Spread of invasive species into Scotland: study

Methods

Horizon Scanning

The approach followed the process previously outlined by (Roy, Peyton et al. 2020, Dawson, Peyton et al. 2023). Five expert groups were established:

1. Marine (led by Elizabeth Cottier-Cook)

2. Freshwater (led by Laurence Carvalho and Iain Gunn)

3. Terrestrial invertebrates (led by Karsten Schönrogge)

4. Terrestrial plants (led by Wayne Dawson and Jodey Peyton)

5. Terrestrial vertebrates (led by Rich Broughton)

The expert groups developed long lists of INNS to include within the horizon scanning by reviewing the list of Prevention Priority Species^[6] alongside the horizon scanning list derived for Britain in 2019^[7] and additional lists, specifically INNS established in countries in close proximity to, but currently absent from, Scotland. The expert groups worked independently to agree the long lists of INNS to consider through the horizon scanning process. However, throughout there were a number of points of clarification raised particularly on the scope. The outcomes of these discussions were shared across the expert groups to ensure consistency of approach (Annex 1: Brief Summary of Horizon Scanning Approach).

The INNS included in the lists were scored for likelihoods of a) arrival, b) establishment and c) magnitude of the potential negative impact on biodiversity and ecosystem services following the methods outlined in previous studies (Dawson, Peyton et al. 2023) and provided as a briefing note to all authors (Annex 1: Brief Summary of Horizon Scanning Approach). The expert groups also assessed the potential negative economic and human health impacts of the INNS, again following the methods outlined⁵. However, more emphasis was given, and detail provided, for INNS considered likely to have impacts on biodiversity and ecosystem services because INNS likely to impact economies and human health are considered by other sectors. Therefore, given limitations in the time available, the experts did not follow a consistent scoring process for human health and economic impacts but did list INNS likely to have such impacts. Confidence levels were considered and information on likely pathways of introduction were compiled following the Convention on Biological Diversity classification of pathways (Harrower, Scalera et al. 2018). The information on the long lists for each expert group was compiled through e-mail exchange and virtual meetings. Each of the five expert groups submitted a ranked list of INNS and associated scores for inclusion in the virtual consensus workshop (Annex 2: Workshop Agenda).

The consensus workshop was held over two days with one day in between to enable the groups to refine and review the scores following the outcomes of the discussions on the first day of the workshop. During day one of the workshop the groups highlighted the high-scoring INNS. Discussions across the groups provided an opportunity to achieve consistency in the scoring approach and informed the discussions within each group during the breakout sessions. On day two of the workshop the participants were presented with the compiled list of all INNS from across the groups. Through further discussions and review, a top 30 list was agreed.

A small group, comprising NatureScot experts and project team members, met to review the top 30 to agree on a top 10 list of INNS. This group focussed on INNS with impacts on biodiversity and ecosystem services but noted that many of the INNS have impacts across multiple categories (human health and economies alongside impacts on biodiversity and ecosystem services), for communication and awareness raising. There was agreement that grouping the flatworms together was pragmatic for communication purposes and similarly the Dreissena species were grouped together. However, it is important to note that although all these species have distinct ecological traits, their pathways of arrival are likely to be similar.

Finally, a long list of 171 INNS was derived based on a review of the scores agreed by the project team and expert groups and included:

• all INNS allocated an overall score of 48 or more (arrival score x establishment score x biodiversity impact score) or;
• biodiversity impact scores of four or five if the arrival and establishment scores were three or more or;
• all species with a biodiversity score of five if the arrival was two or more and the establishment score was greater than three and;
• species allocated five for arrival and establishment for those INNS with impact scores of only two.

Comprehensive Pathways Analysis

Potential pathways of arrival were compiled for a) established non-native species in Scotland and b) the long list of INNS compiled through the horizon scanning exercise and predicted to arrive, establish and impact biodiversity and ecosystem services in the next 10 years. The analysis included all 174 INNS identified through the horizon scanning noting that subsequently three of the INNS were considered likely to be established.

The information on pathways was assessed to rank the pathways of introduction of non-native species introduced into Scotland since 1950 and likely pathways of introduction for non-native species into Scotland in the next ten years. The analysis used the Convention on Biological Diversity (CBD) classification of pathways (CBD 2014) and was based on previous analysis conducted for Great Britain (Booy 2019).

The list of non-native species established in Scotland was based on information from the GB Non-Native Species Information Portal (GBNNSIP). The GBNNSIP contains information on the status in Scotland for many species and using any species marked as either established or established indoors (typically these are household or commodities pests, such as carpet beetles or cereal/flour pests, but with a couple of greenhouse/hot house species) - hereafter both referred to as established in Scotland.

