庞泉沟自然保护区华北落叶松林的自组织特征映射网络分类与排序

自组织特征映射网络（SOM）是新近引入植物生态学的分析方法,对复杂问题和非线性问题具有较强的分析和求解功能。本研究应用SOM分类和排序研究了庞泉沟自然保护区华北落叶松林。研究结果表明,SOM将120个样方分为7个植物群落类型,分类结果具有明确的生态意义;样方和物种在SOM训练图上呈现一定规律的分布;7个群落类型各有其分布范围和界限,揭示了群落间的生态关系。在此基础上,通过引入一种在SOM训练图上可视化环境因子梯度的方法,能够较好地完成样方、物种和环境因子相互关系的分析,揭示了海拔是影响该区华北落叶松林生长和分布的最主要因子。生态分析表明SOM分类和排序是一种有效的梯度分析方法,适用于表征生态特征和探索群落和环境相互关系的研究。     

An algorithm for predictive ordination

Summary An algorithm is described for predictive ordination. The functional form of species response is required, but it need not be Gaussian. Any integrable function is acceptable.The project has been supported by N.S.E.R.C. of Canada funds.     

A numerical analysis of arborescent communities in the rain forest of the Luquillo mountains,Puerto Rico

Summary Cluster analysis and canonical ordination were used as complementary techniques to identify major associations in a lower montane rain forest in Puerto Rico. These techniques were applied to forestry inventories from both secondary and primary communities. Eight major forest types were identified from the analyses and were confirmed by field checks. Distribution of these types was strongly correlated with past land-use and topographic position.Comprehensive ecological surveys are rare in moist tropical forests; however, forestry inventories are common and the results of this study suggest that valuable ecological information can be obtained from these data sets as well.Scientific Journal Series No. 10793 of the Minnesota Agricultural Experimental Station. Plant nomenclature according to Little & Wadworth (1964) and Little, Woodbury & Wadsworth (1974).This research was conducted while the senior author was stationed at the Institute of Tropical Forestry, USDA Forest Service, Rio Piedras, Puerto Rico.     

Classification and ordination in the Indian Peaks area,Colorado Rocky Mountains

Summary A mid-latitude, Northern Hemisphere alpine vegetation in the Colorado Rocky Mountains was suitable for analysis by syntaxonemie and numerical methods, which interpreted the Braun-Bianquet association data, distributed by habitat and vegetation dominance types, correspondingly. Classification and ordination yielded complementary, interpretable results, which combined were more informative than results of either technique alone. Syntaxonomic results can faeilitate the interpretation of numerical results; the syntaxonomic relationships can be clarified by the results of numerical analysis.Alliances and orders of the Braun-Blanquet hierarchy, and their diagnostic taxa groups were derived by numerical methods from an association/taxa matrix. On the basis of this matrix, numerical methods demonstrated some of the environmental gradient complexes controlling the compositional variation. The compositional distinctiveness and uniformity of orders and alliances parallel environmental distictiveness and uniformity.The following attributes of the sampled vegetation and environment were probably responsible for the successful combination of syntaxonomic and numerical results: 1. The sampled area as a geographically limited universe, characterized by relative environmental, biotic, and historical distinctiveness and uniformity, 2. The wide range, high diversity, and clear denfinition of habitat and vegetation dominance types. Because it effectively summarizes and simplifies field data, syntaxonomy can play an important role in building local vegetatio models.Most of the figures in this paper are from a book version of a University of Colorado, Department of Environmental, Population, and Organismic Biology, Ph. D. thesis which was produced at the Institute of Arctic and Alpine Research. I would like to thank Dr. P.J. Webber, the thesis advisor, for his support, for suggesting the numerical methods, and for computer programs which were written by W.F. Reid. Vicki Dow, Marilyn Joel, and Karen Sproul drafted the majority of figures. I am grateful to Gwen Archer for editorial help.     

Determinants of assemblage structure in Neotropical dry forest lizards

We investigated the structure of a lizard assemblage from Seasonally Dry Tropical Forest enclaves in the Brazilian Cerrado biome, by testing the roles of ecological and historical components. We analysed data from 469 individuals, belonging to 18 lizard species, sampled by a combination of pitfall, funnel and glue traps, as well as by haphazard sampling. Null model analyses and Canonical Phylogenetic Ordination analysis, coupled with Monte Carlo simulations, revealed lack of both ecological and phylogenetic structure in microhabitat use. Conversely, these analyses revealed a mean overlap in diet composition significantly smaller than expected by chance and significant historical structure. Structure in diet composition was due to phylogenetic effects corresponding to the most basal divergence of the squamate phylogeny (Iguania/Scleroglossa) and the clades Teiidae and Gymnophthalmidae. Among lizards, evolutionary constraints on microhabitat use appear less than on prey use, suggesting that the availability of historically preferred prey types moderates microhabitat selection. The lack of structure in microhabitat use suggests absence of competitive interactions on the spatial component. On the other hand, food preferences have a deep historical basis and do not reflect current competitive interactions.     

