An inventory of native-alien populations in South Africa

Species can be both native and alien to a given administrative region. Here we present the first consolidated inventory of these ‘native-alien populations’ for South Africa, and provide an overview of the data it contains. To gather data, literature searches were performed and experts were consulted both directly and via an on-line survey. Putative native-alien populations were then scored based on a newly developed protocol. The final inventory contains information on 77 native species from 49 families across nine classes that have formed 132 native-alien populations across the terrestrial, freshwater, and marine environments. The phenomenon is rare when compared to the prevalence of related phenomena, such as alien species introduced from other countries (2033 alien species in South Africa), but is under-reported. However, they pose a specific problem for regulators and managers and their importance will likely increase with global change. These data will be integrated with an existing alien species list and, we hope, will provide a useful foundation to address the issue. We encourage those working on biodiversity to contribute more records.

www.nature.com/scientificdata www.nature.com/scientificdata/ Under the NEM:BA A&IS Regulations, the South African National Biodiversity Institute is mandated to report on the status of biological invasions and their management every three years (see http://iasreport.sanbi.org.za/). As part of the process followed to produce the national status report on biological invasions, a South African alien species inventory has been developed to systematically and consistently record information in line with global data standards 13 . Such actions, and the inclusion of additional information on factors such as pathways of introduction and dispersal, date of introduction, and degree of establishment, increases the usefulness of these inventories 14 . Here we aimed to develop an inventory that: (1) consolidates the available information on native-alien populations in continental South Africa (i.e., excluding the sub-Antarctic Prince Edward Islands), (2) includes additional data on these populations that are vital for research and management; and (3) follows global data standards, and as such can be integrated with the existing alien species list produced as part of South Africa's national status report on biological invasions. In this paper we present this inventory and provide an overview of the data it contains. This represents the first inventory of native-alien populations in South Africa, and is a step towards a greater understanding of native-alien populations and the biosecurity threat they pose.  The number of native species with native-alien populations (black), the number of alien species introduced to South Africa from other countries (grey), and number of native species from South Africa (white) across eight taxonomic groups. Note that the axis is on a log scale. Native-alien population data are from this study, data on other alien species are from the species list of South Africa's national status report on biological invasions 13 , and native species data are from the National Biodiversity Assessment 51 .
www.nature.com/scientificdata www.nature.com/scientificdata/ Methods Data collection. Records for native-alien populations in South Africa were gathered through an online survey, through direct discussions with experts, and through online searches.
An online survey was created using Google Forms (see Supplementary material 1), and experts were consulted through snowball sampling (experts were asked to nominate other experts until no new experts were identified). A request for information was also made during presentations at the South African National Symposium on Biological Invasions (Tulbagh 2019). Individuals who responded to this request were approached for information and were asked to nominate and provide the details of other experts who could provide information on www.nature.com/scientificdata www.nature.com/scientificdata/ this topic. Experts were then consulted via email and using the online survey between July 2020 and May 2021. This online survey included questions on the higher taxon-group, scientific name, common name, native range, location, and references for suspected native-alien populations (see Supplementary material 1 for full online survey and email that was sent to experts). A total of 21 of 29 experts contacted responded.
To augment the data obtained from the experts, the ISI Web of Knowledge and Google Scholar were used to search for scientific publications and grey literature on native-alien populations in South Africa. Searches were performed between February 2019 and May 2021. Terms that have been used to refer to this phenomenon (see Box 1 of Nelufule et al. 4 ) were used as search strings, for example, "domestic exotics", "intra-country established alien species" , "home-grown exotic" , "extralimital species" and "native-alien species" . Additional searches were performed by adding "AND South Africa" to these terms. Relevant papers (i.e. those with information on populations in South Africa) were selected based on the content of their titles and abstracts. A total of 23 published articles, two books and four scientific reports containing information on native-alien populations in South Africa were used for this study.
A draft inventory was produced using the collated information from the literature review and expert consultation. The draft inventory was sent back to the consulted experts for comments on errors and omissions six months after they were initially contacted. Follow-up emails were also sent to experts to encourage those who had not participated in the online survey to do so, and to encourage those who had participated to add any new suspected native-alien populations to the inventory. A request for individuals to consult the draft inventory and provide additional information (e.g. on populations missing from the inventory) was submitted to a South African list server on biological invasions on the 6 th of August 2020 (invasives@wordlink.co.za), at the time that the request was sent out 450 people subscribed to this list server. The online survey was also published in the newsletter of the Entomological Society of Southern Africa in August 2020 15 . classification of native-alien populations. The compiled inventory comprised a list of 176 suspected native-alien populations, of which 139 were collated from the literature and 37 from the experts (Fig. 1). The data collected on suspected native-alien populations came from various sources that used various frameworks and definitions. Therefore, a standardised protocol for classifying native-alien populations was developed and used to evaluate each population and determine whether it is a native, cryptogenic, alien or native-alien 8 . A total of 44 populations did not meet the criteria of a native-alien population during classification, and were excluded from the current inventory. Excluded populations either occurred within their native range (i.e. native populations), had uncertain native ranges (i.e. cryptogenic), were the result of range expansion in response to human-induced environmental change or there was uncertainty on whether the populations were outside their historic native range (Fig. 1).
Structure of the inventory. We followed the data structure used for the species list of South Africa's second national status report on biological invasions 'The status of biological invasions and their management in South Africa in 2019' 13 . In line with this report we tried to ensure the data were FAIR (Findable, Accessible,  www.nature.com/scientificdata www.nature.com/scientificdata/ Interoperable, and Reusable) and Darwin Core Terms were used where appropriate 16 (see Table 1 for a list of terms used). The species names in the inventory were standardised according to the GBIF Backbone Taxonomy (https://doi.org/10.15468/39omei, accessed: 20 May 2020). Dispersal pathways were classified using the classification scheme of the Convention on Biological Diversity 17 , following the guidance provided in Harrower et al. 18 . Introduction status was classified as per the unified framework for biological invasions 19 , environmental impacts were as per the Environmental Impact Classification for Alien Taxa scheme 20 , and socio-economic impacts as per the Socio-Economic Impact Classification for Alien Taxa scheme 21 . For each entry, we included the source from which the data were obtained and, following the guidance outlined in Wilson et al. 22 , a confidence estimate of high, medium, or low, was assigned so that the degree of confidence in every category is explicit (see the metadata for full explanations of these confidence estimates 23 ).

