Country Compendium of the Global Register of Introduced and Invasive Species

GRIIS and the Country Compendium

The Global Register of Introduced and Invasive Species (GRIIS) arose following recognition of the need for a product of this nature in discussions on implementation of the Convention on Biological Diversity (CBD). In 2011, a joint work programme to strengthen information services on invasive alien species as a contribution towards Aichi Biodiversity Target 9 was developed¹⁹. The Global Invasive Alien Species Information Partnership (GIASI Partnership) was then established to assist Parties to the CBD, and others, to implement Article 8(h) and Target 9 of the Aichi Biodiversity Targets. The Conference of Parties (COP-11) welcomed the development of the GIASI Partnership and requested the Executive Secretary to facilitate its implementation (paragraph 22 of decision XI/28). In 2013, the development of GRIIS was identified as a key priority to be led by the IUCN ISSG and Partners built on a prototype initiated almost a decade earlier (Item 4, Report of the Global Invasive Alien Species Information Partnership, Steering Committee, 1st meeting Montreal, 15 October 2013)²⁰.

GRIIS is a database of discrete checklists of alien species that are present in specified geographic units (including not only countries, but also as yet unpublished checklists of islands, offshore territories, and protected areas) (Fig. 1). The GRIIS Country Compendium is a collation and key product that derives and is updatable from the working GRIIS Research Database that underpins this and other GRIIS products (Fig. 1). Individual checklists are published to GBIF through an installation of the Integrated Publishing Toolkit²¹ (IPT) and hosted by the GBIF Secretariat. Exceptions include the Belgium (hosted by the Research Institute for Nature and Forest) and U.S.A checklists (hosted by the United States Geological Survey). Data are published as Darwin Core (dwc namespace) Archive files and the terms and structure follow that standard exchange format²².

The GRIIS Country Compendium is an aggregation of 196 GRIIS country checklists of which 82% have been verified by Country Editors (see¹³), along with revised and additional fields that enable global level analysis and country and taxon comparisons (Tables 2, 3). Checklists for the 196 countries were combined into a single file (Table 3). A field was added to indicate which country the checklist belonged to, and the ISO 3116-1 Alpha-2 and Alpha-3 country codes are included to facilitate dataset integration (see ‘Usage notes’) (Table 2). A field was also added to indicate the verification status of each checklist (Table 2). The ID field was renamed (originally ‘taxonID’ and now ‘recordID’), as the data now represent a country-level occurrence dataset containing multiple records per species, rather than checklist-type data that contains one record per species. In total, the data now include 18 fields as described in Table 2, encompassing taxonomic, location, habitat, occurrence, introduced and invasive alien status (see also Table 1). This publication represents a versioned, citable snapshot of the Compendium (Fig. 1) that is ready for analysis and integration with other data sources (e.g. workflow²³ and ‘Example applications of the Compendium’ outlined further below).

Population of data fields in GRIIS

The methods by which GRIIS is populated were described in 2018¹³ and are summarised in brief here. A systematic decision-making process is used for each geographic unit by species record to designate non-native origin and evidence of impact (see Fig. 2 in Pagad et al.¹³). Comprehensive searches are undertaken for each country. Records are included from the earliest documented to the most recent accessed record prior to the date of the latest published checklist version. Information sources include peer-reviewed scientific publications, national checklists and databases, reports containing results of surveys of alien and invasive alien species, general reports (including unpublished government reports), and datasets held by researchers and practitioners¹³. A log of the changes to each checklist is available on the GBIF IPT²⁴, with the changes to the Belgium checklist available at the INBO IPT²⁵. The most up to date version of each checklist is thus available via GBIF.org, as is a list of all GRIIS checklists at GBIF.org²⁴.

