The aquatic ornamental industry, whilst providing socio-economic benefits, is a known introduction pathway for non-native species, which if invasive, can cause direct impacts to native species and ecosystems and also drive disease emergence by extending the geographic range of associated parasites and pathogens and by facilitating host-switching, spillover and spill-back. Although current UK temperatures are typically below those necessary for the survival and establishment of commonly-traded tropical, and some sub-tropical, non-native ornamental species, the higher water temperatures predicted under climate-change scenarios are likely to increase the probability of survival and establishment. Our study aimed primarily to identify which of the commonly-traded non-native ornamental aquatic species (fish and invertebrates), and their pathogens and parasites, are likely to benefit in terms of survival and establishment in UK waters under predicted future climate conditions. Out of 233 ornamental species identified as traded in the UK, 24 were screened, via literature search, for potential parasites and pathogens (PPPs) due to their increased risk of survival and establishment under climate change. We found a total of 155 PPPs, the majority of which were platyhelminths, viruses and bacteria. While many of the identified PPPs were already known to occur in UK waters, PPPs currently absent from UK waters and with zoonotic potential were also identified. Results are discussed in the context of understanding potential impact, in addition to provision of evidence to inform risk assessment and mitigation approaches.

Alien species, disease emergence, horizon scanning, invasive non-native species, risk screening

The global trade in aquatic ornamental species is extensive, involving over 140 countries (Evers et al. 2019; Hood et al. 2019). Its value is estimated to be in the region of 15–30 billion US dollars annually, with a trade of 1.3 billion ornamental fishes reported (Evers et al. 2019; King 2019). The total value of live ornamental fishes imported into the UK in 2020 was £16.2 million (OATA 2020). The industry includes trade of both freshwater and marine ornamentals, however 76% of the 1244 metric tonnes of live fishes imported into the UK in 2020 were tropical freshwater fishes (OATA 2020). Though generally less well studied, invertebrates (including Mollusca and Crustacea) also represent an important group of traded aquatic ornamentals (Keller and Lodge 2007; Ng et al. 2016).

While the ornamental industry clearly provides economic and social benefits, it is a known pathway for the introduction of non-native species (NNS), pathogens and parasites, which pose a potential threat to aquatic biodiversity if they become invasive (Padilla and Williams 2004; Copp et al. 2005a; Peeler et al. 2011; Hood et al. 2019). Ornamental species are typically kept in closed systems, isolated from open waterways; but deliberate introduction into the wild, often the result of animals overbreeding or getting too large to house, or accidental introduction following escape, is known to occur (e.g. Courtenay 1999; Crossman and Cudmore 1999; Padilla and Williams 2004; Copp et al. 2005b; Duggan et al. 2006; Wood et al. 2022). Introductions of NNS can drive disease emergence by extending the geographic range of associated parasites and pathogens, and by facilitating host-switching or via spillover and spill-back (Peeler et al. 2011). Outbreaks of Koi Herpes Virus (KHV) and Spring Viraemia of Carp (SVC) in UK fisheries, which resulted in substantial mortalities in common carp Cyprinus carpio carpio L. 1758 and economic losses, have been linked to the introduction of koi carp C. carpio koi, an ornamental variety of common carp (Taylor et al. 2010, 2011, 2013).

