单株树下辐射分布的数值模型

以数值求解的方法建立单株树下的辐射分布模型。模型给出不同纬度、不同坡向、不同树形几何参数、不同树下的平面位置和不同时刻的辐射强度。由此编制了标准的Ｃ 和ＦＯＲ－ＴＲＡＮ 数值函数Ｉ＝ ｆ（ψ，α，Ｓ，Ａ，Ｈｂ，Ｈｃ，Ｒｏ，Ｘ，Ｙ，ｔ）。以新疆和田地区的核桃树下所测得的辐射数据进行拟合运算的结果说明，模型能较好的描述单株树下的实际辐射分布。最后，对树下的平均辐射的全年变化的分析表明，对树冠形状近于锥体的单株树，当树高和冠幅为定值时，增加冠高（减少枝下裸干高）反而能提高树下平均辐射     

北京地区侧柏人工林密度效应

密度是影响森林尤其是人工林生长的重要因素,林冠层是森林生态系统与其他系统进行能量和物质交换的重要场所,树木及树冠生长对林分密度的响应关系可以看作是生物对环境变化产生的适应性现象。林分密度效应是生态学和森林培育学的重要研究内容之一。以23块8种不同密度梯度的北京山区侧柏人工幼龄林林分为研究对象分析其树木生长及树冠生长对密度的响应关系,其中树冠指标使用了参照了美国林务局(USDA)的树冠调查指标。研究结果表明:(1)林分平均胸径、平均树高和平均冠幅生长均随密度增大而减小,林分密度大于3000株/hm2时各指标减小的趋势变缓,使用异速生长模型可以很好地拟合这种变化关系;(2)随密度增加,树冠水平方向和垂直方向生长均到显著地抑制作用,树冠外形表现出由饱满冠型向狭长冠型变化的适应性现象;(3)使用树冠二维、三维指标与密度进行相关性分析可知树冠长度、树冠率等指标与林分密度呈负相关关系,树冠圆满度及树冠生产效率与密度表现出极显著正相关关系;(4)采用枝解析的方法研究了树枝长度、材积的平均生长量、连年生长量与密度的关系,结果表明幼龄期各生长量差异不大;(5)在建立冠幅模型时考虑了自变量间的多重共线性问题,所建的胸径单自变量二次方模型能够很好地预测侧柏人工幼龄林冠幅生长过程,模型相关系数R2为0.961。     

干旱胁迫下新疆杨树冠不同高度叶片水分状况与非结构性碳动态北大核心CSCD

水力结构调整及非结构性碳动态对于认知干旱胁迫下植物生存前景和死亡风险至关重要。该研究以新疆杨(Populus bolleana Lauche.)为对象,通过干旱处理分析树冠不同高度叶片水力性状、光合生理特性、不同功能器官间的可溶性糖、淀粉及非结构性碳水化合物(NSC)含量变化,以甄别树冠不同高度叶片对干旱胁迫的生理响应和适应差异。结果表明:(1)在干旱胁迫下,树冠上部的叶水势、叶含水量、枝条含水量普遍低于下部,饱和膨压渗透势及膨压损失点水势在不同树冠高度间差异不显著;(2)干旱处理组植株净光合速率随树冠高度的增加而降低,叶绿素SPAD值在树冠的上部显著低于下部,而水分利用效率在树冠的上部却高于下部;(3)干旱处理组比叶面积在各树冠高度下显著低于对照组,而胡伯尔值在中部及上部高于对照组但差异不显著;(4)干旱处理组,树冠上部叶片淀粉含量显著高于下部,枝条可溶性糖及NSC含量在树冠上部显著高于下部,韧皮部的可溶性糖、淀粉及NSC含量在不同树冠高度间无显著差异,细根的可溶性糖、淀粉及NSC含量在对照组与干旱处理组间的差异不显著。研究发现,干旱处理下,树冠高度的增加会加剧新疆杨枝叶的干旱胁迫,致使树冠上部枝条木质部发生栓塞的风险大于下部,并导致NSC在不同器官间的分配和组分产生差异,但新疆杨植株可通过水分利用效率和形态上的适应性调整减缓干旱胁迫。     

Crown asymmetry in high latitude forests: disentangling the directional effects of tree competition and solar radiation

