力。生物量的调查样地作了202个，共伐树605株，掘取根系54株，并调查了幼树下木生物量样方960个；草本地被物生物量样方上千个。通过获取的大量资料，研究了云杉天然林分的生物量、生产力及其分配。乔木层平均生物量是212．77×10~3·kg hm~(-2)；幼树和下木层生物量是11．395×10~3kg·hm~(-2)；草本地被层生物量分别为2．71×10~3kg·hm~(-2)和1．38×10~3kg·hm~(-2)。林分平均生物量为230．37×10~3kg·hm~(-2)(含枯枝落叶层)。
林分净生产量乔木层为4676kg·hm~(-2)·a~(-1)，折算其光能利用率为1．35％；林分为6838．5 kg·hm~(-2)·a~(-1)，折算其光能利用率为1．63％，林分的总生产力大致为18．85Kcal·m~2·d~(-1)。
为了充分揭示云杉天然林分生产力与生境条件之间的关系，分别将定性与定量的方法相结合，单因子分析与多因子分析相结合，研究了云杉林分生产力与自然分布、林型、海拔、坡向、坡度、坡位、土壤厚度、腐殖质层厚度、土壤质地、土壤含水率、土壤容重等因子之间的相互关系，找出了影响林分生产力的主导因子，并用生态学的观点对其成因进行了分析。
4)在云杉天然林生态系统碳汇功能的研究中，对云杉天然林生态系统中乔木层各器官、林下植被及土壤中的碳含量，以及碳贮量及其空间分布和碳年净固定量进行了系统测定。结果表明：在调查区域，云杉天然林分平均生物量为230．37×10~3·kg hm~(-2)。云杉天然林生态系统各组分的平均碳含量为树干0．5785gc·g~(-1)，树皮0．4712gc·g~(-1)，树枝0．5122gc·g~(-1)，树叶0．4827gc·g~(-1)，树根0．5239gc·g~(-1)，灌木层平均碳含量0．4991gc·g~(-1)，草本层平均碳含量0．4634gc·g~(-1)，地被层平均碳含量0．4321gc·g~(-1)，枯落物层平均碳含量0．3944gc·g~(-1)，土壤碳含量平均值为0．0141gc·g~(-1)，随土层深度增加各层次土壤碳含量逐渐减少。云杉林分生态系统平均总碳贮量为273．79×10~3 kg·hm~(-2)，其中乔木层109．30×10~3kg·hm~(-2)，占云杉林生态系统总碳贮量的39．92％，灌木层5．69×10~3kg·hm~(-2)，占2．08％，草本层1．26×10~3kg·hm~(-2)，占0．46％，地被物层0．60×10~3kg·hm~(-2)，占0．22％，枯落物层0．83×10~3kg·hm~(-2)，占0．30％，林内土壤(0～100cm)碳贮量为156．11×10~3kg·hm~(-2)，占57．01％．云杉林的碳库分布序列为土壤(0～100 cm)>乔木层>灌木层>草本层>枯落物层>地被物层。云杉天然林分平均净生产总量为6838．5kg·hm~(-2)·a~(-1)，碳年总净固量平均为3584．98kg·hm~(-2)·a~(-1)，其中乔木层净生产量为4676 kg·hm~(-2)·a~(-1)，占林分总量的68．38％，碳年平均固定量2552．99 kg·hm~(-2)·a~(-1)，占林分总量的71．21％。研究还表明：云杉天然林分在高山峡谷区的生产力大于台原区；在四种云杉林型中，偏干灌木——云杉林分平均净生产量和碳年净固定量最高，分别为7515．65kg·hm~(-2)·a~(-1)和3779．54kg·hm~(-2)·a~(-1)，其次是中生箭竹——云杉林，分别为7125．75kg·hm~(-2)·g~(-1)和3591．71 kg．hm~(-2)·a~(-1)，再是干性草类——云杉林，分别为6517．0 kg·hm~(-2)·a~(-1)和3272．37kg·hm~(-2)·a~(-1)，湿润苔藓——云杉林平均净生产量和碳年净固量最低，分别只有3672．75kg·hm~(-2)·a~(-1)和1829．46 kg·hm~(-2)·a~(-1)。另外，通过对林分生物量和蓄积量进行多因素立地因子和林分结构因子数量化分析结果：影响林分生物量变化的前10个主导因子是：侧枝夹角、侧枝粗度、针叶密度、土壤平均含水率、林分年龄、叶面积指数、郁闭度、林分密度、海拔梯度和坡度。通过分析认为，在云杉森林生态系统中，林分结构往往是制约林分生产力的主要方面，生境因子相对于结构其位置稍显次要。模型具有较高的复相关系指数，说明有较好的测报作用。
5)在云杉种群增长的研究中，将调查的240个标准地，按照年龄组的序列，分别计算出各年龄组的种群平均生物量，然后按20个年龄组的种群年龄与种群生物量拟合了云杉种群增长的Logistic方程，分别用林分密度等级、海拔梯度、不同云杉林型建立了云杉种群增长的Logistic模型，并运用数量化理论(Ⅰ)方法分别用林分生物量、林分蓄积量构建了不同生境条件云杉种群增长综合评判模型。结果表明：云杉种群生物量增长的基本模式，为“S”型增长，用Logistic方程进行拟合的效果很好。但在不同生境条件下，云杉种群的增长速度和环境容纳量会发生一些有规律的变化。生境条件好，则环境容纳量高，增长速度较快；反之，则低，则慢。影响种群增长的主导因子通过多变量统计分析其序列是树冠冠体结构>土壤水分条件>种群结构>地形因子。
6)在云杉种群调节机理的研究中，分析了云杉种群密度调节的过程和机制，以及个体在不同密度的生境条件下的变化规律，还研究了云杉种群的拥挤效应与云杉密度制约现象，云杉种群中单株增长与密度的关系及其变化规律，另外，还对云杉种群中单株增长与生境因子关系进行了系统研究。结果表明：(1)云杉种群的调节主要是通过密度制约方式进行，在种群呈现自疏时，可以用-3／2幂定律进行描述；(2)生境条件的优劣限制着云杉种群的调控能力，优良的生境往往对密度制约的种群调节起缓冲作用，较差的生境则产生促进的效果，而非密度制约因子(生境条件)对调控能力和调控强度起一种促进或缓冲的作用，这种作用可以由密度竞争效应的乘幂式W=KN~(-α)中的参数α的绝对值来表征，α值偏大，产生促进作用，α值偏小，则起缓冲作用；(3)影响种群对单株增长调节的主导因子是种群空间的光合营养结构(叶面积指数、侧枝夹角、种群密度)>地形因子>土壤营养条件>种群年龄>土壤水分；(4)云杉种群的增长与调节是一个问题的两个方面，两者紧密相关，种群的增长不断受到调节的反馈制约，不断地调整由于增长产生的种群内部的矛盾。对于云杉种群自我调控的方式、涉及到的因素便有了一个比较系统、清楚的认识。这一结果，可为其他木本植物种群自我调控的研究借鉴。
The distribution of picea, stand structure, standing forest productivity and carbon sequestration functions, evaluation of site quality, population growth and adjustment as well were investigated in the high mountain forest region in the northwest of Sichuan province and river area forest regions of Bai Longjiang and Yaojiang in Gansu Province since the 90's with the objective of filling gaps in this field in this paper. At same time, stand structure were investigated based on probability distribution and mathematical statistic methods to seek after the relations between stand diameter, tree height and the number of living tree; the relations between stand productivity, stand structure and habitats were revealed by ecological methods; in addition, we found the judgment system about evaluation of site quality by quantification methods, multi-statistical method and system analysis. The related mathematic models were founded by the analysis of relation between picea population growth and habitats factors based on gaining the Logistic models of various habitats after fitted Logistic models of picea population growth; and the model can reveal the variation of picea population crowding effect and density dependence. It describes four achievements after the systematic studying the relationship of individual growth and habitats factors in this paper
1. The relations among habitats factors were found after studying the degree of correlation of habitats factors by correlation analysis (including simple correlation analysis and canonical correlation analysis) in the part of evaluation of picea site quality. Whereafter, we sought after the dominant factors, which limit greatly the productivity, to evaluate site condition and form type classifications by the multi-statistical method. Finally, some tables including site index list of picea were drew up to describe and evaluate picea site quality quantificationally.
The results showed that: the habitats factors with higher degree of correlation were terrain factors (including landform and exposure) and soil factors (including humus depth and soil moisture), and landform factors and vegetation types have high degree of correlation. the dominant factors that influence greatly the productivity were landform type, exposure, humus depth, vegetation type, soil moisture, cumulus-form of soil parents material and mean soil bulk density. Besides, multiple regression, stepwise regression, quantification method and principal component analysis etc., the statistical methods applied to search for the dominant factors, were evaluated.