In recognition of the threat posed by live non-native fishes, legislation that restricts the keeping of live fishes is in place in the UK. Key legislation includes ‘The Prohibition of Keeping or Release of Live Fish (Specified Species) (England) Order 2014’ (and its predecessors in 1998 and 2003) implemented under the ‘Import of Live Fish (England and Wales) Act 1980’, and ‘The Keeping and Introduction of Fish (England and River Esk Catchment Area) Regulations 2015’. These legislative instruments apply primarily (if not exclusively) to freshwater fishes, prohibiting their keeping in England without a licence, with similar powers applying in Wales, Scotland and Northern Ireland. The original 1998 Order listed only species considered to be of concern at that time, with the 2003 Order extending the list to include some additional species. These orders were perhaps the most advanced of their kind in Europe and North America (Copp et al. 2005a). The 2014 Order took a much wider approach, with the schedule listing all taxonomic Orders that contained freshwater fish species and stating that all non-native freshwater fishes required a licence to be kept with the exception of those (primarily native species) listed in the 2014 Order’s annexes. However, two general licences have been issued permitting the keeping of fishes in garden ponds and/or indoor aquaria. The first, defined here as the ‘garden pond fish list’ (UK Government 2021a) details standard ornamental pond fishes, namely koi carp, goldfish Carassius auratus L. 1758, orfe Leuciscus idus L. 1758, grass carp Ctenopharyngodon idella Cuvier & Valenciennes, 1844 and sturgeon (Acipenser spp.), permitted to be kept in aquaria or secure outdoor garden ponds. The second, defined as the ‘ornamental fish list’ (UK Government 2019), comprises mainly tropical and subtropical genera (and in some cases species), which are considered to pose a low risk of becoming established or invasive in UK waters, and permitted to be kept in indoor aquaria only. Parallel legislation exists in relation to crayfish in the form of the ‘Prohibition of Keeping of Live Fish (Crayfish) Order 1996’, which only permits the keeping of red-clawed crayfish Cherax quadricarinatus Von Martens, 1868 for ornamental purposes under a general licence and for the use of all non-native fishes in fisheries in the form of ‘The Keeping and Introduction of Fish (England and River Esk Catchment Area) Regulations 2015’. Legislation also exists to prevent the introduction of NNS in to the wild (e.g. Wildlife and Countryside Act 1981), and to limit activities associated with specific NNS (e.g. Invasive Alien Species (Enforcement and Permitting) Order 2019). Further, in relation to aquatic animal disease risk specifically, under ‘The Aquatic Animal Health (England and Wales) Regulations 2009’, live aquatic animal imports require certification.

Currently, UK temperatures are typically below those necessary for the survival and establishment of the commonly-traded tropical, and some sub-tropical, ornamental fishes (species on the ornamental fish list). However, elevated water temperatures (a 2 °C increase) forecasted by future climate change scenarios are predicted to increase the probability of survival and establishment for some existing fish species (Britton et al. 2010). Hence there may be an increased risk of pathogen and parasite introductions, transmissions and disease emergence events (Marcos-López et al. 2010). Climate modelling under four different Representative Concentration Pathway (RCP) scenarios, sensu Moss et al. (2010), indicates that global mean surface temperature may increase by between 0.4 and 2.6 °C by the mid-21^st century (Moss et al. 2010; Van Vuuren et al. 2011; Nazarenko et al. 2015). An understanding of the trade in ornamental species, and potential disease threats associated with commonly-traded species that may have an increased risk of establishment in the wild under future climate conditions, is essential to mitigate threats and protect aquatic biodiversity, and ensure the sustainability of an important industry.

The aim of our study was to identify commonly-traded non-native (NN) ornamental fishes, crustacea and molluscs at increased risk of survival and establishment in UK waters under elevated temperatures predicted by climate change forecasting. Further, applying the workflow proposed by Foster et al. (2021), organisms known, or with potential to be, pathogenic or parasitic that have been observed associating with these NN ornamental species at any point in the ornamental fish trade pathway were identified. Applications of the outcomes to inform risk assessment and mitigation measures to protect and sustain the ornamental industry in the long term are discussed.

Records of commonly imported species, such as packing lists that document details of ornamental species imported via the Heathrow Border Control Post (BCP), were not available for use during this study. Therefore, to identify the NN freshwater ornamental species most commonly-imported into UK, a proxy measure was adopted that combined outputs from three different, but complementary methodologies: expert elicitation, eBay retailer search and Google search.

Expert elicitation involved use of the list, provided by the Ornamental Aquatic Trade Association (OATA; https://ornamentalfish.org/), of ornamental species/genera considered by OATA to be the most commonly traded in the UK (by volume). Further, a short list of those NN ornamental species most likely to establish in UK, i.e. species from warm temperate or sub-tropical climatic zones, was provided by the Fish Health Inspectorate (FHI) for England and Wales.