It was noted that recent information may be missing from the GBNNSIP on recently established non-native species. Therefore, occurrence of non-native species was extracted from the NBN Atlas by searching for any species listed in the GBNNSIP as established in Great Britain but not listed as established in Scotland to identify additional species that might be established in Scotland. Using NBN Atlas data from 2000 onwards, species that had more than 20 individual occurrence records, been recorded in five or more distinct 10-km squares and in five or more different years were considered to be sufficiently recorded in time and space to assume that they were likely to be established and so were added to the species list. This resulted in an additional 32 species being added to the list bringing the total to 1096 species classed as being established in Scotland in the GBNNSIP.

The NBN Atlas was also used to provide an estimate of the year of first record for species on the established list. The year of first record was used to separate historic introductions from more recent introductions (species first recorded from 1950 onwards). The final set of non-native species added to the analyses were the 171 species on the long list from the horizon scanning part of the project.

Pathway information for all these species, the established non-native species and the 171 horizon scanning species, were collated from three main sources: the GBNNSIP, the European Alien Species Information Network (EASIN), and a dataset from a chapter on pathways within a thesis: Prioritising the Management of Invasive Non-Native Species (Booy 2019). This resulted in the compilation of comprehensive information on possible pathways for each species. Collating data from these sources required resolving nomenclature mismatches as well as pathway information which used a different pathway classification scheme, in the case of the GBNNSIP data. The pathway classification scheme used in the analysis was that proposed by the Convention on Biological Diversity (CBD) which has been relatively widely accepted and is the closest to a standardised pathway terminology in the literature (CBD 2014, Harrower, Scalera et al. 2018); see Table 1 for the categories used in the pathway analysis. Where required, pathway information was manually checked or added using online sources such as CABI’s Invasive Species Compendium factsheets and the Global Invasive Species Database. In addition to this pathway information, for the 171 horizon scanning non-native species, information from the experts in the groups was also collated. The pathway information from the expert groups was more tailored to Scotland as it tended to be restricted to pathways the expert groups considered likely for arrival of species into Scotland. The other pathway information was from known introductions to other invaded regions or potential pathways including historic pathways and/or pathways for non-native species into Scotland.

For the horizon scanning species, any potentially relevant pathways noted by experts were reviewed by the project team to ensure that only pathways that the experts considered to be applicable to Scotland were included. Additionally, for the 171 horizon scanning non-native species the biodiversity impacts and overall scores from the horizon scanning were also collated for use in the analysis.

To determine the importance of each pathway for the introduction of each non-native species to Scotland, a number of scoring metrics were calculated for each pathway. The simplest of these metrics was the total number of species associated with that pathway. For this, metric species associated with multiple pathways were counted independently for each pathway with which they are associated. This meant that while each pathway could not have a number of species higher than the number of species in the dataset, the total across all pathways was likely to be higher than and not equal to the number of species.

To correct for this, an alternative scoring approach involved scoring the values for each species by weighting the number of pathways that they are associated with, so that the value contributed by each pathway was 1/p, where p is the number of pathways. In this scoring system, a species with one pathway would still contribute a value of 1.0 to the weighted scoring to its pathway, while pathways from a species with four pathways would each contribute a score of ¼ or 0.25 to the weighted score of each pathway. Using this weighted approach, the total across all pathways was equal to the number of species that have pathway information in the dataset.

The derived horizon scanning long list of 171 INNS included additional data that offered a few alternative scoring options, specifically the biodiversity impact scores and overall horizon scanning scores for each species. Totalling the biodiversity score or overall score for species associated with each pathway, allowed the determination of pathways that were associated with the species predicted to have the largest biodiversity impacts and or with the highest overall scores for the horizon scanning.

The contributions of these pathways were compared using different subsets of the species and/or data, e.g., all established species versus only established species that have arrived since 1950, established species versus horizon scanning species to determine if the contributions of pathways had remained or were predicted to remain relatively constant. Kendall’s Rank Correlation were used to test the degree of correlation in the importance of pathways between these different subsets. It is important to note that the process of attributing pathways of introduction for INNS included within the horizon scanning list is based on expert opinion. Such predictions by experts are based on the best available evidence including from other contexts in which the INNS has prior history of biological invasion or on the known introduction pathways for closely related INNS. However, there is inevitably some uncertainty in the information captured but experts were requested to document all potential pathways of introduction and so the approach is likely to be comprehensive in encompassing the breadth of likely pathways.