Invasive Non-Native Species (INNS) can co-transport externally and internally other organisms including viruses, bacteria and other eukaryotes (including metazoan parasites), collectively referred to as the symbiome. These symbiotic organisms include pathogens, a small minority of which are subject to surveillance and regulatory control, but most of which are currently unscrutinized and/or unknown. These putatively pathogenetic symbionts can potentially pose diverse risks to other species, with implications for increased epidemiological risk to agriculture and aquaculture, wildlife/ecosystems, and human health (zoonotic diseases). The risks and impacts arising from co-transported known pathogens and other symbionts of unknown pathogenic virulence, remain largely unexplored, unlegislated, and difficult to identify and quantify. Here, we propose a workflow using PubMed and Google Scholar to systematically search existing literature to determine any known and potential pathogens of aquatic INNS. This workflow acts as a prerequisite for assessing the nature and risk posed by co-transported pathogens of INNS; of which a better understanding is necessary to inform policy and INNS risk assessments. Addressing this evidence gap will be instrumental to devise an appropriate set of statutory responsibilities with respect to these symbionts, and to underpin new and more effective legislative processes relating to the disease screening and risk assessment of INNS.

Alien species, invasive pathogen, opportunistic pathogen, parasite, symbiont

Invasive Non-Native Species (INNS) are “species whose introduction by human activity outside their natural past or present distribution threatens biodiversity”, as defined by the Convention on Biological Diversity (CBD 2010), and are one of the greatest global threats to biodiversity (IPBES global assessment 2019). New introductions of INNS are increasing every year, with no indication that introduction events are decreasing in frequency (Seebens et al. 2020). It is increasingly recognized that invasions are not the product of single species introduction events but can be considered as holobionts (Skillings 2016): i.e., units of biological organisation including the host and all its symbionts (external and internal), including pathogenic species. Therefore, organisms such as viruses, bacteria, fungi, protists and other (micro-)eukaryotic parasites and pathogens may be introduced to new regions along with their invasive non-native host and can be important factors in the invasion process (Peeler et al. 2011; Roy et al. 2017). A broad basis for referring to these organisms as ‘pathogens’ is required: they may not be recognized as pathogens, or cause disease in the INNS host transporting them, but may impact on other related (or unrelated) hosts in their new range. Further, pathogenesis can be very context dependent, as described by the symbiotic continuum concept (Bass and del Campo 2020). Therefore, a biologically informed approach to horizon scanning for such ‘pathogens’ is necessary, to enable effective identification of potentially new and emerging diseases. For the purposes of this paper, to avoid repetition of “parasites/pathogens” to refer to symbionts that take nutritional advantage of their hosts potentially causing disease, we henceforth use “pathogen” as a catch-all term.

In the field of invasion biology, the translocation of non-native pathogens (emerging infectious diseases in public and wildlife health) are increasingly being researched as important environmental driving factors (Ogden et al. 2019; Thakur et al. 2019); however, this is not currently reflected in national and international policy and legislation. For example, co-introduced pathogens are explicitly excluded from the EU Invasive Alien Species Regulation 1143/2014. Instead, potentially invasive co-introduced pathogens are considered as potential impacts of INNS establishment. Although pathogens are currently excluded from much of the legislation surrounding invasive species, there are a few examples where they are included, for example the Ballast Water Management Convention (Hess-Erga et al. 2019).

Understanding and predicting the impacts of INNS is essential to inform risk analysis, for example, via horizon scanning, risk assessments and impact assessments, which underpin many components of INNS policy and management. However, pathogens associated with most (potential) INNS are very poorly known (Roy et al. 2017; Pagenkopp Lohan et al. 2020), except in the few cases where they are recognized under animal disease or human health legislation and are monitored and reported accordingly. In general, INNS risk analyses focus on the environmental and/or cumulative impacts of INNS, without (specific) reference to co-transported pathogens (e.g., Dick et al. 2017).

Knowledge and policy gaps can result in inadequate scrutiny and assessment of the risks associated with the movement of pathogens into new regions and countries (Hulme 2014; Dunn and Hatcher 2015). This has been recognized with a call for the prioritization of empirical research required to cover knowledge gaps about transmitted pathogens (Chinchio et al. 2020). Therefore, a framework for quantifying and documenting our existing knowledge of INNS and associated pathogens which may also become introduced with host movements is vital. This involves conducting literature-based and pathogen screening to fill knowledge gaps where such information is lacking. These data will then lead to the development of invader pathogen profiles, outlining what is known about the invader’s pathogens and those of related taxa. A complexity in this process is the diversity within the pathogen profile of a given species across its invasive range (e.g., Bojko et al. 2018), where any single INNS may have multiple different disease profiles across its native and invasive range, which will also change over time. This potential spatial and temporal variation in the pathogen profile of a given species could potentially drive the need for more specific risk assessments in relation to invasion risks (i.e., not only a particular species, but also from a particular population).