The classification system of picea habitats quality evaluation, according to the principle of classification factors with huge capacity and simple-measuring, was established based on the dominant factors from the studying results. And the classification system was divided into five levels: producing area, habitation region (landform type), habitats group (vegetation type), and habitat form (exposure and soil parent form), and site type (soil moisture and soil depth in A layer). At same time, this paper gives the describing and argumentation to the divisory types and the characteristics of productivity. The classification results were similar to the results of quantity test by system cluster and principal component sequencing.
Further, picea (natural forest) site (habitation) index list, multi-factors quantificational habitats quality judge table, habitats quality levels judge table and picea habitats type response table were drew to make the evaluation of site quality strong visual and efficient then easy for applying in the production practice. And we test those tables and lists, explain how to use them to expend the evaluation system in the production practice.
In order to cognize the site index of forest ecosystem and its space-temporal variation, the characteristics of space distribution about Picea forest ecosystem was studied by satellite remote sensing, seeking after the related inverse model foundation on remote sensing, then analyzing the value and potentiality of this high-science & technology in application by related accuracy judgement. The results show that remote sensing vegetation indexes NDVI and TNDVI and site index from field measuring have linear relation on the whole. The inverse model of NDVI and TNDVI show high-fitting and larger accuracy, so it accounts for the high value of remote sensing in application of determining forest site index.
2. The distribution of Diameter-number (D-N) and tree height-number (H-N) were fitted by five distribution functions using the data of 328- plot in the study on picea stand structure, that is ,we fitted 1600 distribution functions and made X~2 tests which proved D-N and H-N were fitted best by weibull distribution with the acceptance rate of 64%in X~2 tests among five distribution functions--normal distribution, lognormaldistribution, gamma distribution, beta distribution and weibull distribution--becauseweibull distribution has many advantages such as strong-flexibility, easy-solution and so on.Multi-statistical analysis to three parameters of weibull distribution densityfunction--location parameter (a), scale parameter (b) and figure parameter (c)--revealed the relations between parameters (a, b, c) and stand age, density, height of dominant tree and other factors. The analysis showed location parameter and figure parameter were intimately related to the stand density, and scale parameter was closely connected with the height of dominant tree. The rule of those statistical characteristicparameters decided by stand factors--stepwise regression function and multi-linearregression function--can be uses to predict the stand yield, through forecasting thecharacteristic parameters of weibull distribution function, which indicated the detail distributions of medium diameter class and the number of plants in each height grades, based on stand age, density and height of dominant tree. So the rule may serve for building the table on stand yield number and managing design scheme.
The distribution rules of D-N and H-N about mickle generation picea forest and natural spruce-fir mixed forest were studied. The data showed that: The distribution of D-N and H-N of mickle generation picea forest present some discrete discontinuities, and they were fitted badly by the five distribution functions, however the fitted results would be improved after separating generation by weibull function. There were not so good fitted results to the distribution of D-N and H-N of spruce-fir mixed forest indicated too, while it can be improved evidently esp. to birch in the middle and lower layer by separating tree species using weibull function. We explain the phenomenon from succession of community, stand growth and generation alternation.
The correlation between characteristic parameter of D-N and H-N and stand productivity were analyzed, and the results show that figure and scale parameter would affect on the yield, so the characteristic parameters of distribution functions can predict the stand productivity.
From the studying, this paper reveal the rules of distribution of D-N and H-N in the natural picea stand structure, and it also gives some new point of view in the choosing of distribution model, multi-statistical analysis of distribution characteristic parameters and stand factors, the analysis of relations between distribution characteristic and stand generation alternation and succession of community, and the correlation between distribution figure and stand productivity; and those views can promote the study of stand structure.
