FishBase

FishBase is a comprehensive database of information about fish (finfish).^{[citation needed]} As of June 2011^[update], it included descriptions of 32,000 species and subspecies, 291,200 common names in almost 300 languages, 50,400 pictures, and references to 45,800 works in the scientific literature.^{[citation needed]} Fishbase has been cited over 1,500 times.^{[citation needed]}

Overview

In 1987, Daniel Pauly, inspired by the Species Identification Sheets and other products Walter Fischer had generated for the Food and Agriculture Organization in the 1970s, proposed a standardized database for fish species, as a part of the "ICLARM Software Project".^{[citation needed]} The following year, he began to work with Rainer Froese, who had been working on an expert system to identify fish larvae. After a first attempt to build a system using Prolog, Froese switched to DataEase, a relational database for DOS. In 1989 the project received its first grant.^{[citation needed]}

In 1993 the project switched to Microsoft Access, and 1995 the first CD-ROM was released as "FishBase 100". Its initial reviews in scientific journals lauded the scope but criticized the remaining gaps in coverage. Subsequent CDs were released annually, with the FishBase 2004 release needing five CDs, or one DVD. The software needs Windows 98 or later, and is not available on any other computer platform such as Mac OS X or Linux.^{[citation needed]}

FishBase first appeared on the World Wide Web in August 1996, and a webmaster was hired in the following year. In 1999 a new module on ichthyoplankton and detailed data on fish larvae identification and rearing was created (LarvalBase) under the supervision of Bernd Ueberschaer. Eventually, the complete data of the CDs became available online.^{[citation needed]}

As awareness of FishBase has grown among fish specialists, it has attracted over 1,700 contributors and collaborators. In order to preserve its value as a scientific database, FishBase is not allowed to include original data; all of its content must be based on previously-published material.^{[citation needed]}

FishBase Consortium

Since the year 2000, FishBase has been run by the FishBase Consortium. The consortium consists of:

Aristotle University of Thessaloniki, Thessaloniki, Greece
Chinese Academy of Fishery Science, Beijing, China
Fisheries Centre, University of British Columbia, Vancouver, British Columbia
Food and Agriculture Organization, Rome, Italy
IFM-GEOMAR, Kiel, Germany
Muséum National d'Histoire Naturelle, Paris, France
Royal Museum for Central Africa, Tervuren, Belgium
Swedish Museum of Natural History, Stockholm, Sweden
WorldFish Center, Penang, Malaysia

Reception

FishBase has been warmly received and widely cited.^{[citation needed]} A Google Scholar search shows over 1500 citations. Since FishBase reflects the published literature, it also contains doubtful estimates.^{[citation needed]}

Accuracy

The problem of errors in FishBase is well known by the FishBase Consortium and receives considerable attention during each management meeting.^{[citation needed]}
Let's try to rationalize the approach by categorizing them:

FishBase at its origin used synthetic works like FAO catalogues and well known ichthyofaunas such as Fishes of the Northern Atlantic and Mediterranean. Errors may come from the author who synthesized primary sources.^{[citation needed]}
Errors exist in primary sources, even if peer-reviewed and acknowledged sources. Some of the most criticized sources are the occurrences data repositories such as GBIF and OBIS that cannot clean themselves the source databases curated by the database owner (see Robertson 2008).
FishBase encoders mistyped some information. One of the funniest typos in earlier times was the maximal length for the whale-shark as ... 18 cm! It was corrected as soon as depicted. Quality control procedures, including consultation of specialists are continuously set up and run, but the 100% cleaned data is not realistic due to the complexity of the domain itself.^{[citation needed]}
Some pictures are misidentified, but when errors are signaled, they are corrected in the next update.^{[citation needed]} Some of the "misidentifications" are due to the evolution of the continuous evolution of the taxonomy.
The errors in taxonomy and nomenclature have been studied in detail already and concern less that 3% for species validity, and about 5% for names (including the spelling of authors and the date of publication; see Bailly 2010).
Some missing data (i.e., not found in the literature) are estimated with mathematical models, but are flagged as such when they have not been checked and confirmed (e.g., the map in the species summary page). No model is perfect, we correct the entries when errors are signaled.^{[citation needed]}
Errors pointed out by users are corrected in the next bi-monthly update ^{[citation needed]} such as the ones signaled by Cole-Fletcher et al.^[1] on some length-weight parameters. A "Comments and corrections" link at the bottom of the most important pages allows the user to signal mistakes.
Errors are not errors, but:
- reflect the uncertainty of the current knowledge, and some decisions are taken and documented as far as possible, which must be understood in their context (see Bailly 2010), or
- the entry was not understood by the user signaling the mistake (it happens).^{[citation needed]}
Non-published information is rarely used (in particular for taxonomy), but personal communications are used to reference the correction in obvious cases, awaiting for proper publication.^{[citation needed]}

Notes

^ Errors in Length-weight Parameters at FishBase.org S. Cole-Fletcher, L. Marin-Salcedo, A. Rana, and M. Courtney. (2011). Cornell University Library.

References

Bailly, N. 2010. Why there may be discrepancies in the assessment of scientific names between the Catalog of Fishes and FishBase. Version 2: 06 May 2010. [1]

Froese, R. and D. Pauly, Editors. FishBase 2000: concepts, design and data sources (ICLARM, Los Baños, Laguna, Philippines, 2000)

Robertson, R. 2008. Global biogeographical data bases on marine fishes: caveat emptor. Diversity and Distributions 14(6):891-892. [2]

External links

For early roots and possible future extensions see the German page.

Mirror sites:

[Cole-1] Errors in Length-weight Parameters at FishBase.org S. Cole-Fletcher, L. Marin-Salcedo, A. Rana, and M. Courtney. (2011). Cornell University Library.

[1]