treatment provided by
Genera Plantarum 1: 335 (1789).
Scolopia strongly resembles Ludia Comm. ex Juss. , which is confined to Madagascar, East Africa, and several groups of Indian Ocean islands, except that Ludia lacks petals. Sleumer (1972a) noted that species of Scolopia with caducous petals had been erroneously placed within Ludia in prior literature, while a few apetalous species known from inadequate material had been erroneously placed within Scolopia . He found that characters of the anther connective, style and venation did not correspond to generic boundaries, so the presence of petals is the only reliable distinction between the two genera. It seems possible that Ludia evolved within Scolopia , but in the absence of molecular evidence supporting that hypothesis, we maintain the traditional generic circumscriptions.
Fourteen species of Scolopia are herein recognized in Madagascar, one of them newly described and all endemic. These can be divided into two groups: the first consists of five species native to relatively dry habitats, which have single-flowered inflorescences and usually bear spines, while the second comprises nine species found usually in humid and littoral forests, which have more than one flower per inflorescence and lack spines. Flowering specimens are more easily identified than fruiting specimens, as known fruits of all species are similar in appearance (red, subglobose, the individual species’ size ranges fairly large and overlapping too much to make size a useful character), while useful characters such as stamen number, flowering pedicel length, and anther morphology are obscured in fruit. (As far as can be determined, all Malagasy species have anthers with long extended connectives; this is not the case for all species of the genus, and the character helps to distinguish certain Malagasy species from similar species of the Mascarenes.) However, most herbarium specimens collected are fruiting, so the key presented here emphasizes vegetative features and those characters visible in fruit as much as possible.
Leaf shape is a valuable character for identification of species. In the humid and littoral forest species, leaf shape can be quite variable within species and the ranges of variation in related species overlap, but significant average differences exist. As these differences can be difficult to describe verbally, a figure showing “typical” leaf shapes in these species ( Fig. 1 View FIG ) has been provided to facilitate use of the key. The color of the leaves upon drying is not entirely consistent within species, but shows strong trends that are correlated with ecological preferences. Collections from low- to high-elevation humid forests and dry bush almost always dry brown to dark brown, very rarely greenish brown, while most taxa of littoral and sublittoral forests almost always dry a pale color, usually greenish to gray-green or olive on at least one surface, or rarely dark-mottled. Since this pattern is observed from numerous localities and collections made over several decades, it is unlikely to be merely an artifact of varying specimen processing techniques; rather, it seems likely to reflect substantive differences in leaf anatomy or biochemistry correlated with adaptation to different habitats.
No known copyright restrictions apply. See Agosti, D., Egloff, W., 2009. Taxonomic information exchange and copyright: the Plazi approach. BMC Research Notes 2009, 2:53 for further explanation.