Data Records
Structure of dataset. The dataset is archived and available from figshare.com as a comma delimited file (.csv) 23 .
There are two files in figshare namely; "List of native-alien populations in South Africa.csv" , and "metadata.doc". The dataset, contains an inventory of species native to South Africa that have formed native-alien populations in South Africa. The column names of the dataset are shown in Table 1. In the dataset each row below the header represents a record for a single native-alien population 23 . An NA in a cell means that no information was obtained, while DD means data deficient. The metadata contains a full description of the columns in the dataset 23 .  www.nature.com/scientificdata www.nature.com/scientificdata/ Summary of the inventory. We found a total of 77 native species from 49 families across nine classes that have formed 132 native-alien populations in South Africa (Tables 2-4 Native-alien populations are rare when compared to the prevalence of related phenomena-0.1% of native species have formed native-alien populations, and the number of alien species introduced from other countries is 25-fold higher than the number of species with native-alien populations-but it is likely under-reported. Native-alien populations are particularly prevalent in specific taxonomic groups. Most species with native-alien populations were plants, and plants had more recorded native-alien populations than other taxonomic groups. However, fish had the highest percentage of native species with recorded native-alien populations (Fig. 2). All other taxa had a low percentage of native species with recorded native-alien populations (<= 4%) ( Supplementary Fig. 1). The terrestrial environment (n = 101) had a higher number of recorded native-alien populations than freshwater (n = 26) or marine (n = 5) environments .
The pathways of dispersal through which these populations were introduced differed by organism type (Fig. 3). Most fish and mammal populations were released intentionally, while most bird and plant populations escaped from captivity or cultivation. For many groups, the majority of the populations were accidentally introduced, with most gastropod, amphibian and insect populations introduced as contaminants on transported goods, and most reptile populations introduced as stowaways on transport vehicles. Mammals were released for hunting and improving eco-tourism 24,25 . Some individuals of native-alien populations were dispersed through a single pathway while others were dispersed through more than one pathway of dispersal. All species that dispersed through corridors were fish (Fig. 3). Data for degree of establishment was available for 77% of the recorded native-alien populations, but a relatively large number of vertebrate and plant native-alien populations had an unknown degree of establishment (Fig. 4). The majority of native-alien populations (59%) are established, while very few (18%) are invasive (Fig. 4).
There were few records of potential impacts, but for the populations for which they were available, these impacts differ between organism types and were mostly Minor 23 . The highest impact recorded for gastropods and fish was Minor, Moderate for plants, amphibians, insects, and mammals, and Major for reptiles. Hemidactylus mabouia (Tropical house gecko) was recorded as having Major environmental impacts when introduced to another country. The gecko competes with native species for resources in the United States of America (USA) where it has caused the loss of local populations 26 .  www.nature.com/scientificdata www.nature.com/scientificdata/ The majority of species with native-alien populations (74%) are native to multiple provinces and biomes of South Africa, while 24% are native to one province and biome 23 .
South African native species have formed native-alien populations across the terrestrial, freshwater, and marine environments, with native-alien populations more prevalent in some parts of South Africa than others (Fig 5a).
The Western Cape province (n = 58) had the highest number of recorded terrestrial native-alien populations followed by the Eastern Cape province (n = 22) (Fig. 5a). All other provinces had few recorded terrestrial native-alien populations (n < = 7) (Fig. 5a). Most of the native-alien populations that have been introduced to the Western Cape province are native to the KwaZulu-Natal province followed by the Limpopo, Mpumalanga and Eastern Cape provinces (Fig. 5b).
The Fynbos biome (n = 55) had the highest number of recorded terrestrial native-alien populations followed by the Grassland (n = 19), and Albany Thicket biomes (n = 11) (Supplementary Fig. 2). All other biomes had few recorded terrestrial native-alien populations (n < = 8) (Supplementary Fig. 2). Most of the native-alien populations that have been introduced to the Fynbos biome are native to the Savanna and Grassland biomes (Fig. 5c).
Of the biomes, the Fynbos had the greatest number of recorded native-alien populations. The Fynbos (0.06 native-alien populations per km 2 ) and Albany Thicket biomes (0.04 native-alien populations per km 2 ) had a relatively large number of recorded native-alien populations relative to their total area. The Forest (0.01 native-alien populations per km 2 ), Succulent-Karoo (0.01 native-alien populations per km 2 ), Savanna (<0.01 native-alien populations per km 2 ), Nama-Karoo (<0.01 native-alien populations per km 2 ), and Grassland biomes (<0.01 native-alien populations per km 2 ) had the lowest number of recorded native-alien populations relative to their total area.
The Tugela river catchment (n = 12) had the highest number of recorded freshwater native-alien populations followed by the Limpopo river catchment (n = 8) (Supplementary Fig. 3). All other river catchments had few recorded freshwater native-alien populations (n < 6) ( Supplementary Fig. 3). The majority of freshwater fish native-alien populations are native to the Orange and Vaal river catchments (Fig. 5d).
The Agulhas ecoregion (n = 4) had the highest number of recorded marine native-alien populations followed by the Benguela ecoregion (n = 1) 23 . The majority of marine native-alien populations are native to the Delagoa and Natal ecoregions 23 .  www.nature.com/scientificdata www.nature.com/scientificdata/ Technical Validation Record verification. Records were collected, the taxonomy was standardised, where possible, using the GBIF taxonomic backbone, and sent to a taxonomist for a further check. Experts from different fields also confirmed the existence of the records from the online inventory. If any information recorded in the inventory was unclear and could not be verified by the authors, the record of the population was traced back to the original manuscript. All records that could not be verified were excluded from the inventory. www.nature.com/scientificdata www.nature.com/scientificdata/

