Organizing large-scale biodiversity inventories in the tropics: lessons from IBISCA projects

Leponce, M.1, Pascal, O.2, Novotny, V.3,4 & Y. Basset5 (1) Royal Belgian Institute of Natural Sciences, Belgium (Maurice.Leponce@naturalsciences.be); (2) Pro-Natura International, France; (3) University of South Bohemia, Czech Rep.; (4) Czech Academy of Sciences; (5) Smithsonian Tropical Research Institute, Republic of Panama. Background: IBISCA is an international and informal network of biodiversity experts conducting large-scale biotic inventories in various regions of the World (www.ibisca.net). Each IBISCA project is a collective effort addressing a global ecological question. IBISCA-Panama (2003-2004) aimed at estimating the overall arthropod diversity of a lowland rainforest while the Papua New Guinea survey (2012-2014), conducted in the framework of the “Our Planet Reviewed” programme, aimed at assessing the diversity generated by the elevational factor, from sea level up to the tree line. Methods: All projects are multi-taxa (with an emphasis on plants and arthropods), multi-strata and multi-methods. A central database (DB) is at the heart of each project. Results: The data flow follows a 10 step standard process: (1) sampling design which is often a trade-off between sampling effort and representativeness; (2) pre-printing of permanent labels with unique codes for samples and specimens; (3) collection of specimens with standardized mass collection methods; (4) on-site pre-sorting of material by helpers (para-taxonomists, students) to Order level; (5) further sorting to Family level by Taxonomic Working Group (TWIG) leaders and dispatching of specimens to experts; (6) identification of the material to (morpho-)species level by taxonomic experts who send afterwards the results to their TWIG leader; (7) control of the quality of data by TWIG leaders who fill a data entry template and send it to the database administrator; (8) import and cleaning of the data by the database administrator; (9) analysis and publication of the data by participants, either collectively or individually; (10) export of the DB to a public repository of data. Assisted data entry with high tech equipment (barcode scanner, PDA) reduces the risk of errors. Discussion/conclusion: Our experience shows that the main bottleneck in the data flow is the processing of the huge quantity of specimens collected. Solutions include securing enough funds for this critical step, training research technicians (para-taxonomists/ecologists) to assist main investigators and focusing on a limited number of informative yet tractable taxa. An additional benefit is that providing employment to local research assistants supports initiatives of local communities to conserve their forests.