Multivariate approach to the classification and ordination of the forest ecosystem of Nandiar valley western Himalayas

QuestionsDoes the vegetation composition of the forests of Nandiar Valley correlate with climatic, topographic and edaphic variables? Is it possible to identify plant communities through indicator species in relation to environmental gradients? Can this approach of classification and ordination will be helpful for conservation planning?LocationForests of Nandiar Valley, Moist temperate Western Himalayas Pakistan.MethodsEighty stands were selected for quantitative and qualitative characteristic of vegetation between an elevations of 525–3817 m. Species composition was recorded by using 400 m long transects. GPS, climatic, edaphic and topographic data were recorded for each sampling site. The relationship between habitat types, species composition and distribution along with climatic, edaphic and topographic variables were analyzed using TWINSPAN, Cluster analysis and DCA ordination.ResultsSum 325 vascular plants species belonging to 97 families were recorded. Diversity index and species richness was maximum in the moist temperate zone. Classification and ordination showed that the variance in species data was 7.07. Two-ways indicator species analysis classified the vegetation into eight plant communities. Indicator species analysis revealed that slope aspect, wind speed, temperature, dew point, wet bulb, pH, organic matter and phosphorous were the strongest parameters (p ≤ 0.05) determining plant community composition and indicator species in each habitat. The results also show the strength of the environment – species relationship using Monte Carlo procedures. DCA ordination grouped different species having similar habitat and habitats having common species.ConclusionsThe multivariate analysis of the vegetation along with environmental variables of Nandiar valley confirmed the indicators of each sort of vegetation communities/microclimatic zones which could further be used in conservation planning and management not only in studied area but also in the adjacent regions as well as in the areas exhibit similar sort of climatic, edaphic and topographic conditions.     

The Finnish forest site type approach: ordination and classification studies of mesic forest sites in southern Finland

A.K. Cajander's forest site type classification system is based on definition of plant communities typical to certain climatical and edaphical site conditions, but the structure and composition of the tree stands in Finland are considered sensitive to random variation and are therefore not used as primary classification criteria. The system has often received criticism, usually that the effects of the tree stand and successional stage of the stand have been underestimated. Most of the present-day forest stands in Finland represent young successional stages and are subjected to intensive management. This should result in an additional difficulty in the application of the forest site types in the field.The present study is based on three independent data sets representing forests on mineral soil in southern part of Finland. TWINSPAN classification, DCA ordination and canonical correspondence analysis (CCA) techniques were applied in successive stages of the data analysis. It was found that the definition of the intermediately fertile, mesic site types was clearly confused by the effects of the tree species and age of the stand. The analyses also revealed that the succession pathways on mesic forest sites are largely determined by the tree species composition. In stands dominated by Pinus sylvestris, the succession follows the competitive hierarchy model, whereas in stands dominated by Picea abies, severe shading of the tree canopy governs the development of understorey vegetation.Abbreviations CCA Canonical correspondence Analysis - DCA Detrended correspondence Analysis - TWINSPAN Two-way indicator species analysis     

DCCA排序轴分类及其在关帝山八水沟植物群落生态梯度分析中的应用

首次在ＤＣＣＡ排序的基础上使用排序轴分类法（ＯＡＣ）,将其发展为一个新的典范分类方法,并应用这种方法对山西关帝山八水沟的植物群落进行了数量分类,共划分出５类,效果较好。方差分析与多重比较表明,环境与群落结构因子在不同群落之间存在较为显著的差异,表明ＤＣＣＡ排序轴分类法能较好地反映群落与环境的变异。因为ＤＣＣＡ排序轴分类同时结合了植被因子与环境因子,所以它能较好地表达植物群落的环境梯度和结构梯度,反     

