【作者】 相聪伟；

【导师】 张建国；

【作者基本信息】 中国林业科学研究院 ， 森林培育， 2013， 博士

【摘要】 论文以福建卫闽国有林场杉木密度试验林连续观测30年数据为依据，系统研究了杉木人工林生长和材种结构变化的立地及林分密度效应，得出如下主要结论：1.林分发展的初期，直径生长主要受林分初植密度的影响，随林分初植密度增加而减小；在林分发展的中后期，随着自然稀疏的发生，直径呈现加速生长现象，林分密度效应逐渐减弱，而立地效应逐渐突出，随立地指数的增加而增大，相同立地指数小区直径生长有趋于相近的趋势。基于Richards参数与林分因子的相关关系，构建的林分平均直径生长混合效应模型，拟合结果良好（R2=0.9749），并且具有明确的生物学解析性，能够准确预测林分直径的生长过程，推荐在林分直径生长模拟中应用。2.林龄相同时直径分布偏度值随初植密度的增加而增大；林龄不同时，随林分发展低密度小区（A密度（1667株/ha）小区及B1（3333株/ha）小区）直径分布偏度先减小后增加；高密度小区直径分布偏度则先减小后增加，生长末期再减小，30年生时分布在0.5左右。林龄相同时，林分直径分布峰度值随林分密度的增加而减小。林龄不同时，随林分发展峰度先增加后减小。而与之相反，随林分发展变异系数则先减小后增加，同时研究发现林分直径的变异系数与直径生长速率关系密切。基于林分平均直径、偏度值、峰度值的Weibull直径分布参数预估模型，预测结果良好（决定系数R2均在0.9以上），为杉木人工林直径分布结构预测提供了理论依据。3.林龄相同时，枯损量随初植密度及立地指数的增加而增加。初植密度为A密度的小区以及B密度小于16m立地指数小区，30年生时不发生枯损。30年生时，立地指数相同小区立木株数有趋于相近的趋势，其中14m及16m立地指数小区立木株趋于B密度，而高于16m立地指数的小区立木株数趋于A密度，这一发现对杉木人工林经营具有重要的指导作用。本研究中基Richards直径生长混合效应模型与Yoda最大密度线方程构建的林分立木株变化方程：ln N=13.411-1.913ln+13.8272-0.4560SI-17.3264PD0.3553ln1-exp-3.4573PD0.7041t，预测结果良好（R2＝0.9689），同时具有生物学解析性，能够准确描述既定初植密度及立地指数林分中立木株数的变化，为杉木人工林经营过程中不同时期保留密度的确定提供了理论基础。4.林分蓄积及总材积整个林分发展过程中呈现出“S”曲线，林龄相同时，林分蓄积及总材积随立地指数的增加而增加。林分发展后期，立地指数相同小区蓄积量差异逐渐减小，有趋于相近的趋势，而林分总材积则未表现出趋近的现象。初植密度相同时，林分蓄积及总材积年平均生长量峰值林龄在低密度和高密度林分（A、D、E密度），随立地指数的增加而增加。但在B、C密度小区内峰值林龄随立地指数的增加而减小。5.材种出现时间取决于林分初植密度及立地质量，相比较受立地质量影响更大，中径材、大径材出现的时间随立地指数增加而减小，而随初植密度增加各材种出现时间变化不明显。30年生时，立地指数为12m、14m小区以及16m的B1小区中径材出材量未达峰值，其它小区峰值出现时间随立地指数增加而减小。12m及14m立指数小区及16m立地指数的B小区30年生时仍处于中径材生长期，大径材出材量及出材率较低。大于16m立地指数的小区30年生时均处于大径材生长期，各小区中大径材出材率均在0.5以上，且在林分发展后期林分蓄积相近小区，初植密度越大，林分大径材出材量及出材率越小，小径材出材率越大。若以大径材为人工林培育目标，应选择大于16m立地指数的小区，且以低密度造林，轮伐期应在30以上，以促进大径材的生长。更多还原