Light foraging by trees is a fundamental process shaping forest communities. In heterogeneous light environments this behavior is expressed as plasticity of tree growth and the development of structural asymmetries. We studied the relative influence of neighborhood structure and directional solar radiation on horizontal asymmetry of tree crowns in late‐successional high latitude (67–68°N) forests in northern Fennoscandia. We described crown asymmetries as crown vectors (i.e. horizontal vectors from stem center to crown center), which we obtained from canopy maps based on crown perimeter measurements in the field. To disentangle the influence of the two main determinants, inter‐tree competition and directionality of above‐canopy solar radiation at high latitudes, we applied circular statistical models, utilizing cylindrical distributions, to these data consisting of orientations and intensities of crown asymmetry. At the individual tree level, our model predicted crown asymmetry vectors from the current stand structure, and the predictions became better when the intensity of asymmetry (i.e. crown vector length) was higher. Competition was the main determinant of crown asymmetry for 2/3 of trees, and the model predictions improved when we incorporated the directionality of solar radiation. At the stand‐level, these asymmetries had resulted in a small increment of the projected canopy area and an increased regularity of spatial structure. Our circular statistical modelling approach provided a quantitative evaluation of the relative importance of directionality of solar radiation and neighborhood stand structure, showing how both of these factors play a role in formation of crown asymmetries in high latitude forests. This approach further demonstrated the applicability of circular statistical modeling in ecological studies where the response variable has both orientation and intensity.     

Estimating Stem Volume Using QuickBird Imagery and Allometric Relationships for Open Populus xiaohei Plantations

There has been a great deal of interest in studying the crown of trees using remote sensing data.In this study,crownwidth was extracted from high-resolution QuickBird images for open Populus xiaohei plantations.Regression modelsfor predicting the individual stem volumes of Populus xiaohei were established using extracted crown width,as well asestimated tree parameters(i.e.diameter at breast height[DBH]and tree height)as predictors.Our results indicated thatcrown width could be accurately extracted from QuickBird images using a multi-scale segmentation approach with a meanrelative error of 5.74%,especially for wide-spacing stands.Using either extracted crown width alone or with estimatedDBH and tree height can successfully estimate individual stem volume of Populus xiaohei with the R~2 value ranging from0.87 to 0.92 depending on different model forms.In particular,the two second-order polynomial models(model2 andmodel 6),based on QuickBird image-derived crown widths and estimated DBH and tree heights,respectively,were the bestat describing the relationship between stem volume and tree characteristics.     

Estimating Stem Volume Using QuickBird Imagery and Allometric Relationships for Open Populus xiaohei Plantations

There has been a great deal of interest in studying the crown of trees using remote sensing data. In this study, crown width was extracted from high-resolution QuickBird images for open Populus xiaohei plantations. Regression models for predicting the individual stem volumes of Populus xiaohei were established using extracted crown width, as well as estimated tree parameters （i.e. diameter at breast height [DBH] and tree height） as predictors. Our results indicated that crown width could be accurately extracted from QuickBird images using a multi-scale segmentation approach with a mean relative error of 5.74%, especially for wide-spacing stands. Using either extracted crown width alone or with estimated DBH and tree height can successfully estimate individual stem volume of Populus xiaohei with the R2 value ranging from 0.87 to 0.92 depending on different model forms. In particular, the two second-order polynomial models （model 2 and model 6）, based on QuickBird image-derived crown widths and estimated DBH and tree heights, respectively, were the best at describing the relationship between stem volume and tree characteristics.     

基于边际回归的沈阳市绿化树种人工油松冠形模拟

构建城市典型绿化树种树冠外部轮廓预估模型,可以为城市绿化树种的空间配置优化奠定基础。本研究以辽宁省沈阳市典型绿化树种人工油松为对象,基于Crown Window装置获取60株样木树冠形状,利用幂函数、分段抛物线方程和修正Kozak方程选取基础模型,通过再参数化引入树冠结构因子(最大树冠半径)和相邻木竞争变量(相邻木平均树高、平均胸径、平均冠幅、相邻木株数,以及样木与其相邻木平均树冠接触高),构建耦合相邻木竞争及树冠结构因子的油松树冠外部轮廓预估模型。结果表明: 修正Kozak方程R_a²最大、均方根误差(RMSE)最小,模型稳定性良好,选取其为构建油松冠形模型的基础模型。通过再参数化在基础模型中引入最大树冠半径和相邻木平均胸径后,模型的R_a²提高0.0693,MSER为14.4%。分析最大树冠半径和相邻木竞争对油松树冠形状的影响发现,最大树冠半径对树冠形状影响较大,树冠半径随最大树冠半径增大而增大;相邻木平均胸径对树冠形状影响相比最大树冠半径较弱,树冠上半部分随相邻木竞争增强而增大,树冠下半部分随相邻木竞争增强而变小。本研究构建的耦合相邻木竞争及最大树冠半径的油松树冠外部轮廓边际回归模型具有良好的拟合优度,能够合理地模拟及预测沈阳市典型绿化树种人工油松的树冠形状。     

