Our knowledge of fish distribution is ultimately based on and restricted to reported encounters between humans and fishes. It is the traditional task of taxonomists to collect, as thoroughly as possible, the species occurring in a certain area, to preserve specimens in a suitable manner, to properly identify known species, to formally describe species that are new to science, to deposit the collected specimens in a museum for reference, and to publish the results of this exercise. The continuing importance of such work as a precondition to our understanding of biodiversity has been recently stressed (e.g., di Castri and Younès 1994; Froese and Pauly 1994; Froese and Palomares 1995).

However, other types of encounters are also acceptable for occurrence records if they can be or have been verified, or if the chance of a misidentification is remote. Such encounters are underwater observations by divers, verified by an identifiable photograph or a video sequence; angling records verified by local experts and supported by a photo; research vessel surveys where the catch was identified by experts; industrial catches of species that are not easily misidentified; or tagging experiments with well-known species. The OCCURRENCES table is designed to accommodate information from these different sources in a standardized form. The FishWatcher database (see ‘National Databases’, this vol.) is a tool to report such encounters and to make them available to FishBase.

Ultimately, we believe that all reported occurrences of fish, old and new, should be accessible to researchers through FishBase. We believe that the analysis of such a dataset will lead to important insights about fish zoogeography. It will help in the conservation of fish by identifying areas of high diversity or high endemism. At the national level, it will repatriate data stored in other places, assist in resource assessment, and help in establishing protected areas (Froese and Pauly 1994).

In November 2000, the OCCURRENCES table contained over 630,000 records for over 19,600 species (see Fig. 11). These were drawn from twelve museum collection databases, and over 200 references.

We have also drawn occurrence records from other databases such as the fish collection records of the Zoology Department, University of British Columbia, and regional and national surveys, e.g., those documented in Vakily (1994), Künzel et al. (1996) and Pauly and Martosubroto (1996).

The sources mentioned above, although a good source of occurrence records in principle, require a considerable amount of quality checking. Museum collection records, for example, often contain misspellings and names that are no longer valid and, therefore, would need to be matched with and attached to current valid names.

Occurrence datasets contributed to FishBase go through a process of checking and validation outlined below before they are incorporated into the database. The amount of work required varies between datasets, and depends mostly on the number of records and the format used.

Here is our nine-step procedure for incorporating occurrence data of fishes:

1. Import dataset into MS Access format;

2. Match scientific names against FishBase, using the Check-Names procedure (this vol.). Assign names automatically to valid FishBase species, where possible. Send report of synonyms and misspellings to data provider. Ask for references for species not yet in FishBase.

The fields in the OCCURRENCES table are described below:

The Name Used in the publication or, in the case of a museum specimen, the name written on the label or in the catalog is given for reference purposes. This name may be different (synonym, misspelling or misidentification) from the valid FishBase name.

A Catalog No. or collection number is given, if available (default is ‘n.a.’). Where museum names are abbreviated, the full name and address may be found in the GLOSSARY table.

The Picture field is used when the record is documented through a fish picture. This can be secured by double-clicking on the field.

Information on the locality where the specimen has been collected is organized in several fields:

The Locality states the name of the place or water body as given on the label or in the catalog.

The Station field gives the name or code number such as is often used in research vessel surveys (see also the ‘EXPEDITION table’, this vol.).

A gazetteer links locality names with geographic coordinates. The Gazetteer field is a first attempt to standardize locality names in the OCCURRENCES table. So far, it has only been filled for 2,000 locality records. We are looking for existing gazetteers, preferably in digitized format, that could be used for this purpose.

Latitude and Longitude are certainly the best method to describe a locality and are given whenever available. Coordinates are particularly useful because they allow plotting of occurrence points (see ‘The WinMap Software’, this vol).

Country, FAO area and sea or river basin are given as an additional way to classify and access the locality. Assigning historical localities to modern countries is a particularly challenging task.

Altitude, Water depth, Salinity and Temperature describe environmental parameters.

Date, Year and Time of collection are given.

Information on the collected specimen(s) is stated in the following fields: Length and length type used (in case of more than one specimen, the Range is given), Weight in g (in case of more than one specimen, the mean weight is given), Number (of specimens collected or sighted), Life stage (egg; larvae; juvenile; adult; juveniles and adults); and Sex (females; males; mixed).

Representation of the species in the catch as Percent of catch in wet weight is given.

Abundance is classified by five choices derived from those used by birdwatchers: abundant (always seen in some numbers); common (usually seen); fairly common (chances are about 50%); occasional (usually not seen); scarce (very unlikely).

The Bottom and Gear fields record the type of substrate in the collection area and the gear used, respectively. Additional information pertaining to the collection can be given in the Remark field.