Introduced species of all taxonomic groups are considered for inclusion in GRIIS. Habitats include terrestrial, freshwater, brackish, marine and also host (i.e. for species that are not free-living) (Table 2, Pagad et al.¹³). The habitat information in GRIIS (Table 2) is sourced from taxon and region-specific databases such as WoRMS (World Register of Marine Species), FishBase, Pacific Island Ecosystems at Risk, and the USDA Plants Database. Typically, GRIIS records are at the species level, but in some cases, other ranks are more appropriate including infraspecies (including forms, varieties and subspecies). A separate field is provided for hybrids (Table 2). Where species are present and both native to parts of a country and alien in other parts of the country, their introduction status (dwc:establishmentMeans) is included as Native|Alien (Tables 1, 2)²⁶. If there is limited knowledge about the Origin of the species, its introduction status (dwc:establishmentMeans) is included as Cryptogenic|Uncertain (Tables 1, 2).

Two types of evidence are considered to assign a species by country record as invasive (Table 1, see also Pagad et al.¹³): (i) when any authoritative source (e.g. from the primary literature or unpublished reports from country/species experts), describe an environmental impact, and/or (ii) when any source determines the species to be widespread, spreading rapidly or present in high abundance (based on the assumption that cover, abundance, high rates of population growth or spread are positively correlated with impact)^27,28. Each record is assigned either invasive or null in the isInvasive field to reflect the presence of evidence of impact, or absence of evidence of impact (note, not ‘evidence of absence of impact’), for that species by country record (Table 2). In the future this information may be supplemented with impact scores^29,30,31. Finally, a draft checklist is sent to Country Editors for validation and revision (see Technical Validation).

Taxonomic harmonization and normalization

The use of different synonyms across countries to refer to the same taxonomic concept is frequent³². The species in each Country Checklist were thus harmonised against the GBIF Backbone Taxonomy³³. The names in each checklist were matched using a custom script that integrates with the GBIF API³⁴, and the accepted name, taxon rank, status and higher taxonomy (Table 2) were obtained at this stage. Spelling and other errors in assigning species authorship were corrected where appropriate.

To validate the taxonomic harmonisation, every name variant present in the GRIIS Country Compendium was checked against the GBIF Backbone Taxonomy using the API³³. A unique list of names (i.e. acceptedName Usage) was thus produced and the source name retained as ‘scientificName’ (that can differ across countries) (Table 2). Over 95% of names across all kingdoms matched exactly at 98% or greater confidence (Table 4). All names that were below 98% confidence or had a match type other than ‘Exact’ were checked and modified if appropriate to do so. Of the non-matches (n = 253, those with a match type of ‘None’), most were formulaic hybrid names of plants and animals (~62%), which are not officially supported by GBIF³⁵. The remaining non-matches were names of mostly plants (17%), but also animals (8%), viruses (8%) and chromists (3%).

Data summary

There are currently ~23 700 species represented by 101 000 taxon-country combination records, across 196 countries in the GRIIS Country Compendium. All raw numbers are provided to the nearest order of magnitude to reflect the taxonomic uncertainty and dynamic nature of GRIIS (see ‘Known data gaps and uncertainties’). The vast majority of records are at the species level (97.6%), with the remaining present as subspecies (1.7%), varieties (0.6%), genera (0.1%) and forms (<0.1%). In addition, approximately 0.5% of records are either named or hybrid formulae. For the purpose of providing an overview of the content of the Compendium, we counted species using the accepted species name in the ‘species’ field (for this purpose not differentiating by taxonRank) (Table 2).

Most country checklists include between 100–1000 species; 36 checklists have <100 species and 23 checklists with >1000 species (minimum = 5; maximum values = 5649). On average the number of species per country is 516 (721 s.d.), with the fewest records for San Marino, South Sudan and Monaco (<12) and most for the U.S.A, France and Australia (>2900). The taxonomic and habitat coverage in the Compendium is broad, with 55 phyla, 149 classes, 630 orders and 2305 families represented across terrestrial, marine, freshwater and brackish habitats (Fig. 2). Those currently represented include Animalia, Bacteria, Chromista, Fungi, Plantae, Protista (Protozoa), and Viruses. Over 98% of records have establishmentMeans as simply alien, whereas 1% of records are Cryptogenic|Uncertain and 1% Native|Alien (Tables 1, 3).