In this paper we present a workflow to meet these imperatives. This can be applied to INNS already present in a region, those with the potential to arrive, and those already present but yet to establish. For the purposes of this paper, we focus on (potentially) pathogenic symbionts of aquatic INNS of concern to the UK, which may be permanently or transiently associated with one or multiple water bodies. We include all pathogenic symbiont types: viral, microscopic, and macroscopic parasites (including metazoans). The underlying premise can be applied across all habitat types, and all symbionts including pathogens that manipulate behaviors of one or more of their hosts, and symbionts that have no discernible effect on their hosts.

Biology and ecology of pathogens co-transported with INNS

The movement of INNS beyond their native range can result in changes to established host-pathogen relationships, including INNS losing or gaining parasites (Peeler et al. 2011; Dunn and Hatcher 2015; Vilcinskas 2015). The multitude of potential outcomes resulting from relationship changes are summarized in the schematic shown in Figure 1. The enemy release hypothesis (see glossary) states that INNS can lose their pathogens as they move into a new range, which may be due to ecological factors, or for heteroecious parasites, the absence of a secondary host (Colautti et al. 2004). Co-introduced pathogens can potentially infect native species (Keane and Crawley 2002).In some cases, pathogen loss can also increase invasion success by reducing pathogen burden and associated health costs as well as reducing/eliminating competing susceptible native species (Prenter et al. 2014). Furthermore, lack of co-evolution potentially results in the increased susceptibility of native hosts to the invading pathogen (Taraschewski 2006). For example, a study comparing invasive pathogenicity in co-introduced and native hosts suggests that in 85% of cases it is higher in native hosts compared to non-native hosts (Lymbery et al. 2014).

Figure 1.

Potential fates of symbionts (including pathogens) co-transported with INN host species. The left-hand panel represents a hypothetical INNS with a symbiome comprising pathogens A, B and uncharacterized symbionts 1–3. Potential symbionts already in the native system are pathogens C– F and uncharacterized symbionts 4–6. Symbionts can be gained and/or lost by INNS hosts. The main panel on the right presents, with examples, scenarios of gains, losses, and transfers between non-native and native hosts of different species, and outcomes associated with such interactions. Skull and crossbones indicates death/negative effects to native host population. Boxes with gray fill indicate theoretical outcomes for which no empirical evidence was found.

Co-introduced pathogens can have significant effects on both native and invasive host evolution, and also different populations of the same host species (Blakeslee et al. 2019a). For example, Rhithropanopeus harrisii has adapted to parasitism by an introduced castrating rhizocephalan parasite, Loxothylacus panopaei, resulting in much higher pathogen prevalence in its introduced range where the host is naive. This demonstrates the potential consequences of parasite introduction and transmission host populations where they lack an evolutionary relationship (Tepolt et al. 2019). Co-introduced pathogens can also suffer genetic founder effects themselves; this is particularly exhibited in obligate parasitic organisms with complex life cycles. Trematodes infecting the invasive eastern mudsnail (Tritia obsoleta) have been shown to have significantly lower genetic diversity in their introduced range compared to their native range (Blakeslee et al. 2019b).

INNS can also affect native host-pathogen relationships, altering population dynamics and disease transmission. Thieltges et al. (2008) demonstrated that the presence of invasive Crepidula fornicata and Crassostrea gigas significantly reduced the trematode parasite burden of native Mytilis edulis, by interfering with the transmission of free-living infective trematode larval stages and therefore reducing infection of M. edulis. Host-pathogen ecosystem interactions prove complex, creating challenges for the prediction of invasion success at different locations. The invasive amphipod Echinogammarus ischnus has outcompeted the native Gammarus fasciatus at many locations in the Great Lakes and St. Lawrence River in North America through predation and competition (Dermott et al. 1998). However, a native oomycete infects the invasive E. ischnus and causes greater mortality to the invasive host than to the native G. fasciatus, which facilitates the coexistence of the two species in areas of disease prevalence (Kestrup et al. 2011). This relationship is further nuanced in that E. ishnus can also act as a reservoir of the oomycete and facilitate parasite spillback to native amphipods.