3. The biomass and productivity of picea natural forest were studied in detail in the study of stand productivity. We selected 202 investigational plots, cut down 605 trees, and digged 54 tree-roots to survey biomass. The biomass under young tree was investigated in 960 plots, and the biomass of greensward and ground cover in thousands of plots. The biomass and productivity and distribution of picea natural forest will be gained from huge amounts of material. The results showed: the mean biomass in arbor layer is 212.77×10~3kg·hm~(-2); biomass of young tree and undergrowth is 11.395×10~3kg·hm~(-2), the biomass of greensward and ground cover are 2.71×10~3kg·hm~(-2) and 1.38×10~3kg·hm~(-2) respectively. The mean biomass of forest (including forest floor) is 230.37×10~3kg·hm~(-2).
Stand net primary productivity were gained by calculating according to formula, the data showed: net primary productivity in arbor layer is 4676kg·hm~(-2)·a~(-1) converting 1.35% utilization ratio of light energy; that of stand is 6838.5kg·hm~(-2)·a~(-1) converting 1.63% utilization ratio of light energy; and the whole productivity of forest is about 18.85Kcal·m~2·d~(-1)
We combined qualitative and quantificational methods and bound single-factor analysis with multi-factor analysis in order to reveal fully the relation between stand productivity of picea natural forest and habitats. The correlation among picea standproductivity and habitats factors--natural distribution, forest type, altitude, exposure,slope, slope location, soil depth, humus depth, soil texture, soil moisture, and soil bulk density--were studied, then we found out the dominant factors influencing on the stand productivity, finally, we analyzed the reason by ecology. This study is infrequency in the domestic researches for large number of biomass data, and using together of variousmethods--single-variance regression, multiple variance regression, stepwise regression,and quantification method--in analyzing the relation between productivity and habitatsconditions
The carbon storage, space distribution, annual net amount of fixed carbon, and the carbon density of organs from arbor layer, undergrowth vegetation and soil in the picea natural forest ecosystem were measured systematically in the study of carbon sequestration functions. The results show that the mean biomass of picea natural forest is 230.37×10~3kg·hm~(-2); the components of mean carbon density in the picea natural forest ecosystem indicated that the stem is 0.5785gc·g~(-1), the bark is 0.47 12gc·g~(-1), the branch is 0.5122gc·g~(-1), the leaf is 0.4827gc·g~(-1), and the root is0.5239gc·g~(-1). And the mean carbon densities of bush layer, ground cover, forest floor and soil are 0.4991gc·g~(-1), 0.4634gc·g~(-1), 0.4321gc·g~(-1), 0.3944gc·g~(-1) and 0.0141gc·g~(-1) respectively. Besides, the soil carbon density will decrease with soil height increase gradually. The whole carbon storage of picea forest ecosystem is 273.79×10~3kg·hm~(-2), and the carbon storage in arbor layer, brush layer, herb layer, ground cover, forest floor and forest soil in this ecosystem occupy 39.92%, 2.08%, 0.46%, 0.22%, 0.30% and 57.01% of the whole storage respectively, they are 109.30×10~3 kg·hm~(-2), 5.69×10~3 kg·hm~2, 1.26×10~ 3kg·hm~2, 0.60×10~3kg·hm~(-2), 0.83×10~3kg·hm~(-2), 156.11×10~3 kg·hm~(-2) in order. The Caron stock distribution order of picea forest from large to small is soil(0-100 cm) >arbor layer>brush layer>herb layer>forest floor>ground cover. The mean net productive amount of picea natural forest is 6838.5kg·hm~(-2)·a~(-1) accounting for 63.38%; and annual net amount of fixed carbon is 2552.99 kg·hm~(-2)·a(-1), accounting for 71.21% of the whole amount. The results also indicate that the productivity of picea stand in alpin-canyon is higher than that in the hilly-plateau, and the order of the mean net amount of productivity(MNAP) and annual net amount of fixed carbon(ANAFC) from large to small to the four picea forest types is leaning dry shrub-picea forest with 7515.65kg·hm~(-2)·a~(-1) MNAP and 3779.54kg·hm~(-2)·a~(-1) ANAFC, mid-generation arrow bamboo-picea forest with 7125.75kg·hm~(-2)·a~(-1) MNAP and 3591.71kg·hm~(-2)·a~(-1) ANAFC, drying grass-picea forest with 6517.0kg·hm·a~(-1)MNAP and 3272.37kg·hm~(-2)·a~(-1) ANAFC, and moist-bryophyte-picea forest with 3672.75kg·hm~(-2)·a~(-1) MNAP and 1829.46kg·hm~(-2)·a~(-1) ANAFC. Besides, the quantitative analysis of the multi-factor to site and stand structure through stand biomass and stock showed: the former 10 dominant factors influenced the changing of stand biomass are included angle of branch stem, branch stem thickness, needle density, mean soil moisture, stand age, leaf area index, canopy density, stand density, altitude gratitude, and slope. The results indicate that, on the whole, stand structure is the main aspect of limiting the stand productivity, while habitats factor is slightly unimportant than stand structure. The model has high multiple correlation indexes accounting for the better predicted function.
4. 240 sample plots were indagated according to the age-group order, and then the mean population biomass in each age-group was worked out respectively in the study of picea population growth. The population biomass calculated and the 20 age-groups were fitted by the Logistic function to present picea population growth, and the Logistic models were established based on the data of stand density class, altitude gratitude and various picea stand type. In addition, the comprehensive judge model of picea population growth in different habitats condition was built based on the data of stand biomass and stock by the quantitation method (I) The results show: 1 the essential model of picea population biomass growth show "s" , fitted well by the Logistic function, while, population growth speed and carrying capacity of circumstance would vary on the rule-the carrying capacity of circumstance is high with high speed growth in the superior habitats condition, vice versa. 2 Cui-Lowson model did not apply in the study of picea population for the complex calculation; and the accuracy of fitting population growth by grey-system Verhulst model and GM (1.1) model did not get the satisfied results. 3 the statistical analysis of multi-variation showed that the order of dominant factors of limiting the population growth is tree canopy bulk, soil water, population structure, and landform, however, the order of factors will be changed as to population biomass and stock.
