Klaus Riede, Institut für Biologie I (Zoologie), Albertstr.21 a, 79104 Freiburg i. Br.

Bioacoustic analysis of species-specific vocalizations is an integral part of species descriptions in many taxa and has often been used for the diagnosis of cryptic species. Usually - but by no means necessarily - these recordings are deposited in phonotheks, ideally with a cross reference to the recorded specimen. Publications as disks or CDs are the exception, and unpublished, private sound archives might easily get lost. New information technologies for digital sound processing and data exchange allow much more efficient ways of storage, organisation, publication and accessibility of acoustic data, but also signify new problems with respect to data formats, safety and intellectual property rights. Digitized sounds still require considerable storage space, but with decreasing storage costs it will be possible to computerize whole sound archives and make them available to the public via the WorldWideWeb . Such distributed digital archives could be connected to one huge Virtual Phonothek. The value of the Virtual Phonothek can be enhanced by incorporation of descriptive parameters in a standard database format (SQL-standard). Though this additional information might be taxa specific, there is a standard set of acoustical (e.g. carrier frequency, pulse rate) and biological data (e.g. locality, time). Together with the actual sound file, such data sets make up a multimedia database which allows to answer completely novel, otherwise intractable questions. This is exemplified for crickets (Gryllidae: Orthoptera), together with a novel data compression algorithm for long-time song recordings and time series analysis. In the future, more complex searching algorithms will allow pattern recognition within original sound files. Some vertebrate sounds are already available on the WEB, but insect sounds are still an exception. General Availability of invertebrate songs could facilitate taxonomic analysis of these species-rich taxa considerably. Naturally, such a project creates conflicts between data providers and users: how can we assure intellectual property rights and scientific credit to those feeding the data base? I propose a three-stage model: data available for non-commercial public access and further scientific processing, data requiring permission of the author, and data sets restricted to a certain user community. Public availability should be mandatory for all sound recordings forming part of species descriptions. These rules can be integrated into the new biocode, and details for data formats formulated by the International Bioacoustics Council (IBAC).

This abstract is also available on the WWW, together with relevant hyperlinks to current projects, at:

/data/riede/phonothek.html