Résumé
often difficult and even experts sometimes experience difficulties to separate closely related species. Although morphometric data have sometimes been used in the past, they were in most instances restricted to simple data sets used to back up evidence derived from mere comparisons of morphological characters. Two nominal species of Pocillopora, Pocillopora verrucosa (Ellis & Solander, 1786) and P. meandrina, Dana, 1846 have been considered in the past either as one species or as two distinct species. The strategy used to unravel this taxonomic wrangle relies (1) on an analysis in a multidimensional space of morphometric data collected on the whole colonies and on individual branches, and (2) on unimodal statistical tests. Three types of analyses have been carried out: (1) A principal component analysis which has allowed to extract and visualize the most structuring variables, (2) a discriminant analysis which has allowed to classify each specimen within one of the two groups of specimens, and (3) a series of t tests (comparisons of means) to validate the differences identified between each group. The results obtained clearly show that Pocillopora verrucosa and P. meandrina can easily be separated on a morphometric basis. Our results should be confirmed by the combination of investigations focusing on other biological aspects such as ecological, molecular and physiological characteristics.