In the study on Accommodation mechanism, the process and mechanism of accommodation of picea population density were analyzed, and crowding effect and density dependence of picea population were investigated besides the alteration of the individual in the habitats condition of different density were investigated because organs constitute one tree in population, at same time, the variation and the relation between individual and component growth and density in the picea population were studied, and we also investigated systematically the relation of organ biomass and population density and the relation between individual growth and habitats factors. The results show: 1 density dependence accommodate the population growth, and it can be descried by the -3/2 power law in the self-scanting process. 2 habitats conditions limit the accommodation capacity of picea population. Non-density dependence factors will prompt or buffer the capacity and strength the accommodation, which can be characterized by the absolute value of a in the power function of density competition effect W=KN~(-a), in which, if a leans great, the function is prompting, otherwise, is buffering. On the whole, the superior habitats will buffer the density dependence of population accommodation, while inferior habitats will accelerate it. We discussed the accommodation of picea population in the individual and component levels. The results indicated: individual accommodation was the comprehensive behavior of component accommodation, and habitats conditions would prompt or buffer the accommodation of two levels. The order of sensibility of responding to accommodation is flower, fruit, branch, leaf, root, stem, and bark. 4 the dominant factors of influence on individual accommodation were structure of mitsukazu nutrient in the population space (leaf area index, branch stem included angle, and population density), landform, soil nutrient, population age, and soil moisture. 5 the growth and accommodation of pices population are the two sides of one thing, and they are intimately correlate. The population growth will be continuously responded and influenced by the accommodation, and the accommodation will adjust the inner conflict from growth.
The work was divided into four parts which had close relation. The evolution of habitats factor and site quality provided the background of research about picea stand structure, population growth and accommodation, and population productivity, which reveal the essential characteristic of potential productivity and the dominant environmental factors that limited the picea stand productivity; while, the research on the variation of picea stand structure in various habitats type, increase-decrease of productivity and the mechanic of population growth and accommodation will further indicate the form and strength of affect on the stand growth and development.
Obviously, habitats condition effect on the stand structure, and stand structure and habitats condition together limited the picea population growth and accommodated the stand productivity. In our work, the relation between habitats conditions, stand structure and productivity has been analyzed qualitative and quantificationally. In addition, the population biomass in the age-group order were fitted by the Logistic function, and we established the Logistic models of picea population growth, various stand density levels, altitude gratitude, and forest type, then built different habitats conditions. In the comprehensive judge model about picea population growth, the bulk dependence and crowding effect were explained; the variation and the relation between individual, component growth and density were investigated systematically based on the theory of population accommodation. The results of this paper can be used in the practice of picea management, directing artificial reforestation, selection of afforestation density and form, and tending and intermediate falling and other practice to actualize the maximum increasing of stand productivity under bringing out the superior stand structure and good management form condition.
引文
[1]蒋有绪，川西米亚罗、马尔康高山林区生境类型的初步研究，林业科学，1963，8(4)：321-335
[2]管中天，四川松杉植物地理，成都，四川人民出版社，1982
[3]刘翼，杉木地位质量的数量化评价方法，中国林科院研究报告，1982，(1)：29-41
[4]南方14省(区)杉木协作组，杉木立地条件的系统研究及应用，林业科学，1983，(19)：246-254
[5]成子纯，马尾松经营体系模拟系统，北京，中国林业出版社，1991
[6]马明东，刘跃建，云杉天然林地位指数表的编制与应用，四川林业科技，1986，7(3)：10-16
[7] 马明东，楠竹林分立地质量评价及类型划分研究，竹子研究汇刊，1991，10(3)：49-60
[8]马明东，刘跃建，四川綦江森林立地系统研究，论文集，成都，西南交大出版社，1993
[9]唐守正，多元统计分析方法，北京，中国林业出版社，1986
[10]罗积玉．微机用多元统计分析软件，成都，四川人民出版社，1986
[11]马明东，四川西南部高山云杉天然林生产力与立地生态因子关系研究，中国首届植物生态学青年研究，论 文集，北京，中国植物学会，1987
[12]马明东，云杉天然林分生产力及生态条件的关系，国际林联山地森林保护与管理论文集，北京，科学出版社， 1990，9
[13]马明东，云杉适生立地条件的研究，中国林学会第二届造林学术讨论会论文集，1988
[14]马明东，森林生境质量评价及其分类评述，四川林业科技，1993，14(4)：30-35
[15]马明东，云杉天然林分结构、生产力状况及生境质量评价研究，中国技术成果大会，1992，北京，科学技术 文献出版社
[16]马明东，三江流域、岷江上游生态区数量分类，自然资源，1992(2)：62-67
[17] 马明东，罗承德，张健，胡庭兴等，云杉天然林分多因素数量化生境质量评价研究，中国生态农业学报， 2006，(14)2
[18]马明东，罗承德，张健，胡庭兴等，云杉天然林分生境条件数量分类，中国生态农业学报，2006，(14)2
[19]马明东，四川盆地西绿楠木人工林分生物量的研究，四川林业科技，1989，10(3)：6-14
[20]马明东，罗承德，张健，胡庭兴等，四种数学方法对暗针叶云杉林分生境属性比较分析研究，中国生态农 业学报，2006，(14)1196-201
[21]马明东，江洪等，森林生态系统立地指数的遥感分析，生态学报，2006，26(9)：2810-2816
[22] Hampf, Frederick E.Site index curves for some forest species in the eastern United States. United States. ForestService. Eastern Region. Upper Darby, Pa., 1965.
[23] Lamson, N.I. Estimating northern red oak site-index class from total height and diameter of dominant andcodominant trees in central Appalachian hardwood stands. USDA Forest Service Research Paper NE-RP -Northeastern Forest Experiment Station, Aug 1987. (605), 3 p.
[24] The editing group of Forest Site Classification of China. The forest site classification of China. Chinese ForestryPress, Beijing. 1989.
[25] Ma Mingdong, Liu Yue-Jian. Quantitative classification of site condition in natural spruce forest [J], ChineseJournal of Eco-Agriculture[J], 2006,14(2) (in press).
[26] Carvalho, J.P.; Parresol, B.R. A site model for Pyrenean oak (Quercus pyrenaica) stands using a dynamic algebraicdifference equation[J]. Canadian journal of forest research, 2005.35 (1) 93-99.
