The Freshwater Biodiversity Information System (FBIS) Fish Data

The FBIS project is led by the Freshwater Reserach Centre (FRC) in partnership Kartoza Open Source Geospatial Solutions and the South African National Biodiversity Institute (SANBI). The project is funded by the JRS Biodiversity Foundation and the South African National Biodiversity Institute (SANBI). The system accepts and serves data on species occurrence, abundance and associated habitat parameters, for freshwater algae, invertebrates and fish. In addition to data gleaned from scientific papers, reports and university theses, the FBIS serves (1) invertebrate data from the former ‘Rivers Database’ (national SASS (South African Scoring System) data), (2) invertebrate data from the legacy ‘Biobase Database’ (national aquatic invertebrate data), (3) fish data from the Global Biodiversity Information Facility (GBIF) database, which includes fish data records from the South Africa Institute for Aquatic Biodiversity (SAIAB).​​​​​​Through links with other systems (like the Department of Water and Sanitation’s Water Management System and SANBI’s Biodiversity GIS platform), the FBIS offers powerful spatial and temporal data filtering functionality designed to cater for the needs of end users ranging from consultants to conservation planners.

The geographic scope of this project was restricted to the boundaries of the Republic of South Africa, but also included the Kingdom of Eswatini (formerly Swaziland) and the Kingdom of Lesotho (for convenience purposes, this study area is hereafter referred to as ‘South Africa’). Data collection was limited to rivers, dams and freshwater lakes within South Africa. Importantly, estuarine and marine systems were excluded.