As expected for invasive alien species, the Compendium is dominated by plant and invertebrate species and records (Fig. 2a,c). Overall, 16% (n = 16 221) of all records are associated with impact (isInvasive, Table 2), and 22% of species (n = 5199) are associated with evidence of impact in at least one country. Terrestrial habitats have most species and records associated with evidence of impact, followed by freshwater and marine species and records (Fig. 2b,d). Approximately 15% of plant (n = 11,436) and 16% of invertebrate (n = 2300) species by country records are associated with evidence of impact, whereas mammal (42%, n = 644) and amphibian (34%, n = 84) records have higher percentages of evidence of impact at the country level. Over half of the species in the Compendium are recorded from only a single country, with 10.5% present in more than 10 countries (Fig. 3a). As expected, the numbers of ‘isInvasive’ records per species are lower than the total number of records per species, although species with many country occurrence records do tend to have many ‘isInvasive’ status annotations (Fig. 3b).

Example applications of the Compendium

The GRIIS Country Compendium enables the first multitaxon assessment of the most widely distributed subset of introduced species that harm biodiversity and ecosystems. These ‘worst’ invasive alien species (in terms of realised impact across most countries), include wild sage, black rat, common carp, red-eared slider turtle, American bullfrog, common myna, sea walnut and giant African snail (Table 5). To demonstrate key envisaged uses of the Compendium, it has been incorporated in Map of Life enabling: (1) multitaxon mapping of richness patterns in invasive alien species (e.g. Fig. 4), and (2) the combination, at the species level, of alien information along with point observations, expert defined ranges and other types of data describing species distributions worldwide (Fig. 5). This facilitates model-based integration of the Compendium data with environmental data to assess changes and trends in species distributions. The Compendium also enables the consideration of gaps and inferred absences of species at the country scale, useful for the purpose of surveillance, updating country checklists, informing risk assessments and for research on the drivers of invasive alien species spread and establishment^23,36. For example, occurrence records for the red eared slider turtle in East Africa, the Middle East and Australia (Fig. 5) suggest that GRIIS country checklists for the countries concerned may need to be updated, triggering an update to the Compendium. A growing number of publications now cite use of multiple GRIIS checklists, accessed through GBIF, by use of Digital Object Identifiers³⁷. The Compendium will now streamline and facilitate such multi-checklist applications.

The Compendium will further facilitate international research avoiding the need for repeated multi-country data integration by individual researchers. In support of surveillance and information currency, it provides a mechanism for generating automated alerts of risks of new incursions by analysing presence in neighbouring countries, especially through development of further workflows reporting new occurrence data published through GBIF. The Compendium may be used in decision support, contributing to invasion pathway and impact analyses and horizon-scanning exercises, as well as presence-absence inference in risk analysis³⁶. In addition to the GRIIS country checklists available to national governments through the CBD Clearing House Mechanism (CHM) and national CHMs, the GRIIS Country Compendium provides a foundation for long-term reporting on progress towards meeting invasive alien species-relevant goals and targets under the Convention on Biological Diversity, the Sustainable Development Goals and regional or thematic goals and targets. For example, the habitat data included in the Compendium could be disaggregated and used by relevant conventions, such as The Ramsar Convention on Wetlands and the Convention on Migratory Species. The Country Compendium data were also used in the development of the Global Biodiversity Outlook GBO5 and are being used in the IPBES Invasive Alien Species Assessment currently underway.

The Compendium provides the baseline used in the development of the invasive alien species headline indicator that measures trends in introduction events. It provides the key information source for Essential Biodiversity Variable-based indicators of invasion, including rate of spread, impact and status information^6,14. This includes an indicator for reporting against the draft post-2020 Global Biodiversity Framework (CBD/WG2020/3/3) Target 5 and associated Headline Indicator – Rate of Invasive Alien Species Spread (CBD/WG2020/3/3/ADD1), enabling the calculation of baseline values at different scales, including the country level. Aggregated data from GRIIS, now published as the GRIIS Country Compendium, have thus already supported multiple research and key policy activities.