[27] Carmean, Willard H. A comparison of site index curves for northern hardwood Species. North Central ForestExperiment Station (Saint Paul, Minn.); United States. Forest Service. North Central Forest Experiment Station,??Forest Service, U.S. Dept of Agriculture, 1979.
[28] Ingrid Seynave, Jean-Claude Gegout, Jean-Christophe Herve, Jean-Francois Dhote, and Jacques Drapier. Picieaabies site index prediction by environmental factors and understory vegetation: a two-scale approach based onsurvey databases. Canadian Journal of Forest Research[J], 2005.35(1)1669-1678.
[29] Clatterbuck, W.K. Height growth and site index curves for cherrybark oak and sweetgum in mixed, even-agedstands on the minor bottoms of central Mississippi. Southern journal of applied forestry[J], 1987,11 (4)219-222.
[30] Jokela, E.J.; Jack, S.B.; Nowak, C.A. Site index curves for unthinned Norway spruce plantations in New York.Northern journal of applied forestry[J], 1988,5 (4) 251-254.
[31] Boyer, W.D. A generational change in site index for naturally established longleaf pine on a South AlabamaCoastal Plain site[J]. Southern journal of applied forestry, 2001,25 (2) 88-92.
[32] Kimmins, J. P. Forest Ecology (Third Edition). Prentice Hall. New Jersey. 2003.
[33] Nigh, GD.The geometric mean regression line: a method for developing site index conversion equations forspecies in mixed stands[J]. Forest science, 1995,41 (1) 84-98.
[34] Chen, H.Y.H.; Klinka, K.; Kabzems, R.D. Site index, site quality, and foliar nutrients of trembling aspen:relationships and predictions. Canadian journal of forest research[J], 1998,28 (12) 1743-1755.
[35] Ung, C.H.; Bernier, P.Y.; Raulier, F.; Fournier, R.A.; Lambert, M.C.; Regniere, J. Biophysical site indices for shadetolerant and intolerant boreal species[J]. Forest science, 2001,47 (1) 83-95.
[36] O'Hara, K.L.; Oliver, CD. Three-dimensional representation of Douglas-fir volume growth: comparison of growthand yield models with stand data[J]. Forest science, 1988,43(6) 724-743.
[37] Walters, D.K.; Sloan, J.P.; Kurmis, Aspen site index as related to plant indicators.USDA Forest Service generaltechnical report NC - North Central Forest Experiment Station, 1990. (140), p. 337-341.
[38] Jiang Hong, Strittholt James R.,Frost, P.A. and Slosser, N.C. The classification of late serai forests in the PacificNorthwest, USA using Landsat ETM+ imagery[J]. Remote Sensing of Environment, 2004,91:320-331.
[39] Running, S.W.,Peterson,D.L.,Spanner,M,A.andTeuber,K.B. Remote Sensing of coniferous forest leaf area[J].Ecology, 1986,67:273-276.
[40] Jensen, J.R.. Remote sensing of the environment, Prentice Hall. New Jersey. 2000.
[41] Jiang Hong, Apps Michael J. Zhang Yanli, Peng Changhui and Woodard Paul M . Modeling the spatial pattern ofnet primary productivity in Chinese forests [J]. Ecological Modeling 1999,122:275-288.
[42] Chen. J.M. Derivation and validation of Canada-wide coarse-resolution leaf area index maps using high-resolutionsatellite imagery and ground measurements [J]. Remote Sens Ennviron, 2001,80:1-20.
[43] Chen, J.M. and J.Cihlar. Retrieving Leaf Area Index of Boreal Conifer Forests Using Landsat TM Images [J].Remote Sens Environ, 1996,55: 153-162.
[44] Qi, J. Leaf Area Index Estimates Using Remotely Sensed Data and BRDF Models in a Semiarid Region [J].Remote Sens Environ, 2000,73:18-30.
[45] Hong Jiang. The preliminary analysis of net primary productivity in natural spruce forests[J]. Acta Botanica Sinica,1986,28 (5) 538-548.
[46] Jiang Hong, Ma Mingdong. The relation between the natural spruce forest Productivity and the ecologicalenvironment Protection and management of mountain forests, Science Press, Beijing, 1992.
[47] Ruiliang, Pu, and Peng Gong. The application of hyperspectram remote sensing. Beijing, Higher Education Press.2000.
[48] Shugart, H.H. 1998. Terrestrial Ecosystems in changing environments. Cambridge. Cambridge University Press.
[49] Peng Changhui, Jiang, Hong, Apps, Michael J. and Zhang, Yanli. Simulating the Effects of harvesting regimes oncarbon and nitrogen dynamics of boreal forests in Central Canada using a process model [J]. Ecological Modeling.2002,155:177-189
[50]浦瑞良，宫鹏．高光谱遥感及其应用．北京：高等教育出版社，2000．
[51]江洪，紫果云杉天然中龄林分生物量和生产力的研究，植物生态学与地植物学学报，1985，10(2)：146-152
[52]李文华，小兴安岭谷地云、冷杉林群落结构和演替的研究．自然资源，1980(4)：17-29
[53]毕国昌，关于西南高山林区林型分类的几个问题，林业科学，1964，9(1)：86-92
[54]木村允，陆地植物群落的生产量测定法(姜恕等译)，北京，科学出版社，1981
[55]马明东，罗承德等，四川西北部亚高山云杉天然林生态系统碳含量、净生产量和碳贮量的初步研究，植物 生态学报，2007，31(2)：305-312
[56] Woodwell GM.Whittaker R H.Reiners WA.etal.The biota and the World carbon budget Science,1978,199:141-146 Houghton, R.A, Hackler, J.L., Lawrence, K.T. 1999. The US carbon budget: Contributions from land-usechange. Science, 285,574-578.
[57] Post WM. Emanuel WR. Zinke PJ. etal. Soil Pools and World life zone Nature, 1982,298:156-159
[58] Jiang, H., Apps, M.J, Zhang ,Y.L, et al., 1999. Modelling the spatial pattern of net primary productivity inChinese forests, Ecological Modeling, 122,275-288.
[59] Fang ,J.Y, Wang, GG, Liu, G H., et al., 1998. Forest biomass of China: an estimation based on thebiomass-volume relationship. Ecological Applications, 8,1084-1091.