基于线性分位数混合效应的辽东山区红松冠幅模型

冠幅是反映单木生长状态及构建林木生长收获模型的重要变量。本研究以辽东山区大边沟林场10～55年生红松人工林为对象,基于66块固定样地的2763株红松的每木检尺数据,选取冠幅基础模型,采用再参数化的方法引入单木竞争指标(R_d),利用哑变量的方法引入了林分密度、林层变量,构建不同分位点(0.50、0.90、0.93、0.95、0.96、0.99)的冠幅分位数回归模型,并与传统方法进行比较,选取模拟林分最大冠幅的最优分位点。为反映林分中单木冠幅在林木个体之间的差异,建立了基于样地水平的最优分位点的线性混合效应分位数回归冠幅模型,分析各变量对单木冠幅的影响。结果表明: 基于F统计检验,不同林分密度和林层的冠幅模型具有显著差异,在基础模型中引入林层、林分密度和竞争后,模型R_a²提高0.0104,均方根误差降低0.0115,均方误差降低为7.4%;与最小二乘法比较,分位数回归模型能够较好地模拟林分状态下的单木最大冠幅,并选出0.96分位点和0.93分位点作为上林层和下林层的分位数回归模型的最优分位点。引入混合效应的线性分位数回归模型的赤池信息准则、贝叶斯信息准则、HQ信息准则等评价指标优于传统分位数回归,参数标准误显著降低,混合效应的引入很好地解释了样地之间的差异。就上林层和下林层而言,林分密度越大,最大冠幅越小;相对直径越大,最大冠幅越大,其中林分密度对下林层的冠幅影响大于上林层,当林分密度足够大时,冠幅随着胸径的增大先增大后降低。本研究构建的基于混合效应的分位数回归模型能有效提高模型的拟合优度,今后可通过调控林分密度、适度抚育间伐等措施,实现对辽东山区红松人工林的科学营建和可持续发展。     

樟子松人工林树冠表面积及体积预估模型的研究

基于樟子松(Pinus sylvestris var. mongolica)人工林6块固定标准地30株枝解析数据,在分析树冠表面积和树冠体积与林分变量和林木变量的基础上,利用幂函数建立了树冠表面积（CSA）和树冠体积(CV)的预估模型,同时还对林木材积生长量与CSA和CV进行了相关分析。研究结果表明：樟子松人工林树冠表面积和树冠体积随着林木胸径、树高和冠长的增大而增大,林木材积生长量与树冠表面积和树冠体积均明显呈线性关系。不同林分条件的樟子松人工林CSA和CV随林分年龄和胸径的增大而增大,CSA随林分密度的增大而减小,而CV与林分密度相关不紧密。林分树冠表面积和树冠体积预估模型的检验结果表明,两个模型的平均相对误差都在±8%之内,预估精度均大于91%,说明所建模型可以很好地预估樟子松人工林不同林分条件下的林木树冠表面积和树冠体积。     

The effect of two growth forms of Norway spruce on stand development and radiation interception: a model analysis

Summary Development of tree and canopy structure, and interception of photosynthetically active radiation (PAR) were studied in two model stands of Norway spruce consisting of trees with rapid versus slow site capture. The tree models were derived using Burger's (1953) sample tree material, from which two subpopulations of dominant trees were selected using the rate of horizontal site capture of the tree crowns as the criterion of division. The development of stand structure and interception of PAR were simulated in the two model canopies. The simulation period covered the period from tree age 15–80 years. The average development of the trees in the two subpopulations proved to be very different. The rapidly expanding trees were characterized by low mean within-crown needle area density and a long crown. The slowly expanding trees were smaller but had a higher mean within-crown needle area density. Up to approximately 40 years of age the stand of rapidly expanding trees contained more leaf area and intercepted more radiation than the stand of slowly expanding trees, when canopy cover was held constant. After 40 years of age this relationship was reversed due to the subsequent decline of leaf area in the stand of rapidly expanding trees and the increase in leaf area in the stand of slowly expanding trees. The biological relevancy and silvicultural implications of the simulated patterns of tree and stand development are discussed.     