Fields identifying the collectors are: Vessel (name of the research vessel used in the expedition), Collector (person who collected the specimen), and Identifier (person who identified the specimen).

The Type field gives the taxonomic status of the specimen(s), i.e., holotype; syntype; paratype; lectotype; cotype; paralectotype; neotype; paratopotype. Type of storage used for the specimen is also identified in the Storage field.

The Record Type field distinguishes between the different sources of information. It has the following choices: trawl survey; other survey; museum record; type locality; tag/recapture; literature; recapture; fishery; angling record; other survey. Also, a multiple choice field is used to identify the Expedition that generated a record (see also the ‘EXPEDITIONS table’, this vol.).

The Validity field refers to the reliability of the occurrence record with the following choices: requires matching against distributional range; compatible with distributional range; doubtful, outside of distributional range; introduced; aquaculture or aquarium specimen(s).

We plan to add to the OCCURRENCES table fields for the coordinates of museum and show aquaria holding fish ex-situ, and to link these records with WinMap (see ‘The WinMap Software’, this vol.).

When a record refers to a fish that has been ‘tagged’ (or otherwise ‘marked’) and released, then recaptured, it can be entered into the OCCURRENCES table as a ‘Tag/recapture’ record. In this case, the body of the table, as described above is used for the information relating to the tagging site (location, time), and to the released fish (in which case the field for the ‘Catalogue number’ is used for the tag number).

Information on the recaptured fish (location, time, length) is then entered in the appropriate fields, along with the straight line distance between the tagging under the heading Recovery and recovery sites (in km), if available [this distance is also calculated by a built-in routine, using spherical geometry, from the two locations, if both were entered]. Using the straight line distance, the (minimum) swimming speed (in km/day) is then computed, given the number of days between tagging and recapture.

Tag/recapture data for only a few species (see e.g., Scomber australasicus) have been entered so far, mainly to test the ability of the design to accommodate diverse sets of tag/recapture data. Here again, we invite interested colleagues to share suitable data, and to work with us in extracting a maximum amount of insights from them.

You get to the OCCURRENCES form by clicking on the Range button in the SPECIES window and the Occurrences button in the STOCKS range window. Alternatively, you can click on the Occurrence button in the COUNTRY INFORMATION window (accessed by clicking on the Country button, then the CountryInfo button). You get to the FishWatcher table by clicking on the National Databases button in the Main Menu and the FishWatcher button in the NATIONAL DATABASES window.

On the Internet, you can access the OCCURRENCES table by clicking on the Occurrences link in the ‘More information’ section of the ‘Species Summary’ page. The Point Map link in the same section will show all occurrence records for the respective species which have coordinates. In the resulting lists, catalog numbers that are links will open the respective record in the respective museum fish collection. Point maps in the Internet are active, i.e., if you click on an occurrence point it will show the underlying record(s).

di Castri, F. and T. Younès. 1994. DIVERSITAS: Yesterday, today and a path towards the future. Biol. Int. 29:3-23

Froese, R. and D. Pauly. 1994. A strategy and a structure for a database on aquatic biodiversity, p. 209-220. In J.-L. Wu, Y. Hu, and E.F. Westrum, Jr. (eds.) Data sources in Asian-Oceanic Countries. DSAO, Taipei, 1994. CODATA, Paris.

Froese, R. and M.L.D. Palomares. 1995. FishBase as part of an Oceania biodiversity information system, p. 341-348. In J.E. Maragos, M.N.A. Peterson, L.G. Eldredge, J.E. Bardach and H.F. Takeuchi (eds.) Marine and coastal biodiversity in the tropical island Pacific region. Vol. 1. East-West Center, Honolulu, Hawaii.

Künzel, T., A. Darar and J.M. Vakily. 1996. Composition, biomasses, et possibilités d'exploitation des ressources halieutiques djiboutiennes. Rapport du projet DEP/GTZ. Direction de l'Elevage et des Pêches, Ministère de l'Agriculture et du Développement Rural, Djibouti, et Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ) GmbH, Eschborn, Germany. Tome 1: Analyse 63 p.; Tome 2: Données 296 p.

Pauly, D. and P. Martosubroto, Editors. 1996. Baseline studies of biodiversity: the fish resources of western Indonesia. ICLARM Stud. Rev. 23, 321 p.

Vakily, J.M. 1994. Sierra Leone Fishery Surveys Database System (FiDAS). IMBO, Freetown/ICLARM, Manila. Vol. 1. User Manual; Vol. 2 Technical Reference Handbook.

Rainer Froese, Rodolfo B. Reyes, Jr. and Emily Capuli