Usage Notes
The goal of this inventory is to ensure that native-alien populations are correctly classified, separated from other alien populations, included in alien species inventories, and confirmed following a standardised framework. The inventory could be used as a template to assist countries to collate information on within-country invasions that follow global data standards. The collated data can be used to report on the state of biological invasions, and inform the monitoring, and management of these invasions, and is required to track progress towards biodiversity targets (e.g. the Convention on Biological Diversity's Aichi Target 9 (www.cbd.int/sp/targets/), and the post-2020 targets 27 ). As the taxonomic names in the inventory were standardised using the GBIF taxonomic backbone, the data from it can be easily integrated with the existing alien species list produced as part of South Africa's national status report on biological invasions. For further information Takalani Nelufule can be contacted through email at: takalani.nelu@gmail.com.
Limitations. We found several limitations when creating the inventory of native-alien populations in South Africa, for example, the lack of a standardised protocol for collecting information on these populations, lack of precise location data for the native range and for native-alien populations, and a lack of information on the biogeographical barrier that separates native and native-alien populations. The experts who collected the records of native-alien populations did not follow a standardised protocol or definition, and a clear description of the precise location (e.g. distance and direction from nearest town or a coordinate of the centroid of the population) of the native range and native-alien population were not available in most cases. Without information on the precise location of the native-alien population it will be difficult for researchers and managers to find these populations. Information on the biogeographical barriers that separate the native-alien population from its native range, the date of introduction, and impact, were also not available for the majority of the native-alien populations. This information is vital for the management of these biological invasions and is useful for the prediction of future trends. To overcome these limitations, and improve the understanding of, and quality of data for, native-alien populations, protocols and global data standards were implemented when compiling the inventory presented here. We suggest that these protocols and data standards be used in future to create lists of native-alien populations to allow for comparisons across taxa and regions. There are several classes for which no native-alien populations were recorded, including Arachnida, Chilopoda, Crustacea, and Annelida. It is likely, however, that for some of these taxa native-alien populations do exist in South Africa, but have not been recorded as invertebrates are generally understudied 28 . In addition, it is likely that many native-alien populations may not have been reported as the phenomenon has not been well described and its importance has not been fully appreciated 29 . For example, some native Arachnid and Annelid species are offered for sale in the pet trade in South Africa 30,31 and could have formed native-alien populations by being released irresponsibly in areas beyond their native ranges. These groups require research attention if we are to better understand the scope of the native-alien population phenomenon in South Africa.

code availability
No custom computer code or algorithms were used to process or generate the data presented in this manuscript.