[60] Fang, J.Y., Chen, A. P., Peng, C.H ,et al., 2001. Changes in forest biomass carbon storage in China between 1949and 1998. Science, 292,2320-2322.
[61] Jiang, H. Apps, M.L., Peng, C.H., et al., 2002. Modeling the influence of harvesting on Chinese boreal forestcarbon dynamics. Forest Ecology and Management, 169,65-82.
[62] Lei peifeng(雷丕锋)，Xiang Wenhua(项文化)，Tian Dalun(田大伦)，2004，Carbon storage and distribution in Cinna momum camphor Plantstion．[J]Chinese Journal of Ecology(生态学杂志)23(4)：25-30(in chinses)
[63]Fang Xi(方晰)，Tian Dalun(田大伦)，2003，Productivity and Carbon Dynamics of Masson Pine Plantation．[J] Journal of Central South Forestry University(中南林学院学报)23(2)：11-15(in chinese)
[64]Zhou Guomo(周国模)，Jiang Peikun(姜培坤)，2004,Density Storage and Spatial Distribution of Carbon in Phyllostachy Pubescens Forest．[J]Scientia Silvae Sinical(林业科学)40(6)：20～24(in chinses)
[65]Hong Jiang(江洪)，1986，The preliminary analysis of net primary productivity in natural spruce forests[J]．Acta Botanica Sinica(植物学报)28(5)538-548(in chinses)
[66]陈灵芝，长白山西南坡鱼鳞云杉林分结构的初步研究，植物生态学与地植物学丛刊，1963，1(1-2)：69-80
[67] 尚玉昌，种群的特征、动态和数量调节，陆地生态学译报(连载)，1985-1987
[68]董文泉，数量化理论及其应用，长春，吉林人民出版社，1979
[69]熊惠，川西高山林区的森林土壤，西南高山林区综合考察报告，1963，107-119
[70]詹昭宁等编，森林收获量预报，北京，中国林业出版社，1986
[71]董鸣，缙云山马尾松种群数量动态初步研究，植物生态学与地植物学学报，1986，10(4)：283-293
[72]蒋有绪，川西亚高山暗针叶林群落特点及其分类原则，植物生态学与地植物学丛刊，1963，1(1-2)：42-50
[73]蒋有绪，川西亚高山森林植被的区系种间关系和群落排序，植物生态学与地植物学丛刊，1982，6(4)：281-301
[74]葛剑平，小兴安岭原始红松林的结构与动态，东北林业大学，研究生论文集，1987
[75] 周德彰，杨玉坡，四川亚高山云冷杉林采伐迹地生态因子的变化，林业科学，1984，20(2)：132-138
[76]杨玉坡，西南高山地区冷杉、云杉林冠下天然更新的初步观察，林业科学，1956，4：337-354
[77]杨玉坡，中国的云杉，四川林业科技，1983，4(2)：28-33
[78]杨启修，四川亚高山针叶林植被的群落结构与组成特点，四川林业科技，1980，1(4)：31-37
[79]周德彰，氮磷钾矿物质营养元素对云杉幼苗生长的影响，四川林业科技，1982，3(2)：25-26
[80]林鹏编著，植物群落学，上海科学技术出版社，1986
[81]马钦彦，对崔-Lowson氏种群增长模型的探讨，生态学报，1985，5(3)：282-290
[82]万晶秀、梁婷，逻辑斯蒂曲线的一种拟合方法，生态学报，1983，3(3)：288-296
[83] 中国植被编写组，中国植被，北京，科学出版社，1980
[84]王本楠，崔-Lowson种群模型的显示解及拟合实例，生态学，1987，6(2)：27-30
[85]刘家岗，单种群密度变动的崔-Lowson模型基本成立，中国科学院林土所编，1986，64-68
[86]浓国彤主编，森林培育学，北京：中国林业出版社，2001，9
[87]张万儒，关于西南高山地区云、冷杉林下土壤形成过程的若干资料，土壤学报，1964，12(1)
[88]宋达康，张万儒主编，森林与土壤，北京：科学出版社，1981
[89]朱鹏飞，李德融，四川森林土壤，成都，四川科技出版社，1990
[90]史立新，全金锡，川西米亚罗地区暗针叶林采伐迹地早期植被演替过程的研究，植物生态学与地植学学报， 1988，12(4)
[91]佐藤大七郎等著(聂绍基译)，陆地植物群落的物质生产，北京，科学出版社，1986
[92]吴中伦，川西高山林区主要树种的分布和对天然更新造林树种规划的意见，林业科学，1959，6
[2]管中天，四川松杉植物地理，成都，四川人民出版社，1982
[3]刘翼，杉木地位质量的数量化评价方法，中国林科院研究报告，1982，(1)：29-41
[4]南方14省(区)杉木协作组，杉木立地条件的系统研究及应用，林业科学，1983，(19)：246-254
[5]成子纯，马尾松经营体系模拟系统，北京，中国林业出版社，1991
[6]马明东，刘跃建，云杉天然林地位指数表的编制与应用，四川林业科技，1986，7(3)：10-16
[7] 马明东，楠竹林分立地质量评价及类型划分研究，竹子研究汇刊，1991，10(3)：49-60
[8]马明东，刘跃建，四川綦江森林立地系统研究，论文集，成都，西南交大出版社，1993
[9]唐守正，多元统计分析方法，北京，中国林业出版社，1986
[10]罗积玉．微机用多元统计分析软件，成都，四川人民出版社，1986
[11]马明东，四川西南部高山云杉天然林生产力与立地生态因子关系研究，中国首届植物生态学青年研究，论 文集，北京，中国植物学会，1987
[12]马明东，云杉天然林分生产力及生态条件的关系，国际林联山地森林保护与管理论文集，北京，科学出版社， 1990，9
[13]马明东，云杉适生立地条件的研究，中国林学会第二届造林学术讨论会论文集，1988
[14]马明东，森林生境质量评价及其分类评述，四川林业科技，1993，14(4)：30-35
[15]马明东，云杉天然林分结构、生产力状况及生境质量评价研究，中国技术成果大会，1992，北京，科学技术 文献出版社
[16]马明东，三江流域、岷江上游生态区数量分类，自然资源，1992(2)：62-67
[17] 马明东，罗承德，张健，胡庭兴等，云杉天然林分多因素数量化生境质量评价研究，中国生态农业学报， 2006，(14)2
[18]马明东，罗承德，张健，胡庭兴等，云杉天然林分生境条件数量分类，中国生态农业学报，2006，(14)2
[19]马明东，四川盆地西绿楠木人工林分生物量的研究，四川林业科技，1989，10(3)：6-14
[20]马明东，罗承德，张健，胡庭兴等，四种数学方法对暗针叶云杉林分生境属性比较分析研究，中国生态农 业学报，2006，(14)1196-201
[21]马明东，江洪等，森林生态系统立地指数的遥感分析，生态学报，2006，26(9)：2810-2816
[22] Hampf, Frederick E.Site index curves for some forest species in the eastern United States. United States. ForestService. Eastern Region. Upper Darby, Pa., 1965.