Detrended correspondence analysis: An improved ordination technique

Summary Detrended correspondence analysis (DCA) is an improvement upon the reciprocal averaging (RA) ordination technique. RA has two main faults: the second axis is often an arch or horseshoe distortion of the first axis, and distances in the ordination space do not have a consistent meaning in terms of compositional change (in particular, distances at the ends of the first RA axis are compressed relative to the middle). DCA corrects these two faults. Tests with simulated and field data show DCA superior to RA and to nonmetric multidimensional sealing in giving clear, interpretable results. DCA has several advantages. (a) Its performance is the best of the ordination techniques tested, and both species and sample ordinations are produced simultaneously. (b) The axes are scaled in standard deviation units with a definite meaning, (c) As implemented in a FORTRAN program called DECORANA, computing time rises only linearly with the amount of data analyzed, and only positive entries in the data matrix are stored in memory, so very large data sets present no difficulty. However, DCA has limitations, making it best to remove extreme outliers and discontinuities prior to analysis. DCA consistently gives the most interpretable ordination results, but as always the interpretation of results remains a matter of ecological insight and is improved by field experience and by integration of supplementary environmental data for the vegetation sample sites.This research was supported by the Institute of Terrestrial Ecology, Bangor, Wales, and by a grant from the National Science Foundation to R.H. Whittaker. We thank R.H. Whittaker for encouragement and comments, S.B. Singer for assistance with the Cornell computer, and H.J.B. Birks, S.R. Sabo, T.C.E. Wells, and R.H. Whittaker for data sets used for ordination tests.     

以林相图为基础的不同取样方式对植被排序分析结果的影响

以6种不同方式对林相图中同一样带取样,采用CA、DCA和CCA 3种排序方法,研究了取样方式对排序轴的解释效果、物种与环境因子、环境因子之间、环境因子与坐标轴之间关系的影响.结果表明,样方大小和形状的变化在不同程度上改变了排序结果.大样方和长方形样方都增强了排序轴的解释效果,并对双序图中稀有种、独特种的位置有较大的影响;环境因子中土壤因子对样方的大小和形状都很敏感,坡度、经纬度只对样方大小敏感,坡位、海拔、温度和降水则对样方形状敏感;随着样方面积的增加,海拔、温度和降水的作用降低,而坡向的作用增加.     

海南岛热带草地的数量分类和排序研究

本文用一些数量分类和排序的方法对海南岛鹦歌岭热带草地进行了分类和排序。所用的方法包括两种多元等级聚合分类--最近邻体法(NN)和最远邻体法(FN),极点排序(PO)和主分量分析(PCA)排序。结果表明：把19个样地分为三大类型、9个群落,其分布格局与坡度、放牧强度和土壤肥力密切相关。所用的四种方法在热带草地的研究中均有一定的适用性。     

On the variation explained by ordination and constrained ordination axes

Abstract. Total inertia (TI), the sum of eigenvalues for all ordination axes, is often used as a measure of total variation in a data set. By use of simulated data sets, I demonstrate that lack-of-fit of data to the response model implicit in any eigenvector ordination method results in polynomial distortion ordination axes, with eigenvalues that normally contribute 30–70% to TI (depending on data set properties). The amount of compositional variation extracted on ecologically interpretable ordination axes (structure axes) is thus underestimated by the eigenvalue-to-total-inertia ratio. I recommend that the current use of total inertia as a measure of compositional variation is discontinued. Eigenvalues of structure axes can, however, be used with some caution to indicate their relative importance. I also demonstrate that when the total inertia is partitioned on different sets of explanatory variables and unexplained variation by use of (partial) constrained ordination, (35) 50–85% of the variation ‘unexplained’ by the supplied explanatory variables represents lack-of-fit of data to model. Thus, the common interpretation of ‘unexplained variation’ as random variation (‘noise’) or coenoclinal variation caused by unmeasured explanatory variables, is generally inappropriate. I recommend a change of focus from the variation-explained-to-total inertia ratio and ‘unexplained’ variation to relative amounts of variation explained by different sets of explanatory variables.     

Frugivorous butterfly species in tropical forest fragments: correlates of vulnerability to extinction

Insects exhibit a variety of population-level responses to forest fragmentation, ranging from population increase to extinction. However, the biological attributes that underlie differences in extinction vulnerability among insects have been little-studied. Using the frugivorous butterfly community of tropical dry forest in Venezuela, we studied body size, population density and colonization ability as attributes that might underlie the range of responses of insects to forest fragmentation. The study was carried out in a set of forest fragments in the reservoir Lago Guri, formed by the damming of the Caroni River in eastern Venezuela. Results show that larger butterfly species were more vulnerable to extinction from habitat fragments than smaller ones. Rarer species were not more vulnerable to extinction, showing that rarity may not be an important correlate of vulnerability to extinction amongst insects. Contrary to expectation, faster-flying species were more and not less vulnerable to extinction from small habitat fragments. We speculate on the possible reasons for the observed patterns in extinction vulnerability using additional observations on behavioural patterns and larval host plant distributions of some of the butterfly species.     