不同生境栓皮栎天然更新幼苗植冠构型分析

栓皮栎存在于秦岭南坡的多种林分中,生活在不同生境中的个体往往形成不同的树冠形态和构型特征。为了说明不同生境条件下栓皮栎幼苗的植冠构型变化,采用典型抽样法,对秦岭南坡3种生境中(林冠下、林隙、林缘)的栓皮栎天然更新幼苗的侧枝、叶片特征及其空间分布进行了调查分析,结果表明:3种不同生境中栓皮栎幼苗植冠形态发生了明显的可塑性变化,(1)林冠下的幼苗明显为开阔型树冠,林隙和林缘处的幼苗树冠相对紧密;(2)幼苗的1级侧枝密度与分枝角度在3种生境下均差异显著(P0.05);从Ⅰ到Ⅳ层,林冠下幼苗的分枝角度在冠层内变化幅度不到5°,而林缘处幼苗的分枝角度变化高达40°;发生5个以上1级侧枝的概率以林冠下最大,为0.6;(3)从林缘、林隙到林冠下,幼苗的叶长、叶宽、单叶面积、叶面积指数逐渐降低,数量叶密度和比叶面积则逐渐增大,与其它两种生境相比,林冠下幼苗的叶片逐渐向树冠上层集中,且以更高序的侧枝为主要着生枝条;(4)林隙中栓皮栎幼苗的树高、地径明显优于林缘和林冠下,缩短了苗木进入主林层的时间,林隙对栓皮栎种群更新有利。在今后栓皮栎林的经营中,可以通过适当间伐来增加林隙数量,为森林更新和结构的优化的提供有利条件。     

Development of biomass proportions in Norway spruce (Picea abies [L.] Karst.)

This study tests the hypotheses that (1) the above-ground structure of Norway spruce (Picea abies [L] Karst.) is derivable from the functional balance theory, and that (2) crown ratio is a key source of structural variation in trees of different age and social position. Twenty-nine trees were measured in three stands (young, middle-aged, and mature), with three thinning treatments (unthinned, normal, and intensive) in the two older stands. There was a strong linear relationship between the total cross-sectional area of branches and that of stem at crown base. Foliage mass was linearly related with stem basal area at crown base. Also an allometric relationship was found between foliage mass and crown length. The mean length (weighted by basal area) of branches obeyed an exponential function of crown length. The parameters of most of these relationships were independent of slenderness (tree height/breast height diameter) and tree age However, total branch cross-sectional area per stem cross-sectional area in the young trees was greater than in the older trees. The young trees also had slightly shorter branches than predicted by the mean branch length equation. This was probably caused by branch senescence which had not yet started in the young stand. The older trees had a relatively long lower crown segment which was growing slowly and senescing. It was proposed that a segmented crown structure is characteristic of shade tolerant tree species, and that the structural model could be further developed by making the two segments explicit.     

A model for simulating transpiration of Eucalyptus salmonophloia trees

Understanding the water relations of Eucalyptus trees plays an important role in finding solutions to dryland salinity in southern Australia. A model for studying structure–function relationships in isolated tree crowns (radiation absorption, transpiration and photosynthesis, RATP) was parameterized to permit the seasonal transpiration course of a Eucalyptus salmonophloia tree to be quantified. Model responses to different parameterizations were tested in a sensitivity analysis. Predictive quality was mostly affected by the accuracy of information about leaf area density and stomatal responses to air vapor pressure deficit, and to a lesser extend by foliage dispersion. Assuming simple, non‐synergistic influences of changes in photosynthetic active radiation and air vapor pressure deficit on stomatal transpiration control, the model was able to simulate the daily water uptake of E. salmonophloia trees with reasonable predictive quality during an entire season. In order to more precisely simulate short‐term (i.e. diurnal) water use dynamics, the model must be extended to account for hydraulic and chemical controls of stomatal regulation of crown energy balance.     

