Watch-lists are compilations of alien species that are absent from a region of interest but likely to become invasive if introduced. They have appeared under various names in the literature, including black list-warning lists, warn lists, and alert lists, and they serve as practical instruments for decision-making, resource prioritization, and the development of preventive strategies and contingency plans. They can direct surveillance toward high-risk species both at borders, to prevent illegal, intentional, and accidental introductions, and post-border, so that newly arrived alien species can be managed quickly.

Their value is illustrated by a ranked watch-list published for Great Britain in 2014: within two years, seven of the top ten ranked species had been recorded in that region. Such lists are also central to international commitments. The 196 countries that are party to the Convention on Biological Diversity must meet Target 6 of the Kunming-Montreal Global Biodiversity Framework, which requires them to prevent the introduction and establishment of priority invasive alien species and to reduce rates of introduction by at least 50 percent by 2030.

Yet producing a watch-list has traditionally been time-consuming, data-intensive, and expensive, because it requires curating information on invasion and impact history, distribution, introduction pathways, and environmental suitability for large numbers of species. This burden has made watch-lists difficult for resource-poor countries to develop. The problem has intensified over the past decade as accessible data on alien species have exploded. When a watch-list was developed for South Africa in 2014, it drew on the Global Invasive Species Database, which contained 884 species, and on roughly 500 million occurrence records from the Global Biodiversity Information Facility (GBIF). Today, the Global Register of Introduced and Invasive Species (GRIIS) lists around 24,000 alien species, 27 times the number available in 2014, and GBIF now holds over 3 billion occurrence records, a six-fold increase. Community science platforms such as iNaturalist have been increasingly incorporated into GBIF, adding further volume. A method that was simple and rapid a decade ago now requires computational tooling to manage the data effectively and cost-efficiently.

WatchListR, which consists of Linux and R scripts, follows steps similar to the 2014 South African approach but automates them at scale. It takes as input a list of invasive terrestrial and freshwater species from GRIIS, processed one kingdom at a time, and a list of species already present in the target country; an optional input is a list of high-risk countries, such as regular trading partners. The software first resolves taxonomic names against the GBIF Backbone Taxonomy, which assigns unique identifiers known as GBIF Species Keys. Because all taxonomic synonyms share the same Species Key under a single accepted scientific name, this step ensures that records submitted under different names are correctly tallied to the same species. Taxonomic synonyms and homonyms are a major caveat when searching public databases. The developers illustrate the issue with the cactus Opuntia maxima: a GBIF search currently returns three separate listings with different authorities, two of which are synonyms of Opuntia ficus-indica and share its Species Key, while the third, Opuntia maxima Mill., is a distinct accepted taxon and a homonym with its own key. WatchListR searches the taxonomic backbone for every input name, groups synonyms under single Species Keys, distinguishes homonyms by their unique keys, and records final taxon names according to the currently accepted name. The same synonym-resolution process is applied to the list of species already present in the country, which can include endemics, agricultural species, and invasive species that have already arrived.

Species present in the country are then removed from the invasive species list, leaving a list of invasive species that are absent or not yet documented. Next, the software downloads all available GBIF records for the absent invasive species using their Species Keys, excluding records with geospatial issues or missing coordinates and records originating from zoos, botanical gardens, or fossil deposits. Each record’s coordinates are assigned to a Köppen-Geiger climate zone, allowing broad-scale climate matching between the target country’s zones and those in which each invasive species is known to occur. To manage the enormous volume of distribution data, WatchListR uses a data-partitioning approach in which the dataset is divided into smaller subsets, processed in parallel, and merged at the end. For each species, the software records the total number of occurrence records, the cumulative number and proportion of records within all Köppen-Geiger zones present in the target country, the proportion in each individual zone, and the proportion found in the country itself. An optional final step assesses propagule pressure by identifying countries with high climatic similarity, high trade volumes contributing 5 percent or more of the value of goods imported, and 20 or more known invasive species, using trade and tourism data from FAOSTAT and UN Comtrade.