Many diseases initially thought to be caused by one primary agent are now known to be the result of interactions between multiple symbionts, the host, and their environment; resulting in the pathobiome concept (Bass et al. 2019). Each INNS individually co-transports its own symbiome, making it difficult to predict its effect on the invaded ecosystem. A survey of symbionts of the invasive green crab Carcinus maenas in its native and invaded range showed many co-transported parasites persisted within the host at its invasion territory (Bojko et al. 2018). The latest approximation suggests this species is associated with ~82 known symbionts, many of which are pathogenic and pose risks to native ecosystems and the bioeconomy (Bojko et al. 2020).

The combination of hosts and their symbionts is of more immediate concern than considering the simple transposition of a pathogenic agent, such as a single virus or bacterium. Co-introduction of symbionts with an INNS is more likely to result in pathogen establishment because the co-evolved biological system is already in place to facilitate transmission (Peeler et al. 2011). Generalist pathogens are the main cause for concern since they can utilize native hosts more readily (Peeler et al. 2011). Symbiotic co-invaders may also present parasitic traits in new locations. For example, Aphanomyces astaci, the oomycete agent of crayfish plague, is a non-pathogenic symbiont of many invasive North American crayfish species (Tilmans et al. 2014); however, the introduction of A. astaci into Europe has resulted in large-scale mortalities in native crayfish populations, including: Austropotamobius pallipes, Astacus astacus and Astacus leptodactylus. In some cases, their local extinction is possible and has been noted in many regions (Mrugula et al. 2014). Box 1 details examples of known co-transported pathogens and their effects, co-transported symbionts and how symbiome research can help to assess invasion risks.

Box 1.

Download as

CSV

XLSX

1) Co-transportation of pathogens The invasive Asian eel (Anguilla japonica) brought with it the parasitic swim-bladder nematode Anguillicoloides crassus, which has caused high mortalities in native European eels (Anguilla anguilla) and significantly affected the sustainability of future European populations (Peeler et al. 2011). The OIE (World Organisation for Animal Health) – listed pathogen, Bonamia ostreae has caused decimation of native oysters (Ostrea edulis) in Europe, when it arrived with cultivated American populations of O. edulis for aquaculture in the late 1970s (Peeler et al. 2011). The ornamental trade has been implicated in the introduction of the chytrid Batrachochytrium dendrobatidis; a pathogenic agent partly responsible for the global decline of amphibians and species extinctions (Fisher and Garner 2007). The trade of freshwater molluscs has long caused concern about the potential for snail-mediated zoonotic diseases as they can act as intermediate hosts for parasites of significance to humans and livestock (Ng et al. 2016), e.g., angiostrongyliasis in humans caused by the parasitic nematode Angiostrongyliasis cantonensis co-introduced with the invasive snails Pomacea canaliculata and Pomacea maculata. Symbionts co-transported with INNS may be known pathogens which impact on wildlife in an expanded range (e.g., white spot syndrome virus; Mrugała et al. 2015), or their pathogenic potential may only be revealed when presented with new and susceptible hosts (e.g., the impact of Aphanomyces astaci on native white-clawed crayfish in Europe; Tilmans et al. 2014). 2) Co-transportation of commensal organisms The killer shrimp, Dikerogammarus villosus, invaded the UK in 2010, carrying the gregarine protists Uradiophora longissima and Cephaloidophora mucronata characterised from Polish freshwaters (Ovcharenko et al. 2009; Bojko et al. 2013). Gregarines are common commensal organisms of invertebrates that cover a wide symbiotic to parasitic spectrum (Rueckert et al. 2019) and undergo sexual reproduction in the animal gut, releasing spores into the environment that are consumed by other organisms. Uradiophora longissima and C. mucronata appear to be commensal organisms that have co-invaded with their host and do not exhibit any controlling effect upon the killer shrimp population (Bojko et al. 2013). Further molecular and histological studies will better identify commensal species by screening native and invasive populations of high-risk groups, such as the Amphipoda. 3) Invading symbiomes Assessing the symbiome of an organism requires the use of multiple tools, including both visualisation (microscopy) and diagnostic (molecular detection) techniques. By understanding the symbiome, we can explore co-infection and approach the invasion from a pathobiome perspective (Bass et al. 2019). The symbiome of the demon shrimp Dikerogammarus haemobaphes, a European invader originating from the Ponto Caspian region, has been shown to include viruses, bacteria, protists (including microsporidia) and metazoans (Bojko et al. 2019; Bojko and Ovcharenko 2019), identifying risks coupled with the invasion process (Bojko et al. 2015; Allain et al. 2020; Subramaniam et al. 2020). For example, the microsporidian parasite, Cucumispora ornata, has been shown to reduce the activity of D. haemobaphes and increase its rate of mortality, initiating population control at invasion sites and lowering he direct impact of the host on local biodiversity and the environment. In tandem, this parasite is also capable of infecting native Gammarus pulex, constituting a wildlife risk (Bojko et al. 2019). Metabarcoding and metagenomic techniques provide us with a capacity to easily pre-screen native species before they may become translocated. Metabarcoding of the UK invasive Homarus americanus cuticle revealed 170 associated bacterial taxa, suggesting that these microbial symbionts may have the capacity to invade with their host (Meres et al. 2012). Without technologies like these being used to advance invasion science, we remain in the dark about the complete symbiome and its associated risks.

