A protocol for online documentation of spider biodiversity inventories applied to a Mexican tropical wet forest...

Spider community inventories have relatively well-established standardized collecting protocols. Such protocols set rules for the orderly acquisition of samples to estimate community parameters and to establish comparisons between areas. These methods have been tested worldwide, providing useful data for inventory planning and optimal sampling allocation efforts. The taxonomic counterpart of biodiversity inventories has received considerably less attention. Species lists and their relative abundances are the only link between the community parameters resulting from a biotic inventory and the biology of the species that live there. However, this connection is lost or speculative at best for species only partially identified (e. g., to genus but not to species). This link is particularly important for diverse tropical regions were many taxa are undescribed or little known such as spiders. One approach to this problem has been the development of biodiversity inventory websites that document the morphology of the species with digital images organized as standard views. Their main contributions are the dissemination of phenotypic data for species difficult to identify or new with the assignment of species codes, allowing species comparisons between areas regardless of their taxonomic status. The present paper describes a protocol to produce these websites almost automatically. This protocol was successfully applied to 237 species from a tropical primary forest in Mexico. The time and infrastructure required for the documentation of these species are discussed. Taxonomic information in terms of identification challenges, possible new species, and potential nomenclatural issues is described. In addition, the conventional community parameters (e. g., inventory completeness, species richness estimations, sampling intensity) are also calculated and compared through time and between methods. An optimized version for sampling allocation effort per season is presented and compared with protocols optimized for other tropical forests.