The design and objective of a community study imply the selection of the appropriate ordination technique in terms of species response models and weighting options. In this paper, we start from the observation that existing two‐table ordination techniques and related measures of niche breadth inevitably weight a sample in proportion to its abundance. We introduce a new multivariate method, which gives a more even weight to all sampling units, including those which are species poor or individual poor. We use this new method of analysis which we call OMI (for Outlying Mean Index) to address the question of niche separation and niche breadth. The Outlying Mean Index, or species marginality, measures the distance between the mean habitat conditions used by species (species centroid), and the mean habitat conditions of the sampling area (origin of the niche hyperspace), and OMI analysis places species along habitat conditions using a maximization of their mean OMI. Therefore, the position of the species depends on their niche deviation from a reference, which represents neither the mean nor the most abundant species, but a theoretical ubiquitous species that tolerates the most general habitat conditions (i.e., a hypothetical species uniformly distributed among habitat conditions). We demonstrate that OMI analysis is well suited for the investigation of multidimensional niche breadths in the case of strong limiting factors (e.g., meteorological conditions) or strong driving forces (e.g., longitudinal stream gradient). Furthermore, the analysis helps in finding which ecological factors are most important for community structure and organization and provides a separation of species based on their niche characteristics.

Abstract. This paper aims at proposing efficient vegetation sampling strategies. It describes how the estimation of species richness and diversity of moist evergreen forest is affected by (1) sampling design (simple random sampling, random cluster sampling, systematic cluster sampling, stratified cluster sampling); (2) choice of species richness estimators (number of observed species vs. non‐parametric estimators) and (3) choice of diversity index (Simpson vs. Shannon). Two sites are studied: a 28‐ha area situated in the Western Ghats of India and a 25‐ha area located at Pasoh in Peninsular Malaysia. The results show that: (1) whatever the sampling strategy, estimates of species richness depend on sample size in these very diverse forest ecosystems which contain many rare species; (2) Simpson's diversity index reaches a stable value at low sample sizes while Shannon's index is affected more by the addition of rare species with increasing sample size; (3) cluster sampling strategies provide a good compromise between cost and statistical efficiency; (4) 300 ‐ 400 sample trees grouped in small clusters (10–50 individuals) are enough to obtain unbiased and precise estimates of Simpson's index; (5) the local topography of the Western Ghats has a major influence on forest composition, the steep slopes being richer and more diverse than the ridges and gentle slopes; (6) stratified cluster sampling is thus an interesting alternative to systematic cluster sampling.

A crucial step in understanding the origin and maintenance of biological diversity is the assessment of its distribution over space and time and across environmental gradients. At the regional scale, two important attributes of species can be assessed that provide insight into speciation processes: species geographical and environmental ranges. The endemic tree flora of the Western Ghats is an interesting case for analyzing broad‐scale biodiversity patterns because of the steep environmental gradients that characterize this tropical region of India. We analysed species geographical and environmental ranges by Canonical Correlation Analysis of point data from herbarium collections. We performed partial analyses to discriminate spatial and environmental correlates of species distribution, and evaluate the contribution of higher taxonomic ranks to these ranges. We identified different levels of organization in the distribution of endemism: 1) general features, such as the concentration of endemic species in the southern part of the Western Ghats, and the decrease in endemic species richness along the altitudinal and the dry season length gradients, and 2) patterns specific to genera or families, such as species niche separation along the environmental gradients. Our analyses enabled us to formulate hypotheses about the diversification of the endemic tree flora of the Western Ghats. They also confirm the value of Canonical Correlation Analysis as the suitable method for collection data analysis.