A list of incoming aquatic INNS of concern to the UK was compiled from the lists of Roy et al. (2014) and a GBNNSS horizon scanning exercise (GBNNSS 2019). Seventy-seven aquatic INNS were identified from these lists (see Table 1). The literature searches were completed between August-October 2020.

To perform the literature search, both PubMed and Google Scholar were used to develop the best methodology (Figure 2). Figure 2 illustrates the workflow options and key considerations for choosing which database to search. Each has different characteristics that may preferentially suit different investigations. Both are subscription-free. The search terms used in this paper are given below; these can be adapted as required. This process can be used/adapted for non-aquatic species and with respect to any geographic region.

Figure 2.

Workflow for investigating existing data relating to symbionts (including pathogens) of current and potential INNS. The bullet points in each box indicate key considerations for each step of this customisable process. The list of factors in gray text influence whether PubMed or Google Scholar (or both) would be more appropriate for the particular species being researched.

Using PubMed; 34 of the 77 aquatic INNS were found to have no relevant literature relating to any known symbiotic species or pathogens. At genus level this number falls to 23; however, the relevance of symbionts and potential pathogens associated with the genus-level compared to the target species is uncertain but aids prediction. Symbiont and pathogen information extracted from the literature search for each species is listed in Suppl. material 1.

There were nine taxa for which species-level symbiont/pathogen data were published in 20+ papers; Neogobius melanostomus, Homarus americanus, Oncorhynchus gorbuscha, Carassius gibelio, Micropterus salmoides, Castor canadensis, Marsupenaeus japonicus, Myocaster coypus, and Ondatra zibethicus. The importance of these species in aquaculture, fisheries and human health is likely to explain their dominance within the literature. Homarus americanus, Marsupenaeus japonicus and Oncorhynchus gorbuscha are all highly valuable aquaculture species. Carassius gibelio and Micropterus salmoides are associated with the ornamental trade and recreational angling respectively. Castor canadensis, Myocaster coypus and Ondatra zibethicus carry multiple pathogens of human importance (see Suppl. material 1).

The results from PubMed and Google Scholar show some similarity. For taxa with little relevant literature, Google Scholar was more likely to return relevant data. As shown in Table 1, only 26 of the 77 aquatic INNS returned no relevant literature through Google Scholar in comparison to 34 from PubMed. For taxa with more literature; PubMed returned a larger proportion of useful papers in fewer results, and although these were usually also identified in Google Scholar, significantly more manual sifting of results in order to find these papers was required. For example, Marsupenaeus japonicus had 65 relevant papers selected from PubMed, but only 28 were identified from the first 100 Google scholar results despite a vastly larger overall return. This is likely to be because PubMed allowed for a more targeted search. We found using both PubMed and Google Scholar in parallel gives the most comprehensive picture.

PubMed search tools enabled a more accurate search as highly structured search criteria could be applied to just the title and abstract of papers, allowing a more focused search. However, the library of literature available in PubMed is smaller than on Google Scholar, and data from some figures and tables is not screened, sometimes leading to the omission of useful information. Google Scholar returned a significantly higher number of publications; the library of literature is much larger and it also scans grey literature and academic thesis repositories. However, Google Scholar also returns a much higher rate of irrelevant results which require significant manual sifting, in part because it scans the references of articles, and because the search cannot be narrowed by abstract. It is also important to scrutinize the source of literature from Google Scholar as it includes non-peer reviewed literature which may not always be suitable depending on the remit of the literature search. Haddaway et al. (2015) provides evidence to show that Google Scholar can be a powerful resource when used alongside other search methods; but is best used as a complementary tool.