[23] Lamson, N.I. Estimating northern red oak site-index class from total height and diameter of dominant andcodominant trees in central Appalachian hardwood stands. USDA Forest Service Research Paper NE-RP -Northeastern Forest Experiment Station, Aug 1987. (605), 3 p.
[24] The editing group of Forest Site Classification of China. The forest site classification of China. Chinese ForestryPress, Beijing. 1989.
[25] Ma Mingdong, Liu Yue-Jian. Quantitative classification of site condition in natural spruce forest [J], ChineseJournal of Eco-Agriculture[J], 2006,14(2) (in press).
[26] Carvalho, J.P.; Parresol, B.R. A site model for Pyrenean oak (Quercus pyrenaica) stands using a dynamic algebraicdifference equation[J]. Canadian journal of forest research, 2005.35 (1) 93-99.
[27] Carmean, Willard H. A comparison of site index curves for northern hardwood Species. North Central ForestExperiment Station (Saint Paul, Minn.); United States. Forest Service. North Central Forest Experiment Station,??Forest Service, U.S. Dept of Agriculture, 1979.
[28] Ingrid Seynave, Jean-Claude Gegout, Jean-Christophe Herve, Jean-Francois Dhote, and Jacques Drapier. Picieaabies site index prediction by environmental factors and understory vegetation: a two-scale approach based onsurvey databases. Canadian Journal of Forest Research[J], 2005.35(1)1669-1678.
[29] Clatterbuck, W.K. Height growth and site index curves for cherrybark oak and sweetgum in mixed, even-agedstands on the minor bottoms of central Mississippi. Southern journal of applied forestry[J], 1987,11 (4)219-222.
[30] Jokela, E.J.; Jack, S.B.; Nowak, C.A. Site index curves for unthinned Norway spruce plantations in New York.Northern journal of applied forestry[J], 1988,5 (4) 251-254.
[31] Boyer, W.D. A generational change in site index for naturally established longleaf pine on a South AlabamaCoastal Plain site[J]. Southern journal of applied forestry, 2001,25 (2) 88-92.
[32] Kimmins, J. P. Forest Ecology (Third Edition). Prentice Hall. New Jersey. 2003.
[33] Nigh, GD.The geometric mean regression line: a method for developing site index conversion equations forspecies in mixed stands[J]. Forest science, 1995,41 (1) 84-98.
[34] Chen, H.Y.H.; Klinka, K.; Kabzems, R.D. Site index, site quality, and foliar nutrients of trembling aspen:relationships and predictions. Canadian journal of forest research[J], 1998,28 (12) 1743-1755.
[35] Ung, C.H.; Bernier, P.Y.; Raulier, F.; Fournier, R.A.; Lambert, M.C.; Regniere, J. Biophysical site indices for shadetolerant and intolerant boreal species[J]. Forest science, 2001,47 (1) 83-95.
[36] O'Hara, K.L.; Oliver, CD. Three-dimensional representation of Douglas-fir volume growth: comparison of growthand yield models with stand data[J]. Forest science, 1988,43(6) 724-743.
[37] Walters, D.K.; Sloan, J.P.; Kurmis, Aspen site index as related to plant indicators.USDA Forest Service generaltechnical report NC - North Central Forest Experiment Station, 1990. (140), p. 337-341.
[38] Jiang Hong, Strittholt James R.,Frost, P.A. and Slosser, N.C. The classification of late serai forests in the PacificNorthwest, USA using Landsat ETM+ imagery[J]. Remote Sensing of Environment, 2004,91:320-331.
[39] Running, S.W.,Peterson,D.L.,Spanner,M,A.andTeuber,K.B. Remote Sensing of coniferous forest leaf area[J].Ecology, 1986,67:273-276.
[40] Jensen, J.R.. Remote sensing of the environment, Prentice Hall. New Jersey. 2000.
[41] Jiang Hong, Apps Michael J. Zhang Yanli, Peng Changhui and Woodard Paul M . Modeling the spatial pattern ofnet primary productivity in Chinese forests [J]. Ecological Modeling 1999,122:275-288.
[42] Chen. J.M. Derivation and validation of Canada-wide coarse-resolution leaf area index maps using high-resolutionsatellite imagery and ground measurements [J]. Remote Sens Ennviron, 2001,80:1-20.
[43] Chen, J.M. and J.Cihlar. Retrieving Leaf Area Index of Boreal Conifer Forests Using Landsat TM Images [J].Remote Sens Environ, 1996,55: 153-162.
[44] Qi, J. Leaf Area Index Estimates Using Remotely Sensed Data and BRDF Models in a Semiarid Region [J].Remote Sens Environ, 2000,73:18-30.
[45] Hong Jiang. The preliminary analysis of net primary productivity in natural spruce forests[J]. Acta Botanica Sinica,1986,28 (5) 538-548.
[46] Jiang Hong, Ma Mingdong. The relation between the natural spruce forest Productivity and the ecologicalenvironment Protection and management of mountain forests, Science Press, Beijing, 1992.
[47] Ruiliang, Pu, and Peng Gong. The application of hyperspectram remote sensing. Beijing, Higher Education Press.2000.
[48] Shugart, H.H. 1998. Terrestrial Ecosystems in changing environments. Cambridge. Cambridge University Press.