A list of ornamental live-fish retailers was constructed from an eBay search carried out on 8 October 2020 using the term ‘live fish’. Search results were filtered for NN fishes that fell under the water type categories of ‘fresh’, ‘pond’, ‘all water types’ and ‘not specified’. Species were recorded from all listings between 9 September and 8 October 2020, inclusive. The total number of listings for each NN fish species was used as a proxy measure of trade volume. A separate eBay search using the term ‘live invertebrates’ was carried out on 13 October 2020. Search results were filtered for NN invertebrates that fell under the water type categories of ‘fresh’, ‘pond’, ‘all water types’ and ‘not specified’. Initial results indicated that significantly fewer invertebrate species were listed compared to fish species. Therefore, all NN invertebrate species listings returned by the search, with no restrictions on the date, were recorded. The number of listings per NN invertebrate species was not recorded and species listed multiple times were recorded only once.

A Google search was carried out on 20 October 2020 using the term ‘fish species for cold or unheated aquaria”, and this provided information on popular ornamental fish species likely to be traded in the UK. Although returning primarily temperate species, it also included tropical fish species with wide temperature tolerances, which therefore do not require heated aquaria, e.g. Endler’s livebearer Poecilia wingei Poeser, Kempkes & Isbrücker, 2005, and zebra danio Danio rerio Hamilton, 1822 (López-Olmeda and Sánchez-Vázquez 2011). In the search for cold or unheated aquaria species, the most popular NN fish species listed in the first 20 websites or blogs (See Suppl. material 1: List S1) were used to represent commonly-traded species. All species mentioned were recorded only once.

A master list of species’ common and scientific names was developed. If the species scientific name was absent in the eBay listing or on the website/blog, then it was searched for (using the common name) on FishBase (www.fishbase.se/search.php) for all fish species or via a google search for the invertebrate species. Where fishes and invertebrates were not identified to species level, the entry was removed from the master list. Species recorded via any of the methodologies were collated into the single master list (See Suppl. material 1: Table S1).

The master list was refined by removal of species, based on the following criteria: i) the NN species is present on the ‘garden pond fish list’ or is not present on the ‘ornamental fish’ list under The Prohibition of Keeping or Release of Live Fish (Specified Species) (England) Order 2014; and ii) the NN species is recorded as present within UK waters on the Global Biodiversity Information Facility (GBIF; www.gbif.org). Although climate change may increase the risk of some of these NNS, either increasing their current range or establishing new populations as a result of further introductions, the associated pathogen risk was considered to exist already because the species is already present.

Trade in live aquatic ornamental species is vast, benefitting from globalisation and improved transport in recent decades. Over 140 countries are involved in the international trade of more than 1500 fish and 300 aquatic invertebrate species (Weir et al. 2012; Hood et al. 2019). Our study provides an assessment of freshwater ornamental trade in the UK in which commonly traded species and their likelihood of establishment in UK waters under current and future climate conditions, with their potential pathogens and parasites also identified..

These data on commonly traded species are a snapshot in time, which potentially limits accuracy and prevents the assessment of seasonal and annual variations. That said, the six species identified were listed amongst the 30 species reported to predominate the global trade in ornamental freshwater organisms in a relatively recent review: Endler’s livebearer, goldfish, zebra danio, neon tetra Paracheirodon innesi Myers, 1936, angel fish Pterophyllum scalare Schultze, 1823, and discus Symphysodon aequifasciatus Pellegrin, 1904 (Dey 2016).