【Abstract】 In this paper,30-years continuing data of Cunninghamia lanceolata density experimentalplantation in Shaowu, Fujian Province was used. The effects of site quality and stems densityon stand growth and timber assortment structure were studied systemic. And we had theconculsions that:(1) At the stand developing early stage, stand diameter growth be mainly effected byplanting density, and diameter decreased with planting density increasing. At the standdeveloping middle and later stage, because of the self-thinning, the diameter growthaccelerated, the effects of planting density weakened and the effects of density strengthened,the diameter increased with site index. Mean diameter in plots with same site indexapproaching to no difference.The fittign results of stand mean diamter growth mixed model, which based on correlationof Richards parmeters and stand factor, is good (R2=0.9749). This mixed model with goodbiological analyticity can predict the diameter growth, in different plots with different densityand site index, so should be used to diameter growth at stand level in the future.(2) At the same stand age, the diameter skewness increased with planting density; and atdifferent stand age, diameter skewness in the lower site index plots (A-density plots and B1plot) decreased at early developing stage, and then increasing at later developing stage.However, in high site index plots, diameter skewness decreased at early developing stage, thenincreasing with stand age, and decreasing at later developing stage. At30years old, theskewness scattered about0.5.At same stand age, diameter kurtosis decreasing with planting density increasing. Atdifferent stand age, diamter kurtosis first increased and then decreased with stand developing.On the contrary, diameter variable coefficient first decreased and then increased with standdeveloping. At the same time, the results suggested that diameter variable coefficient wasclosely related to diameter growth speed. The fitting results he Weibull diameter distribution model, wich based on the correlationbetween parameters and mean diameter, skewness and kurtosis is good, and the R2in plotsused to testing were above0.9. This model could be used to stand diameter distribution as basicmodel, in the future.(3) There was no mortality or feww in plots with planting density less than1667trees/ha,or plots with3333trees/ha planting density and site index less than16m, at30years old. Thestocking stems number in plots with same site index approaching to no difference, at30yearsold, and approaching to B-planting density in plots with site index less than16m, to A-plantingdensity in the plots with site index greater than16m. This results was very important forCunninghamia lanceolata plantation management.The mortality models usually predicted mortality based on the previous data, and thesemodels are diffcult to used for management measure established. In this paper, the stockingtrees model:ln N=13.411-1.913ln13.82720.4560S I-17.3264P D0.3553ln1exp3.4573PD0.7041t,which based on Richards diameter growth mixed effects model and Yoda maximumsize-density line could predict the stems numbers accuracy. This models with good biologyanalyticity, could describe the dynamic changes of stem number in plots with given plantingdensity and site index, and could provid theoretical basis for retain stem number determinationat different developing stage in Cunninghamia lanceolata plantation.(4) At the same stand age, stand volume and gross volume increased with planting densityand site index. At later developing stage, the volume difference between plots with same siteindex lessened, and approached to no difference, but the gross volume not. In the plots withsame planting density, peak value of mean annual increment of stand volume and gross volumein lower and higher planting density (A, D, E planting density) increased with site index, anddecreased with site index in middle planting density (B, C planting density).(5) Different timber assortment appearance time were determined by planting density andsite index, and more by site index comparatively. Middle-size and large-size timber appearance time decreased with site index increasing, however, there was no trend in appearance time withplanting density increasing.30years old, in the plots with12and14m site index and B1plot,the middle-size timber volume didn’t reach to peak value, the middle size timber volume reachto peak value earlier in high site index plots.At30years old, plots with12m and14m site index and B1plot with16m site inex werestill in the period of middle size timber growth, and the large size tibmer outturn rate was verylower. But the plots with site index larger than16m all entered to large size timber growthperiod, at30years old, and middle size and large size timber outturn rate were highter than0.5.At late stand developing stage, the stand with similar volume, the proportion of large sizetimber decreased with planting density and the proportion of small size timber increased withplanting density. So, the aim of plantation is middle and large size timber, the plantation shouldbe built in the plots with site index higher than16m, planted with lower planting density, androtation shold longer than30years old.更多还原