Where there is a knowledge gap regarding the symbionts and pathogens of the target species, expert advice may be highly beneficial. This is likely to be the case for many known and potential INNS in most countries. Collaborative expert-elicitation is also a highly valuable tool within the field of biological invasion policy and has been implemented in numerous successful studies (Booy et al. 2017; Hughes et al. 2020; Peyton et al. 2019; Roy et al. 2014, 2017, 2018). These methods have been refined to ten guiding principles to consider within expert-elicitation to increase the effectiveness of this tool (see Roy et al. 2020).

When assessing the reliability of reports of co-transported pathogens in the literature, it is important to consider the methods used for their identification. Genetic signatures of pathogens may be associated with particular host samples in the literature, but these do not necessarily represent infections of those hosts; for example they could be passing through the gut and/or infecting host food items. Visualization techniques such as histopathology or in situ hybridization can be used to more precisely determine host-pathogen relationships initially inferred from molecular-only data. The use of such complementary techniques is recommended for research seeking to fill knowledge gaps such as those identified in this paper.

Invasion biology needs more robust methods for reliably evaluating the risks associated with INNS introductions (Kumschick et al. 2015). One of the most important factors to consider as part of risk assessments is evaluating the symbiome of INNS. Therefore, there is a need to better understand symbionts associated with INNS in order to evaluate the potential threat of emerging co-invasive pathogens as part of the INNS risk assessment processes (including horizon scanning). The workflow proposed in this paper uses a tested set of search terms in both PubMed and Google Scholar to thoroughly scan any available literature. This workflow aims to allow comprehensive data gathering of pathogens potentially co-transported with INNS, and constitutes a simple yet powerful methodology for the robust and standardized assessment of symbionts associated with INNS. As such, it provides a crucial step towards addressing the knowledge gaps regarding co-transportation of symbionts, facilitating integration of such knowledge into INNS risk assessment.

While limitations exist with respect to INNS data, the increasing use of histological, eDNA, and molecular diagnostics also offer new opportunities for monitoring INNS, potentially enabling the capture of pathological data more easily. Innovative modelling approaches, such as those using evolutionary trait-based frameworks (Barwell et al. 2020), can also inform horizon scanning and risk assessment to identify potentially impactful pathogens.

The introduction of INNS is widely recognized as important in both introducing known pathogens and a driver for the emergence of new pathogens (Peeler et al. 2011). There is a need at both international and national level for a collaborative approach to the assessment of INNS, efficient resource use and the formulation of guidance and risk assessment tools to both prevent and control the introduction of INNS and their symbionts. INNS do not recognize political boundaries so their effective management, particularly within the marine environment, requires transboundary co-ordination and collaboration.

Improved awareness raising, in particular across key sectors and stakeholder groups, will be important for managing the threat of INNS and their symbionts. The proposed amendments to risk assessment processes should aid in the more appropriate identification of INNS risk, but this will also need to be incorporated into other aspects of risk analysis including horizon scanning, risk management and prioritization. Further, robust and standardized prevention and mitigation approaches are needed globally to implement suitable actions once a species has been prioritized. For example, pathway management, border checks (to include molecular based screening for symbionts) and quarantine for intentionally introduced INNS, and routine monitoring and rapid response following detection of unintentionally introduced INNS. The use of molecular based tool sets is increasingly becoming a go to option for the detection of INNS and will be a necessity for the detection symbionts they may carry. Explicit consideration of symbionts and potential for disease emergence should also be made within assessments undertaken prior to the translocation of both INNS and native species for conservation or assisted colonization purposes such as for aquaculture.

This issue is now more pressing than ever: climate change could act synergistically with other stressors, to increase the impacts of invading pathogens. Rising water temperatures may mean more INNS and their pathogens are able to survive and establish in the UK. Furthermore, the increasing global demands on aquaculture production, mean that impacts arising from emerging aquatic diseases are increasing in frequency, and have increasingly diverse and serious economic implications.

RF was supported by a PhD studentship co-funded by the NHM and Cefas. DB was supported by Defra projects FB002 and FC1215. HER is supported by the Natural Environment Research Council award number NE/R016429/1 as part of the UK-SCAPE program delivering National Capability and through the Defra-funded GB Non-Native Species Information Portal. JB was supported by the Global Challenge Research Fund at Teesside University to develop outreach for understanding disease transmission via biological invasions.