The full analysis for an invasive species list of approximately 4,000 taxa takes under two hours; the Mauritius run was performed on the Centre for High Performance Computing Lengau Cluster, with each job using 23 CPU cores. An accompanying R Shiny dashboard allows users to upload their final watch-list through a graphical interface, produce summary graphics, and filter the data to their specifications. The pipeline and dashboard are freely available on GitHub with step-by-step deployment instructions.

In the Mauritius case study, conducted as part of a United Nations Development Programme contract with NCC Environmental Services to build the island nation’s biosecurity capacity, the team considered terrestrial and freshwater plants and animals only, excluding marine species, which Köppen-Geiger zones cannot describe, and kingdoms such as fungi, bacteria, and viruses. For plants, 11,114 GRIIS records were consolidated and filtered, removing 740 indigenous and introduced species already present, leaving 2,481 invasive plant species; after synonym searches the input list grew to 3,229 names, yielding 162,603,988 GBIF distribution records for climate matching against the three climate zones present in Mauritius. A total of 2,902 plant taxa occur in those zones, all of which could pose an invasion threat but which together form too large a list for meaningful biosecurity planning. Of these, 1,868 had more than 1,000 distribution records, suggesting they are widely invasive with high propagule pressure, and 328 of those had more than 40 percent of their records within one of Mauritius’s three climate zones. The authors recommend that all 2,902 taxa be placed on the watch-list, with the 328 most widely recorded taxa as the focus of biosecurity efforts, since they are both likely to be introduced and likely to establish.

For animals, 2,242 GRIIS records were processed down to 722 invasive species after removing 23 species already present, and 705 names were used to query GBIF, yielding 137,360,927 distribution records. Climate matching identified 680 animal taxa in Mauritius’s climate zones, which was reduced to 188 priority species using the same thresholds of more than 1,000 records and greater than 40 percent climate match. The analysis also assessed which priority species occur in high-risk trading and tourism partners: India, China, South Africa, Malaysia, the United Arab Emirates, France, Indonesia, and Kenya. Between 45 and 80 of the priority plant species occur in these countries, with China hosting 79, India 68, and Indonesia and South Africa 63 each, suggesting that surveillance of cargo and transport vessels from ports of origin in these countries would be important. Among the animals, 27, 20, 19, and 18 of the priority species occur in China, India, Malaysia, and Indonesia, respectively.

Six plant species, including Hellenia speciosa, Muntingia calabura, Oldenlandia corymbosa, Pteris vittata, Sieruela rutidosperma, and Sphagneticola trilobata, occur in three or four of these countries, as do the Asian common toad (Duttaphrynus melanostictus) and the Asian honey bee (Apis cerana), marking them as species whose transport pathways warrant particular attention. The authors note that GRIIS data collection is skewed toward better-resourced countries and that community science records carry biases, but they argue that a watch-list identifying species that are invasive elsewhere, present in climatically similar regions, and likely to arrive along known trade routes is invaluable for planning biosecurity operations, especially where prevention can reduce the spread of invasive species relatively cheaply.

Subject of Research: Automated software for developing watch-lists of potentially invasive alien species to inform biosecurity decisions

Article Title:  WatchListR : a tool for developing watch-lists of invasive species to inform biosecurity decision-making

Article References: Ivey, P. J., Faulkner, K. T., Miller, J., & van Steenderen, C. J. M. (2025). WatchListR: a tool for developing watch-lists of invasive species to inform biosecurity decision-making. NeoBiota, 104, 163-177. https://doi.org/10.3897/neobiota.104.163164

Image Credits: AI Generated

DOI: 10.3897/neobiota.104.163164

Keywords: invasive species, biosecurity, watch-list, WatchListR, GBIF, GRIIS, climate matching, Köppen-Geiger, Mauritius, propagule pressure, biodiversity, Kunming-Montreal Global Biodiversity Framework