Tree form in a mixed dipterocarp forest in Indonesian Borneo

The form of tropical trees was studied with reference to the production structure of the component individuals of a tropical rain forest stand in Sebulu, East Kalimantan in Indonesian Borneo, since the production structure as a physical or bio-economical basis of tree form still remains obscure in tropical rain forests. The pipe model theory successfully explained the crown shapes of different trees, and its parameter, designated as specific pipe length, suggested an increase in the cost of leaf mass growth with an increase in crown size. A mathematical model consisting of exponential functions of aboveground height was applied for describing stem form, and its properties were examined through changes in its coefficients and by adopting an assumption of the geometrical similarity of individual stem form as a criterion for comparing differences in stem form among individual trees. Furthermore, the cost of buttersses was discussed using the relation between bole- and buttress weight calculated from the mathematical model.     

Application of the Functional-Structural Tree Model LIGNUM to Sugar Maple Saplings (Acer saccharum Marsh) Growing in Forest Gaps

LIGNUM is a functional-structural model that represents a treeusing four modelling units which closely resemble the real structureof trees: tree segments, tree axes, branching points and buds.Metabolic processes are explicitly related to the structuralunits in which they take place. Here we adapt earlier versionsof LIGNUM designed to model growth of conifers for use withbroad-leaved trees. Two primary changes are involved. First,the tree segment for broad-leaved trees consists of enclosedcylinders of heartwood, sapwood and bark. Leaves consistingof petioles and blades are attached to the segments. Secondly,axillary buds and rules governing their dormancy are includedin the model. This modified version of LIGNUM is used to simulatethe growth and form of sugar maple saplings in forest gaps.The annual growth of the model tree is driven by net productionafter respiration losses are taken into account. The productionrate of each leaf depends on the amount of photosyntheticallyactive radiation it receives. The radiation regime is trackedexplicitly in different parts of the tree crown using a modelof mutual shading of the leaves. Forest gaps are representedby changing the radiation intensity in different parts of themodel sky. This version of LIGNUM modified for use with broad-leaf,deciduous trees and parameterized for sugar maple, yields goodsimulations of growth and form in saplings from different forestgap environments. Copyright 2001 Annals of Botany Company LIGNUM, functional-structural tree model, tree architecture, sugar maple, modelling growth and form     

Competition-dependent modelling of foliage biomass in forest stands

Currently, foliage biomass is estimated based on stem diameter or basal area. However, it is questionable whether the relations between foliage and stem observed from plantations of a single tree species can be applied to stands of different structure or species composition. In this paper, a procedure is presented to simulate foliage and branch biomass of tree crowns relative to crown size and light competition. Crowns are divided into layers and segments and each segment is divided into a foliated and an unfoliated fraction. Depending on the competitive status of the segment, leaf area density, specific leaf area and foliated branch fraction are determined. Based on this information, foliage biomass is calculated. The procedure requires a crown shape function and a measure to characterise competition for light and space of each individual segment within the canopy. Simple solutions are suggested for both requirements to enable an application with data that can be measured non-destructively in the field; these were stem position, tree height, crown base height, crown radii and some general crown shape information. The model was parameterised from single trees of Norway spruce and European beech and partly evaluated with independent data close to the investigation plot. Evaluations showed that the model can attribute the ecology of the different crown forms. Modelled foliage distribution for beech and spruce as well as total needle biomass of spruce agreed well with measurements but foliage biomass of beech was underestimated. The results are discussed in the context of a general model application in structured forests.     

Ground monitoring the light–shadow windows of a tree canopy to yield canopy light interception and morphological traits

Monitoring the light–shadow windows of a tree via a grid system on the ground was performed on sunny summer days at high spatial resolution using a custom‐built, inexpensive scanner. The measurements were taken with two goals: (1) to quickly and remotely quantify the overall, short‐wave solar radiation (300–1100 nm) intercepted by the tree canopy, and (2) to yield such crown geometric traits as shape, size and the number of theoretical canopy leaf layers (leaf layer index, LLI) in relation to the section orthogonal to sunbeam direction (sun window). The ground readings at each measurement over the day were used to project a digitized shadow image. Image processing was applied and the intercepted radiation was calculated as the difference from the corresponding incoming radiation above the canopy. Tree‐crown size and shape were profiled via computer imaging by analysing the different shadow images acquired at the various solar positions during the day. It is notable that these combined images yielded the crown features without having to parameterize such canopy characteristics as foliage extension and spatial distribution.     