汾河河口湿地植被数量分类与排序

在群落样方调查基础上,采用双向指示种分析（TWINSPAN）和除趋势对应分析（DCA）对汾河入河口湿地植被群落进行了数量分类和排序。TWINSPAN将78个样方分为18个群丛,并论述了各群丛的群落学特征。DCA排序结果反映了植物群落类型与环境梯度水分之间的关系,表明影响群丛分布格局的主导生态因子为土壤水分。     

Influences of transformation of cover values on classification and ordination of lake vegetation

Summary Numerical ordination (Program ORDINA) and classification (Program TABORD) have been applied to vegetation data from a lake in south Sweden.Full percentage values and different transformations of the estimated percented cover scale have been tested on relevés with nymphaeids (+ elodeids) and on relevés containing obly helophytes. The used scales have been: percentage (P), Hult-Sernander-Du Rietz (HSD), Log (%+1) (LP), 1+ log% (L), 100+ % (HP) and Presence/Absence (P/A). The main aim has been to find if there is an optimal transformation regarding both quantitative and qualitative information in vegetation data.In the classification of helophytes a dataset with 87 relevés and 20 species have been used applying P-and L-scales and P/A. The similarity between the different scales has been compared in relation to common relevés between every two corresponding clusters. L and P/A give very similar results, more similar than L and P or PA and P. It is also interesting that the result with P-scale is more similar to L than to P/A. The P-scale neglects with the species with a low cover value, whilst the L-scale takes notice of both quantity and quality.In the classification of 90 nymphaeid relevés containing only 6 species P, HSD, L, HP and P/A were applied. The L-scale (and partly PH-scale) gave a result similar to the most quantitative P-and HSD-scales. P/A-values gave a very deviating result. In the ordination of helophytes only 38 relevés with 18 species, taken along a transcect, were used and P, P/A and 4 different transformations were applied. The similarity between order of relevés in the graphs and real order of relevés along the transect was used as an objective test of the effectiveness of the scale applied. The P-scale gave a bad result, whilst P/A gave the best result. The intermediate scales L and HP gave acceptable results.In the ordination of nymphaeids (+ elodeids) 90 relevés, where only pattern was studied, the L-scale gave the best result most similar to P-and HSD, but applying the two extremes, P-scale and P/A, resulted in large reduction of information.In the submitted ordinations and classifications the applied L-scale (and HP-scale) always gave ecologically interpretable results, but the extreme transformation were less reliable.This work has been supported by grants from the University of Lund, for which I am duty grateful.Nomenclature follows Lid (1974).I am grateful to Professor Nils Malmer, head of the Department of Plant Ecology, Lund, Sweden, Dr. Eddy van der Maarel, University of Nijmegen, Holland, and my colleague Fil. Kand. Stefan Persson for valuable advice and discussions. I also thank my wife Fil. Kand. Eva Waldemarsson Jensén, Institute of Ecological Botany, Uppsala, for valuable discussions, Mr. Tommy Olsson for valuable technical advice, Dr. R.S. Clymo for advice with DIVINF and PCA and correction of the language, Mrs. Mimmi Varga for drawing the figures and Mrs. Brita Billstein for typewriting the text and the tables.     

Classification of natural forest communities of coastal British Columbia, Canada

In comparison to countries with a tradition of vegetation studies, a comprehensive, hierarchical classification of plant communities in the province of British Columbia has not yet been developed. Such a classification is needed for systematic ecological studies and coordinated conservation of vegetation. As the culmination of fifty years of detailed surveys, tabular and multivariate analyses of 3779 relevés of natural, old-growth, submontane, montane and subalpine forest communities in coastal British Columbia were used to develop a hierarchy of vegetation units according to the Braun-Blanquet approach. At the highest level, we distinguished seven orders: Quercus garryana, Pseudotsuga menziesii-Mahonia nervosa, Tsuga heterophylla-Rhytidiadelphus loreus, Tsuga mertensiana, Thuja plicata-Tiarella trifoliata, Populus trichocarpa, and Pinus contorta-Sphagnum. Diagnostic table, ordination, and elimatic and edaphic regimes were used to show floristic affinities among the orders and to interpret their relationships to regional environmental gradients. Plant communities of each order were briefly characterized by their floristic composition, physiognomy, succession, and environment. The synopsis for all delineated vegetation units (order, suborder, alliance, and association) of coastal British Columbia is included.     

Comparisons of three ordination techniques

Summary Simulated coenoclines were used to test performance of several techniques for ordinating samples by species composition: Wisconsin polar or Bray-Curtis ordination with Euclidean distance (ED) and the complements of percentage similarity (PD) and coefficient of community (CD) as distance measures, Principal components analysis, and polar and non-polar or indirect use of Discriminant function analysis. In general the Bray-Curtis technique gave the best ordinations, and PD was the best distance measure. Euclidean distance gave greater distortion than PD in all tests; CD may be better than PD only for some sample sets of high alpha and beta diversity and high levels of noise or sample error. Principal components ordinations are increasingly distored as beta diversity increases, and are highly vulnerable to effects of both noise and sample clustering. Discriminant function analysis was found generally unsuitable for ordination of samples by species composition, but likely to be useful for sample classification.