Access to robust ornamental trade data, in particular with respect to species traded and import origin, is fundamental to fill knowledge gaps and inform risk screenings and the risk analysis process (Copp et al. 2016; Chan et al. 2019). A comprehensive understanding of spatial and temporal trade patterns to species level will increase capacity to identify high risk links, facilitating targeted and cost-effective surveillance. Species-level information will also support analyses from a conservation point of view, particularly for marine species that are wild sourced, and ensure there is increased transparency in the source, quantity and sustainability of trade for each species (Andersson et al. 2021). In the UK, aquatic species imports from third party countries are electronically recorded on the ‘Import of Animals Food and Feed System’ (IPAFFS). Historically, non-susceptible aquatic species imported from the EU for ornamental use were not recorded on any system. As of January 2021, all aquatic imports must be recorded on IPAFFS, but records on this database are categorised by international commodity codes (www.uktradeinfo.com/find-commodity-data/help-with-classifying-goods/). However, the purpose of these codes is to enable the application of appropriate tariffs to imported goods and they do not provide sufficient resolution to identify consignments to species level, so cannot be used to inform disease susceptibility or invasive potential. Import data at the species-level for the UK currently exist in paper format only (on the invoices that accompany all other relevant import certification), though there are periods for which species-level electronic data have been available for freshwater fishes, including ornamental varieties, imported to England between 2000 and 2004 inclusive (see Copp et al. 2007). Limitations with respect to access to detailed live ornamental import data are not unique to the UK (Rhyne et al. 2012, 2017; Leal et al. 2016; Pinnegar and Murray 2019; Biondo and Burki 2020). Creation of an import data App, or extension of an existing, established App, is an opportunity to capture detailed import data, integrating species trade information with other crucial information such as invasiveness potential, associated disease threats and conservation status. Such a system would enhance capacity for real-time monitoring and analysis, at the point of exporter application for import, allowing trade in high risk species to be tracked and incidences of illicit trade, such as the import of prohibited species, to be detected in a timely manner (Rhyne et al. 2017).

Legislative instruments restrict the keeping of many temperate species but do allow the keeping of numerous commonly traded tropical and sub-tropical species. The application of SDMs indicated that, while establishment of commonly traded species if released into the wild is unlikely in the UK under current conditions, predicted temperature increases associated with climate change may increase risks of survival and establishment. The mean increase in temperature suitability of 2.4% and 1.8% for fish and invertebrates, respectively, by 2050 under RCP 4.5 demonstrated in our study may seem a small increase in ‘risk’, but RCP 4.5 represents a moderate climate-change scenario, and temperature increases may be greater than this scenario predicts. Although a broad scale indication of the change in suitability under climate change is provided, careful interpretation of SDM outputs may be required. For instance, the red shiner Cyprinella lutrensis Baird & Girard, 1853, is widespread across the USA and has been identified, using the Fish Invasiveness Screening Kit, as posing a medium risk of being invasive in England and Wales (Copp et al. 2009). However, the red shiner had a relatively low suitability score (0.10) under current day conditions in our study (Table 2). The Chinese muddy loach Misgurnus mizolepis Guenther, 1888 had a suitability score of 0.26 under current day conditions, yet there has been only one record of a reproducing population in the UK (in southern England), which was subsequently eradicated (Zięba et al. 2010). Although outside the scope of our study, there will be value in assessing species suitability at a finer spatial resolution, for example accounting for differences in conditions across the entirety of the UK (Thrush and Peeler 2013), for example, the notably warmer conditions in the South compared to the North UK, affecting the risks of survival and establishment of sub-tropical and tropical species (see Suppl. material 1: Figs S1, S2). Further, extension of the SDMs to incorporate environmental factors such as rainfall, habitat type or elevation (Logez et al. 2012), species' life-history traits such as size or fecundity (Copp and Fox 2007; Liu and Olden 2017), and consideration of the likelihood of release of each species (i.e. potential propagule pressure) will be of value. For example, inclusion of elevation will account for species with native distributions at higher altitudes within the tropical zone that may be more adapted to temperature conditions more similar to the temperate zone species due to Rapoport’s latitudinal rule to altitude (Stevens 1992).

