Huitfeldt Kaas: Freswhater fish distribution in Norway 1918

Study Extent

The study area covers lakes on the Norwegian mainland. The original source also to some degree covers rivers - these have been excluded from this dataset.

Sampling

The original source data were collected by Hartvik Huitfeldt-Kaas while working as a fisheries inspector for the Norwegian government in the period from 1902. The main method for data harvest was questionnaires distributed to local fisheries experts by local administrators in the various hundreds (former administrative units) throughout the country. In addition, supplementary information was gathered through personal travels throughout the country. Also, questions on establishment means were included in the questionnaires. The information was published in Norwegian in 1918 and in German in 1923 (see bibliographic citations). The current dataset is based upon transcription of the Norwegian text using the open-source Citizen Science transcription tool "Dugnadsportalen" developed by the Norwegian GBIF Node (https://dugnad.gbif.no/nb_NO/ - see https://github.com/gbif-norway/dugnad/ for source-code documentation). In the original text, occurrences of various species of freshwater fish in individual water bodies are organized by species and geographic area (municipality and county). The water bodies are identified by name, and due to archaic naming, these have been manually matched to a gazetteer of current Norwegian lakes. Please note that the original text also includes some references to rivers, streams and general watersheds. These are not transcribed in the current project. Lakes are in addition to name (not unique) identified by a locationID of Norwegian lakes maintained by The Norwegian Water Resources and Energy Directorate (NVE:vatnLnr). This can be used to match individual waterbodies with hydrological and other characteristics (e.g. area, perimeter, catchment). See https://www.nve.no.

Quality Control

Occurrences have been manually matched to lakes. Obvious errors in terms of missing location information and duplication of occurrences were removed before publishing. Documentation available on https://github.com/gbif-norway/data_huitfeldt-kaas_1918. All transcriptions are open to public scrutiny (https://dugnad.gbif.no/nb_NO/) and a visual representation of the data including feedback form for errors can be found on https://shiny.vm.ntnu.no/users/andersfi/data_huitfeldt-kaas_1918/. The transcription tools is designed for Citizen Science type of projects, but this dataset origin from a test run of the tool, and all transcriptions were conducted by trained personnel.

Method steps

1. The transcription process starts with manually matching historical lake names with current names as present in a gazetteer of Norwegian lake names based upon The Norwegian Water Resources and Energy Directorate's lake database ("innsjødatabasen").
2. Data on fish occurrences are given a unique ID at the time of transcription that constitutes the occurrence ID (uuid). Raw data are available from the Dugnadsportalen server (https://dugnad.gbif.no/nb_NO/project/huitfeldt-kaas/export.json) and downloaded and parsed for further processing in R. Post-processing documented in R script available at https://github.com/gbif-norway/data_huitfeldt-kaas_1918.
3. Occurrences are matched to lake, and observations from each lake are given an eventID (UUID). As such, observation of fish from a given lake in the source text is regarded as one sampling event. The eventIDs are predefined (one eventID for each Norwegian lake).
4. Duplicated occurrences (i.e. one taxa transcribed several times for one lake) are removed. Duplicates occur in raw data due to instances of duplicated transcription. This is a wanted feature of the transcription tool as it is designed for multiple users and Citizen Science style projects, and intended to be used for verification purposes. However, this feature is not utilized in the current project.
5. Absence information is inferred when a species not is mentioned in the text for each individual lake. These are recorded as occurrenceStatus="absent". OccurrenceIDs (UUID) for absences are predefined for each taxon/lake combination and assigned. This step is needed due to the organization of the original data which not are organized by sampling site (i.e. lake), but by species.
6. XY coordinates (in WGS84) for lake centroids were obtained from the Norwegian lake database directly (downloaded from http://nedlasting.nve.no/gis/ 2019-01-26) using the "polygon centroid" tool in QGIS 2.18.16. Additional information common for all observations (e.g. recordedBy) added and dataset split into event (core) and occurrence (extension) tables

Mainland Norway