Known data gaps and uncertainties

One of the principal objectives of GRIIS is to provide a geographically and taxonomically comparable source of information on invasive alien species. The intent is to collate the taxonomy and use the same approach to designating every taxon as alien, Native|Alien, Cryptogenic, invasive or not (Table 1). These factors have previously been the cause of major uncertainty in collations of information on alien species. Nonetheless, a number of known gaps and uncertainties remain in the Compendium.

Taxonomy

Sources of taxonomic uncertainty include multiple accepted names (acceptedNameUsage) for the same species concept – including author variations, lexical variants and potentially different taxonomic interpretation from the source. For example, the bay barnacle is associated with both Balanus improvisus and Amphibalanus improvisus. There are also multiple scientificName variants for individual ‘acceptedNameUsage’ cases (Table 2). No attempt has been made to edit or match species authorities or reference literature defining the taxon concepts. The species-level prevalence of lexical variants, synonyms and infraspecific names used across countries in the scientificName field is 20%, i.e. ~ 4740 species with more than one scientific name used, ~1520 species with more than 2 names, 14 species with more than 10 names and a maximum of fifteen scientificName variants for a single species. Overall, 4.5% of unique species names are associated with an infraspecific name. The average percentage of such cases across species groups is ~14%, with the lowest percentage for birds (~9%) and the highest for plants (~ 27%) (microorganisms not considered).

The taxonomic information in the Compendium thus includes sources of uncertainty in common with all interspecific inventories of taxonomic information³⁸, including different usage of synonyms across countries that could result in the inclusion of the same species as multiple taxa³⁹. A shortage of editors specializing in particular taxonomic groups within countries is a recurring problem, as is the time necessary for editors to conduct and return reviews of draft checklists. Nonetheless, the most appropriate approach for harmonising taxonomy is determined by the intended use of the dataset. The Compendium provides both harmonised and un-harmonised taxonomic data in different fields to enable users to select the appropriate data for their use case. The ‘acceptedNameUsage’ cases provided are not necessarily the most taxonomically up to date usage, or necessarily the preferred usage of a particular country, but this field is included to enable comparisons across countries. For instance, Map of Life uses harmonized taxonomic data excluding all infraspecies level records (i.e., subspecies, hybrids, varieties) to integrate GRIIS data with other types of data describing the distribution of species (e.g., species occurrence data mediated through GBIF, or expert range maps) (Fig. 5).

Completeness

Invasive alien species checklists are inherently dynamic, and indeed the value they embody is in the changes in the inclusion and exclusion of species over time. Alien and invasive alien species databases and lists are by nature dynamic for multiple reasons, including the population and range dynamics of alien species, the time lag that often occurs between establishment and spread of alien species outside of their native ranges, and time lags between such events and the quantification and documentation of their impacts⁴⁰. Other biases include those well known for alien species, including geographic and habitat bias (such as marine invasions being generally less well known than terrestial ones⁴¹ and taxonomic gaps, language barriers to information access and collation, and inconsistent use of invasion terminology across information sources^40,42). In addition, potential sources of uncertainty and error include (i) misidentification of particular species in particular countries, (ii) species listed as being present that have subsequently become extinct or been eradicated, (iii) species listed as present based on in-country occurrence records but where the species has not established, and (iv) cases where particular taxonomic groups are undergoing rapid or regular taxonomic revision.

GRIIS aims to be a list, as complete as possible, of those alien species both confirmed to be present and also associated with evidence of impact in one or more countries, i.e. of ‘true presences’ of introduced species with evidence of impact. It is less complete as a list of true presences of alien species for which no evidence of negative impacts on biodiversity and ecosystems exists. In other words its focus is on the subset of alien species most likely to be a policy and management priority⁷. GRIIS should be regarded and appreciated as an evidence-based information repository designed for the purpose of improving and monitoring change in alien and invasive status and properties. Acknowledging the uncertainties and the dynamic properties of these data (in some cases inherent and in other cases by design), by providing a transparent, traceable process of data collation, the GRIIS Compendium provides a platform for delivering, refining and updating information on invasive alien species needed for global reporting, a baseline for ongoing updates, and a platform for sustainable delivery of this information (Fig. 1).

Source: Ecology - nature.com