Model simulations of spatial distributions and daily totals of photosynthesis in Eucalyptus grandis canopies

Summary A simulation model for radiation absorption and photosynthesis was used to test the hypothesis that observed nonuniform distributions of nitrogen concentrations in young Eucalyptus grandis trees result in greater amounts of daily assimilation than in hypothetical trees with uniform N distributions. Simulations were performed for trees aged 6, 9, 12 and 16 months which had been grown in plantations under a factorial combination of two levels of fertilization and irrigation. Observed leaf N distribution patterns yielded daily assimilation rates which were only marginally greater (<5%) than for hypothetical trees with uniform distributions. Patterns of assimilation distribution in individual tree crowns closely resembled those for absorbed radiation, rather than for N. These conclusions were unaffected by three choices of alternative leaf area density distributions. The simulation model was also used to calculate hourly and daily rates of canopy assimilation to investigate the relative importance of radiation absorption and total canopy nitrogen on assimilation. Simulated hourly rates of carbon assimilation were often lightsaturated, whereas daily carbon gain was directly proportional to radiation absorbed by the tree crown and to total mass of N in the leaves. Leaf nitrogen concentrations determined photosynthetic capacity, whereas total leaf area determined the amount of radiation absorbed and thus the degree to which capacity was realized. Observed total leaf area and total crown N were closely correlated. The model predicted that nitrogen use efficiences (NUE, mol CO₂ mol^–1 N) were 60% higher for unfertilized than for fertilized trees at low levels of absorbed photosynthetically active radiation (PAR). Nitrogen use efficiency was dependent on fertilizer treatment and on the amount of absorbed PAR; NUE declined with increasing absorbed PAR, but decreased more rapidly for unfertilized than for fertilized trees. Annual primary productivity was linearly related to both radiation absorbed and to mass of N in the canopy.     

基于地基激光雷达的落叶松人工林枝条因子提取和建模

地基激光雷达(TLS)可以实现从森林中无破坏收集数据。本文基于地基激光雷达数据通过点云处理软件以人机交互的方式获取了26株落叶松样木的1266组枝条信息,包括着枝高度、弦长、枝长、着枝角度、基径和弓高。枝条可提取的最大相对着枝高度的平均值为0.83。在所提取的枝条因子中,提取精度依次为着枝高度＞弦长＞枝长＞基径(基径大于20 mm的枝条)＞弓高,将树冠分为4部分后分析发现,随着冠层高度的增加,枝条密度呈升高趋势,枝条提取率和提取精度呈下降趋势。此外,由于枝条基径提取精度较低,以弦长、着枝高度、胸径和树高为自变量构建基径预测模型。对不同基径的实测值、提取值与模型预测值对比分析发现,枝条基径的预测精度大于提取精度。对于造材来说,最有价值的部分是树木中下部,本方法能够较准确地提取树木胸径树高和相对着枝高度0.8以下的枝条属性信息,提供构建木材质量模型所需要的参数。     

Primary ecological succession in vascular epiphytes: The species accumulation model

Epiphytes are integral to tropical forests yet little is understood about how succession proceeds in these communities. As trees increase in size they create microhabitats for late‐colonizing species in both small and large branches while maintaining small tree microhabitats for early colonizing species in the small and young branches. Thus, epiphyte succession may follow different models depending on the scale: at the scale of the entire tree, epiphytes may follow a species accumulation model where species are continuously added to the tree as trees increase in size but at the scale of one zone on a branch (e.g., inner crown: 0–2 m from the trunk), they may follow the replacement model of succession seen in terrestrial ecosystems. Assuming tree size as an indicator of tree age, I surveyed 61 Virola koschnyi trees of varying size (2.5–103.3 cm diameter at breast height) in lowland wet tropical forest in Costa Rica to examine how epiphyte communities change through succession. Epiphyte communities in small trees were nested subsets of those in large trees and epiphyte communities became more similar to the largest trees as trees increased in size. Furthermore, epiphyte species in small trees were replaced by mid‐ and late‐successional species in the oldest parts of the tree crown but dispersed toward the younger branches as trees increased in size. Thus, epiphyte succession followed a replacement model in particular zones within treecrowns but a species accumulation model at the scale of the entire tree crown.