Using a Master Species List of freshwater fishes (see below for details on how this list was developed), a comprehensive freshwater fish data set was built, using data from multiple sources, in order to compile all known, accessible records of freshwater fish occurrences in South Africa. Specific details pertaining to the scope of the data set, as well as the collection and analyses of the data, are provided in the sections below. All data were uploaded onto the Freshwater Biodiversity Information System (FBIS) database (Dallas et al., 2021; freshwaterbiodiversity.org). The FBIS database is an online, open-access biodiversity database developed by the Freshwater Research Centre (FRC) in collaboration with the Kartoza and the South African National Biodiversity Institute (SANBI) and funded by the JRS Biodiversity Foundation. The FBIS database will serve as the long-term storage facility for the data collated during this study. Coverage includes all freshwater fish species known to occur in the wild (species held in captivity were excluded) within South Africa. For the purpose of this chapter, only formally described species recognised using the GBIF taxonomic backbone were included. All formally described sub-species were also included, listed under their parent taxa. Species in the process of being described, or that belong to groups that are currently under taxonomic review (such as those belonging to the genera Enteromius, Galaxias and Sandelia) were omitted completely from the data collection and analyses of this chapter, pending resolution of their taxonomic status. Freshwater fish species are defined as species that spend all, or a critical part, of their lives in either freshwater or brackish environments (IUCN Freshwater Fish Specialist Group, 2015; Arthington et al., 2016). For the purpose of this study, only primary freshwater fish (fishes living in inland waters, with little or no tolerance of saltwater), secondary freshwater fish (fishes relatively tolerant of salt water, but normally living in inland waters) and diadromous/catadromous fish (fishes that, in the course of their life cycle, regularly migrate between inland waters and the sea, or vice versa) species were included. Importantly, both marine peripheral and sporadic marine species were omitted entirely from both the data collection and analysis (see Skelton (2001) for ecological-tolerance category definitions). Based on these criteria, a preliminary freshwater fish species list was compiled using the South African Institute of Aquatic Biodiversity’s (SAIAB) database and the RHP (River Health Programme) (2015) Rivers Database (Dallas et al., 2007). This preliminary list of taxa was then updated using the freshwater fish field guide by Skelton (2001), as well as subsequent scientific papers, including Skelton (2002), Skelton (2016), Ellender and Weyl (2014) and Weyl et al. (2020). From these sources, a preliminary Master Species List for freshwater fishes was compiled (See Table S. 2). This contained 297 fish species (comprising both native and non-native fishes) known to occur in the whole of southern Africa (defined by Skelton (2001) as the region “south of the northernmost watershed of the Cunene, Okavango and Zambezi rivers, but excluding Lake Malawi and the Shire River above the Kapachira Falls”; Skelton, 2001). All 297 fish species were included in this preliminary Master Species List to ensure that any recently translocated or introduced species from one of South Africa’s neighbouring countries were not mistakenly omitted from the data search. 2.2.3. Data Collection Occurrence Data Published Data A thorough literature search was undertaken for each species included on the preliminary Master Species List, starting in November 2018 and culminating in January 2020. The literature search was conducted using Google Scholar. To ensure that relevant papers were identified, the following search term combination was used for all native freshwater fishes: “Genus species” and “South Africa” (where “Genus species” represents the scientific name of each taxon). A different search term was used for non-native species, to further refine the search. The following search term was used: "Genus species" and "South Africa" and "study" OR "survey" OR "sample" OR "site" and "map" OR "coordinate" OR "coordinates" OR "latitude" OR "longitude" OR "GPS"” (where “Genus species” represents the scientific name of each taxon). Relevant articles were identified based on their titles, abstracts, methods and results sections. Only the first 500 search results returned by Google Scholar were assessed for relevancy, due to time constraints. This number was selected based on an initial trial phase, which determined that after 500 search results, Google Scholar mostly outputted irrelevant, obscure articles. Articles (including peer-reviewed publications, theses and published reports) were deemed to be relevant if information from the title, abstract or methods section indicated that the article contained occurrence data for the specific freshwater fish taxon in South Africa. If an article was deemed to be relevant, the article was catalogued in a Microsoft Excel (365 ProPlus) spreadsheet and the PDF version of the article was downloaded. Data from each relevant article were extracted once all species searches were complete. Detailed explanation of the data extraction process is provided below. If an article was not deemed to be relevant, this was also catalogued in Microsoft Excel and a reason for disregarding the article was logged. Databases Occurrence data for freshwater fishes in South Africa were also collected from 15 national databases curated within FBIS, and from one international database, the Global Biodiversity Information Facility (GBIF), which included the South African Institute for Aquatic Biodiversity (SAIAB) database. (Table 2. 1). Table 2. 1. List of databases accessed to search for freshwater fish occurrence records in South Africa. # Name of Database 1 Global Biodiversity Information Facility (GBIF)* 2 South African National Parks Fish Database 3 Limpopo Department of Environmental Affairs Fish Database 4 Rivers Database 2015 5 Mpumalanga Tourism and Parks Agency Fish Database 6 Cape Nature State of Biodiversity Database 7 Clean Stream Biological Services Fish Database 8 Inkomati-Usuthu Catchment Management Agency Database 9 Nepid Fish Database 10 Ecotone Freshwater Consultants Database 11 Ecosun Database 12 Department of Water and Sanitation Regional Fish Database 13 Gauteng Department of Agriculture and Rural Development Database 14 Freshwater Research Centre Monitoring Database 15 The Biodiversity Company SASS and Fish Database * The Global Biodiversity Information Facility’s (GBIF) database. These data included all data uploaded to GBIF by South African and International research institutions like SAIAB, as well as citizen science data from iNaturalist that were classified by GBIF as “Research Grade” data. Unpublished Data Several freshwater fish experts and conservation organisations from across the country were identified and contacted to request any additional unpublished data that may exist in analog form, or on private hardware (See Table S. 3). Additionally, several presentations were delivered at conferences, workshops and meetings to appeal for data. In October 2019, the FBIS database was launched online and opened for the upload of additional data by approved, registered users. These data were validated by experts and downloaded on 22 September 2022, and also incorporated into this study. 2.2.4. Data Extraction Each article that was deemed to be relevant during the literature search was inspected for GPS coordinates. In cases where GPS coordinates were not provided, the paper was temporarily archived and catalogued, and an email was sent to the listed corresponding author asking whether these were available and if so, requesting a copy. Where site-specific coordinates were provided, these were converted to decimal degrees and entered into the dataset. Available metadata, including collector details, physico-chemical data, sampling method and abundance measures were also documented for each record. See Table S. 4 for details of the full list of parameters included during data extraction. All data were then uploaded into the FBIS. 2.2.5. Biological Information Species Diversity and Origin Once data collection was complete, species with no known occurrence records in South Africa were removed from the preliminary Master Species List. This list was then sent via email to various freshwater fish experts (See Table S. 3) in South Africa for comment. After several iterations, a subset of 134 species were included in the first Master Species List for freshwater fishes in South Africa (See Table S. 5). The GBIF database (GBIF.org, 2018) was used to ensure that the most recent, accepted scientific name of each species was recorded. All previous scientific names and subspecies names were also included as attributes in the Master Species List. In addition, the taxonomic classification (Kingdom, Phylum, Class, Order, Family, Genus and Species) was downloaded for each species using the GBIF database (GBIF.org, 2018) and cross-referenced with current information available on the International Union for the Conservation of Nature (IUCN) Red List of Threatened Species website (IUCN Red List, 2020). Common names for each species were sourced using the SAIAB database, the freshwater field guide of Skelton (2001), or the IUCN Red List of Threatened Species website (IUCN Red List, 2020). Information pertaining to species origin was determined using historic distribution records from the IUCN Red List of Threatened Species website (IUCN Red List, 2020), the GBIF database (GBIF.org, 2018), Skelton (2001) and Ellender and Weyl (2014). Each species was classified as being either ‘native’ (taxon occurring within its natural range – past or present – and dispersal potential) or ‘non-native’ (alien or extralimital taxa) (Dallas et al., 2021). Ecological Tolerance Category Information pertaining to the ecological tolerance of each species was determined using Skelton (2001). Each species was classified as being either ‘primary freshwater’; ‘secondary freshwater’; or ‘diadromous’. This classification was confirmed with relevant fish experts in the country (Skelton, personal communication, 01 April 2020; Chakona et al., personal communication, 14 September 2020). Endemism The endemism category of each species was determined using definitions provided on the FBIS (Dallas et al., 2021). Species were classified based on the following definitions: Widespread – occurring in more than one Freshwater Ecoregion and not restricted to southern Africa; Subregional endemic – endemic to southern Africa; National endemic – occurring in more than one Freshwater Ecoregion, but restricted to South Africa; Regional endemic level 1 – endemic to South Africa, but occurs in more than one primary catchment within a single Freshwater Ecoregion; Regional endemic level 2 – endemic to one primary catchment in South Africa; Micro-endemic 1 – endemic to less than five rivers, within one primary catchment in South Africa; Micro-endemic 2 – endemic to one river in South Africa. Conservation Status The IUCN Red List of Threatened Species website (IUCN Red List, 2020) classifies species into six main categories based on their extinction risk. Each species was classified as either ‘Not Evaluated’, ‘Data Deficient’, ‘Least Concern’, ‘Near Threatened’, ‘Vulnerable’, ‘Endangered’ or ‘Critically Endangered’.

The FBIS project is led by the Freshwater Reserach Centre (FRC) in partnership Kartoza Open Source Geospatial Solutions and the South African National Biodiversity Institute (SANBI). The project is funded by the JRS Biodiversity Foundation and the South African National Biodiversity Institute (SANBI).