In total, 155 PPPs were found to be associated with the screened ornamental fishes and invertebrates. Despite following a standardised approach for each host species, the number of PPPs identified in the literature may be skewed by the research effort applied to a species and affected by the use of different accepted names or synonyms. One of the key drivers of impacts associated with NN aquatic species (Peeler et al. 2011) are PPPs, with disease emergence events resulting from NNS introductions being well documented (Taraschewski 2006; Peeler et al. 2011; Lymbery et al. 2014). Disease emergence can be driven solely by a switch in geographical range, because a new environment may favour increased PPP virulence, or by host switching and pathogen/parasite spillover (Peeler et al. 2011; Foster et al. 2021). Thermal tolerances of a PPP may also determine the likelihood and impact of disease emergence resulting from co-transportation, particularly for PPPs that can survive outside a host or have free-living stages (Barber et al. 2016).

Even if the long-term survival of an ornamental species is not supported by future UK temperatures, the host species may persist long enough to transmit a PPP to a native susceptible host or introduce a free-living stage which can survive. Temperature may also determine the likelihood of PPPs causing disease and morbidity in infected hosts. For example, KHV is thought to only cause clinical signs and mortality between 16 ˚C and 25 ˚C (OIE 2019), whereas outbreaks of VHS rarely occur above 15 ˚C (Baillon et al. 2020). While beyond the scope of this study there will be value in building on the present study by examining the environmental tolerances of the PPPs and the likely impact of climate change. Indeed, PPPs introduced via the ornamental pathway may cause wider impact, for example causing mortality and yield loss in aquaculture systems and affecting human health, although strong biosecurity and health and safety precautions can mitigate against such risks.

The PPPs that cause mortalities or clinical expression of disease in the traded host species are more likely to be detected via visual inspection or quarantine at border control posts or other stages in the ornamental trade pathway (Table 3). However, PPPs that live symbiotically with a host, or those PPPs that have sub-clinical or latent infection stages, provide a greater challenge to detection (Gomez et al. 2006; Becker et al. 2014). Traditionally, the testing of host species for PPP presence often requires destructive sampling, which limits the number of specimens and ultimately reduces the probability of PPP detection. However, new methodologies that incorporate molecular techniques at border control posts, such as environmental DNA, may present a good non-lethal option (Trujillo-González et al. 2019b; Brunner 2020) for improved PPP detection. Measures such as heat ramping (gradually increasing the water temperature over a minimum period) can be used during quarantine to detect latent infections (Eide et al. 2011); however, this may not appropriate for all PPPs, when surveillance aims to target multiple PPPs and where there is a need to adhere to animal welfare laws and guidance.

Though the remit of our study was to undertake a high-level screening to identify all PPPs associated with commonly traded ornamental species, rather than novel threats per se, we note that while some of the identified PPPs are known in the UK, others are not. Some PPPs are already known within the ornamental fish trade industry, and do not cause widespread impact or can be successfully treated to minimise impact. However, the abundance and diversity of PPPs increases potential for future disease outbreaks under changing environmental conditions. Even where a PPP has not yet been implicated in any mortality events, the changing climate and alterations to host communities (e.g. due to species introductions) may provide the perfect storm for disease emergence into the future. Next steps should aim to assess the risk associated with each PPP, focussing on the interplay between the PPP, all potential hosts and changing environmental conditions.

In conclusion, the ornamental fish trade is largely free from serious and untreatable diseases. However, through screening of a small subset of ornamental freshwater species, our study highlights the abundance and diversity of PPPs present in ornamental species commonly traded in the UK. An understanding of hazards associated with PPPs, in particular under changing ecological and environmental conditions, is crucial to determine and communicate risks and enhance risk awareness amongst stakeholders and the general public, thereby enabling mitigation through management actions (Britton et al. 2011) to ensure a safe and sustainable ornamental aquatics industry into the future (Copp et al. 2016).

This work was funded by Department for Environment, Food & Rural Affairs, Project FB002. We would also like to express our gratitude to Alasdair Scott and anonymous reviewers for their helpful advice and comments on the manuscript.