Infodoc : Sciences, Techniques et Culture

Vegetation studies in species rich neotropical forests are greatly hampered by high species numbers, identification difficulties, and large sample areas necessary to adequately cover species rich woody plant communities as well as by logistic, financial, personal, and temporal limitations. In the present study, we assessed the representativeness and usefulness of 65 plants families as indicator groups for such studies. Generally speaking, we found that the species associations in any family provided substantially more information about the relationship between vegetation plots than would be indicated by the percentage value of species number of the respective family. Thus, the loss of information from not considering the whole flora was more than outweighted by the possibility to incr[...]

Vegetation studies in species rich neotropical forests are greatly hampered by high species numbers, identification difficulties, and large sample areas necessary to adequately cover species rich woody plant communities as well as by logistic, financial, personal, and temporal limitations. In the present study, we assessed the representativeness and usefulness of 65 plants families as indicator groups for such studies. Generally speaking, we found that the species associations in any family provided substantially more information about the relationship between vegetation plots than would be indicated by the percentage value of species number of the respective family. Thus, the loss of information from not considering the whole flora was more than outweighted by the possibility to increase sample sizes, and by easier and more complete identification. This effect varied greatly between plant families (mainly in relation to biogeographic differences) as well as
with total species number (being more pronounced in species rich communities). Usefulness of the 65 plant families was evaluated based on seven criteria : a) specialization of individual species to habitat ; b) diversity of life forms in family ; c) recognizability of morphospecies in the field ; d) minimum area ; e) wide ecological and geographical distribution ; f) family easily recognized physiognomically ; g) well known and stable taxonomy. These criteria revealed considerable differences between families but varied regionally and between habitat, so that no indicator families could be recommanded as useful for all studies. We provide guidelines for selecting indicator groups for vegetation studies in the neotropics. A further analysis revealed that full floristic samples loose fairly little information if the plants are only identified to genus level, raising the possibility to base vegetation classification studies on comparatively easy genus
identification.