DORSA will provide database access to important collections of Orthoptera (crickets and grasshoppers) in German research collections. DORSA will function as a prototype for an internet-based link between an existing Global Species Register (OSF, see below) and a specimen-based database, including geo-referenced localities and multimedia information such as pictures, video-files and especially songs. Specimen-based information from different institutions will be integrated within a single database, creating one "Virtual Museum Collection", accessible by an internet-based user interface. DORSA will be accessible from any part of the world, which is important for potential users in species-rich, but resource-poor developing countries with incipient biodiversity infrastructure. The "Virtual Museum Collection" will help to improve classical taxonomic work such as description of new taxa. Further important functions of this database, including distribution map generation and retrieval of pictures and songs from determined specimens, will be especially useful for ecologists, conservationists and applied entomologists, facilitating species identification by looking at pictures or by comparative analysis of songs. In particular, species-specific songs of Orthoptera are an important diagnostic character for species determination and monitoring. The development of "Rapid assessment tools" based on songs will be accelerated by direct use of databased songs (e.g., feeding of neural networks with songs from the database). A tool for automatic song recognition will be developed as part of a Ph.D. thesis.
The relatively small insect order Orthoptera has been selected for a number of critical reasons:
- a Global Species Register of Orthoptera is already available on the Internet. This "Orthopteran Species File" (OSF: Otte & Naskrecki 1997) contains all described species and will serve as a taxonomic backbone for DORSA.
- German research institutions harbour important collections of Orthoptera, including material from tropical countries dating back to the last century. Most of this material is not databased, or only available as a simple flat file.
- Germany has a long tradition in Orthoptera bioacoustics, which resulted in several important sound archives ("phonotheks") at Universities and in private hands. These should be integrated into museum collections and administered by an appropriate multimedia database.
- The same tradition produced a number of specialists with sufficient expertise to help with taxonomic determination, supervision of accompanying masters and Ph.D.-thesis and additional fieldwork.
- Many Orthoptera are agricultural "pests" because of mass outbreaks. Therefore, DORSA could be important for applied entomology and pest control, especially in tropical and subtropical countries.
The rich and important German insect collections contain material about 150 years old, in part from tropical localities which might have been altered or destroyed in the meantime. The process of databasing should cover the major museum collections, but should also include several important bioacoustic collections at universities and in private hands. The total number of specimens is estimated about 500.000, and the number of types and paratypes probably around 5.000. Therefore, data capture should focus on documentation of type specimens (including secondary types). In addition, several German research groups collected considerable numbers of tropical Orthoptera by canopy fogging in recent times (and probably will continue) also within the BIOLOG project. This interesting material contains many undescribed taxa and should be integrated into museum collections and databased with high priority.
The multimedia functionality of the database can easily be adapted for the administration of all other storage-intensive data-sets, as for example chromosome pictures, isozyme data, 3-D-data of scanned animals and DNA sequences.
2.2. Previous work and experience
The project staff has many-years experience with databasing of entomological collections (K.-H. Lampe; BIODAT) or taxonomy of Orthoptera and bioacoustics especially with species rich tropical fauna (K. Riede, S. Ingrisch). Further experience concern linking between geographical information and zoological databases with publication in world wide web (Global Register Of Migratory Species - GROMS), Riede 1998a).
An inquiry on the existance and condition of animal sound collections in Germany was done in collaboration with Dr. G. Peters, ZFMK Bonn, in 1998. For Orthoptera, there are at least seven comprehensive sound libraries in private hand and at universities (including the sound libraries of Riede and Ingrisch). The valuable sound library of A. Faber recorded in the fifties and thought to be lost, was recently acquired by the Staatliches Museum fur Naturkunde Stuttgart.
Riede (1993) used sound analysis in a tropical rain forest of Equador to estimate the bioacoustic diversity and the number of cricket species in a given area. This work was extended in the doctoral thesis of Nischk (1998, 1999), who prooved by collecting and individual sound recording that the species specific sounds are really a reliable tool to identify the cricket species. Thus, bioacoustic studies are a fast and non-invasive method for the registration of the Orthoptera fauna in a given area. An extention of this method away from a local frame failed so far due to the lack of a broad database of sound recordings of identified species. In grasshoppers and katydids the structure of the sounds is much more complicated than in crickets, but also those sounds are classifiable with certainty (summary in Riede 1998b).
A pilot version of a JAVA Mapserver is already running within the project "The Global Register of Migratory Species (GROMS)". It was developed in collaboration with the Geographical Institute of the University Bonn (Riede 1998a, Fitzke & Friebe in press). It is platform independend, searcheable and figures the results on a zoomable world map (Fig. 1).
Fig. 1: Pilot version of the JAVA mapserver website (Mapserver).
3. Database software - criteria and adaptation
At present, several software tools have been developed for collection management. Two powerful programmes with Internet access (BIOLINK: Australia; SPECIFY: US) are available as beta-versions. Both are very complex and probably too complicated to be handled by technical staff. German programmes for the handling of zoological collections have been developed by the Senckenberg Institution (SESAM) and by the ZFMK (BIODAT). The latter contains various tools supporting ergonomic and efficient data input. However, both German programmes do not provide Internet access, which has to be developed on an SQL Client-Server-system level. Available programmes should be tested before application, which could also combine a German data-entry module interacting with BIOLINK or SPECIFY, respectively. For the storage of specimen and multimedia information, the database SYSTAX will be used in collaboration with other subprojects of EDIS. However, the database SYSTAX must be adapted to the requirements of DORSA and the other EDIS subprojects first. Adaptation of SYSTAX will be done by Dr. J. Hoppe of the Institute of Systematic Botany and Ecology, University Ulm.
4. Working plan
The tie of the GIS interface with SYSTAX will be done parallel to the registration of specimen information. The geographical coordinates can be stored as simple numeral data and independent of the special requirements of the different EDIS subprojects. A Beta- / preliminary Version of a JAVA mapserver (Fig. 1) will be further developed and adapted for DORSA in cooperation with the EXSE work group at the Geographical Institute of the University Bonn. As a prerequisite of EDIS, a link to SYSTAX has to be developed. Moreover, it would be useful if that project specific GIS tasks become possible, as for example the presentation of the geographical distribution of higher taxa as genera or families, or the number of species described from or found in a selected area.
Registration of specimen information
The registration of specimen information will start with databasing of type specimens and the re-evaluation of published errors concerning the type material.