[49] Peng Changhui, Jiang, Hong, Apps, Michael J. and Zhang, Yanli. Simulating the Effects of harvesting regimes oncarbon and nitrogen dynamics of boreal forests in Central Canada using a process model [J]. Ecological Modeling.2002,155:177-189
[50]浦瑞良，宫鹏．高光谱遥感及其应用．北京：高等教育出版社，2000．
[51]江洪，紫果云杉天然中龄林分生物量和生产力的研究，植物生态学与地植物学学报，1985，10(2)：146-152
[52]李文华，小兴安岭谷地云、冷杉林群落结构和演替的研究．自然资源，1980(4)：17-29
[53]毕国昌，关于西南高山林区林型分类的几个问题，林业科学，1964，9(1)：86-92
[54]木村允，陆地植物群落的生产量测定法(姜恕等译)，北京，科学出版社，1981
[55]马明东，罗承德等，四川西北部亚高山云杉天然林生态系统碳含量、净生产量和碳贮量的初步研究，植物 生态学报，2007，31(2)：305-312
[56] Woodwell GM.Whittaker R H.Reiners WA.etal.The biota and the World carbon budget Science,1978,199:141-146 Houghton, R.A, Hackler, J.L., Lawrence, K.T. 1999. The US carbon budget: Contributions from land-usechange. Science, 285,574-578.
[57] Post WM. Emanuel WR. Zinke PJ. etal. Soil Pools and World life zone Nature, 1982,298:156-159
[58] Jiang, H., Apps, M.J, Zhang ,Y.L, et al., 1999. Modelling the spatial pattern of net primary productivity inChinese forests, Ecological Modeling, 122,275-288.
[59] Fang ,J.Y, Wang, GG, Liu, G H., et al., 1998. Forest biomass of China: an estimation based on thebiomass-volume relationship. Ecological Applications, 8,1084-1091.
[60] Fang, J.Y., Chen, A. P., Peng, C.H ,et al., 2001. Changes in forest biomass carbon storage in China between 1949and 1998. Science, 292,2320-2322.
[61] Jiang, H. Apps, M.L., Peng, C.H., et al., 2002. Modeling the influence of harvesting on Chinese boreal forestcarbon dynamics. Forest Ecology and Management, 169,65-82.
[62] Lei peifeng(雷丕锋)，Xiang Wenhua(项文化)，Tian Dalun(田大伦)，2004，Carbon storage and distribution in Cinna momum camphor Plantstion．[J]Chinese Journal of Ecology(生态学杂志)23(4)：25-30(in chinses)
[63]Fang Xi(方晰)，Tian Dalun(田大伦)，2003，Productivity and Carbon Dynamics of Masson Pine Plantation．[J] Journal of Central South Forestry University(中南林学院学报)23(2)：11-15(in chinese)
[64]Zhou Guomo(周国模)，Jiang Peikun(姜培坤)，2004,Density Storage and Spatial Distribution of Carbon in Phyllostachy Pubescens Forest．[J]Scientia Silvae Sinical(林业科学)40(6)：20～24(in chinses)
[65]Hong Jiang(江洪)，1986，The preliminary analysis of net primary productivity in natural spruce forests[J]．Acta Botanica Sinica(植物学报)28(5)538-548(in chinses)
[66]陈灵芝，长白山西南坡鱼鳞云杉林分结构的初步研究，植物生态学与地植物学丛刊，1963，1(1-2)：69-80
[67] 尚玉昌，种群的特征、动态和数量调节，陆地生态学译报(连载)，1985-1987
[68]董文泉，数量化理论及其应用，长春，吉林人民出版社，1979
[69]熊惠，川西高山林区的森林土壤，西南高山林区综合考察报告，1963，107-119
[70]詹昭宁等编，森林收获量预报，北京，中国林业出版社，1986
[71]董鸣，缙云山马尾松种群数量动态初步研究，植物生态学与地植物学学报，1986，10(4)：283-293
[72]蒋有绪，川西亚高山暗针叶林群落特点及其分类原则，植物生态学与地植物学丛刊，1963，1(1-2)：42-50
[73]蒋有绪，川西亚高山森林植被的区系种间关系和群落排序，植物生态学与地植物学丛刊，1982，6(4)：281-301
[74]葛剑平，小兴安岭原始红松林的结构与动态，东北林业大学，研究生论文集，1987
[75] 周德彰，杨玉坡，四川亚高山云冷杉林采伐迹地生态因子的变化，林业科学，1984，20(2)：132-138
[76]杨玉坡，西南高山地区冷杉、云杉林冠下天然更新的初步观察，林业科学，1956，4：337-354
[77]杨玉坡，中国的云杉，四川林业科技，1983，4(2)：28-33
[78]杨启修，四川亚高山针叶林植被的群落结构与组成特点，四川林业科技，1980，1(4)：31-37
[79]周德彰，氮磷钾矿物质营养元素对云杉幼苗生长的影响，四川林业科技，1982，3(2)：25-26
[80]林鹏编著，植物群落学，上海科学技术出版社，1986
[81]马钦彦，对崔-Lowson氏种群增长模型的探讨，生态学报，1985，5(3)：282-290
[82]万晶秀、梁婷，逻辑斯蒂曲线的一种拟合方法，生态学报，1983，3(3)：288-296
[83] 中国植被编写组，中国植被，北京，科学出版社，1980
[84]王本楠，崔-Lowson种群模型的显示解及拟合实例，生态学，1987，6(2)：27-30
[85]刘家岗，单种群密度变动的崔-Lowson模型基本成立，中国科学院林土所编，1986，64-68
[86]浓国彤主编，森林培育学，北京：中国林业出版社，2001，9
[87]张万儒，关于西南高山地区云、冷杉林下土壤形成过程的若干资料，土壤学报，1964，12(1)
[88]宋达康，张万儒主编，森林与土壤，北京：科学出版社，1981
[89]朱鹏飞，李德融，四川森林土壤，成都，四川科技出版社，1990
[90]史立新，全金锡，川西米亚罗地区暗针叶林采伐迹地早期植被演替过程的研究，植物生态学与地植学学报， 1988，12(4)
[91]佐藤大七郎等著(聂绍基译)，陆地植物群落的物质生产，北京，科学出版社，1986
[92]吴中伦，川西高山林区主要树种的分布和对天然更新造林树种规划的意见，林业科学，1959，6
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |