______Risk analysis CMTV-like virus April 2013 1

This document should be referred to as: Rijks JM, Sptizen-van der Sluijs A, Leuven RSEW, Martel A, Kik M, Pasmans, F, Zollinger R, Verbrugge LNH, Gröne A (2012). Risk analysis of the common midwife toad-like virus (CMTV-like virus) in the Netherlands. NVWA, Min EZ report 60000784-2012

______Risk analysis CMTV-like virus April 2013 2 Table of contents

Executive summary 5

Part 1 - General background information for risk analysis 9 Theme 1 - The virus 10 Theme 2 - Distribution of CMTV-like virus 12 Theme 3 - Known susceptible host species 14 Theme 4 - Host life stages affected 16 Theme 5 - Dispersal characteristics of host species 18 Theme 6 - Distribution of susceptible host species in the Netherlands 20 Theme 7 - Distribution of potential hosts across the border 22 Theme 8 - Accompanying amphibian species 24 Theme 9 - Water body types with CMTV-like virus 26 Theme 10 - Month of outbreak and water temperature 28 Theme 11 - Water quality at outbreak sites 30 Theme 12 - Inventory of trade of amphibians in Belgium 32 Theme 13 - Inventory of hobby kept amphibians in Belgium 34 Theme 14 - Legal re-introductions and translocations in the Netherlands 36 Theme 15 - Illegal introductions 38 Theme 16 - Flyways 39

Part 2 - Risk assessment 40 1. Method 40 2. Results 42 2.1 Probability of introduction 42 2.1.1 Introduction through human activities 42 2.1.1.1 Legal trade through Schiphol 42 2.1.1.2. Amphibian trade in Belgium 44 2.1.1.3 Entry of CMTV-like virus imported via live amphibians into nature 45 2.1.2. Introduction through natural dispersal 46 2.1.2.1 Introduction via infected amphibians crossing borders into the Netherlands 46 2.1.2.2 Introduction via migrating birds 46 2.2 Likelihood of establishment 47 2.2.1 Susceptible hosts are present 47 2.2.1.1 Susceptible species 47 2.2.1.2 Susceptible populations 47 2.2.1.3 Susceptible individuals 47 2.2.2 Dutch environment in general appears suitable 49 2.2.2.1 Temperature 49 2.2.2.2 Suitable sites 49 2.2.3 The virus has strategies to maintain itself 49 2.2.3.1 Probable lengthy survival outside the host 50 2.2.3.2 Reservoirs 50 2.3 Probability of spread 50 2.3.1 Spread through human activities 50 2.3.1.1 Legal re-introductions and translocations of threatened native species 50 2.3.1.2 Other human-mediated transfer of potentially infected hosts 52 2.3.1.3 Dispersal by fomites (equipment, transfer of water or sediment) 52 2.3.2 Natural spread 52 2.3.2.1 Dispersal of (sub-)clinically infected hosts 52 2.3.2.2 Birds or other animals 53

______Risk analysis CMTV-like virus April 2013 3

2.4 High risk areas 53 2.5 Impact 54 2.5.1 Ecological impact 54 2.5.1.1 Amphibians 54 2.5.1.2 Effects on biodiversity 55 2.5.2 Socio-economic impact 56 2.6 Risk classification using the ISEIA protocol 57 2.6.1. Expert consensus scores 57 2.6.2 Dispersion potential or invasiveness 57 2.6.3 Colonisation of high conservation value habitats 58 2.6.4 Adverse impacts on native species 58 2.6.5 Alteration of ecosystem functions 58 2.6.6 Risk classification 59

Part 3 - Risk management options 61 3.1. Prevention of introduction 61 3.1.1 Make sure that CMMTV-like virus can be detected in imported captive specimens at the border 3.1.1.1 Obtain sufficient sight on amphibians imported 61 3.1.1.2 Improve the capacity to detect CMTV-like virus in imported specimens at the border 61 3.1.2 Reduce the risk that CMTV-like virus enters nature via imported kept amphibians and their waste 63 3.1.2.1 Obtain good sight on amphibian and vivarium waste disposal behavior 63 3.1.2.2 Make sure that people are aware of risky behavior and know how to minimize it 63 3.1.3 Make sure that CMTV-like virus imported via natural dispersal can be detected early on at its site of introduction, before it occurs widespread 63 3.1.3.1 Obtain a better understanding of the relative importance of introduction via natural dispersal 63 3.1.3.2 Ensure monitoring for early detection and source tracing 63 3.2 Elimination 65 3.2.1 Assisted elimination – currently only applies to captive populations and small free-living populations 65 3.2 1.1 Encourage destocking and disinfection for elimination of infection in captive settings 65 3.2.1.2 Explore options for elimination in the case of (small) infected free-living populations 65 3.2.2 Natural elimination – learn from it for larger scale assisted elimination 66 3.3 Control 68 3.3.1 Prevent human mediated spread of virus within the Netherlands 68 3.3.1.1 Prevent inadvertent to new sites via re-introduction projects 68 3.3.1.2 Prevent inadvertent spread to new sites by public or field biologists 68 3.3.2 Explore the possibility to take advantage of natural barriers to limit natural dispersal 68 3.3.3 Try to avoid further impact on threatened species and high conservation value habitats 68 3.3.4 Make sure the correct data is collected to be able to predict the effectiveness of control And elimination measures 69

Glossary 71 List of abbreviations 72 Acknowledgements 73 References 73 Contributors 80

______Risk analysis CMTV-like virus April 2013 4 Executive summary

The term CMTV-like virus is used to indicate the common midwife toad virus (CMTV, GenBank accession no. FM213466.1) and viruses with partial major capsid genes showing 99.8% to 100 % sequence homology. CMTV-like virus infection is an emerging infectious disease in wild amphibians in the Netherlands. One possibility is that CMTV- like virus is a recently introduced exotic pathogen.

Study aim and set-up The aim of this study is to perform a risk analysis for CMTV-like virus in the Netherlands. General background information from the literature, field surveys and interviews, was compiled in 16 themes (Part 1). Subsequently, an assessment was performed to identify 1) the possible routes of introduction of CMTV-like virus into the Netherlands, 2) the likelihood of its establishment, 3) the likelihood of its spread, and 4) the possible ecological and social impact of CMTV-like virus. An ecological risk classification score was obtained using the Belgian Invasive Species Environmental Impact Assessment (ISEIA) protocol (Part 2). Finally, opportunities for management were evaluated, as well as aspects that require further understanding for appropriate decision-making (Part 3).

Part 1 - General background information for risk analysis (themes 1-16) CMTV-like virus is an amphibian ranavirus (Th.1). The virus has been detected in restricted areas in Spain in 2007 and 2008, in Belgium in 2010, and in the Netherlands in 2010, 2011 and 2012. In the Netherlands, the virus was first detected in a high conservation value habitat, the National Park Dwingelderveld (Th.2).

Ranaviruses infect amphibians, reptiles or fish, or several of these (Th.1). CMTV-like virus has caused disease and mortality in half of the amphibian species that are native to the Netherlands. Among these species is the vulnerable common spadefoot Pelobates fuscus . The susceptible amphibian species belong to the order of frogs ( Anura ) or salamanders ( Caudata). In addition, CMTV-like virus infections have been observed naturally in amphibian species exotic to the Netherlands, in particular the American bullfrog Lithobathes catesbianus and poison dart frogs Dendrobates spp. The infected amphibians can show disease, but infection can also be subclinical. Evidence for infection of fish is inconclusive: one specimen found dead tested positive by PCR but autolysis hampered pathological investigation. Therefore, it remains unclear whether this case represented a true infection or contamination from the aquatic environment (Th.3). Multiple life stages of amphibians were shown to be infected in the Netherlands. The fact that not only larvae and juveniles but also (sub-)adult amphibians died during outbreaks, is suggestive of disease emergence in immunologically naïve populations (Th.4).

A number of the susceptible host species are common species found in most of the country (Th.5), as well as in the neighbouring countries Belgium and Germany (Th.6). Many are good dispersers, meaning they may move distances up to several kilometres per year (Th.7). This information is relevant to the potential of the virus to spread. Further, many of the susceptible host species are common and characteristic accompanying species for each other, suggesting frequent opportunity for inter-species exposure (Th.8).

Themes 9 – 11 report environmental factors that are associated with disease occurrence. The water bodies where CMTV-like outbreaks have been detected are both natural and man-made and vary in soil type, size, altitude, and uses. However, they were generally permanent, small to medium sized and unconnected to other water bodies (Th.9). Mass mortalities due to CMTV-like virus occurred in the period May to September 2011. Monitoring at infected pools in the National Park Dwingelderveld showed that water temperature at monitored sites was on average ± 19°C, and it followed mean daily air temperature. However, mortality can also occur in water that is several degrees colder (Th.10). Measurements provided no clear evidence for poor water quality at sites with CMTV-like virus associated amphibian mortality (Th.11).

Themes 12 – 16 cover aspects relevant to translocation of infection. For understanding of the ways how amphibians could enter the Netherlands via Belgium, data were obtained on amphibians provided for trade by importing dealers in Belgium. Most of these amphibians were captive bred or farmed, but some were wild-caught. Several thousands of amphibians are imported into Belgium from outside the EU each year. In addition, numerous species of exotic and native amphibians are kept and exchanged between Dutch and Belgian hobbyists. There is virtually no knowledge

______Risk analysis CMTV-like virus April 2013 5 regarding ranavirus infections occurring in such animals (Th.12 &13). Within the Netherlands, recently three native endangered amphibian species have been legally translocated and re-introduced into new areas. These legally authorized biological conservation activities have a phased implementation (activity identification, procurement of stock, growth of stock in captivity, release at site and monitoring at site); however, the risk of transfer of CMTV-like virus is not yet fully taken into account in these implementation steps (Th.14). There are other (un-)intentional releases of amphibians in the Netherlands besides these authorized re-introduction activities. Most releases are untraceable, and the frequency of occurrence of such events is unknown. However, some releases have been documented, such as the release of alpine newts and common midwife toads from France and elsewhere into the area where CMTV-like virus disease events are occurring in the Netherlands. The significance of these releases for the introduction of CMTV-like virus infection is undetermined. (Th.15). Finally, birds have been suggested as cause of spread of ranavirus. Migratory birds have flyways along the axe Spain-Belgium-the Netherlands. However, there is no definite proof that dispersal of the virus via birds occurs (Th.16).

Part 2 - Risk assessment §2.1 Probability of introduction from abroad - CMTV-like virus is present in the Netherlands at least since 2010, but there is no evidence suggesting that it is an endemic virus. Two possible pathways for introduction of CMTV-like virus into the Netherlands are human-mediated introduction or introduction via natural dispersal of infected or contaminated animals. With regards to human mediated introduction, it has to be noted that numerous amphibian species from all over the world are traded. The traceability of imported specimens is poor, and data is lacking on prevalence of CMTV-like infection in these amphibians. Further, the detection of CMTV-like virus infection during import via the national airport Schiphol is currently highly unlikely, unless the animals are clinically ill or dead at the moment of inspection. However, the chance that animals arrive clinically ill or dead is rather low due to short flight and total transportation times. To the best of our knowledge, live amphibians are not imported for food production in the Netherlands; many are imported to serve as pets. Wild caught specimens of unknown health status are likely to be regularly introduced into facilities where captive-bred species are kept. There is significant captive-bred exotic amphibian trade, also within the EU, but the volume is not quantified and there is poor insight into the occurrence of ranavirus outbreaks in such collections. If ranavirus occurs in captive amphibians, there is a chance that it is introduced into nature. It is relevant that CMTV-like virus exists in a free-living amphibian population of American bullfrogs across the border in Belgium, as this could lead to introduction via natural dispersal. Documented evidence for introduction of CMTV-like virus via natural dispersal of infected amphibinans or eventually via other animals is still lacking. To conclude, there are several possible routes of introduction of CMTV-like virus, but species, numbers, and prevalence of CMTV-like infection in imported specimen is unknown.

§2.2 Risk of establishment - CMTV-like virus has already persisted in the area where it was initially detected in the Netherlands. The area where it is present includes a Natura 2000 area, National Park Dwingelderveld. For establishment, the virus requires susceptible hosts and suitable environmental conditions, and needs to be able to maintain itself from year to year. Species susceptible to CMTV-like virus include common and widespread amphibian species, of which the populations so far appear immunologically naïve. The environmental conditions of water bodies in which CMTV-like virus thrives, are diverse and commonly found throughout several water types in the Netherlands. The water temperature of these water bodies is likely to be favourable to ranavirus replication in the summer and virus survival in the winter. Finally, CMTV-like virus probably has several mechanisms to maintain itself after initial introduction. These include interspecies transmission, co-evolution, lengthy survival outside the host under certain conditions, and intra-species reservoirs. Taking all information together, the conclusion is that conditions for establishment are present in the Netherlands, and so the probability of establishment is high.

§2.3 Risk of spread - The virus has been detected in a restricted area in the Netherlands, initially at sites located in the Province of Drenthe, but in 2012 it was also detected in the Province of Overijssel. One possibility is that it naturally spread after initial introduction. Short distance natural dispersal via (sub-)clinically infected hosts is very likely. A number of susceptible amphibian host species are good dispersers, up to 15 km a year. However, human-mediated introductions cannot be excluded. Short to long distance dispersal of CMTV-like virus within the Netherlands via human-mediated activities is highly likely, as this has been documented for other ranaviruses elsewhere.

______Risk analysis CMTV-like virus April 2013 6 §2.4 Area at risk - The area at risk is the whole of the Netherlands, in habitats where amphibians live. Very high risk areas are areas in provinces where threatened species occur. For the common spadefoot these are located in the provinces of Drenthe, Overijssel, Noord Brabant, Gelderland, and Limburg. For the common midwife toad these are located throughout the entire country.

§2.5 Impact - A negative effect of emerging CMTV-like virus on amphibian numbers is seen at outbreak sites. Elsewhere, the emergence of ranavirus infection at sites was transient, catastrophic, or persistent with recurrent mortality events and populations declines of on average more than 81%. Follow-up at some CMTV-like virus infected sites suggests that catastrophic and persistent effects on populations may also be occurring. However, the events are too recent for final judgement and long-term monitoring will be required to corroborate these observations. Further, a number of characteristics of the CMTV-like virus suggest that it has the potential to cause endangered species to go (locally) extinct. There is an urgent need for more understanding into possible effects on fish. Other ecological and socio-economic effects are very poorly documented.

§2.6 Score - Current - The risk classification of the CMTV-like virus for the current situation in the Netherlands is B2, using the ISEIA protocol. This indicates a non-native species with a restricted distribution range and moderate environmental hazard (i.e., ecological risk) that should be placed on a watch list. Future – If no measures are taken, the future scenario shows a potential increase in the invasion of the CMTV-like virus resulting in a widespread distribution and viral disease of amphibians in a significantly increasing number of high conservation value habitats, resulting in highly adverse direct and indirect impacts on biodiversity and alterations of ecosystem functions at a wider spatial scale. In particular, the expected increase in number of high conservation value habitats with adverse effects would lead to an increase in the total risk score by one point according to the ISEIA protocol. Combined with the widespread distribution this will result in a reclassification to a higher environmental risk (A3 - Black list classification).

Part 3 - Risk management and risk communication Two important assumptions underlying the proposed risk management and risk communication measures are that 1) CMTV-like virus is not widespread but emerging and 2) there are no other ranaviruses in the Netherlands.

§3.1 Prevention of introduction – To prevent the introduction three leads are provided. 1) The first lead is to make sure that CMTV-like virus can be detected in imported captive specimens at the border. This requires more insight into the species and numbers of amphibians traded commercially or exchanged internationally, and the prevalence of infection in the imported specimen and in natural populations in other countries. Further the capacity to detect and confirm cases at the border must be increased. 2) The second lead is to reduce the risk that CMTV-like virus enters nature via imported kept amphibians or their waste. This requires obtaining good sight on amphibian and vivaria waste disposal behavior, establishing how the risk of introducing the virus can be mitigated, and communicating this to the public. 3) The third lead is to arrange that CMTV-like virus imported via nature can be detected early on at its site of introduction, before it occurs widespread. To achieve this, monitoring through investigation of unusual mortality events in amphibians was recommended. Further, specific attention for CMTV-like virus in amphibians along the Dutch-Belgian border near Hoogstraten was suggested.

§3.2 Elimination –Assisted virus elimination, i.e., elimination undertaken by humans, is currently only recommended in captive populations and in small free-living populations. Humane, timely and thorough destocking and disinfection of premises is the most secure option if CMTV-like virus is found in captive amphibians. Acceptable elimination measures for use in water bodies are virtually unidentified. Therefore some guidance in taking the decision to opt for elimination or not in small free-living amphibian populations is recommended, and practical advice for best practice should be provided. The need to obtain permission to destock should be known to the public and the process speedy. Technical alternatives for achieving virus elimination from natural sites and side-effects should be investigated, and official registration of disinfectants for such use encouraged. The effect of water body management interventions on the occurrence of CMTV-like virus infections should be monitored, even those not primarily undertaken with the view of elimination. The mechanisms and factors underlying natural elimination, i.e., elimination achieved by natural processes in the wild, are likely to provide information useful for assisted elimination on a large scale. These can be identified through long-term monitoring and should then be translated into practical measures for use in the field.

______Risk analysis CMTV-like virus April 2013 7 §3.3 Contr ol - Possibilities to control CMTV-like virus focus on minimizing spread and impact. Special effort should be made to prevent sites where threatened species are found from becoming infected. Projects legally authorized to translocate and re-introduce threatened species need to take the necessary precautions to avoid introducing disease in the process. Veterinary advice to minimize this risk throughout the project’s implementation is required. Education of the public to refrain from releasing amphibians into nature may reduce spread via intentionally released live animals. Education of field biologists to properly clean and disinfect equipment, shoeing and clothes should reduce spread via fomites, in particular if disinfection is made more practical. Absolute containment of wild animals in an area is not possible, but possibly there are barriers in the landscape and land uses that can limit natural dispersal by amphibians of which temporary use could be made. Vaccination of threatened species is not an option at this stage, because there are no vaccines available for CMTV-like virus or other ranaviruses. Chemotherapy too is not available. Finally, epidemiological models are very useful to properly assess the effectiveness of control measures. The information in this document can be used to define which data is missing to make such models, so that the collection of such data could then be a priority if the route of control and elimination are taken.

To conclude:

1. Risk assessment 1a) CMTV-like virus was detected for the first time in the Netherlands in a pond in National Park Dwingelderveld in 2010, and there is no evidence suggesting it is an endemic virus. There are several potential routes of introduction, both human-mediated and via nature, but it is currently not possible to determine probability of introduction via these routes.

1b) CMTV-like virus has been detected in a restricted area of the Netherlands only. It has maintained itself in amphibians in water bodies of Dwingelderveld since 2010. Suitable hosts and environmental conditions are present, and the probability that it can establish itself is high.

1c) Initially observed only in the Province of Drenthe, CMTV-like virus infections were detected in the Province of Overijssel in 2012. The probability that CMTV-like virus will spread is high, either short distances (up to 15 km a year) through infected amphibians, or short to long distances via humans and possibly birds.

1d) The area at risk is the whole of the Netherlands, in habitats where amphibians live.

1e) So far, CMTV-like virus has caused catastrophic and persistent effects on populations at infected sites. Other ecological and socio-economic effects are very poorly documented.

1f) Currently the CMTV-like virus is assessed to be on the ‘Watch List’ (B2), being a non-native species with a restricted distribution range and moderate environmental hazard. In the future, without interference, the distribution of CMTV-like virus is likely to be more widespread, with more adverse effects on high conservation habitats. This would place it on the ‘Black List’ (A3).

2. Risk management and communication 2a) To prevent the introduction of CMTV-like virus, three measures are recommended: - make sure that CMTV-like virus can be detected in imported captive specimens at the border. - reduce the risk that CMTV-like virus enters nature via imported kept amphibians or their waste. - arrange that CMTV-like virus imported via nature can be detected early on at its site of introduction, before it occurs widespread.

2b) Destocking and disinfection of premises should be used for virus elimination in captive populations. Options for elimination of virus from free-living populations should be explored further, because there is still little experience with such measures. Monitoring the effect of water body management interventions on CMTV- like virus presence and a better understanding of the mechanisms and factors leading to ‘natural elimination’ can contribute to this.

2c) Control focuses on limiting the spread of the infection and the impact on threatened species through education of the public and field biologists. Epidemiological models could help to assess the effectiveness of ______suggested control measures. ______Risk analysis CMTV-like virus April 2013 8 Part 1 General background information for risk analysis (themes 1 – 16)

General background information is provided as 16 theme pages (Th.). The relevance of theme for each component of the risk assessment (introduction, establishment, spread or impact, or several) is indicated first. Then after defining theme concepts and providing trends and patterns, the significance of the findings for CMTV-like virus infection mentioned.

Sources of information - The sources of information were literature, field data, interview data and expert opinion. Literature - Literature was searched using Scopus, Web of Science and Pubmed. No limit in year of publication was added. ‘Ranavirus was used as search term, and yielded 250 records of papers and book chapters. Documents were screened for suitability for this risk assessment by checking for relevant information. Additionally, for information on host species, the library of RAVON (17,000 records) was accessed using ReferenceManager. Field data - In 2011, an extensive survey in the aftermath of the 2010 outbreak of ranavirus in National Park (NP) Dwingelderveld was conducted. This included post-mortem investigation of mortality events nationwide 1, as well as monitoring of a number of ponds in NP-Dwingelderveld regularly during the period June – October 2011 2. Per visit per pond, the number of sighted individual amphibians were noted, as well as the species, their condition (healthy, sick or dead), life stage (adult, sub-adult, juvenile or larvae). Additionally characteristics of the water body were noted, such as water temperature, and water quality was measured. In 2012,as part of this assignment, two field visits (24/8/12 and 4/9/12) were made to this NP to obtain an update. Interview data and expert opinion - For information on amphibian trade via Schiphol, DWHC consulted the Netherlands Food and Consumer Product Safety Authority and a board member of the ATA (Animal Transportation Association). For information on legal re-introduction and translocation programmes, DWHC consulted amphibian re-introduction/translocations files at Dutch Ministry of Economic Affairs, Natuurbalans – Limes Divergens BV, and RAVON. To get an idea about the trade of amphibians in Belgium, UGent consulted the FAVV (Federal Agency for the Safety of the Food Chain, Belgium), research laboratories dealing with amphibians, and animal importers. About 40 research labs can perform research on amphibians. Three of these labs have indicated that they, sometimes, import amphibians. In Belgium, 124 merchants are registered as allowed to import amphibians. The five most important amphibian importers were contacted for this survey. In addition, a questionnaire was sent to private amphibian collections in Belgium.

______Risk analysis CMTV-like virus April 2013 9 Theme 1 – The virus

Theme relevant to pathogen introduction, establishment, spread and impact

Definition – CMTV - Common midwife toad virus (CMTV; GenBank accession no. FM213466.1)3 is a double- stranded DNA virus belonging to the genus Ranavirus, family Iridoviridae . The ranavirus genus - The identification of the different virus species belonging to the genus Ranavirus is on-going. To date six (6) species are officially recognized and there are also a number of tentative species, i.e., ranavirus isolates that may be classified as distinct ranavirus species in the future. The six recognized species are Ambystoma tigrum virus (ATV), Bohle irido virus (BIV), Frog virus 3 (FV3/TFV), Epizootic haematopoietic necrosis virus (EHNV), European catfish virus (ECV/ESV), and Santee-Cooper ranavirus (SCRV, LMBV) 4. CMTV is a tentative species, with an intermediate position in evolution between EHNV-like viruses and FV3-like viruses 5. CMTV-like virus - Besides CMTV, a number of CMTV- like ranaviruses have been described 6-8. Genetic material of these viruses, amplified by PCR and then sequenced, shows 99.8% to 100 % sequence homology with CMTV partial major capsid protein gene (GenBank accession no. FM213466.1). Definition When referring to CMTV and such CMTV-like viruses collectively, we will further use the term CMTV-like virus.

Patterns and trends in other ranaviruses – Ranaviruses are infective as naked particles and even more so as enveloped particles. An envelope is acquired when the virion buds through the host cell membrane 4. Transmission with and without direct contact - Transmission is horizontal. It can be through direct through direct contact with infected individuals, cannibalism and necrophagy of infected individuals, or exposure to water and fomites containing virus 9-16 . There is no conclusive proof for vertical transmission to date, though it has been suggested 17 .High infectivity - ATV needs about 1 s of direct contact for transmission via intact or damaged skin 10 . FV3 released in water by infected specimens infected others within 3 hours of exposure 16 . Replication temperature fits poikilothermic hosts - Ranaviral replication takes place at 12-32°C, with viral protein synthesis occurring within hours of cell infection. Cell death can occur as quickly as a few hours following infection, by either necrosis or apoptosis 18 . Survival in the environment is enhanced by cold temperatures - Water - Virions remain stable in water for extended periods at 4°C 4. For two amphibian and two reptilian ranaviruses, it was demonstrated that about 10% of the initial virus load remained infectious for 2 months in water at +4°C and for 3weeks to a month at +20°C 19 . Cell-free EHNV remained infective after 97 days in water (no detail provided on temperature) 14 . Sediment – Wood frogs ( Lithobates sylvaticus ) could be infected by exposure to sediment from a site where a ranavirus die-off was occurring 13 . For two amphibian and two reptilian ranaviruses it was demonstrated that about 10% of the initial virus load remained infectious for over a month in soil at +4°C 19 . BIV survives desiccation for up to six weeks at temperatures up to 42°C 4, but ATV degrades quickly 9. In tissue - EHNV remained infectious in tissue of dead fish for more than seven days at +4°C, and more than two years at -20°C and -70°C 14 . Virus Inactivation - Heat – Ranaviruses are inactivated within 30 minutes at 60°C 4; ENHV was also inactivated after 24 hours at 40°C 14 . Ultra- violet-irradiation – FV3 is inactivated by UV-irradiation 4. Ultraviolet sterilizing units were effective in inactivating water-suspended iridovirus isolated from frogs at a flow rate comparable to that used in nurseries in aquaculture (5000 L/h). With transmittance reduced to 27.7%, the minimum effective dose was 2.6 × 10 4 uW.sec/cm². 20 . Extreme pH : One hour of pH<4 or pH>12 inactivated EHNV 14 . Disinfectants – In absence of organic matter, sodium hypochlorite (200 mg / l) is effective 14, 21 ; Ranavirus is also inactivated by 70% ethanol 14 . Surface application for one (1) minute or more of chlorhexidine 0.015% (0.75% Nolvasan® 2.0), 0.180% sodium hypochlorite (3.0% bleach 6.0), or potassium peroxymonosulfate 0.204% (1% Virkon S® 20.4) reduced ranavirus load by 8 log 10 , and was 21 recommended for disinfecting contaminated equipment .

Significance for CMTV-like virus - Little research has been done specifically on CMTV-like viruses, other than the complete genome sequencing by Mavian et al ., 2012 5. Likely data on transmission modes, infectivity, replication temperature, survival and inactivation is based on extrapolation of knowledge with regards to other ranaviruses.

______Risk analysis CMTV-like virus April 2013 10 Figure 1 – Electron Microscopy (EM) photograph of an array of CMTV-like virus. Ranaviruses have a distinctive icosahedral shape, approximately 150 nm in diameter, visible by EM as paracrystalline arrays in cytoplasma of cells 22 . The genome of CMTV (106,878 bp) contains 104 open reading frames encoding putatively expressed proteins 5.

______Risk analysis CMTV-like virus April 2013 11 Theme 2 - Distribution of CMTV-like virus

Theme relevant to pathogen introduction, establishment, spread and impact

Definition - The distribution of the virus is its occurrence in space and time.

Patterns and trends - CMTV-like virus has only been detected in Europe, namely in Spain 3, 23 , Belgium 8 and the Netherlands 1, 6 (Figure 2a, left). Spain 2007 & 2008 - The very first detection (and description) was in northern Spain in 2007, when amphibian mortality occurred at a site in National Park (NP) Picos de Europa 3. This was followed in 2008 by amphibian mortality at a second site in the NP Picos de Europa, only 1 km away from the first 23 . In between these two events 23 and in subsequent years (Balseiro A., pers. comm. 2012) no further CMTV-associated mortality was observed. Belgium 2010 - In Belgium, CMTV-like virus was not detected in any of the amphibian mortality events investigated between 2007 and 20106. However in 2010 it was found in healthy larvae of free-living exotic species in Hoogstraten (Figure 6, Theme 6). The Netherlands 2010, 2011 & 2012 - In the Netherlands, CMTV-like virus was first detected in 2010, at NP Dwingelderveld (Drenthe province; Figure 2a, right)6. Subsequently, the situation in NP Dwingelderveld was monitored, and a nationwide request was put out to the public to report significant amphibian mortality events. The public reported 20 significant amphibian mortality events outside NP in 2011. At 16 sites no CMTV-like virus or other ranavirus was detected. At the remaining four (4) sites, however, mortality was shown to be associated with CMTV-like virus infection 1. These sites were Wijster, Darp, Fluitenberg and Ijhorst (Drenthe province). These are all located within 25 km of each other and NP Dwingelderveld, where CMTV-like virus associated mortality continued in 2011 (Figure 2b) and 2012 1, 2 . In 2012, CMTV-like virus associated amphibian mortality occurred at a sixth site, Staphorst (Overijssel province), located only 7 km to the south of IJhorst (Figure 2a, right)1. Retrospectively, three amphibian mass mortality events were reported to have occurred in the summer or autumn of 2009 and 2010 to the east-north- east of NP Dwingelderveld; ranavirus infection may have been the cause but no material remained or has since been submitted to ascertain this 1. Finally, CMTV(-like virus) was detected in 2010 in exotic hosts belonging to a captive collection of which the location is not disclosed 7.

Significance - Phylogenetic analyses shows ranaviruses generally group according to their geographic and their host class origin 24 . Because CMTV-like virus was detected in several European countries, it has been suggested that it may have a wide geographic distribution across the European mainland 5. Distribution suggests disease emergence in the Netherlands Regardless of whether this is the case or not, the nationwide surveillance based on investigation of unusual mortality events in the Netherlands in 2011 points towards focal occurrence of severe disease and mortality due to CMTV-like virus in the Netherlands, i.e., a disease hotspot of about 500 km 2. This suggests infectious disease emergence 1.

______Risk analysis CMTV-like virus April 2013 12 Figure 2a - Left – Countries (green surfaces) where CMTV-like virus has been detected to date (December 2012). CMTV-like virus has a focal distribution in each country (red dots) and is not known to occur widespread. Right – The sites where CMTV-like virus has been detected in the Netherlands, per year of first detection.

Figure2b – Ponds and pools monitored or visited after a mortality report, Dwingelderveld 2011. Those monitored were Kolenveen (A), Bezoekerscentrum (C & D), Achterlandseveen (E), and Smitsveen (F). Those visited were Drostenvbeen (H), Noorsterveen (I) and Holveenslenk (J). CMTV-like virus infected hosts were found at all sites except D and H.

______Risk analysis CMTV-like virus April 2013 13

Theme 3 - Known susceptible host species

Theme relevant to pathogen introduction, establishment, spread and impact

Definition - A host species is considered susceptible to CMTV-like virus infection and disease if individuals belonging to the species were found to have post-mortem lesions consistent with ranavirus infection associated with ranavirus genetic material 25 showing 99-100% sequence homology 25 with gen bank Accession No. FM213466.1. A host species is considered susceptible to CMTV-like virus infection if ranavirus genetic material showing 99- 100% sequence homology with gen bank Accession No. FM213466.1 was detected in internal organs of individuals belonging to the species. A host species is considered possibly susceptible to CMTV-like virus infection if ranavirus genetic material showing 99-100% sequence homology with gen bank Accession No. FM213466.1 was detected in individuals belonging to the species without confirmation of infection.

Patterns and trends - There are eleven (11) host species susceptible to CMTV-like virus infection and disease: - The edible frog ( Pelophylax klepton esculentus, family Ranidae )6 ⋅ The marsh frog ( Pelophylax ridibundus, family Ranidae )1 ⋅ The pool frog ( Pelophylax lessonae, family Ranidae) 6 ⋅ The smooth newt ( Lissotriton vulgaris, family Salamandridae )6 ⋅ The common toad ( Bufo bufo, family Bufonidae )1 ⋅ The common spadefoot toad ( Pelobates fuscus, family Pelobatidae )1 ⋅ The common midwife toad ( Alytes obstetricans, family Alytidae )3, 23 ⋅ The Alpine newt, subspecies cyreni ( Mesotriton alpestris cyreni, family Salamandridae) 3, 23 ⋅ The Golfo Dulce poison frog ( Phyllobates vittatus, family Dendrobatidae )7 ⋅ The black-legged poison frog ( Phyllobates bicolor, family Dendrobatidae )7 ⋅ The green and black poison dart frog (Dendrobates auratus, family Dendrobatidae )7. The first three are collectively named water frogs. For the first eight host species, infection and disease was detected in wild populations. These eight host species are native to the Netherlands (though the specific subspecies of the eighth species that was affected in Spain is not the subspecies present in the Netherlands). For the last three host species, infection and disease was detected in captive aquarium-kept frogs in the Netherlands. These last three host species are all exotic species for the Netherlands There are two (2) host species susceptible to CMTV-like virus infection. These are the common frog (Rana temporaria, family Ranidae )1 and the American bullfrog (Lithobates catesbeianus, family Ranidae) 8. Evidence for disease in the infected common frog examined was inconclusive 1, and the American bullfrogs were clinically healthy tadpoles 8. The American bullfrog, exotic to Europe, may be a carrier of ranaviral disease 8. There is one (1) host species possibly susceptible to CMTV-like virus infection , the ten-spined stickleback ( Pungitius pungitius, family Gasterosteidae) . This is a species of fish that is native to the Netherlands. One specimen of this species, originating from a pond where an outbreak of CMTV-like virus disease had occurred, was found to be positive by PCR-test. However, this specimen was too decomposed for post-mortem examination and due to the pooling of multiple organs in the sample tested, it is not possible to know if the virus had infected the fish or simply occurred on the skin, gills or in the lumen of the intestines of the fish 1, 2 . Difference in host species susceptibility – The current data is too restricted for conclusions in this regard. However, there is some evidence for the occurrence of subclinical infections.

Significance – CMTV-like virus has a broad host range. Multiple native amphibian species susceptible - At least half the native amphibian species present in the Netherlands are susceptible to CMTV-like virus infection and disease (Figure 3). Confirmed infection with CMTV-like virus has been demonstrated in the two orders of the amphibian class, Anura (frogs and toads) and Caudata (salamanders and newts) . Exotic amphibian species - Exotic species of amphibians can also be infected, sometimes associated with disease. Species belonging to other taxonomic classes - It is still unclear if hosts from other classes, such as fish, can be infected. Subclinical infections occur.

______Risk analysis CMTV-like virus April 2013 14 Figure 3 – Overview of current knowledge on CMTV-like virus host species

______Risk analysis CMTV-like virus April 2013 15 Theme 4 – Host life stages affected

Theme relevant to pathogen establishment, spread and impact

Definitions - Amphibians have several life stages. After spawning/fertilisation, embryo’s develop into larvae, which in turn develop into juveniles and then (sub)adults. Metamorphosis is the change in the form and in the habits of an animal during normal development after the embryonic stage. It includes the full development from the larval stage up till the moment the juveniles go on land. Species with aquatic hibernation are present year round in water bodies.

Patterns and trends – CMTV-like virus affects both pre and post metamorphic life stages. Life stages involved in mass mortality events - Mass mortality events associated with CMTV-like virus in Spain involved only larvae and juveniles, whereas in the Netherlands adult life stages also succumbed 1, 3, 6, 23 . Life stages in which mortality was observed were in the aquatic phase, and often in or following periods of high visibility (Figure 4). The numbers of animals observed to die ranged from several tens (+ in Figure 4) to hundreds or even thousands (+++ in Figure 4). The larvae observed to die were in different stages of development. So far no spawn has been tested, but spawn death was reported for Darp. Life stages involved in the aftermath of an outbreak – During the monitoring in NP Dwingelderveld in 2011, disease and mortality were observed predominantly in the adult stages of the water frogs (Table 1) 2. Life stages involved in subclinical infections - The sub-clinically infected American bullfrogs in Belgium were larvae 8.

Table 1 – Amphibian host species and their life stages found sick or dead during the pool monitoring in NP Dwingelderveld from May to October 2011. CMTV-like virus was confirmed in many of the water frogs but not in the great crested newts.

Significance and discussion – Host life stages affected in the Netherlands suggestive of immunologically naïve populations - CMTV-like virus outbreaks involve larvae, juveniles and (sub)adults, and possibly spawn too. The fact that life stages are collectively affected, including adult life stages, is suggestive of disease emergence in immunologically naïve populations. For comparison, mass mortality in adults is reported from the United Kingdom (UK) where ranavirus disease is emerging since 1985 26 , while outbreaks with ranaviruses in the United States of America (USA) typically involve larvae and recently metamorphosed juveniles only 27 . CMTV-like infection in spawn has not been demonstrated as yet but cannot be excluded at sites known to harbour the virus, even though there is some evidence that amphibians are least susceptible to ranavirus infection during embryo stage 28 . It is still unknown if vertical transmission occurs with ranaviruses 9, 22 . Aquatic phase - The stages affected are observed in the aquatic phase. This can be observation bias (visibility) or an effect of increased transmission opportunities during the aquatic phase. During the aquatic phase of the life cycle, there is congregation of individuals, likely to increase direct contact transmission opportunities. Further, water is a ranavirus vehicle 16 and it is likely there will be more indirect transmission.

______Risk analysis CMTV-like virus April 2013 16 Figure 4 – Life stages involved in CMTV-like virus associated mass mortality events in the Netherlands plotted against life stage visibility (phenophase). The events were NP Dwingelderveld (NP), IJhorst (IJ), Wijster (W), Fluitenberg (F), Darp (D) and Staphorst (S). Stages observed to die in numbers of hundreds to thousands are indicated by +++, stages observed to die in numbers of tens by +. The phenophases were compiled using the phenological frequency diagrams provided per species in Creemers R & van Delft J, 2009 29 .The blue shades indicate the aquatic phenophase (the darker the shade the greater the relative proportion of sightings), the green the terrestrial phenophase.

______Risk analysis CMTV-like virus April 2013 17 Theme 5 - Distribution of susceptible host species in the Netherlands

Theme relevant to pathogen establishment, spread and impact

Definition – The spatial distribution maps of species of amphibians in the Netherlands are based on the observation of the species in a given square kilometre. A species occurring widespread is often also common, but not necessarily.

Patterns and trends –Distribution and occurrence of native susceptible species (Figure 5) - The water frogs , consisting of edible frogs, marsh frogs and pool frogs, can be detected in nearly the whole of the Netherlands and currently the national distribution is known on 5 x 5 km base 30 . Pool frogs are found in all provinces except Zeeland and Flevoland. The edible frog is the commonest water frog in the Netherlands and can be found in all provinces 31 . Similar to the edible frog, the marsh frog inhabits all provinces in the Netherlands. The common frog and the common toad are also present in all provinces, and are among the most common amphibian species in the Netherlands 32, 33 . The smooth newt is the most common of the four Dutch newt species, ranging throughout almost the entire country 34 . The subspecies of the Alpine newt (Mesotriton alpestris cyreni ) affected in Spain is not present in the Netherlands, but Ichthyosaura (or: Mesotriton ) alpestris is present and inhabits the provinces Limburg, Noord- Brabant, Zeeland, Gelderland and Drenthe. The common spadefoot , considered a threatened species in the Netherlands 35 , occurs in the provinces Drenthe, Overijssel, Gelderland, Noord-Brabant and Limburg. Its range has decreased by 74% since 1950 36 . The natural range of the common midwife toad in the Netherlands is restricted to the southernmost part of the province of Limburg, east of the river Meuse. This species is listed on the Red List as ‘vulnerable’ and its range has decreased with 42% since 1950 37 . Distribution and occurrence of exotic susceptible species - Established populations of American bullfrog are rare and the policy is to eliminate them. The map presented in figure 6 shows the situation up to 2007: one American bullfrog was sighted in 2002 in an Amsterdam garden pond; a few lived in open air terrariums in Limburg, at least up to 2003; one was captured in ‘het Wormdal’, Limburg province in 2006. Since then, there have been new findings. In particular, some American bullfrogs have successfully reproduced in Baarlo, Noord-Brabant province 29 , and there are 6 sites in and around Baarlo where bullfrogs occur not yet figuring on the map. One of these sites was successfully eliminated in 2010 38 . Another specimen was reportedly shot in Sint-Oedenrode, Noord Brabant province, near the valley of the Belgian- Netherlands small river Dommel (2009). The origin of this latter specimen is unknown 39 .

Significance –Common and widely distributed susceptible species - Common native host species susceptible to CMTV-like virus have a nationwide distribution. Endangered species - The endangered common spadefoot occurs within the CMTV-like virus disease hotspot. Potential source - Introduced Alpine newts and common midwife toads occur in the CMTV-like virus disease hotspot.

______Risk analysis CMTV-like virus April 2013 18 Figure 5 - Distribution of susceptible amphibian host species in the Netherland, period 1996 – 2007 (Source: Creemers & van Delft, 2009 29 ). Red dots are natural populations, blue dots are introduced populations, maintaining themselves. Larger dot sizes indicate a higher percentage of occupied squared kilometres. A grey background indicates the presence of elevated sandy soils or the hilly area in South Limburg.

______Risk analysis CMTV-like virus April 2013 19 Theme 6- Distribution of potential hosts across the border

Theme relevant to introduction

Definition - Distribution maps indicate the area of spread of a host species, and do not reflect numbers.

Patterns and trends – Native susceptible species - As in the Netherlands, native species susceptible to CMTV- like virus infection have a widespread distribution in Belgium and Germany. In particular the edible frog, the common frog, the common toad and the smooth newt are distributed more or less throughout Flemish Belgium 40 and Germany 41 . Exotic susceptible species - Further, the exotic American bullfrog occurs in the wild at least in Belgium (Figure 6).

Significance – General - Given the presence of many susceptible host species along the borders, CMTV-like virus is not likely to stop at the border, either way. Specific potential source - The most northern reproducing population of the American bullfrog shown on the map is the population from which the American bullfrog larvae were taken that were infected with CMTV-like virus 8. This population is located close to the Dutch border (< 5 km).

______Risk analysis CMTV-like virus April 2013 20 Figure 6 - Distribution of the American bullfrog Lithobates catesbeianus in Flemish Belgium (1996-2012). In black the IFBL-square-kilometer-blocks where the American bullfrog has been observed. Those encircled by red represent regions where the American bullfrog is reproducing in the wild. IFBL is the mapping grid used by the Instituut voor Floristiek van België en Luxemburg . (Source: Robert Jooris, Hyala, Natuurpunt).

______Risk analysis CMTV-like virus April 2013 21 Theme 7 - Dispersal characteristics of host species

Theme relevant to pathogen introduction and spread

Definition - Biological dispersal refers to species movement away from an existing population or the parent organism.

Patterns and trends Dispersal distance native species - All post metamorphic life stages disperse, in particular juveniles. Pool frogs can be highly mobile. Within a season (March – October/November) they disperse several kilometres 42 . In Austria, they proved to be able to disperse distances of 15 km in ten days 43 . The edible frog colonizes new water bodies easily and in Germany, dispersal distances of approximately 2 km are recorded, mainly of juveniles 41 . Literature states that marsh frogs are relatively sedentary 44 . The animals stay close to the water and rarely move over land 41 . However, they have been able to colonize the province Flevoland within a period of 30 years, which is indicative of high dispersal ability 45 . The greatest dispersal distances recorded for juvenile and adult common frogs are 2 – 4 km 46-49 . During the migration of the common toad from hibernation sites towards reproduction sites, the toads may disperse several kilometres 50 . The smooth newt is reasonably mobile, and colonizes new habitats quickly 51 , these newts may disperse several hundreds of meters 47 .The Alpine newt has an action radius of 400 m.47 The species colonizes new ponds rapidly when these are situated close to already occupied ponds. Both young and adult individuals disperse throughout the reproductive season 52 . This is in contrast to the common spadefoot, which has a limited distribution capacity, with the animals hibernating relatively close to their reproduction sites 47, 53, 54 . Comparable site fidelity is shown by the common midwife toad , where the distance between their summer habitat and reproduction site measures less than 100 m. 47, 55 . Dispersal distance exotic species - Post-metamorphic stages of the American bullfrog are capable of dispersing long distances and are adept at colonizing new sites (>1200 m) 56 . A review of the dispersal distances of amphibians based on mark-and-capture studies and displacement studies quotes three studies in which adult American bullfrogs moved maximum distances of 1600, 914 or 966 m {{}}.

Significance – Potential for dispersal via hosts - Amphibian hosts susceptible to CMTV-like infection can disperse distances up to 15 km, mostly as juveniles or (sub-)adults (Figure 7). The effect of CMTV-like virus disease on dispersal and dispersal distance is unknown. It seems reasonable to assume little effect of CMTV-like virus on the dispersal of sub-clinically infected individuals.

______Risk analysis CMTV-like virus April 2013 22 Figure 7 – Host species classified according to dispersal speed

______Risk analysis CMTV-like virus April 2013 23 Theme 8 - Accompanying amphibian species

Theme relevant to pathogen establishment, spread and impact

Definition - Accompanying amphibian species are those that occur remarkably often with a given other species. Common accompanying species are those often seen in its proximity, characteristic accompanying species are those that have significant overlap in their area of occurrence with it. Characteristic accompanying species cannot be determined for rare species.

Patterns and trends - The Alpine newt is accompanied by all species of Dutch amphibians, some of the characteristic species reflecting its occurrence in poor, humid elevated sandy soils with some forest cover in the south of the Netherlands 52 . The smooth newt, having a large distribution range and broad habitat choice, is accompanied by all species of Dutch amphibians, the most important characteristic ones being the common frog, the common toad, and the three water frog species 34 .The common frog is accompanied by all native amphibian species, especially the smooth newt, the common toad and water frogs 32 . The common toad is accompanied by all species of Dutch amphibians, of which the most important species are the smooth newt, common frog and the three water frog species 33 . Characteristic species accompanying the marsh frog are the edible frog and the natterjack toad ( Epidalea calamita )45 . The pool frog is accompanied by the crested newt ( Triturus cristatus ) and moor frog ( Rana arvalis ), all indicative of good quality waters 30 . The most important common and characteristic accompanying species of the edible frog are the smooth newt, the pool frog, the common frog, and the common toad 31 . The common midwife toad is accompanied commonly by the Alpine newt, the smooth newt, yellow-bellied toad ( Bombina variegata ) and the common frog, as well as by all other widely distributed amphibian species. It is the only species that has the yellow-bellied toad as characteristic accompanying species 37 . Rare susceptible species - The common spadefoot is rare, so characteristic accompanying species cannot be determined. Besides commonly occurring species, common accompanying species are the great crested newt and the moor frog 36 .

Significance: Inter-species transmission - Concurrent CMTV-like virus infection in multiple species is observed in single water bodies in the field, both in Spain 23 and in the Netherlands (Dwingelderveld 2010, Wijster 2011, Fluitenberg 2011, Staphorst 2012)1. It is therefore most likely that inter-species transmission of CMTV-like virus occurs. Favorable conditions - Common susceptible host species are both common and characteristic accompanying species of each other (Figure 8, left side). This suggests frequent occurrence of conditions favorable to inter-species exposure and transmission events. Endangered species – Common susceptible host species are also common accompanying species for endangered species.

______Risk analysis CMTV-like virus April 2013 24 Figure 8 - Amphibian species accompanying known susceptible host species. The percentage represents the probability of that a host species is found in the same square kilometre as the susceptible species (common accompanying species). A green background indicates that the accompanying species is among the five accompanying species having the highest area-of-occurrence overlap with the susceptible species (Source: compiled based on data in Creemers & Van Delft, 2009 29 ).

______Risk analysis CMTV-like virus April 2013 25 Theme 9 - Water body types with CMTV-like virus

Theme relevant to establishment

Definition - A water body is any significant accumulation of water, covering the Earth or another planet.

Patterns and trends: Water body types – The water body types varied (Figure 9). Spain - The sites in Spain where CMTV was detected were a permanent drinking trough (fishing not possible) and a pool , both located at high altitude 23 . NP Dwingelderveld - The 2010 mass mortality occurred in a shallow, 3000 m 3 artificial pond adjacent to the visitor centre of National Park Dwingelderveld (site C) 6. The sites that were monitored in 2011 and where incidental CMTV-like virus cases were detected were pools (e.g., sites A, E and F)2. IJhorst - The mortality in 2011 occurred at a garden pond 1.Wijster - The mortality in 2011 occurred at a garden pond (170 m 2) with water plants 1.Darp - The mortality in 2011 occurred at garden pond , measuring 10 x 10 m.1. Fluitenberg - The mortality in 2011 occurred in a garden pond was constructed in 1991. Staphorst - The site at which the common spadefoots died in 2012 concerns a natural swimming pool with sandy shores, it is approximately 2 – 2.5 m. deep and at the bottom there is some waterweed growing 1. Presence of fish - NP Dwingelderveld - Ten-spined sticklebacks occur in site C. IJhorst – The pond contained ide Leuciscus idus melanotus ( family Cyprinidae). Wijster – The pond had frozen the winter before the 2011 outbreak and all fish had died. The fish had been replaced in the spring by ide and goldfish Carassius auratus auratus (family Cyprinidae). Some juvenile goldfish died at the same time as the CMTV-like virus infected amphibians but the cause of death of the fish was not determined. Darp - The pond contains fish. Fluitenberg - The pond contained ide until 2006. In the period 2007-2012, the pond harboured a healthy school of minnows ( Pimephales promelas auratus, family Cyprinidae ). Minnows are exotic fish species originating from North America.

Significance and discussion – CMTV-like virus infections occur in a variety of water types - The water bodies were pools and ponds, natural or man-made, located at different altitudes and different soil types. All were permanent. The landscape setting found to play a role in the USA (high catchment positions in wetlands57 ) may possibly apply to Spain but not to the Netherlands. Generally in relatively small water bodies with standing water - In general they were small- to medium-sized and not directly connected to other water bodies. They do not to dry up naturally. Fish – Fish species were often present in these water bodies. The fish species could be native or exotic, wild or cultured, and were often from the carp family. In water bodies with standing water, fish—if they are not refractory to infection 58 with CMTV-like virus—are unlikely to be natural dispersal vectors .

______Risk analysis CMTV-like virus April 2013 26 Figure 9 – Water bodies where CMTV-like virus associated mortality occurred.

______Risk analysis CMTV-like virus April 2013 27 Theme 10 - Month of outbreak and water temperature

Theme relevant to establishment

Definition - Water temperature was measured using a HANNA, HI 98311 in 0.1°Celsius (Figure 10a). Mean daily air temperature was obtained from KNMI (Hoogeveen). Mean water temperature per site is the mean of the water temperatures measured during all the 17 monitoring moments (between June and October 2011).

Patterns and trends – Outbreaks occur late spring and in the summer – The initial mass mortality events at all eight (8) known CMTV-like outbreak sites occurred between May and September . At the two sites in Spain, the CMTV outbreaks occurred in August (2008) and September (2007)23 . At the six sites in the Netherlands where CMTV-like virus outbreaks with mass mortality occurred, these outbreaks were detected between May and September: May (2011, IJhorst), July (2011, Wijster; 2012, Staphorst), August (2011, Fluitenberg) and September (NP Dwingelderveld, 2010; Darp, 2011) 1, 6 . Timing mortality in outbreak aftermath - At the water bodies in NP Dwingelderveld that were monitored from June to October 2011, the first confirmed ranavirus disease and mortality cases of the year were detected in June and July (Table 2). Mortality continued to be observed up to September 2. In 2012, dead infected common frogs were reported in NP Dwingelderveld as early as March in 2012, but it is unclear if ranavirus disease contributed to their death (cf. Theme 3) 1. Water temperature and air temperature (physical stressors) – NP Dwingelderveld 2011 - Water temperatures were monitored simultaneously with amphibian populations in NP Dwingelderveld in 2011. The water temperatures measured ranged between 9.9°C (site F on October 24 th ) and 25.4°C (sites C and E on August 5 th ). Water temperature followed the air temperature, water temperature being roughly 3 to 4°C higher than the mean daily air temperature (Figure 10, b and c). The mean water temperature in the month of at which the first sighting of a PCR-positive animal was found, measured on average 19.1 ± 1.5°C. This temperature is not significantly different from the mean water temperature of the four sites (A,C,E and F) during the study period (19.0 ± 3.4°C). The mean daily air temperature during the study period was 15.0 ± 2.6°C. Air temperature IJhorst outbreak May 2011 - The water temperature at the start of the outbreak was not measured, but the ambient air temperature at the day of the first case being 7.5°C. Air temperature NP Dwingelderveld March 2012 - The water temperature was not measured where the common frogs were found, but the ambient air temperature on 27/3/12 was 9.3°C.

Significance and discussion – CMTV-like virus outbreaks tend to occur in the spring and summer months - This is consistent with findings other ranaviruses elsewhere 22, 27 . Host life cycle factors (Th.4) and effects of temperature on virus replication rate (Th.1) and on host immunity (this theme) could contribute to this observation. Relation mortality and (water) temperature, possible effects on host immune function – The relation between water temperature and ranavirus infections in poikilothermic hosts is still not fully understood. In Maine, USA, there was no relation between temperature and onset of ranavirus-associated mortality, but sites that experienced mortality were markedly warmer compared to those with unaffected populations 59 . In Tennessee, USA, there was a significant positive association between water temperature and ranavirus prevalence only in fall, and not in the other three seasons 60 . Experimentally, temperature strongly influenced both time to death and mortality rate in larval Sonoran tiger salamanders ( Amblyoma tigrum ) exposed to via Amblyoma tigrum virus (ATV) 61 . Most salamanders survived when exposed at 26°C, nearly all died 10°C, and all died at 18°C. The experiment showed that the salamanders’ immune function was negatively affected in the lower temperature range 61 . In conclusion, water temperature is a probable determinant in CMTV-like virus-associated mortality in amphibians, but its relation to mortality and its importance relative to other environmental factors is not fully understood.

______Risk analysis CMTV-like virus April 2013 28 Table 2 – Water temperature data obtained in the period June to October 2011

Figure 10 – Left - Water temperature was measured using a HANNA thermometer. Right – Graphs of water temperature measured during monitoring moments in NP Dwingelderveld at Koleveen (Site A, top) and Smitsveen (Site F, bottom) and mean air temperature. All other monitored sites showed the same fluctuations.

______Risk analysis CMTV-like virus April 2013 29 Theme 11 - Water quality and habitat use at outbreak sites

Theme relevant to establishment

Definition – Water quality was analyzed by measuring pH, NO3, NH4, O-PO4, Na, K, Cl, Al, Fe, Zn, t-P, t-S, Mn, Si, Mo, and Ca. Acidity and conductivity were measured in the laboratory on the day following the sampling using a pH electrode with a TIM800 pH meter and an ABU901 Autoburette from Radiometer Copenhagen.

Patterns and trends – Water quality (chemical stressors) - From the two sites in Spain , no information on the water quality is available. Also, no water quality data is available from the NP Dwingelderveld , before or during the 2010 outbreak 6. However, in 2011 and 2012 information was collected, and no significant changes in the chemical composition were present 2. Water quality proved to be sufficient and nutrients resemble values expected for pools (pers. comm. Prof. J. Roelofs) 2. Table 3 gives an overview of all measured variables in water bodies at the monitored locations. The pH ranged between 3.9 – 7.6 (mean ± sd; 5.3 ± 0.8) for the pools and 6.5 – 8.1 for the garden pond (site G). The Staphorst site is a natural swimming pool with sandy shores (Figure 11). The water quality is measured regularly. Oxygen levels, EGV and acidity all gave normal values in 2012 before the outbreak (pers. comm. H. Hop – district water board Groot Salland). Habitat use by cattle – There was no evidence for cattle in the surroundings of the water bodies. Habitat use by humans - Several but not all of the infected sites were garden ponds. Further, t he initial outbreak site in NP Dwingelderveld was used for educating the public on amphibians and excursions with fishing nets were organised for the children. The Staphorst site is used by the public for swimming (Figure 11).

Significance and discussion – No clear evidence for chemical stressors, no cattle presence – In the Netherlands, no strikingly abnormal features were detected with regards the water quality of the water bodies monitored in 2011. There was no indication for high aluminum. The acidity in a number of pools was rather low. Possibly conductivity was slightly low and ammonium slightly high. Elsewhere high aluminum has been associated with sites with ranavirus infection 59 . Low acidity (pH <4.5) increased the likelihood of a site to be unaffected by ranavirus elsewhere 59 . Low conductivity had been the best predictor for ranavirus-affected sites (no ranavirus at sites >60 µS/cm) in a study elsewhere 59 . Further, elsewhere, cattle and nitrogenous fertilizers have been associated with higher prevalence of ranavirus infection and more severe ranavirus-associated disease 62, 63 . Possible explanations put forward were immune-suppression due to elevated un-ionized ammonia concentrations and high nitrate and nitrite levels (water quality) and more transmission opportunities due to less vegetation and thus more clustering of individuals 62, 63 . Sometimes but not always habitat use by humans – Human exploitation occurred at a number of sites but not at all of them. Elsewhere, anthropogenic disturbance of habitat has been associated with higher prevalence of ranavirus infection 64 . One possible explanation is that the host’s acquired immune response was negatively affected, another that there is more risk of pathogen introduction 64 .

______Risk analysis CMTV-like virus April 2013 30 Tables 3 a, b, and c – Water quality characteristics (mean ± standard deviation and ranges) for water bodies in NP Dwingederveld (site A-F) and IJhorst (Site G).Values in µmol/l.

Figure 11 – Was disturbance a stressor at Staphorst?

______Risk analysis CMTV-like virus April 2013 31 Theme 12 - Inventory of trade of amphibians in Belgium

Theme relevant to introduction

Definitions - Farm bred = wildlife farming. Definition of wildlife farming according to CITES: modified form of wild harvest, habitat managed to enhance and maximize natural recruitment for the specific purpose of harvesting the increased production, generally confined geographically and applied to segments of the population occurring on privately-owned lands, monitoring impact of management on wild population(s) not critical.

Patterns and trends – Numbers imported as registered by FAVV - The number of imported amphibians via airports and harbours registered by the Federal Agency for the Safety of the Food Chain (FAVV) increased from 2,612 in 2011 (January – December) to 4,170 in 2012 (January – October). The countries of origin of these animals are summarized in Table 4. Additionally, in 2012, 190 wild caught Kassina and Hyperolius species (family Hyperoliidae ) imported from Burundi were confiscated. Numbers and species imported for research - Three of 40 research labs sometimes performing research on amphibians indicated that they import amphibians. African clawed frogs Xenopus laevis and tropical clawed frogs Silurana tropicalis are the species most frequently used for research in these three laboratories. In 2011, 550 X. laevis and 500 S. tropicalis were imported from the USA for research use. It is not known how these animals were brought into Belgium. The participating research labs did not import amphibians in 2012 due to breeding at their own facilities. Amphibians traded by importing dealers - In Belgium, 124 merchants are registered as allowed to import amphibians. About 4,977 amphibians were offered in 2012 at the Belgian market by the three (3) of the 5 most important amphibian dealers. Table 5 gives an overview of the different species and their numbers.

Table 4 - Numbers of amphibians imported in Belgium via airports and harbours

Significance – The above provides some insight into the volume of the trade in amphibians in Belgium.

*

______Risk analysis CMTV-like virus April 2013 32 Table 5 – Numbers of amphibians in animal shops in Belgium in 2012

______Risk analysis CMTV-like virus April 2013 33 Theme 13 - Inventory of hobby kept amphibians in Belgium

Theme relevant to introduction

In 12 captive collections (including two zoos) in Belgium, who responded to the questionnaire sent out by UGent, 1,385 amphibians are kept. The table provides an overview of amphibian species kept by respondents. This list is heavily biased towards urodelan amphibians. The response of urodelan keepers was much higher compared to that of the anuran keepers. Despite dendrobatid frogs being the most widely kept amphibians in captivity, they are severely underrepresented in the listing.

Table 6 - Captive kept amphibian species in Belgium in 2012.

______Risk analysis CMTV-like virus April 2013 34

______Risk analysis CMTV-like virus April 2013 35 Theme 14 - Legal re-introductions and translocations in the Netherlands

Theme relevant to introduction and spread

Definition – Re-introduction - Re-introduction is an attempt to establish a species in an area which was once part of its historical range, but from which it has been extirpated or become extinct 65 . Translocation - Translocation is the deliberate and mediated movement of wild individuals or populations from one part of their range to another 65 . Amphibian re-introduction/translocation activities require exemptions from the Flora and Fauna Act (FFW), the law that protects native wildlife species in the Netherlands, in order to be legal. These may be exemptions to capture and remove individuals (article 9 of the FFW) or egg masses (article 12) from the wild; to transport and keep eggs and individuals in captivity (article 13); to release them back into nature in the Netherlands (article 14); and to handle individuals at sites during subsequent monitoring of the population.

Patterns and trends – Three amphibian species recently re-introduced or translocated - From the global perspective, the conservation status of the 16 amphibian species present in the Netherlands is of least concern (IUCN; www.iucn.org ). From the national perspective, however, half of these species are near-threatened (1), vulnerable (2), endangered (3) or critically endangered (1) (Dutch Red list, 2007). Legal re-introductions and translocations since the adoption of the FFW ten years ago have concerned the critically endangered species (yellow-bellied toad Bombina variegata, (starting 2004) and two of the endangered species, the common spadefoot Pelobates fuscus and the European tree frog Hyla arborea, (starting 2009). Implementation pattern - Examination of the submitted re-introduction projects for these three species showed that all have an identification phase (Step 0 in Figure) and an implementation phase (Steps 1-4 in Figure). Documents resulting from the identification phase form the basis on which the Dutch Ministry of Economic Affairs (EZ) makes decisions regarding the law exemptions. Typically, partial egg masses or young larvae are either removed from the wild in the Netherlands or obtained from captive adult stock (Step 1). They are then raised in captivity (Step 2) until they are of an age deemed appropriate for release at the re-introduction site (Step 3). The best age for release varies per species and is a trade-off between survival chances (increases with age) and inclination to stay in place (decreases with age for species that are not sedentary). Initially, step 2 was performed at one site in Nijmegen. Recently, step 2 is occurring at several sites, and the tendency is to move away from controlled conditions (plastic basins) to more natural conditions (nets in outdoor ponds). An under-exposed feature of these re-introduction programs is that captive parent stocks are being set-up. Pathogen-screening and hygiene - Recent re-introduction projects include testing for Batrachochytrium dendrobatidis prior to release, and project documents include the ‘Hygiene protocol for handling amphibians in the field’ 66 . However, structural screening for the presence of ranavirus before release into nature is missing.

Significance – Risk per implementation step - Species being re-introduced can be susceptible for CMTV infection. With nationwide exemptions, infected stock may be procured from areas where CMTV occurs. Also, under current exemptions, captive parent stock may be legally imported from other European countries without any disease control. Infection could thus be translocated to the sites where the stock is grown in captivity. The length of the stay in captivity is in favour of CMTV detection, but raising stock in outdoor ponds rather than under controlled conditions makes detection and especially elimination more difficult. If not detected, infection could be translocated to the site of release. Regions at risk - Currently, this would imply introduction of the virus into the provinces of Noord-Brabant and Limburg, also home to the endangered fire salamander Salamandra salamandra, the vulnerable palmate newt Lissotrition helveticus and the vulnerable CMTV susceptible common midwife toad Alytes obstreticans .

Figure 12- Summary of the ten amphibian translocation projects since the ‘Flora and fauna wet’ was adopted.

______Risk analysis CMTV-like virus April 2013 36

______Risk analysis CMTV-like virus April 2013 37 Theme 15 - ‘Unofficial introductions’

Theme relevant to introduction and spread

Definition - Unofficial introductions – By ‘unofficial introductions’ are meant all human activity-mediated release of amphibians into nature, which are not legal re-introductions or translocations.

Patterns and trends - Unfortunately many unofficial introductions have occurred either with amphibians that were collected abroad (mainly France) or from natural Dutch populations. Introductions have occurred with nearly all amphibian species, the focus here is on known susceptible host species (for maps, cf. Th.5). Two less common host species with widespread unofficial introductions - Alpine newt - The Alpine newt has been introduced to many sites in the Netherlands, even as early as the 1900s in Drenthe. Introductions occurred mostly around cities. In Drenthe it was introduced in Rheebruggen in the 70s of the last century, and also around 1995 at the Utrechtse Heuvelrug, the Veluwe and Arnhem 52 . Common midwife toad - The midwife toad only occurs naturally in the southern part of Limburg, but can be found throughout the country, mostly in or in the close vicinity of cities. They include both offspring from specimens originating in France or Limburg 37 . Since the 70s of the last century a population survives in Rheebruggen (Drenthe), and these toads were collected in Northern France 67 . Common widespread host species unofficially introduced on Wadden Sea islands – Water frogs - Unofficial introductions are known from the 1930s in Urk and Texel. In Texel around 1968 approximately 50 frogs were released, this introduction failed. In the eighties another introduction occurred in Texel, and this introduction succeeded. In Ameland in 2003 and Schiermonnikoog in 2009, water frogs popped up as well, all introduced. In the beginning of July 1971 over 10,000 larvae were released in Rotterdam and in November of that year, 2,000 Bulgarian water frogs were released near Rijsoord. It is unknown if these larvae grew successfully and if they mixed with the native frogs 30 . Common toad - The species was unsuccessfully introduced in Texel between 1979 –1985 68 . To our knowledge, there is no more information on introductions of the common toad in the Netherlands. Smooth newt - Even though the smooth newt is widely distributed in the Netherlands, it has been introduced in the first half of the 20th century on all Wadden islands, apart from Texel at which it occurs naturally. It is unknown whether there are more recent introductions 34 . No known unofficial introductions Common spadefoot - No information is known about unofficial introductions of spadefoots in the Netherlands. Exotic species - Amercian bullfrog : cf. Th.5.

Significance - Widespread unofficial introductions Introduced Alpine newts and common midwife toads occur in the CMTV-like virus disease hotspot (cf. Th.5).

______Risk analysis CMTV-like virus April 2013 38 Theme 16 - Flyways

Theme relevant to introduction and spread

Definition - A flyway is the entire range of a migratory bird species (or groups of related species or distinct populations of a single species) through which it moves on an annual basis from the breeding grounds to non- breeding areas, including intermediate resting and feeding places as well as the area within which the birds migrate 69 .

Patterns and trends – There are flyways along the CMTV-like virus axe Spain-Belgium-Netherlands.

Significance – Hypothetical dispersal via birds – It has been hypothesized that birds may act as vectors in the spread of ranavirus infections, but to date there is no scientific evidence for this.

Figure 13 – Eurasian Flyways. In red the south-eastern European flyway, in blue other migration flyways (source: SE European Bird Migration Network, http://www.seen-net.eu/ )

______Risk analysis CMTV-like virus April 2013 39 Part 2 Risk assessment

Method Using the compiled background data and further information obtained as mentioned in Part 1, assessment was made of the possible routes of introduction of CMTV-like virus into the Netherlands, the likelihood of its establishment and spread, and the possible social and ecological impact of CMTV-like virus. Risk was then scored as detailed in next paragraph.

Risk assessment scoring method - Dispersion potential, invasiveness and ecological impacts - The Belgian Invasive Species Environmental Impact Assessment (ISEIA) protocol 70 was used to assess risks associated with dispersion potential, invasiveness and ecological impacts. The risk assessment was carried out by an expert team. This team consisted of six (6) persons: Rob van Leuven, (RU), An Martel (U Gent), Annemarieke Spitzen-van der Sluijs (RAVON), Ronald Zollinger (RAVON), Marja Kik (DWHC) and Jolianne Rijks (DWHC). Each expert completed an assessment form independently, based on the content of a knowledge document. Following this preliminary individual assessment, the entire project team met, elucidated differences in risk scores, discussed diversity of risk scores and interpretations of key information. The results of these discussions were presented in a draft report. Further discussion led to agreement on consensus scores and the level of risks relating to the four sections contained within the ISEIA protocol (Table 7). The ISEIA protocol contains twelve criteria that match the last steps of the invasion process (i.e., the potential for spread, establishment, adverse impacts on native species and ecosystems). These criteria are divided over the following four risk sections: (1) dispersion potential or invasiveness, (2) colonisation of high conservation habitats, (3) adverse impacts on native species, and (4) alteration of ecosystem functions. Section 3 contains sub-sections referring to (i) predation / herbivory, (ii) interference and exploitation competition, (iii) transmission of diseases to native species (parasites, pest organisms or pathogens) and (iv) genetic effects such as hybridisation and introgression with native species. Section 4 contains sub-sections referring to (i) modifications in nutrient cycling or resource pools, (ii) physical modifications to habitats (changes to hydrological regimes, increase in water turbidity, light interception, alteration of river banks, destruction of fish nursery areas, etc.), (iii) modifications to natural successions and (iv) disruption to food-webs, i.e. a modification to lower trophic levels through herbivory or predation (top-down regulation) leading to ecosystem imbalance. Each criterion of the ISEIA protocol was scored. Possible scores are 1 (low risk), 2 (medium risk) or 3 (high risk). Definitions for low, medium and high risk, according to the four sections of the ISEIA protocol are given in table 7. If knowledge obtained from the literature review was insufficient, then the assessment was based on expert judgement and field observation leading to a score of 1 (unlikely) or 2 (likely). If no answer could be given to a particular question (no information) then no score was given (DD - deficient data). Finally, the highest score within each section was used to calculate the total score for the species. Consensus on the risk score of each section was reached using a hierarchical method where evidence from within the Netherlands was given priority over evidence derived from impacts occurring outside the Netherlands. It was also considered that the suitability of habitats in the Netherlands may change due to e.g. water temperature rise due to climate change. Moreover, consideration was given to the future application or non- application of management measures that will affect the invasiveness and impacts in the Netherlands. Subsequently, the Belgian Forum Invasive Species (BFIS) list system for preventive and management actions was used to categorise the species of concern 70 . This list system was designed as a two dimensional ordination (Environmental impact * Invasion stage; Figure 14). It is based on guidelines proposed by the Convention on Biological Diversity (CBD decision VI/7) and the European Union strategy on invasive non-native species. Environmental impact of the species was classified based on the total risk score (global environmental risk) which is converted to a letter / list: score 4-8 (C), 9-10 (B - watch list) and 11-12 (A - black list). This letter is then combined with a number representing invasion stage: absent (0), isolated populations (1), restricted range (2), and widespread (3). Socio-economic and public health impacts - Potential socio-economic and public health impacts did not form a part in the risk analysis according to the ISEIA protocol. However, these potential risks should be considered in an integrated risk analysis. Socio-economic risks were examined as part of the literature study and in discussions with project partners. Socio- economic risks occurring at present or in the future dependent on alterations in habitat suitability and management interventions were considered.

______Risk analysis CMTV-like virus April 2013 40 Table 7 - Definitions of criteria for risk classifications per section used in the ecological risk assessment protocol (Branquart et al. , 2009 70 ).

1. Dispersion potential or invasiveness risk The species does not spread in the environment because of poor dispersal capacities and a Low low reproduction potential. Except when assisted by man, the species doesn’t colonise remote places. Natural dispersal Medium rarely exceeds more than 1 km per year. However, the species can become locally invasive because of a strong reproduction potential. The species is highly fecund, can easily disperse through active or passive means over distances > 1km / year and initiate new populations. Are to be considered here plant species High that take advantage of anemochory, hydrochory and zoochory, insects like Harmonia axyridis or Cemeraria ohridella and all bird species. 2. Colonisation of high conservation habitats risk Population of the non-native species are restricted to man-made habitats (low conservation Low value). Populations of the non-native species are usually confined to habitats with a low or a medium Medium conservation value and may occasionally colonise high conservation habitats. The non-native species often colonises high conservation value habitats (i.e. most of the sites High of a given habitat are likely to be readily colonised by the species when source populations are present in the vicinity) and makes therefore a potential threat for red-listed species. 3. Adverse impacts on native species risk Data from invasion histories suggest that the negative impact on native populations is Low negligible. The non-native is known to cause local changes (<80%) in population abundance, growth or Medium distribution of one or several native species, especially amongst common and ruderal species. The effect is usually considered as reversible. The development of the non-native species often causes local severe (>80%) population declines and the reduction of local species richness. At a regional scale, it can be considered as a factor for precipitating (rare) species decline. Those non-native species form long standing High populations and their impacts on native biodiversity are considered as hardly reversible. Examples: strong interspecific competition in plant communities mediated by allelopathic chemicals, intra-guild predation leading to local extinction of native species, transmission of new lethal diseases to native species. 4. Alteration of ecosystem functions risk Low The impact on ecosystem processes and structures is considered negligible. The impact on ecosystem processes and structures is moderate and considered as easily Medium reversible. The impact on ecosystem processes and structures is strong and difficult to reverse. Examples: alterations of physico-chemical properties of water, facilitation of river bank erosion, prevention High of natural regeneration of trees, destruction of river banks, reed beds and / or fish nursery areas and food web disruption.

Figure 14 - BFIS list system to identify species of most concern for preventive and mitigation action (Branquart et al. , 2009 70 ).

______Risk analysis CMTV-like virus April 2013 41 Results

2.1 – Probability of introduction

Possible recent introduction - CMTV-like virus is already present in a restricted area of the Netherlands (Th.2), where it is associated with disease emergence 1, 6 . The cause of this disease emergence is not fully understood but the most likely possibility is that CMTV-like virus was recently introduced from abroad, either once, or on multiple occasions into the same area. Elsewhere, in the U.K., another ranavirus is causing disease among amphibians since the mid-1980s. Phylogenetic analyses comparing multiple isolates of ranaviruses provides evidence that this virus was probably introduced into the U.K. from the American continent 24 . Potential geographical origin - Other countries where CMTV-like virus has been detected incidentally and locally are Spain and Belgium (map in Th.2)3, 8 . The global distribution of CMTV-like virus is unknown at the moment. Possibly it has a wider occurrence 5, that has not been detected due to under-surveillance 71 . Assuming introduction from abroad, more insight into possible pathways is helpful for tracing origin, for preventing future introductions of CMTV-like virus into susceptible host populations, or both.

2.1.1 Introduction through human activities Among the possible pathways of introduction from abroad (Figure 15), there are those associated with human activities. The focus has been on obtaining 1) data with regards to the legal trade in amphibians through Schiphol airport, 2) data with regards to the legal trade in amphibians in Belgium and from Belgium to the Netherlands, and 3) data on release of imported known susceptible hosts into nature. Entry of CMTV-like virus via fomites into the Netherlands follows the same mechanisms as depicted for spread in §2.3.

2.1.1.1 Legal trade through Schiphol Schiphol airport is one of the two airports in the Netherlands from where live animals are imported from abroad, including from outside the European Union, and the main airport for import of wild animals. The other airport, Maastricht, focuses on transport of equines. Rotterdam harbour is the main port of entry of animal products. ∗ Numbers and species imported via Schiphol airport – An ‘umbrella’ CN-code Worldwide, trade in animals and products is regulated using CN (combined nomenclature; Council Regulation [EEC] No 2658/87 of 23 July 1987 on the tariff and statistical nomenclature and on the Common Customs Tariff). Live amphibians fall under the CN-code 0106 ‘other living animals’ (specifically CN-code 0106 9000 ‘All live animals other than mammals, birds and reptiles that have been named or considered elsewhere. Live frogs, to be kept alive in vivaria, or to be slain for human consumption, fall under this CN code’). This ‘umbrella’ CN code provides the veterinary inspection no insight into whether amphibians being imported with a shipment or not. Only 4 shipments of amphibians were identified via CN-code in TRACES Import of animals and products into the EU are tracked through the Trade Control and Expert System (TRACES) database. Between 2006 and 2012, there have been over 11,000 CN-code 0106 shipments from outside the EU into the EU via the Netherlands. Only four (4) of these 11,000 shipments specified that the animals being shipped were ‘amphibians and reptiles’, providing health certificates to the veterinary inspection. These shipments concerned between 150 and 350 animals, of 1-12 different species of frogs and reptiles, or of unspecified species. All four shipments entered the EU from abroad (Singapore, Suriname) via Schiphol, and had other European countries (Spain, Czech Republic) as final destination. It is quite possible that amphibians with the Netherlands as destination were present in some of the other 10,996 shipments with CN-code 0106. For example, if a shipment contains mostly live insects, it may be registered as ‘insects’, while amphibians travel on the same certificate. A number of CN-code 0106 9000 documents were opened randomly in TRACES, but none had amphibians on them. Search in the TRACES database targeting the more likely importers and countries of origin also did not provide more data . Purpose of trade It follows that no overview of the purposes for which amphibians are imported was obtained. However, frogs are not being imported to harvest frog legs in the Netherlands (Antwoord 31/10/12 op kamervraag, NVWA). Remark - For a subgroup of amphibians traded, namely the endangered species, some information on numbers traded may be obtained from Customs, as Customs do ‘Convention on International Trade in Endangered species of Wild Fauna and Flora (CITES)’ checks. CITES documents accompanying shipments indicate the purpose (e.g., T=for commercial purpose, Z=for zoos) and the source (e.g.,W=wild-caught) of animals transported. CITES export

______Risk analysis CMTV-like virus April 2013 42 Figure 15 – Possible pathways of introduction from abroad

______Risk analysis CMTV-like virus April 2013 43 and re-export permits may be detailed to number and species level. Information may be available at the organization (Ministry of EZ, DR-Loket) that provides the CITES documents (lead not pursued in this study). ∗ Transport conditions and health check – Casualities during transport limited by IATA regulations IATA Live Animal Regulations are set annually by the International Air Transport Association (IATA) and accepted as guidelines in respect of transportation of animals by air by CITES and the OIE. The EU has adopted these regulations as the minimum standard for transporting animals in containers, pens and stalls. IATA Live Animal Regulations have evolved greatly, including container requirements. Airlines such as KLM inform shippers about these requirements and actively coach the shippers throughout the whole process (even visiting the facilities of regular shippers). Death of animals during air transport is very rare. Short transport time Transport duration is generally kept short. For example, considering a trader in Suriname ships amphibians to a trader in the Netherlands, the time frame is likely to resemble the following: a 3-hour drive to airport to arrive 3 – 5 hours before the embarking, say that is at 17.45 in the evening. The flight would arrive around 9 hours later at Schiphol, 08.00 hours local time, undergo its health check at 11.00 and be picked up for delivery in the Netherlands at 12.00, i.e., a delivery from trader to trader within 24 hours. Subclinical cases may go unnoticed. Health check involves visual inspection The health check is performed by the NVWA and consists of verification of the documents, in particular if these are correctly completed, whether the contents of the shipment are conform the certificate (verification often done in collaboration with Customs and the airline), and whether the animals appear healthy. The latter is done by visual health inspection of a sample of the animals. At Schiphol, all boxes of shipments are usually opened for inspection, unless the shipping concerns ornamental fish. In case of ornamental fish, a sample consisting of at least 10% of the shipment and a minimum of 2 boxes is inspected. Importing parties can verify the health conditions required for certification on the NVWA site (Veterinary Import Online [IVO], http://wisdom.vwa.nl/ivo/Start.do). Ranavirus infection and health certification The health certificate used in the context of CN-code 0106 shipments poses no requirements with regards to ‘infection with ranavirus’ for amphibians, therefore specific certification is absent. This is different for EHNV and live ornamental fish (VILS012001A, version 1.0.0). For ornamental fish it must be certified that the animals were in quarantine (Beschikking 2008/946/EG), or that the country of origin of the fish is free of EHN and that EHN is notifiable in that country, that known susceptible species are from areas where EHN does not occur, and that no vaccination against EHNV was provided.

To conclude, it appears that for live amphibian traded, the numbers and species imported from outside the EU and inside the EU through Schiphol can currently hardly be traced. Further, there is no specific control for ranaviruses in amphibians via health certificate requirements. It seems unlikely that infection with ranavirus in amphibians will currently be detected at Schiphol, because traders are unlikely to put individuals from a batch of amphibians clinically infected with ranavirus on transport, transport time is so short cases incubating the disease may not develop into clinical cases within it, and sub-clinical cases will not be detected by visual inspection. Present requirements for documentation are too limited to allow answers to the following question: how many animals in what status of health of what species are imported?

2.1.1.2 Amphibian trade in Belgium Amphibians may also be imported into the Netherlands from European countries by other means than by air. For this reason an inventory of trade was made of trade in amphibians in Belgium and from Belgium to the Netherlands, as well as an inventory of privately kept amphibians and their possible exchange to the Netherlands. ∗ Inventory of trade of amphibians in Belgium - A wide variety of amphibian species is imported, sold en kept in captivity in Belgium. Several thousand imported from outside the EU per year The Federal Agency for the Safety of the Food Chain (FAVV) registered 2,612 amphibians imported into Belgium from outside the EU in 2011 and 4,170 in the first ten months of 2012. They arrived from Africa (Togo, Nigeria), Asia (Indonesia), South America (Peru), Central America (Nicaragua, Panama) and North America (US). Also at least 190 amphibians (Kassina and Hyperolius species imported from Burundi) were confiscated at customs (Th.12). The number of amphibians imported is likely to be greater than these figures. The laboratories imported in 2011 550 African clawed frogs ( Xenopus laevis ) and tropical clawed frogs ( Silurana tropicalis ) from the US for research use. The fact that the laboratories indicate a number of amphibians imported from the US that is greater than that mentioned by the FAVV again highlights the lack of traceability of amphibians with the current international registration system as detailed previously (§ 2.1.1.1). Traded amphibians are often bred but some are wild-caught. Among the amphibians offered by the three (3) Belgian importers in 2012, 52.8% were captive

______Risk analysis CMTV-like virus April 2013 44 bred , 5.4% were farm bred, 10.4% were wild-caught, and 31.4% were of unknown source (Th.12). It is remarkable that a third of the animals were of unknown source, including two species (the Australian green tree frog Litoria caerulea and the Amereega trivittata ) that were imported from or through the Netherlands. Further, at least 58% of the amphibians were captive or farm bred, indicating the importance of captive breeding for the supply of exotic amphibians. The three consulted dealers do not sell imported amphibians directly to animal shops in the Netherlands. ∗ Inventory of hobby kept amphibians – Many more captive kept species than native wild species - The inventory (Th.13). provides insight into the diversity of amphibian species kept in captivity, already over six times the number of native amphibian species in the Netherlands. Captive kept species include native species - Among these are at least three (3) species that occur natively in the Netherlands and are susceptible to CMTV-like virus (the common midwife toad Alytes obstetricans , the smooth newt Lissotriton vulgaris and the common spadefoot toad Pelobates fuscus ). Kept amphibians are often bred but some are wild-caught The numbers presented in the inventory of hobby kept amphibians provide similar insight into the relative importance of sources as the inventory at the traders: 60% were captive-bred, 13% were wild caught and 27% were of unknown source, reinforcing the previous observation that captive breeding is the most common source of amphibians kept in captivity. Ranavirus situation in kept amphibians is virtually unknown Virtually nothing is known about the presence of ranavirus in captive collections in Europe, though CMTV-like virus infection is described in a captive collection in the Netherlands (Th.3) 7 and ranavirus infection was found in red tailed knobby newts ( Tylototriton kweichowensis ) from three hobby collections in Belgium and the Netherlands 72 . In the latter case, the newts had been imported from abroad by a Belgian shop. A minority had shown limited skin ulceration but otherwise they seemed healthy and they had been sold 72 . This example shows how hobbyists in the Netherlands can be supplied by shops outside the Netherlands and demonstrates how ranavirus infection may be missed during import and spread, thus presenting a potential source for infection of wild animals. Deaths among captive amphibians are not often investigated, so the frequency of occurrence of CMTV-like virus infections in kept specimen remains unknown. Exchange of kept amphibians across the border - Exact data regarding exchange of amphibians between Belgium and the Netherlands are not available since registration of animal movements is generally not required. However, keepers of amphibians in both countries have very frequent contacts and country borders do not play any limiting role regarding the exchange of amphibians between keepers of both countries. Based on the questionnaire, an estimated minimum of 50% of the private owners in Belgium exchange animals with keepers in the Netherlands. Captive offspring of Belgian amphibian breeders is widely distributed among Dutch keepers (and vice versa), both for anurans (e.g. during ‘Dendrobatidae Nederland’ meetings) and for urodelans (e.g. during ‘Salamandervereniging’ meetings).

To conclude , the data on amphibian trade in Belgium confirms 1) the difficulty of tracing imported live amphibians and 2) the difficulty of detecting ranavirus infections during the actual import without quarantine, and 3) the frequent import of amphibians from Belgium into the Netherlands. Further, it highlights 4) the diversity of amphibian species being kept— both exotic and native species—, and 5) the fact that the market is largely supplied by these captive bred animals while too little is known about ranavirus infections in captive collections. It is currently not possible to quantify the risk of introducing CMTV-like virus through amphibians imported from Belgium.

2.1.1.3 Entry of CMTV-like virus imported via live amphibians into nature In this paragraph we hypothesize on routes of entry of imported CMTV-like virus imported via live hosts into nature. ∗ Disposal of imported infected amphibians - CMTV-like virus from abroad could enter nature via the (un-) intentional release of an imported infected host (exotic or native species). There is evidence for the release of susceptible host species into nature. For example, of American bullfrogs (Th.3 and Th.5), common midwife toads (imported from France and released in Rheebruggen, which lies within the CMTV disease hotspot; Th.5 and Th.15), and water frogs (imported from Bulgaria; Th.15). But there is no data on the CMTV-like virus status of the source populations or the released specimen. ∗ Via secondary cases - Further, imported infected hosts or their waste products could infect specimen of native susceptible host species that are kept as pet in captivity in the same facility (e.g., Salamandridae , Bufonidae , Genus Pelobates). The latter may be disposed of or even be legally introduced into nature via garden ponds. Evidence for the occurrence of such disposal is the presence of introduced populations of Alpine newts in or close to cities (Th.15). Again there is no data on the CMTV-like virus status of the released specimen.

______Risk analysis CMTV-like virus April 2013 45 ∗ Unintentional transport – Sometimes amphibians are unintentionally transported with other goods, such as vegetables. The route is hard quantify. ∗ Via fomites - Finally, aquarium or vivarium water and sediment may contain infectious virus and be disposed of in nature. In a study in the context of invasive water plants, 2% of the 230 persons with aquaria or garden ponds that responded to a questionnaire confirmed they disposed of water plants in Dutch open waters. Motives given were that it was a pity to throw away living plants, and that it nice to be able to see beautiful plants in open waters. In a study in New Zeeland, at least two of 43 aquarium holders disposed aquarium wastes into out-door ponds or storm-waters drains (risk for entering nature) 73 .

To conclude, there are multiple routes through which human-mediated imported CMTV-like virus could in principle enter nature, but quantification is difficult.

2.1.2. Introduction through nature Introduction of CMTV-like virus from abroad through nature can in theory occur via infected hosts from neighbouring countries that cross the border into the Netherlands through natural dispersal, or via migrating birds.

2.1.2.1 Introduction via infected amphibians crossing borders into NL American bullfrog - CMTV-like virus infection occurs in wild populations of American bullfrogs in Belgium 8, less than 5 km from the Dutch border, and less than 15 km from the centre of Breda (Th.3 and Th.6). American bullfrogs have good dispersal abilities (Th.7)39 . On the Dutch side of the border, in Sint-Oedenrode, Noord Brabant province, an American bullfrog was reportedly destroyed in 2009 (Th.5), but it was not submitted for examination so it could not be determined if its origin was the Belgian wild population. Though there is active removal of exotic American bullfrogs from nature in both countries, other susceptible species sharing their habitat may have been infected by them and not detected to date. Native species - Many of the susceptible host species, that are native, common and widespread species in the Netherlands are also native, common and widespread species in Germany and Belgium (Th.6). However, to date—possibly due to under-surveillance—there have been no reports of CMTV-like virus in native species from these two countries (Th.2).

To conclude , CMTV-like virus has been shown to occur in at least one amphibian species, which is a good disperser, close to the Dutch border (Hoogstraten).

2.1.2.2 Introduction via migrating birds Introduction of CMTV-like virus from abroad could in theory occur via migrating birds (Th.16). Birds have been suggested as vehicle for ranavirus spread, by virus carriage on their feet or bills, or by regurgitation of ingested infected material within a few hours of feeding 74 . The presence of the amphibian chytrid fungus Batrachochytrium dendrobatidis on feet of duck ( Anatidae) has been demonstrated 75 , but to date there is no hard evidence ranavirus could indeed be spread in such a way.

To conclude, the relevance of this route of introduction is unknown.

Overall conclusion pathways of introduction

At the moment it is not possible to draw a quantitative conclusion because precise information on imported amphibian species, numbers and prevalence of CMTV-like virus infection in these is largely lacking. However, with regards to human-mediated import of infected specimen, it is clear that: 1) numerous species from all over the world are traded and kept together; 2) traceability of specimens imported is poor; 3) captive-bred exotic amphibian trade is significant and elusive, at least in terms of diseases occurring; 4) wild caught specimens of unknown health status are regularly introduced into facilities where captive-bred species are kept; 5) disease detection during import is unlikely. With regards to natural dispersal as possible route of introduction, it is noted that CMTV-like virus infection was found to occur in a free-living population of amphibians across the Belgium border and that spread into the Netherlands has to be considered.

To obtain quantitative data on species and numbers imported on short notice would require obtaining data from customs, amphibian dealers and amphibian keepers; to obtain full and continuous insight into this requires a means to trace the amphibians that are imported. To obtain data on the health status of imported specimen ______requires surveillance for CMTV-like infection in the wild abroad______and in captive collections. To obtain data on Risk analysis CMTV-like virus April 2013 46 amphibian disposal frequency would require obtaining data from amphibian keepers.

2.2. – Probability of establishment

CMTV-like virus has persisted locally for at least 2 years - CMTV-like virus is known to have persisted for two years in NP Dwingelderveld 1, 2 . The area where CMTV-like virus is already present has a diameter of approximately 25 km (map in Th.2) 1. Requirements for establishment - This section provides an overview of virus-, host- and environment-related factors, and their interplay, which may influence (further) establishment of CMTV-like virus. CMTV-like virus requires susceptible hosts for replication. This means that to establish itself, the virus needs to have access host populations that belong to susceptible species, that have poor innate immunity to the virus and that have not yet acquired sufficient immunity, or that have enough susceptible individuals in them. Further, the virus needs suitable environmental conditions. Finally, it needs to be able to maintain itself in time.

2.2.1 Susceptible hosts are present 2.2.1.1 Susceptible species Broad host range - CMTV-like virus has a broad host range, infecting at least 9 of the 16 native Dutch amphibian species, and possibly infecting fish (Th.3) 1, 3, 6, 8, 23 . Common species - Water frogs, common frogs, common toads and smooth newts are susceptible to infection and among the most common species in the Netherlands. They have a wide distribution and inhabit a great diversity of habitat types (Th.5). They are each other’s accompanying species (Th.8). There is field evidence for inter-species transmission of CMTV-like virus (Th.8). Differences in species susceptibility to CMTV-like virus are likely but have yet to be determined. Some species may be reservoir hosts, but these still need to be determined.

To conclude , susceptible species occur commonly throughout the Netherlands, suggesting establishment of infection could occur in suitable habitats for amphibians nationwide.

2.2.1.2 Susceptible populations Immunologically naïve populations likely - The CMTV-like virus-associated massive mortality observed at NP Dwingelderveld (2010)6, IJhorst (2011), Wijster (2011), Fluitenberg (2011), Darp (2011) and Staphorst (2012) 1 affected all life stages including adult specimens (Th.4). This suggests that the affected populations were immunologically naïve or at least largely susceptible to CMTV-like virus infection 6. Currently no evidence for co- evolving CMTV-like virus in the Netherlands – Evidence from elsewhere with other ranaviruses shows newly introduced ranavirus strains are generally more pathogenic than endemic strains76, 77 . Virulence appears to be related to the genetic similarity of the introduced strain with coevolved strains and with host specificity 22 . For example, ATV (one of the other ranaviruses) appears to co-evolve with local tiger salamander populations in North America. Disease emerged when novel strains were introduced into populations that had not co-evolved with these strains 77, 78 . There is no evidence for the presence of co-evolving CMTV-like virus in the Netherlands; however, this cannot be excluded for certain due to under-surveillance and under-detection. No evidence for presence of other ranaviruses that could possibly lead to cross-protection - There is some cross-protection between ranaviruses. Antibody raised against the major capsid protein (MCP) of one ranavirus species often cross-reacts with other members of the genus, as MCP is highly conserved 79 . For this reason, CMTV-like virus may have difficulty establishing itself if there are other ranaviruses present which infect the susceptible host species. No studies have been performed to determine the level of acquired immunity against CMTV-like virus in amphibians in the wild in the Netherlands or elsewhere. However, to date there is no evidence for other ranaviruses infecting amphibians in the wild in the Netherlands.

To conclude , to date there no evidence for previous exposure to CMTV-like virus or another ranavirus and therefore Dutch amphibians are considered to be susceptible to infection.

2.2.1.3 Susceptible individuals Annual recruitment in water bodies - Many amphibian species reproduce in pools and ponds 80 . The (sub)adults congregate, and offspring is produced. Based on what is currently known about ranavirus transmission (Th.1), this offspring is likely to be susceptible. In other words, hosts are pooled during the aquatic phase and some to all are likely to be susceptible . Effects of relative host density/life stage –The life stage predominantly present in the water depends on the time of the year (Th.4). The stages affected in the CMTV-like virus associated mass

______Risk analysis CMTV-like virus April 2013 47 Figure 16 - Probability of establishment

______Risk analysis CMTV-like virus April 2013 48 mortality events in the Netherlands were often stages present in peak numbers (relative high density) in the aquatic phase at the time (Th.4, Figure 4). Experimentally, host density was a determinant of ranavirus infection 12 . Adults – Mostly (sub-)adult amphibians return to the same water body each year, a few will colonize new ponds (Th.7). If the (sub-)adults have been decimated in the previous year, and they are not yet replaced by others, this will affect the numbers of offspring produced in that water body. Other factors – The number of susceptible hosts available is influenced by predation and other causes of amphibian death.

To conclude , at the level of water bodies there is annual recruitment of offspring likely to be susceptible to CMTV- like virus infection. Recruitment of susceptible animals is important for a pathogen to maintain itself in a population.

2.2.2 Dutch environment in general appears suitable 2.2.2.1 Temperature Replication - The suitable replication temperature of CMTV-like virus is not known but likely to be within the range of 12°C-32°C as found for other ranaviruses (Th.1). This is consistent with water temperatures at sites where CMTV- like virus disease and mortality have been observed (Th.10). Based on mean daily air temperatures (KNMI, http://www.knmi.nl/klimatologie/grafieken/jaar), such water temperatures are common between May and September in the Netherlands. Survival - On the other hand, survival of the virus outside the host is best at low temperatures (Th.1). Low temperatures occur in winter. Whether survival of CMTV-like virus is long enough to overcome Dutch winters in absence of sub-clinically infected hosts remains to be determined.

To conclude , temperatures likely to be suitable for virus replication occur for at least 6 months a year. In the remaining part of the year, temperatures are favourable for virus survival outside of the host (Th.1 and §2.2.3.1).

2.2.2.2 Suitable sites Water body type - The water bodies involved in the CMTV-like virus outbreaks in Spain and in the Netherlands varied in nature. However, they were generally not connected to other surface water bodies (standing water), permanent, small and rather shallow (Th.9). Such water bodies are suitable breeding sites for many amphibian species. Further, ranavirus dose may be more likely to build up in small still standing water bodies. Landscape, land use, water quality, pesticides and predation – Elsewhere, increased prevalence of ranavirus infection was associated with 1) landscape factors such as high watercatchment areas 57 , 2) land use factors such as grazing by cattle 62, 63 and intensive human exploitation 64 , and 3) water quality parameters, in particular low conductivity; possibly low pH and high aluminum 59 . Further, reduced survival due to experimental ranavirus infection was associated with 4) presence of pesticides or 5) predators or both 81 . Conductivity and pesticides were not measured in the context of CMTV-like virus. Otherwise, apart eventually from human disturbance, the existing data do not suggest that such factors are playing a role in the emergence of CMTV-like virus in the Netherlands (Th.9 and Th.11). In contrast, in terms of perspectives for establishment of the virus, cattle and human disturbance are commonplace in the Netherlands.

To conclude, multiple water bodies in the Netherlands would be suitable for CMTV-like virus infection to occur. There is no clear evidence for enhancement of emergence by landscape, chemical or biological stressors known from the literature, further supporting the concept that CMTV-like virus is likely to be an exotic virus. Some stressors are definitely present in the Netherlands, e.g., cattle and human disturbance, and could enhance establishment

2.2.3 The virus has strategies to maintain itself Ranaviruses can cause severe disease and high mortality, in particular up to 100% of larvae can die, this stage being the most susceptible developmental stage 28, 82 . Pathogens that cause high mortality are at risk of burning themselves out. However, ranaviruses have several mechanisms to be able to persist at infected sites. Two have been previously mentioned, namely the broad host range (§ 2.2.1.1) and co-evolution (§ 2.2.1.2). Further there is the probable lengthy survival in the environment and the possibility of intra-species reservoirs.

2.2.3.1 Probable lengthy survival outside the host Specific data on survival of CMTV-like virus in the environment is lacking. However there is evidence that ranaviruses in general can remain infective for extended periods of time in water, sediment and tissues, in particular when

______Risk analysis CMTV-like virus April 2013 49 temperatures are low, i.e. + 4°C or below zero (Th.1). This suggests long survival periods of ranaviruses in ponds in winter and that amphibians can be (re-)infected when they return to an infected pond after hibernation.

2.2.3.2 Reservoirs Sub-clinical or sub-lethal infections with CMTV-like virus occur 8, but there is no information on duration of infectiousness of such cases. Subclinical 14, 83, 84 or sub-lethal infections 9, 11, 85 have also been described for other ranaviruses, and are considered likely to contribute to maintenance or re-introduction of the ranavirus infection at sites when the next breeding season starts 22, 80 . Lower doses of ranavirus are more likely to lead to sub-lethal infections 86 .

Overall conclusion probability of establishment This means that to maintain itself, it needs to have access host populations that 1) belong to susceptible species, 2) haveThe ecologypoor innate of ranaviruses immunity to involves the virus, a complexand 3) have inter notaction yet acquiredof reservoir sufficient species, immunity, transmission i.e., are routes, naïve, environmental or have enoughpersistence, susceptible stressors, individuals and host in immunitythem 22 . Nevertheless the required susceptible hosts, the suitable environment and mechanisms for persistence are all present in the Netherlands: 1) CMTV-like virus can infect a broad range of amphibian species occurring in the Netherlands, and many of these are common and found nationwide; 2) the populations affected appear immunologically naïve; 3) offspring is likely to be susceptible and recruited annually; 4) temperature appears to be favourable for replication in hosts in the summer and for virus survival in the winter; 5) the type of water bodies affected are commonly found throughout the Netherlands; and 6) the virus probably has multiple mechanisms to maintain itself after initial introduction. Therefore without interfering, the risk of permanent establishment of CMTV-like virus is high (e.g., in Drenthe Province).

2.3 Probability of spread

This section concerns possible routes of spread of CMTV-like virus from areas in the Netherlands where it is established in host species to areas where it is not yet present. Distinction is made between spread through human activities and spread through nature (Figure 17). It applies to the known disease hotspot, but also to possible future or not yet detected sites of establishment.

2.3.1 Spread through human activities 2.3.1.1 Legal re-introductions and translocations of threatened native species Currently three native amphibian host species are involved in re-introduction or translocation projects. These are the common spadefoot, the yellow-bellied toad (Bombina variegata ) and European tree frog (Hyla arborea ) (Th.14). ∗ Common spadefoot - The common spadefoot is susceptible to CMTV-like virus infection and disease 1. For re- introduction purposes, this species is harvested as spawn from the wild. Until 2012, spawn was removed only from sites in Limburg and Gelderland. But in spring 2012, as part of a large-scale attempt to strengthen the remaining small natural populations, it became legal to collect spawn from sites nationwide. An experimental study, with other ranavirus species and other host species, has shown that embryo stages (fertilized spawn) can be infected with ranavirus. However, embryo stages have lower susceptibility to infection than larvae 28 . No spawn was found in 2012 in Staphorst, where CMTV-like virus was detected later in that season. Nonetheless spawn was taken from other sites within the CMTV-like virus disease hotspot, including Valthe where the largest common spadefoot population of the Netherlands is found. Spawn was raised to larvae in basins by Natuurbalans-Limes Divergens BV (in Nijmegen) and by RAVON (at Artis). At both culture sites some mortality occurred during the growth; dead larvae from Nijmegen were submitted for testing to exclude ranavirus infection. The 2012 larvae were released in the provinces of Noord-Brabant and Limburg. As far as is known, CMTV-like virus is currently not present in these provinces. ∗ Yellow-bellied toad - Contrary to the common spadefoot, yellow-bellied toads are not procured from the wild, but captive bred. Risk of infection during captivity by cross contamination or parent stock import cannot be excluded. However no cases of CMTV-like virus infection in this species have been described to date and its susceptibility to CMTV-like virus is unknown.

______Risk analysis CMTV-like virus April 2013 50 Figure 17 - Potential routes of spread

______Risk analysis CMTV-like virus April 2013 51 ∗ European tree-frog - Like the common spadefoot, the European tree frog is procured from the wild, but CMTV- like virus has not been detected in the areas where European tree frog larvae are procured. Risk of infection during captivity by cross contamination cannot be excluded. However no cases of CMTV-like virus infection in this species have been described to date and its susceptibility to CMTV is unknown.

To conclude , there is currently no evidence that CMTV-like virus has been dispersed through legal re-introduction or translocation projects. However, ranavirus infection risk is currently not being taken into account in the re-introduction programs. With nationwide exemptions from the Dutch Flora- en fauna Act (FfWet), stock can be legally procured from areas where CMTV-like virus occurs. Further, under current exemptions, captive parent stock of native species may be legally imported from other European countries without any disease control. Transferring newly procured stock directly to outdoor ponds may be good for animal welfare but will hamper early detection of disease and will make elimination difficult and devastating (previously used individual basins provided a better ‘quarantine’). If not detected, infection can be translocated to the site of release. Currently, this would imply introduction of the virus into the provinces of Noord-Brabant and Limburg. These provinces are home to three other amphibian species that are classified as endangered on the Dutch Red list, one of which is susceptible to CMTV-like virus (the common midwife toad Alytes obstetricans ).

2.3.1.2 Other human-mediated transfer of potentially infected hosts We hypothesize that besides legal re-introduction and translocations, CMTV-like virus could be spread from an infected area to another through transfer of infected (pet-trade) specimens between garden ponds. It seems likely that such transfers occur in the Netherlands on a regular basis, and perhaps also covering greater distances. Further, for example when garden owners who construct a pond may be impatient for amphibians to reach their pond at their own pace, and transport egg masses or vegetation from nearby ponds or ditches. With the current data these routes are hard to quantify.

2.3.1.3 Dispersal by fomites (equipment, transfer of water or sediment) Transmission of ranaviruses through fomites has been reported for fish 14 and amphibians 10 . Ranaviruses can remain infectious for a while in water and sediment. Harp & Petranka 13 showed wood frogs ( Lithobates sylvaticus ) could be infected by exposure to sediment from a site where a ranavirus die-off was occurring. Ranaviruses have thus the potential to be dispersed by fomites, as boots, fishing gear, farm equipment and boats, cf. § 1.1.

To conclude , legal re-introduction or translocation projects present a risk. Other routes of spread are poorly quantified as yet, but represent a certain risk.

2.3.2 Natural spread 2.3.2.1 Dispersal of (sub-)clinically infected hosts Good dispersers among susceptible host species -In the Netherlands, some of the host species susceptible to CMTV-like infection are good dispersers, being both commonly occurring species (§ 1.3) and dispersing hundreds of meters up to 15 kilometres (§ 1.7). These host species are water frogs, common toads and smooth newts. Role (sub)-clinical infected specimens? - Severely diseased individuals may be reluctant to move and not disperse. However, this probably does not apply to individuals that are either sub-clinically infected, or in early stages of disease or recovering. At the sites monitored in Dwingelderveld in 2011, incidental mortality was observed in different life stages while the remaining individuals appeared healthy 2. Also, during the CMTV-like virus outbreak in Staphorst in 2012, healthy looking juvenile common spadefoot toads were moving from water to terrestrial habitat, while other juvenile spadefoot toads at the site were dying. Some of the healthy looking individuals may have been sub-lethally or sub-clinically infected. Fish - In water bodies with standing water, fish—if they can be infected with CMTV-like virus— are unlikely to be dispersers in natural spread .

To conclude , short distance dispersal via (sub-)clinically infected hosts appears to be a very likely route of spread.

2.3.2.2 Birds or other animals 13 - For details see 3.1.2.2.

______Risk analysis CMTV-like virus April 2013 52 Overall conclusion with regards to pathways of spread

The overall conclusion with regards to pathways of spread is that short to long distance dispersal of CMTV-like via human-mediated activities, and especially short distance dispersal via (sub-)clinically infected hosts are very likely, though poorly quantified.

2.4 High risk areas

A high risk area is an area at risk of being occupied by CMTV-like virus (the infection is currently absent but likely to establish itself if introduced). As concluded in 2.2, it is likely that CMTV-like virus can spread over the whole of the Netherlands. This implies the whole of the Netherlands is a high risk area, in habitats where amphibians live.

Zones where threatened species occur are very high risk areas. Based on the current knowledge of ranaviruses and specifically of CMTV-like virus, very high risk areas on short notice are the sites with the spadefoot and common midwife (Th.5). Common spadefoot - The outbreak in Staphorst is only 55 km away (as the crow flies) from Valthe. Here lives the most viable population of spadefoot in the Netherlands of today. From Staphorst it is an even shorter distance to the population around Zwolle and to the populations along the Overijsselse Vecht 87 . In Overijssel several isolated populations occur, all small and vulnerable for stochastic events. This indicates that especially in the Northern provinces the spadefoot is potentially threatened by the arrival of CMTV-like virus. Further, due to the ease by which ranaviruses are likely to be transported (§ 2.1 & 2.3.) over long distances, there is a possibility that more southern populations of spadefoot in Gelderland, Noord-Brabant and Limburg may be wiped out by the introduction of CMTV. Common midwife toad - In the eighties of the last century, it was feared that the common midwife toad would disappear completely from the Netherlands. In order to prevent this over 500 ponds were constructed. Subsequently a conservation plan 88 and an action plan 89 were implemented. The species prospered because of this, but it is still considered ‘vulnerable’ 35 .

Overall conclusion with regards to high risk areas

Given that the broad host range includes common and widespread species, and that types of sites and the environmental conditions at the sites at which currently CMTV-like ranavirus has been detected are widely present in the Netherlands, all Dutch water bodies suitable for amphibians, and their surroundings, are high risk areas.

Very risk areas are areas in provinces where threatened species occur. For the common spadefoot these are located in Drenthe, Overijssel, Noord Brabant, Gelderland, en Limburg province. For the common midwife toad these are located throughout the Netherlands.

______Risk analysis CMTV-like virus April 2013 53 2.5 Impact

2.5.1 Ecological impact Data on the damage occurring in terms of amphibian mortality is largely described in Th.4. Here the focus is on the probable long term impact if CMTV-like virus establishes itself permanently and spreads through high risk areas.

2.5.1.1 Amphibians The probable long- term effects on amphibian population sizes, and the likelihood that species could become extinct, are discussed. ∗ Impact on population size. Other ranaviruses elsewhere – Ranaviruses have the potential to impact on amphibian populations 11 . If ranavirus spreads in wild populations of amphibians it may have severe consequences on the infected populations of amphibians, other susceptible species and their predators 90 . A major cause of mass mortality in the USA and the UK – In the USA, amphibian mortality events that occurred in the period 1996 – 2001 were investigated, and it was found that over half of the mortality events were associated with ranavirus infection and in most of these events ranavirus infection was the sole cause27 . Mortality was very high and recruitment was consistently negligible or zero. Certain sites monitored showed annually recurrent mortality. Despite the impact on numbers at the sites, the events were not associated with host species declines, possibly because all of the events were occurring in fairly widespread and abundant species 27 . In the UK, significant local declines in numbers of adult common frogs have been recorded since the mid-1980s and associated with ranavirus infection 26, 91, 92 . Impact transient, catastrophic or persistent - A study on the long-term impact of ranaviral disease (FV-3) in the UK showed that common frog populations at sites responded in three different manners to the emergence of disease: emergence was transient, catastrophic, or persistent with recurrent mortality events. At sites with transient emergence (16/38, 42%), mortalities were limited to the initial mass mortality. At sites with catastrophic emergence (4/38, 11%), no frogs were seen at the pond since the initial ranavirus-associated mortality. At sites with persistent emergence (18/38, 47%), at least one outbreak was recorded every two years. The populations that experienced these recurring mortality events showed median declines of 81% in the number of adult frogs, whereas comparable uninfected populations showed no change in population size over the same time period 92 . Also, the larger the initial frog population at site experiencing persistent emergence was, the greater was the relative decline 92 . Possible explanatory factors for these three different outcomes include differences in virulence or pathogenicity of circulating virus lineages 11 and differences in the immune response of the host populations (§2.2). In general, when populations in affected ponds rebound, it is not clear whether this reflects the presence of surviving immune animals or re-colonisation by naïve animals 93 . CMTV-like virus in the Netherlands –CMTV-like virus can also cause high mortality. At sites with mass mortality events, the numbers that died were tens to thousands, and these often included adult specimens (Th.4). NP Dwingelderveld –In the aftermath of the first detected outbreak, a number of pools and ponds (Th.2) were visited with regular interval in NP Dwingelderveld, in total on 17 occasions between May and October 2011. Amphibians were searched along the border of the water body, using a net. It was found that 2.1% of all encountered amphibians—corresponding to 6.4% of the encountered water frogs—were sick or dead specimen. Many of the dead specimens were confirmed to be infected and diseased with CMTV-like virus (Th.4, Table 1). In 2012, in the context of this study, the same ponds in NP Dwingelderveld were visited twice (24/8/12 and 4/9/12). Comparison with data obtained in 2011 was possible for sites A, C, D, E and F (Table 8). The pH at the sites ranged from 4.65 to 6.77 (mean 5.49 ± 0.37). The closest comparable date with pH measurement in 2011 was 17/8/2011; then it ranged from 4.05 to 5.94 (mean 4.96 ± 0.37). On both days in all ponds amphibians were encountered. The species encountered were mainly water frogs, but also crested newts, smooth newts and one common frog and one moor frog. Numbers encountered and proportions found sick or dead were comparable to 2011 (Table 8). IJhorst – The IJhorst was also visited on 24/8/12 and 4/9/12. During the two visits to the garden pond in IJhorst, no amphibians were observed. The owner said that in 2012 no amphibians were seen at all. Interpretation – Interpretation is seriously hampered by under-surveillance in all years but 2011. Taking this into account, and the fact that information on amphibian population size in NP Dwingelderveld prior to the outbreak is anecdotal, NP Dwingelderveld data could indicate persistent emergence. And the IJhorst may be an example of catastrophic emergence.

______Risk analysis CMTV-like virus April 2013 54 Table 8 – Amphibian host species found sick or dead during monitoring at pools in NP Dwingelderveld on comparable dates in 2011 and 2012

∗ Host species extinction risk– Density of hosts plays a role in ranavirus infection (§2.2.1.3) 12 . Mathematical models and epidemiological theory predict that pathogens do not drive their hosts to extinction when transmission is density dependent. The host may suffer from local extinction, but overall the pathogen is lost before the host species goes extinct 94 . However, this does not apply 1) when the pathogen has a multiple hosts some of which can act as reservoir hosts and in which the pathogen can remain viable, 2) when the pathogen alters population fitness, i.e., pathogen-driven host selection as may occur in sites with recurrent mortality events, and 3) when anthropogenic translocation events cause repeated introductions into populations where the pathogen would normally fade out off 94 . These three conditions are met for CMTV-like virus, and therefore in particular host species classified as vulnerable may be at risk of extinction. Common spadefoot - The spadefoot population in Staphorst that was infected with CMTV is a relatively large population. In Staphorst in June 2012 hundreds of dead or dying spadefoot larvae and smooth newts were found. Sick or dead adult spadefoot toads were not present, which is not odd since this life stage only occurs in the water during a few weeks in spring. The impact of the mortalities may jeopardize the sustainable survival of this population, because in the Netherlands the species occurs in only 38 areas, and in most of these there is only a single reproduction pond available. The Staphorst location was visited again in September and no amphibians were seen.

To conclude, the impact on common amphibian populations is likely to vary between localized extinction, persistence or recovery. Small populations of endangered species, such as the common spadefoot, may be prone to (local) extinction.

2.5.1.2 Effects on biodiversity Amphibians are an important component of the aquatic as well as terrestrial food web. Their loss can influence predatory birds, macro-fauna, fish and water quality. The latter is because tadpoles and larvae are herbivorous. Therefore (local) removal of amphibians in an ecosystem is likely to have severe implications for the stability of this ecosystem. Effects of population declines or extinction of rare species on the local ecosystem – Literature on the effects of amphibian population declines or extinction of rare species on the structure and function of ecosystems is uncommon. One study removed tadpoles of two rare species from ponds and examined the effect of species loss on the local ecosystem. This study found that while amphibians can influence food web dynamics, the influence of the removal on system processes in temporary forest ponds was limited 95 . Another study was a review that focused on neotropical highland streams, where impacts will likely be greatest. It concluded that there was evidence for ecological effects of catastrophic amphibian declines: “Amphibian declines will have large-scale and lasting ecosystem-level effects, including changes in algal community structure and primary production, altered organic matter dynamics, changes in other consumers such as aquatic insects and riparian predators, and reduced energy transfers between streams and riparian habitats ’96 . The authors also indicated that effects of loss of one species of amphibian causes effects in aquatic as well as in terrestrial ecosystems due to habitat and functional differences between larvae and adults in most amphibians species 96 . Fish – The risk of CMTV-like virus infection is

______Risk analysis CMTV-like virus April 2013 55 not limited to amphibians. CMTV-like virus possibly infected fish at the site of the first outbreak (Th.3.). Elsewhere, an outbreak involving another ranavirus provided clear evidence that it was affecting both fish and amphibian hosts in that water body 97 . Also, experimentally, other ranaviruses isolated from frogs elsewhere were able to cause infection in fish 14, 98 .

To conclude, the impact of large scale and continuing mortalities on ecosystem stability is likely but still uncertain, as no studies on this topic have been conducted. The potential infectivity of CMTV-like virus for fish needs to be determined.

2.5.2 Socio-economic impact 2.5.2.1 Current CMTV-like virus outbreak The socio-economic impact of the emerging CMTV-like virus in its area of occurrence was not assessed systematically. However, the outbreaks had direct consequences at the sites used for human recreation (Th.11). Educational activities for children were suspended at the pond at the visitor centre in NP Dwingelderveld. At Staphorst the public was distressed by the sight of dead toads and worried if there were any implications for the health of pet animals and themselves. Physical measures are taken to try to prevent reoccurrence. Research, in particular surveillance activities and diagnostic tests, have also already cost several hundred thousand Euros.

2.5.2.2 Other Scientific literature on the financial costs of the emergence of a ranavirus is lacking. Literature data on the societal costs of the emergence of a ranavirus are also lacking. At the emotional level - The idea that the human use of amphibian habitat, not only nature reserves, but also ditches and garden ponds, could extirpate species, is a thought that is likely to cause much distress among people with nature affection. This number is most likely very high considering the number of people with membership of nature conservations agencies. Mosquitoes - In wetlands it was shown that mosquitoes avoided laying eggs in habitats which contained larval salamanders and tadpoles. In addition, the survival in the presence of salamanders was low 99 . Hence, the suppression of amphibian populations by disease most likely affects mosquito populations, which are inconvenient to people (and possibly vectors of disease).

Overall conclusion with regards to impact

A negative effect of emerging CMTV-like virus on amphibian numbers is seen at outbreak sites. Elsewhere, the effect of emerging ranaviruses on amphibian populations has been transient, catastrophic, or persistent with recurrent mortality events. Follow-up at some infected sites suggests that this may also be the case for populations infected by CMTV-like virus. However, due to under-surveillance, no firm conclusions can be drawn.

With regards to risk of extinction of host species, a number of characteristics of CMTV-like virus suggest that it may cause endangered species to go (locally) extinct. Other effects on biodiversity are very poorly documented.

Other amphibian ranaviruses elsewhere are able to infect fish. It is still unclear CMTV-like virus does. Given the economic value of fish, more insight into potential consequences for fish species present in the Netherlands is desirable.

______Risk analysis CMTV-like virus April 2013 56 2.6 Risk classification using the ISEIA protocol

2.6.1. Expert consensus scores The risk classifications attributed to the CMTV-like virus for criteria of the ISEIA protocol were determined based on an evaluation of all available information in this report by a multidisciplinary panel of experts. They were all medium or high (Table 9). The total risk score attributed to this species was 10 out of a maximum risk score of 12. This score results in an overall classification to category B, indicating a moderate environmental risk on the basis of current knowledge.

Table 9 – Risk classifications attributed to CMTV-like virus

ISEIA section Consensus score Risk classification Dispersion potential or invasiveness 3 High risk

Colonization of high conservation value habitats 2 Medium risk

Adverse impacts on native species 3 High risk

Alteration of ecosystem functions 2 Medium risk Global environmental risk 10 B-list category

2.6.2 Dispersion potential or invasiveness Classification – High risk . Currently, the CMTV-like virus appears to be locally present in a restricted range and is already locally invasive in the Netherlands (Th.2 and §2.2). Ranaviruses show a high replication ability after infection of host cells (Th.1). Clinically infected amphibians were found in urbanized areas, rural areas as well as remote areas (Th.9). It is probable that sub-lethal and sub-clinical infections occur (§2.2.3.2). Introduction to remote areas will most likely be restricted to human-assisted spread or passive dispersion via infested amphibians and/or eventually via other animals (§2.1). This includes spread via waterfowl or other animals carrying mud from one water body to another or predators of infected amphibians that vomit elsewhere their stomach content. However, documented evidence for dispersal of CMTV-like virus and other ranaviruses elsewhere via other animals than amphibians is still lacking (§2.1.2.2 and §2.3.2.2).

There is multiple evidence that the CMTV-like virus can easily disperse through the above mentioned means over distances > 1 km per year. Firstly, clinical (and so possibly sub-clinical) infection is recorded in half of the native amphibian species and several susceptible species are known to disperse > 1 km per year and even up to 15 km per year (Th.3 and Th.7). In addition, available data over the year 2011 show that the CMTV-like virus disease reports were clustered close to the location of first record in 2010, NP Dwingelderveld, and that in 2012 the new site Staphorst was located at a distance of 7 km south of the nearest site known to have infected amphibians in 2011, namely IJhorst (Th.2). This spread of the virus may be caused by passive dispersion via natural means and/or multiple human assisted introductions. Retrospective proving of causality between certain vectors and the emergence of this viral disease in this area is no longer possible because empirical data are lacking.

2.6.3 Colonisation of high conservation value habitats Classification – Medium risk . This criterion addresses the potential of the CMTV-like virus to spread and to cause viral diseases in habitats with high conservation values (irrespective of its dispersal capacities). According to the ISEIA protocol, habitats with a high conservation value are those where disturbance by man is minimal, thus allowing specific natural communities and threatened native species to occur, such as listed in the European Union Habitats Directive.

The available data for the Netherlands show that clinical (and possibly sub-clinical) infections of amphibians with CMTV-like virus have been recorded in several water bodies in one area with high conservation values: i.e., NP Dwingelderveld. This nature reserve contributes to the Dutch ecological main structure (EHS) and is designated as Natura 2000 area according to the European Habitats Directive and Birds Directive 100 . Relevant habitat types for international classification of this area are acid moorland pools (H3160), peat bogs (H7110B and H7120) and water

______Risk analysis CMTV-like virus April 2013 57 bodies with pioneer vegetation (H7150). These habitat types are very important for biodiversity conservation, including several threatened amphibian species. For the great crested newt ( Triturus cristatus, H1166), an Annex II protected species of the European Union Habitats Directive, there are species conservation goals explicitly mentioned for this Natura 2000 area 101 .

In addition, spread and emerging of disease has also been recorded at several sites with a low to medium conservation value (Th.2). Although these water bodies lack a formal conservation status, some of them largely contribute to habitat for endangered amphibian species in the Netherlands, such as the common spadefoot.

So, it can be concluded that the current spread and adverse impacts of the CMTV-like virus are confined to sites with a low or medium conservation value and are recorded occasionally in high conservation value habitats, resulting in a medium risk score. It should be explicitly noted that available data on current spread and effects of this virus in the Netherlands allow multiple interpretations. We interpreted viral infections of amphibians in several water bodies within one Natura 2000 area as occasionally occurring and not as often occurring in high conservation value habitats.

2.6.4 Adverse impacts on native species Classification – High risk . Several risk criteria of the ISEIA protocol related to this section are not straightforward applicable in risk assessments of exotic viruses. This section mainly addresses the potential of the CMTV-like virus to cause effects on amphibians.

The CMTV-like virus shows a large host range, among which several common amphibian species and some endangered species, such as the common spadefoot toad and common midwife toad. Mortality at infected sites suggests that the endangered great crested newt may also be a host for the virus, but this is still unconfirmed. In the Netherlands, clinical infections and mass mortality are documented for all life stages of amphibian species. Effects on species of other taxonomical groups, such as reptiles and fish, cannot be excluded based on evidence from abroad (§2.5.1).

The potential risk of regional extinction of endangered amphibian species in Netherlands is relatively high due to small and isolated populations of such species that are sensitive to the CMTV-like virus. Some common amphibian species with high dispersal ability are also sensitive to the virus and sub-clinical infected individuals could easily spread this virus.

A study on the long-term impact of ranaviral disease in the United Kingdom showed that populations of the common frog showed different responses to introduction of a closely related virus (FV-3-like virus). The disease was transient, catastrophic or persistent with recurrent mortality events. Declines of more than 80% in the number of adult frogs and reduction of genetic variability were reported at sites where disease persisted 92, 102 . There was no sign of recovery of the common frog population over the ensuing five years. Other affected species included common toads, midwife toads and natterjack ( Bufo calamita ). In contrast, pool frogs were scarcely affected.

Taking into account the high negative impact of the CMTV-like virus on amphibian species observed at several locations in the Netherlands as well as severe effects of this virus and closely related viruses on amphibian populations in other countries in temperate regions, potential risks causing severe population declines of amphibians and the reduction of local species richness is therefore highly likely in the Netherlands. At a regional scale, the virus can be considered as a factor precipitating endangered amphibian species decline, resulting in a high risk classification in this category: adverse impact on native species.

2.6.5 Alteration of ecosystem functions Classification – Likely, medium risk . Risk classification within the section alteration of ecosystem functions should be based on four sub-criteria: 1) Modification of nutrient cycling or resource pools; 2) Physical modifications of habitat; 3) Modification to natural succession; and 4) Disruption to food webs. Viruses can not directly alter these ecosystem functions, but may cause indirect effects in case of strong declines of their host populations. This impact category is only poorly documented, therefore according to the ISEIA protocol, an adapted risk scoring system based on expert judgment has been applied. Available literature and data on effects of CMTV-like or related viruses focuses only on documenting and identifying causes of disease, mass mortality or population declines of amphibians.

______Risk analysis CMTV-like virus April 2013 58 Documented evidence for indirect alteration of ecosystem functioning observed within the Netherlands, countries with a similar climate (e.g., Belgium, United Kingdom and Denmark) or other climate regions (e.g., Spain) is lacking. The scoring system is adapted to scores 1 = unlikely (low risk) or 2 = likely (medium risk).

Amphibians are known to play a number of important roles in functioning of ecosystems 96 . Evidence for alteration of ecosystem functions presented here originates from studies quantifying impacts of strong amphibian population declines by other types of environmental deterioration. Ongoing research in Central America, where declines of amphibian populations are occurring in streams, suggests that these losses will have wide-ranging consequences for ecosystem functioning. Such effects will most likely be large-scale and lasting including changes in algal community structure and primary production, altered organic matter dynamics, changes in other consumers such as aquatic insects and riparian predators, and reduced energy transfers between streams and riparian habitats 96 .

Larval mosquitoes were less abundant in pools with higher densities of larval salamanders in studied ponds in the USA 99 . In addition, experiments on mosquito oviposition and survival showed that mosquitoes avoided ovipositing in habitats containing larval salamanders and tadpoles and had low survival in the presence of salamanders. These data indicate that predation by larval salamanders may influence the breeding distribution of mosquitoes by imposing selective pressure on ovipositing adults. It was concluded that measures to protect amphibians may contribute to controlling mosquito production in wetlands, potentially minimizing disease risk to humans 99 .

It was considered that similar impacts on ecosystem functioning as described above could also occur in the Netherlands, because the CMTV-like virus is also expected to cause strong declines of local populations (>80%) of several sensitive amphibian species (Section 2.6.4). Therefore, expert judgement resulted in a risk score of 2: likely for this impact parameter.

2.6.6 Risk classification Classification – B2, watch list . According to the Belgian Forum on Invasive Species (BFIS) list system, the risk classification of a non-native species corresponds to the consensus score (ISEIA, 2009; Table 8) combined with the current distribution within the country in question.

The risk classification of the CMTV-like virus for the current situation in the Netherlands is B2. This indicates a non- native species with a restricted distribution range and moderate environmental hazard (i.e., ecological risk) that should be placed on a watch list (Figure 18).

Figure 18 - CMTV-like virus risk classification for the current situation, according to the BFIS list system.

______Risk analysis CMTV-like virus April 2013 59 A re-grading of this risk classification (Table 10) will be required in case of future situations where no management interventions follow to prevent introductions of CMTV-like virus in the Netherlands and secondary spread within the country. Moreover, climate change may cause habitat alterations.

In case of absence of management interventions, the CMTV-like virus is expected to distribute wider. Spread and emerging of disease in other areas with high conservation values (Natura 2000; EHS: Dutch ecological main structure) can potentially occur if the virus was accessed by multiple vectors. Such vectors include visitors of nature areas with mud on foot wear or sampling equipment, legal or illegal (re-)introductions of amphibians, dumping of terrarium animals, mud on water fowl and natural dispersal of clinical and subclinical infested amphibians.

According to climate change scenarios developed by the Dutch Royal Meteorological Institute the average air temperature in the Netherlands potentially will increase during winter by 0.9-2.3 oC and 1.8-4.6 oC over the years 1990-2050 and 1990-2100 103 . The average summer temperature may increase over these periods with 0.9-2.8 oC and 1.7-5.6 oC, respectively. As air and water temperature of shallow water bodies are strongly correlated (Th.10), temperature of breeding and nursery sites of amphibians are also expected to increase over the coming decades during winter as well as summer periods. This may affect multiple virus, host and environment-related factors and their interactions. Based on the contemporary body of knowledge it is not possible to predict whether these climate- related changes will enhance or impede establishment. Suitable replication temperatures for other ranaviruses lie in between 12-32°C (Th. 1). Hypothetically, water temperatures, shown to be a few degrees higher than ambient air temperatures (Th.10), may be warm enough for virus replication earlier and later in the season, while not becoming too high. On the other hand, higher temperatures during winter and more freezing and thawing events may have a negative effect on virus survival outside of the host (Th.1; §2.2). Scenarios could be examined using epidemiological models. For example temperature optimal for CMTV-like virus replication is unknown. Also, it is not determined how important virus survival in the environment in winter is for persistent emergence compared to for instance sub- clinically infected hosts (§2.2).

The adverse effects of the CMTV-like virus on native amphibian species in the current situation already resulted in a high risk classification. A wider distribution of this virus in the future scenario without management intervention will likely result in a further increase of adverse impacts on sensitive species at a larger spatial scale and can be considered as a factor precipitating regional species decline. Scientific documentation of alterations of ecosystem functions by declines of amphibian populations is not available for the Netherlands and countries in similar or other climate zones, leading to assessment of this risk criterion by expert judgement (§ 2.6.5). Alterations of ecosystem functions in a future situation are judged to be likely at a regional scale due to the increase of adverse impacts on populations of sensitive amphibian species.

Table 10 - Theoretical classification CMTV-like virus according to potential future situation without management measures. ISEIA section Consensus score Risk classification Dispersion potential or invasiveness 3 High risk Colonization of high conservation value habitats 3 High risk Adverse impacts on native species 3 High risk Alteration of ecosystem functions 2 Likely, medium risk Global environmental risk 11 A-list category

Therefore, the future scenario shows a potential increase in the invasion of the CMTV-like virus resulting in a widespread distribution and viral disease of amphibians in a significantly increasing number of high conservation value habitats, resulting in highly adverse direct and indirect impacts on biodiversity and alterations of ecosystem functions at a wider spatial scale. In particular, the expected increase in number of high conservation value habitats with adverse effects would lead to an increase in the total risk score by one point according to the ISEIA protocol (Table 10). Combined with the widespread distribution this will result in a reclassification to a higher environmental risk (A3 - Black list classification).

______Risk analysis CMTV-like virus April 2013 60 Part 3 Risk management and communication

Major assumptions underlie the risk management and communication proposals. The most important assumption is that CMTV-like virus does not occur widespread in the Netherlands as of now. This is based on the focal distribution of the 2011 outbreaks, for which no particular triggering factors could be found. We also assume that CMTV-like virus in the Netherlands is a single virus, and there are no other ranaviruses. More surveillance including at non-outbreak sites and virus sequencing would confirm or infirm these assumptions.

Most likely the economic costs of restoring a disrupted ecosystem are higher than the costs of preventive or curative measures.

3.1. Prevention of introduction

Measures to prevent the introduction of CMTV-like virus from abroad into areas in the Netherlands that are presumed naïve were elaborated based on risk assessment results (§2.1). Note that many measures will be useful also with regards to import risk of other pathogens 106 . Priority, feasibility and an indication of costs of measures were based on expert judgment (Table 11).

3.1.1 Make sure that CMTV-like virus can be detected in imported captive specimens at the border

3.1.1.1 Obtain sufficient sight on amphibians imported Currently there is no good sight on numbers, species, origin and destination of amphibians imported: a) Quantitative data on amphibian trade (legal or illegal) is lacking. b) Traceability of imported specimens is poor. c) There is significant captive-breeding of species exotic to the Netherlands for trade or sales within Europe and elsewhere. This study made clear that to prevent the introduction of CMTV-like virus via traded amphibians and possibly other poikilothermic hosts, more information is needed on (illegal) trade, the pet industry and international exchange of animals among hobbyists. ∗ Obtain good real-time sight on amphibians imported from outside the EU into Europe and the Netherlands in particular (numbers, species, origin, destination). ∗ Obtain better sight on amphibians imported from within the E.U. into the Netherlands via animal dealers and via hobbyists (volume, species, origin, destination).

3.1.1.2 Improve the capacity to detect CMTV-like virus in imported specimens at the border Currently health status monitoring opportunities of imported and traded specimens are poor: a) Disease detection during import is unlikely. b) Wild caught specimens of unknown health status are regularly introduced into facilities where captive-bred species are kept. c) Captive-bred exotic amphibian trade is elusive in terms of diseases occurring in the facilities concerned. This requires improved capacity for detecting CMTV-like virus at the border. Wildlife trade is a multibillion dollar industry 107 and stakeholders should be involved in the process in order to ensure effectiveness of chosen preventive measures 107 . ∗ Increase the knowledge and awareness about host species that can be infected with CMTV-like virus - Imported specimens of species belonging to the amphibian families Alytidae, Bufonidae, Dendrobatidae, Pelobatidae, Ranidae, Salamandridae should be considered as possible hosts of CMTV-like virus 80, 109 . The situation regarding fish is at the moment unclear (Th.3). It is very likely that amphibian species from other families are also susceptible, because the virus effectively caused disease in nearly all the amphibian species found in the NP-Dwingelderveld and the surrounding area. The virus was also capable of infecting captive Dendrobatidae , which are frogs native to a completely different continent, South America. Therefore this knowledge needs to be updated continuously. ∗ Increase the knowledge and awareness about the range of occurrence of CMTV-like virus – The global distribution of CMTV-like virus is unknown. Therefore, additional data is required to assess whether CMTV- like virus is a European virus or originates from some other continent (Th.2). International exchange of information

______Risk analysis CMTV-like virus April 2013 61 on ranavirus disease should help to answer this question and help to identify areas which may present a risk that the virus is introduced together with imported animals. Such information should be made widely available. ∗ Increase the capacity to detect and confirm cases at import - Diagnosis is usually based on a number of findings (see textbox). Suspect clinical cases can be detected by veterinary authorities, traders and hobbyists. Information on clinical signs and possibilities for laboratory testing including postmortem analysis should be given. Development of a rapid detection test for use in the field would be beneficial. Further education about biosecurity measures, proper disposal of dead animals and disinfection of the environment should be made readily available. ∗ Quarantine - Quarantine periods required to detect ranavirus may differ per species but should cover the length of incubation period 105 . The incubation period of CMTV-like virus is unknown but other ranaviruses have incubation periods that can be as long as a month 86, 105 . To be effective the quarantine period has therefore to be quite long, which has the cost consequences. Water temperature during quarantine would need to be defined, because it affects the duration of the incubation period and clinical manifestation 104 . An alternative to quarantine is the requirement to restrict the procurement of specimens from certified CMTV-like virus–free collections. There exist no certification schemes for amphibians, therefore generally accepted principles should be followed and adapted accordingly).

Diagnosis of CMTV-like virus

1. Clinical signs – Possible clinical signs include hemorrhages, edema, erythema, erratic swimming

behavior in larvae, lethargy, and sudden death 1, 3, 6, 23 . These together with the numbers of dead animals and the involvement of multiple life stages in outbreaks are suggestive but not pathognomonic of CMTV-like infection 1. In other words, such clinical signs can also be seen in other diseases and are not specific enough for diagnosis.

2. Post-mortem examination and histopathology – Microscopy includes the examination of formalin- fixed, paraffin-embedded and H&E stained tissue sections. Lesions include foci of necrosis in liver, kidney, spleen, pancreas, gastro-intestinal tract or skin or several of these, associated with round intracytoplasmic basophilic inclusions 3, 6, 23 . Histopathological examination in combination with laboratory tests allows distinction between infected specimens and specimens contaminated by the environment.

3. Laboratory tests - Methods used in diagnosis of CMTV-like infection have been virus isolation in cell culture and PCR with sequencing of products 3, 6, 7, 25 . Organs of choice in dead animals are liver, spleen and kidney. PCR tests on tissues provide information on the presence of genetic material of the virus in organs. In live animals, tail clips and swabs collected from the cloaca or oral cavity can be taken. Swabs will only be positive if the virus is present in mucosal epithelium, saliva, faeces or urine. Real-time PCR was observed to be more sensitive than conventional PCR 110 . Cytopathogenic effects in cell culture can be used to confirm the virus is infective. Iridoviral particles can be detected by electron microscopy ( Th.1) 3, 6 . Immunohistochemistry has been performed, and is useful for understanding pathogenesis 23 . For other ranaviruses enzyme-linked immunosorbent assays (ELISA)

have been described, in particular antigen-capture ELISA for EHNV 104 . Sensitivity and specificity of tests are to be taken into account in interpretation. Note: There are official reference laboratories for fish ranaviruses.

4. Limitations – Detection level - In particular in sub-lethally infected specimens, virus levels may be below detectable levels by virus isolation, an observation made for ATV in tiger salamanders 9. PRC, generally considered very sensitive, was not as sensitive as virus isolation in one ranavirus study. In vivo examination - PCR tests on tail clips (ante-mortem) underestimate the true prevalence of ATV infection in the first days of infection, but increase with days after exposure 111 .

______Risk analysis CMTV-like virus April 2013 62 3.1.2 Reduce the risk that CMTV-like virus enters nature via imported kept amphibians or their waste

3.1.2.1 Obtain good sight on amphibian and vivarium waste disposal behavior Currently there is lack of data on amphibian and vivarium waste disposal behavior. Behavior of the general public that increases the likelihood of the disease being introduced into nature must be identified. This is required to determine the most risky behavior, and so to provide relevant practical advice.

3.1.2.2. Make sure that people are aware of risky behavior and know how to minimize it It is important that people are aware that the virus can be introduced into nature through the release or disposal of (sub-)clinically infected amphibians or their waste material, and that they are provided with information on how to minimize this risk. ∗ Provide education for the public ∗ Identify possible technical solutions for minimizing infection via contaminated water and other vivarium waste products

3.1.3 Make sure that CMTV-like virus imported via natural dispersal can be detected early on at its site of introduction, before it occurs widespread.

3.1.3.1 Obtain a better understanding of the relative importance of introduction via natural dispersal Prevention of introduction of CMTV-like virus from abroad through natural routes may be difficult, if amphibians, birds and predators effectively and regularly introduce it. To date, no information indicates that this is the case. However this situation may evolve. ∗ Consolidate international cooperation within Europe with regards to ranavirus infections - for updates on CMTV-like virus occurrence, and ranavirus surveillance in general. ∗ Further investigate the role of (migrating) birds and predators in ranavirus transmission - for a better understanding of long distance introduction through natural routes.

3.1.3.2. Ensure monitoring for early detection and source tracing The aim is to detect introduction through nature at an early stage, so that there may be a possibility to eliminate the virus before it becomes widespread. ∗ Ensure monitoring of unusual mortality in amphibians in the Netherlands - The aim is to detect introduction through nature at an early stage, so that there may be a possibility to eliminate the virus before it becomes widespread. Source tracing at sites of new introduction will improve introduction prevention measures. ∗ Establish monitoring for CMTV-like virus near Hoogstraten - There is a known local risk near the Belgian border. The CMTV-like virus is found there in the American bullfrog, an exotic species, for which the Dutch policy is removal from nature. The risk for the Netherlands is due to the likelihood of inter-species transmission, with major risk for threatened species located in the south of the Netherlands. Monitoring of CMTV-like infection in amphibians at the Dutch border is relevant. In addition, there is a need for close international cooperation to fully understand the issue and assess its relative importance.

______Risk analysis CMTV-like virus April 2013 63 Table 11 – Measures to prevent the introduction of CMTV-like virus from abroad into Dutch nature. Priority, feasibility and costs are based on expert judgment only.

3.1 Options for prevention Priority Feasibility Costs 3.1.1 Make sure that CMTV-like virus can be detected in imported captive specimens at the border 3.1.1.1 Obtain sufficient sight on amphibians imported * Obtain good real-time sight on amphibians imported into the High Fair. Likely to be Low, once put in place Netherlands via animal dealers (numbers, species, origin, destination) long-term

* Obtain better sight on amphibians imported into the Netherlands via High Fair Moderate hobbyists (volume, species, origin, destination)

3.1.1.2 Improve the capacity to detect CMTV-like virus in imported specimens at the border * Increase the knowledge and awareness about host species that can be High Good Low, basis laid in this infected with CMTV-like virus report, regular updating needed * Increase the knowledge and awareness about the range of occurrence Moderate Fair, if international Moderate of CMTV-like virus cooperation and funding is achieved * Increase the capacity to detect and confirm cases at import Very high Good Low-high

* Quarantine High Moderate Quarantine costs. Pass on to consumer?

3.1.2 Reduce the risk that CMTV-like virus enters nature via in imported captive amphibians or their waste

3.1.2.1 Obtain good sight on amphibian and vivarium waste disposal High Good Moderate behavior

3.1.2.2 Make sure that people are aware of risky behavior, and know how to minimize this * Provide education to the public High Moderate, needs to Moderate, be continuous continuous * Identify the possible technical solutions for minimizing infection via High Good Pass on to consumer contaminated water and other vivarium waste products 3.1.3 Make sure that CMTV-like virus imported via natural dispersal can be detected early on at its site of introduction, before it occurs widespread. 3.1.3.1 Obtain a better understanding of the relative importance of introduction via natural dispersal

* Consolidate international cooperation within Europe with regards to High Good Low ranavirus infections * Further investigate the role of (migrating) birds and predators in Low Good Moderate ranavirus transmission

3.1.3.2 Ensure monitoring for early detection and source tracing * Ensure monitoring of unusual mortality in amphibians in the Very high Good Moderate Netherlands

* Establish monitoring for CMTV-like virus near Hoogstraten Very high Good Moderate

______Risk analysis CMTV-like virus April 2013 64 3.2 Elimination

This section deals with the current knowledge on the possibilities to eliminate CMTV-like virus infection from (small) populations, i.e., measures to prevent establishment of the virus. These possibilities were elaborated based on risk assessment results (§2.2) and literature. Priority, feasibility and an indication of costs of measures were based on expert judgment (Table 12).

3.2.1 Assisted elimination – currently only applies to captive populations and small free-living populations Assisted elimination is elimination that is achieved by measures undertaken by humans. Distinction is made between captive populations (such as stock of endangered species raised for re-introduction into the wild) and free-living populations.

3.2.1.1 Encourage destocking and disinfection for elimination of infection in captive settings In captive populations infected with CMTV-like virus the aim is to eliminate infection, to avoid establishment and spread. The most secure option is destocking and disinfection of premises, similar to what has been recommended for the elimination of other ranaviruses in captive facilities 105 . The rationale behind destocking is to eliminate infected specimens, including those sub-clinically infected, and the rationale behind disinfection is to inactivate virus present in the environment of these infected hosts (see §2.2.3.). For disinfectants used on contaminated equipment and containers, see section §3.3. ∗ Ensure destocking is done humanely, timely and thoroughly . Good information and conditions for humane destruction need to be available. Euthanized stock should be properly disposed of (e.g., burning, or buried and limed 105 ). Further, current knowledge suggests apparently healthy specimens could be sub-clinically infected and therefore, until further notice, destocking of these too is recommended.

3.2.1.2 Explore options for elimination in the case of (small) infected free-living populations Acceptable elimination measures for use in water bodies are virtually unidentified. If the infected population is small, elimination of CMTV-like infection may be preferable over control. Elimination could be considered for infected populations in small man-made water bodies such as garden ponds, or local populations at sites where CMTV-like virus has been newly introduced and rapidly detected, before becoming widespread (cf. 3.1.3.2.). * Ensure new areas to which the virus has spread are detected as early as possible – cf. 3.1.3. * Provide help with the decision to opt for elimination or not in natural settings, and advice for best practice - The decision to opt for elimination or control needs to be timely and informed. Guidelines could contain criteria to aid decision-making and practical advice for best practice. * Make sure the steps to be taken to obtain permission to destock are well-known to the public and permissions are given in a timely fashion - Permission to destock is likely to be necessary, and the process should be outlined and facilitated for timely action. * Investigate the technical alternatives for achieving virus elimination from natural sites and their side-effects; Encourage registration of disinfectants for inactivation of ranavirus in the environment at Ctgb - Elimination of CMTV-like virus in free-living populations is difficult. Measures as destocking, disinfection and physical alteration of sites have consequences for other flora and fauna, and are not always effective. The problem with destocking in natural settings is that amphibians are mobile and full destocking may not be achieved. The problem with disinfectants in natural settings is that they do not generally work in presence of organic material, such a pond mud. Further, none has been registered for such use at Ctgb (‘College voor toelating gewasbeschermingsmiddelen en biociden’), and it is very unlikely companies will invest in registration because there is no profit to be made. For other ranaviruses, there is some experimental evidence that virus is inactivated when water bodies dry up 10 . There may also be other physicochemical ways than desiccation to inactivate virus (UV, ozone, cf. Th.1) that possibly lead to elimination of virus from the site. Measures applicable to wild settings may come from measures developed for the amphibian trade 107 . * Make sure that effect of water body management interventions on the occurrence of CMTV-like virus is properly monitored - Experience with measures that could eliminate CMTV-like virus from free-living populations is very limited:

______Risk analysis CMTV-like virus April 2013 65 - To the best of our knowledge, destocking and disinfection were put to practice only once with the aim of eliminating CMTV-like virus from a site with a free-living population. This was at the first site where CMTV-like virus was detected in Spain (2007). Here tadpoles were euthanized, and the site was disinfected with sodium hypochlorite (bleach, cf. § 3.3.). The site was a permanent drinking trough, so it was small and man-made (cf. photo in Th.9). CMTV-like virus infection has not occurred at this site again (Balseiro A., pers. comm. 2012). However, the next year (2008), infected amphibians were found in a natural pool in close proximity to the first site. It is unclear if this was due to spread from the first site prior to elimination efforts, or introduction from a different source. This second site could not be disinfected (cf. photo in Th.9). The site was monitored for occurrence of CMTV-like virus infections in amphibians in the subsequent years but the infection has not been detected again (Balseiro A., pers. comm. 2012), indicating the virus did not establish itself in the second site. Taken together, the usefulness of the destocking and disinfection at the first site remains unknown. - In Belgium, there were efforts to remove American bullfrogs from nature because they are invasive exotic species. The efforts were not specifically directed towards eliminating the CMTV-like virus infected populations and the effect of American bullfrog removal on the local occurrence of CMTV-like virus infection was not monitored. Therefore it is unknown whether the efforts to remove of American bullfrogs from the environment were effective for clearing CMTV-like infection from the area or not. - In the Netherlands two water bodies underwent physical modifications for other reasons. The pond in Wijster (outbreak in 2011) underwent complete renovation in the spring 2012 because it leaked, and no dead amphibians were reported in the summer of 2012. At Staphorst (outbreak in 2012), soil of the pond was dug out and overturned and some of it was moved a short distance in the autumn of 2012. This is done every 3 to 5 years because it is a swimming pond. The effect on the occurrence of CMTV-like virus infection at Staphorst is still unknown. Because there is so little experience with eliminating CMTV-like virus from natural sites, effects of such interventions on the presence of CMTV-like virus should be monitored, even if they not primarily undertaken for elimination.

3.2.2 Natural elimination–learn from it for larger scale assisted elimination Natural elimination is elimination that is achieved by natural processes in free-living populations. The water bodies used by amphibians are often not connected to other surface water bodies, and in the case of another ranavirus, FV3, emergence of infection was transient in 42% of the ponds investigated 92 . Understanding more of the mechanisms and the factors involved in natural elimination of infection from sites may provide information on how to assist elimination of infection in large areas and populations. * Establish long-term monitoring for understanding the factors that contribute to transient emergence at sites - CMTV-like virus is (at least) present in the Netherlands in an area of 25 km diameter (ca. 500 km2), and highly likely to establish itself and spread without intervention (§2.6.). The area is too large to consider elimination on short term, but—assuming transient emergence occurs at some sites—it does provide an opportunity to learn about natural elimination processes. A better understanding of mechanisms and factors causing CMTV-like virus disappearance rather than maintenance could be gained by long term monitoring of outbreak sites and their surroundings. More insight into the role of different amphibian species and the role of other classes (fish, reptile, invertebrate) in the epidemiology would be gained. * Translate to acceptable eradication measures / environmentally friendly biosecurity measures for use in water bodies.

______Risk analysis CMTV-like virus April 2013 66 Table 12 – Options and recommendations for elimination of CMTV-like virus infection in captive and free-living populations. Priority, feasibility and costs are based on expert judgment only.

3.2 Options and recommendations for elimination Priority Feasibility Costs 3.2.1 Assisted elimination - currently only applies to captive populations and small free-living populations 3.2.1.1 Encourage destocking and disinfection for elimination of infection in captive settings. * Ensure destruction and disinfection can be done humanely, timely and Very high Good Low (excluding the thouroughly. It is important that persons confronted with CMTV-like virus possible value of infections in captive settings have ready access to guidelines for humane destocked destocking and good disinfection. animals)

3.2.1.2 Explore options for elimination in the case of (small) infected free-living populations and monitor the effect of interventions. * Ensure new areas to which the virus has spread are detected as early cf. 3.1.3 as possible

* Provide help with the decision to opt for elimination or not in natural High Moderate Low settings, and advice for best practice.

* Make sure the steps to be taken to obtain permission to destock are High Good Low well-known to public and speedy.

* Investigate the technical alternatives for achieving virus elimination from Medium If economic. Low to high. natural sites and their side-effects. Industry?

* Encourage registration of desinfectants for inactivation of ranavirus in High Low Moderate, for the environment at Ctgb. industry * Make sure that effect of water body management interventions on the Very high Fair Low occurrence of CMTV-like virus is properly monitored. Even if such measures are not primarily designed to eliminate CMTV-like virus, they may provide valuable leads for elimination.

3.2.2 Natural elimination - learn from it for larger scale assisted elimination

* Establish long-term monitoring for understanding the factors that High Long-term Moderate contribute to transient emergence at sites.

* Translate to acceptable eradication measures / environmentally friendly Medium If economic. Low to high. biosecurity measures for use in water bodies. Industry?

______Risk analysis CMTV-like virus April 2013 67 3.3 Control

This section deals with the current knowledge on the possibilities to control CMTV-like virus infection, i.e., measures to prevent spread to new areas and diminish the impact on amphibian populations. They were elaborated based on risk assessment results (§2.3-§2.5) and literature. Priority, feasibility and an indication of costs of measures were based on expert judgment (Table 13).

3.3.1 Prevent human mediated spread of virus within the Netherlands Human mediated spread of virus within the Netherlands is highly likely without intervention. This can be by translocation infected animals or through fomites such as landing nets.

3.3.1.1 Prevent spread to new areas via re-introduction projects Officially authorized re-introduction and translocation activities concern threatened amphibian species. These re- introduction activities do not yet fully take into account the occurrence of CMTV-like virus infection in the Netherlands and elsewhere in Europe, while the impact of an introduction of CMTV-like virus via such activities into the destination site could undo the intended effect of these activities, or worse. Re-introduction projects are implemented in steps, and disease risk can be minimized in each step. For example, procuring stock from wild populations in the zone known to harbor CMTV-like virus infections in amphibians is best avoided. And quarantine is advisable for new stock (cf. 3.1.1.2.), as is the use of separate equipment for each tank. Biologists performing re-introduction and translocation activities should obtain veterinary advice for disease prevention during the entire project and take action accordingly. If infection occurs despite the preventive measures, elimination of the infected population is advisable (cf. 3.2.1.1).

3.3.1.2 Prevent inadvertent spread to new sites by public or field biologists ∗ Increase awareness of the public (see 3.1.2.2.) - The public is often not aware of the risk of spreading infection by moving (sub-clinically) infected specimens, and needs to be educated with regards to this (see 3.1.2.2.; Figure 19 left). This is particularly relevant at sites known to be infected. ∗ Increase awareness of field biologists - Field biologists are an important group to target when it comes to the risk of dispersal by fomites 112 . ∗ Ensure disinfecting in the field is made practical - Disinfection protocols for ranaviruses are based mostly on the work of Langdon (1989) 14 , Miocevic et al. (1993) 20 and Bryan et al. (2009) 21 . Based on these Phillot et al .(2010) 112 provided an overview of disinfectants to use for ranavirus disinfection (Table 14). Information on hygiene protocols is already provided by RAVON, but field application can be improved. It would be preferable that all products recommended are officially registered for such use at the Ctgb (cf. also 3.2.1.2; Figure 19, right). For compliance to protocols and effectiveness, the application should be practical.

3.3.2 Explore the possibility to take advantage of natural barriers to limit natural dispersal Based on current knowledge, dispersal by (sub-clinically) infected amphibians is highly likely. Further,dispersal by birds or other animals (vectors) may be possible. * Obtain a better understanding of the relative importance of the species in natural dispersal - Which species are important herein and at which rate? (cf. 3.1.3.1.) * Identify possible natural barriers - Absolute containment of wild animals in an area is not possible. Also, current conservation efforts are directed towards linking populations of wild animals to improve their fitness. On the other hand, the fact that effort is put into infrastructure to link populations also means there are barriers in our landscapes and land uses that can limit natural dispersal by amphibians. Temporary use could be made of them. * Explore the benefits and downsides of containment

3.3.3 Try to avoid further impact on threatened species and high conservation value habitats All habitats where amphibians live in the Netherlands are at risk of being infected, including high conservation value habitats (§2.4.). The virus represents a threat for host species on the Dutch Red list. Vaccination of threatened

______Risk analysis CMTV-like virus April 2013 68 species is not an option at this stage. There are no vaccines available for CMTV-like virus or other ranaviruses 104, 105 , and chemotherapy is not available 104 . * * Prevent spread to areas where threatened species are found (see 3.1, 3.3.1 and 3.3.2.) - Particular focus should be on preventing sites where threatened species are found from becoming infected (e.g., the South of the Netherlands). * Gain additional information on the risk for threatened species that occur in the current area of occurrence - Additional information should be gained on the impact of CMTV-like virus on threatened species. Not only for species already known to be susceptible, but also for other threatened poikilothermic species present in the area of occurrence of CMTV-like virus, such as the amphibian predating grass snake Natrix natrix .

3.3.4 Make sure the correct data is collected to be able to predict the effectiveness of control and elimination measures Epidemiological models are very useful to properly assess the effectiveness of control and elimination measures. The quantitative parameters for the disease modelling lack partially. The information in this document can be used to define which data is missing to make such models, so that the collection of such data could then be a priority if the control and elimination are undertaken.

Figure 19 – Left , Poster board warning the public not to remove amphibians from the infected site C in NP Dwingelderveld. Right - Disinfection of wetsuit and equipment.

______Risk analysis CMTV-like virus April 2013 69 Table 13 – Control options, measures to prevent spread to new areas and diminish the impact on (threatened) amphibian populations. Priority, feasibility and costs are based on expert judgment only.

3.3 Options and recommendations for control Priority Feasibility Costs 3.3.1 Prevent human mediated spread of virus within the Netherlands 3.3.1.1 Prevent spread to new areas via re-introduction projects High Good Low 3.2.1.2 Prevent inadvertent spread to new sites by public or field biologists * Increase awareness of the public (see 3.1.2.2) Very high Moderate, needs to Moderate, be continuous continuous * Increase awareness of field biologists Very high Moderate, requires Low, continuous change in behaviour * Ensure disinfecting in the field is made practical Very high Good Moderate

3.3.2 Explore the possibility to take advantage of natural barriers to limit natural dispersal

* Obtain a better understanding of the relative importance of the species Low Fair Moderate in natural dispersal * Identify possible natural barriers Low Fair Moderate

* Explore the benefits and downsides of containment Low Fair Moderate

3.3.3 Try to avoid further impact on threatened species and high conservation value habitats

* Prevent spread to areas where threatened species are found see 3.1, 3.3.1 and 3.3.2 * Gain additional information on the risk for threatened species that occur High Fair Moderate in the current area of occurrence 3.3.4 Make sure the correct data is collected to be able to predict the Very high Good Low effectiveness of control and elimination measures

Table 14 – Overview by Phillot et al.(2010) 110 of disinfectants to use for inactivate ranavirus.

______Risk analysis CMTV-like virus April 2013 70 Glossary

Biosecurity measures a set of preventive measures designed to reduce the risk of transmission of infectious disease.

CMTV-like virus common midwife toad virus (CMTV)3 and viruses with partial major capsid genes showing 99.8% to 100 % sequence homology (GenBank accession no. FM213466.1) 25 .

Co-evolution the evolution of two or more interdependent species, each adapting to changes in the other.

Cross-protection the protection against infection by a pathogen given to the host by its prior infection with a related pathogen

Density-dependent the likelihood of transmission increases with increasing host density.

Hazard a biological, chemical or physical agent in, or a condition of, an animal or animal product with the potential to cause an adverse health effect 113 .

Hazard identification the process of identifying the pathogenic agents which could potentially be introduced in the commodity considered for importation 113 .

Infectivity the ability to produce infection

Pathogen an agent of disease (microorganism capable of causing disease)

Pathogenicity the ability of a pathogen to cause disease in a host organism

Pathognomonic a specifically distinctive or characteristic of a disease or pathologic condition; denoting a sign or symptom on which a diagnosis can be made.

Phenology the study of recurring phenomena (patterns of animal activity), such as animal migration, especially as influenced by climatic conditions.

Ranavirus a virus belonging to the Ranavirus genus

Risk the likelihood of the occurrence and the likely magnitude of the biological and economic consequences of an adverse event or effect to animal or human health 113 .

Risk analysis the process composed of hazard identification, risk assessment, risk management and risk communication 113 .

Risk assessment the evaluation of the likelihood and the biological, economic and public health consequences of entry, establishment and spread of a hazard within the territory of an importing country 113 .

Risk communication the interactive transmission and exchange of information and opinions throughout the risk analysis process concerning risk, risk-related factors and risk perceptions among risk assessors, risk managers, risk communicators, the general public and other interested parties 113 .

Virulence the ability (of a pathogen) to overcome bodily defenses (of a host); the ability to cause rapid, severe and destructive disease. In the narrow sense: the increased mortality rate in infected individuals 86

______Risk analysis CMTV-like virus April 2013 71 List of abbreviations

ATV Ambystoma tigrum virus

BIV Bohle irido virus bp base pairs

BFIS Belgian Forum on Invasive species

CMTV Common midwife toad virus

Ctgb College voor toelating gewasbeschermingsmiddelen en biociden

DWHC Dutch Wildlife Health Centre

ECV European catfish virus

EHS Ecological Main Structure

EHNV Epizootic haematopoietic necrosis virus

EU European Union

FAVV Federal Agency for the Safety of the Food Chain (Belgium)

FfWet Flora- and fauna Act

FV3 FV3

IATA International Aviation Transport Association

ISEIA Invasive Species Environmental Impact Assessment

NVWA Nederlandse Voedsel en Waren Autoriteit

NP National park

OIE World Organization for Animal Health

RAVON Reptile, Amphibian and Fish Conservation Netherlands

RU Radboud University Nijmegen

SCRV Santee-Cooper ranavirus

Th. Theme

______Risk analysis CMTV-like virus April 2013 72 Acknowledgements

We would like to thank the NVWA for financial support of this study and Trix Rietveld for critical comments on an earlier version of the report. Many persons contributed to this study. In particular we would like to acknowledge Lenie Algra-Verkerk, Fons van Asten, Ana Balsiero, Roschong Boonyarittichaikij, Wilbert Bosman, Przemyslaw Busse, Raymond Creemers, Ben Crombaghs, Jeroen van Delft, Anjolieke Dertien, Rinus Gerlofsma, Xavier Harduin, Jelger Herder, Robert Jooris, Marc Schils, Ger van der Sluijs, Marisca Stege, Raymond Tilburg, Matt Whiles, Peter Wohlsein, Maarten Zeylmans van Emmichoven. Last but not least we would like to acknowledge the owners or the managers of the sites where outbreaks occurred in the Netherlands, and Dick Willems, without whom CMTV-like virus may have gone undetected for a long time.

Photos Title page – Joran Janse Electron microscopy picture – Peter Wohlsein Water bodies Spain – Ana Balsiero Water body and disturbance Staphorst - Marc Schils Thermometer, poster board – Jolianne Rijks Disinfection of wetsuits and equipment – Annemarieke Spitzen-van-der-Sluijs All other pictures - Jelger Herder, www.digitalnature.org

References

1. Rijks, J. M. et al . Ranavirus disease emerging in wild amphibians in the Netherlands. (in prep.).

2. Spitzen-van der Sluijs, A. S., Kik, M., Rijks, J. M., Martel, A. & Gröne, A. The development of a ranavirus outbreak in a Dutch heathland, a case study. (in prep.).

3. Balseiro, A. et al . Pathology, isolation and molecular characterisation of a ranavirus from the common midwife toad Alytes obstetricans on the Iberian Peninsula. Dis. Aquat. Organ. 84 , 95-104 (2009).

4. Jancovich, J. K., Chinchar, V. G., Hyatt, A., Miyazaki, T. & Zhang, Q. Y. in Virus Taxonomy: Ninth Report of the International Committee on Taxonomy of Viruses. (eds King, A. M. Q., Adams, M. J., Carstens, E. B. & Lefkowitz, E.) 193-210 (Elsevier Academic Press, 2012).

5. Mavian, C. et al . The genome sequence of the emerging common midwife toad virus identifies an evolutionary intermediate within ranaviruses. J. Virol. 86 , 3617-3625 (2012).

6. Kik, M. et al . Ranavirus-associated mass mortality in wild amphibians, the Netherlands, 2010: a first report. Vet. J. 190 , 284-286 (2011).

7. Kik, M., Stege, M., Boonyarittichaikij, R. & van Asten, A. Concurrent ranavirus and Batrachochytrium dendrobatidis infection in captive frogs (Phyllobates and Dendrobates species), The Netherlands, 2012: A first report. Vet. J. 194 , 247-249 (2012).

8. Sharifian-Fard, M. et al . Ranavirosis in invasive bullfrogs, Belgium. Emerg. Infect. Dis. 17 , 2371-2372 (2011).

9. Brunner, J. L., Schock, D. M., Davidson, E. W. & Collins, J. P. Intraspecific reservoirs: Complex life history and the persistence of a lethal ranavirus. Ecology 85 , 560-566 (2004).

10. Brunner, J. L., Schock, D. M. & Collins, J. P. Transmission dynamics of the amphibian ranavirus Ambystoma tigrinum virus. Dis. Aquat. Organ. 77 , 87-95 (2007).

______Risk analysis CMTV-like virus April 2013 73 11. Cunningham, A. A., Hyatt, A. D., Russell, P. & Bennett, P. M. Emerging epidemic diseases of frogs in Britain are dependent on the source of ranavirus agent and the route of exposure. Epidemiol. Infect. 135 , 1200-1212 (2007).

12. Greer, A. L., Briggs, C. J. & Collins, J. P. Testing a key assumption of host-pathogen theory: density and disease transmission. Oikos 117 (11) , 1667-1673 (2008).

13. Harp, E. M. & Petranka, J. W. Ranavirus in wood frogs ( Rana sylvatica ): potential sources of transmission within and between ponds. J. Wildl. Dis. 42 , 307-318 (2006).

14. Langdon, J. S. Experimental transmission and pathogenicity of epizootic haemotopoietic necrosis virus (EHNV) in redfin perch, Perca fluviatilis L., and 11 other teleosts. Journal of Fish Diseases 12 , 295--310 (1989).

15. Pearman, P. B., Garner, T. W. J., Straub, M. & Greber, U. F. Response of the Italian agile frog ( Rana latastei ) to a ranavirus, frog virus 3: a model for viral emergence in naïve populations. Journal of wildlife diseases 40 , 660-669 (2004).

16. Robert, J., George, E., De Jesus Andino, F. & Chen, G. Waterborne infectivity of the Ranavirus frog virus 3 in Xenopus laevis . Virology 417 , 410-417 (2011).

17. Duffus, A. L., Pauli, B. D., Wozney, K., Brunetti, C. R. & Berrill, M. Frog virus 3-like infections in aquatic amphibian communities. J. Wildl. Dis. 44 , 109-120 (2008).

18. Chinchar, V. G. Ranaviruses (family Iridoviridae): emerging cold-blooded killers. Arch. Virol. 147 , 447-470 (2002).

19. Nazir, J., Spengler, M. & Marschang, R. E. Environmental persistence of amphibian and reptilian ranaviruses. Dis. Aquat. Organ. 98 , 177-184 (2012).

20. Miocevic, I., Smith, J., Owens, L. & Speare, R. Ultraviolet sterilisation of model viruses important to finfish aquaculture in Australia. Aust. Vet. J. 70 , 25-27 (1993).

21. Bryan, L. K., Baldwin, C. A., Gray, M. J. & Miller, D. L. Efficacy of select disinfectants at inactivating Ranavirus. Dis. Aquat. Org. 84 , 89-94 (2009).

22. Gray, M. J., Miller, D. L. & Hoverman, J. T. Ecology and pathology of amphibian ranaviruses. Dis. Aquat. Organ. 87 , 243-266 (2009).

23. Balseiro, A. et al . Outbreak of common midwife toad virus in alpine newts ( Mesotriton alpestris cyreni ) and common midwife toads ( Alytes obstetricans ) in northern Spain: a comparative pathological study of an emerging ranavirus. Vet. J. 186 , 256-258 (2010).

24. Hyatt, A. D. et al . Comparative studies of piscine and amphibian iridoviruses. Arch. Virol. 145 , 301-331 (2000).

25. Mao, J., Hedrick, R. P. & Chinchar, V. G. Molecular characterization, sequence analysis, and taxonomic position of newly isolated fish iridoviruses. Virology 229 , 212-220 (1997).

26. Cunningham, A. A. et al . Pathological and microbiological findings from incidents of unusual mortality of the common frog (Rana temporaria). Philos. Trans. R. Soc. Lond. B. Biol. Sci. 351 , 1539-1557 (1996).

27. Green, D. E., Converse, K. A. & Schrader, A. K. Epizootiology of sixty-four amphibian morbidity and mortality events in the USA, 1996-2001. Ann. N. Y. Acad. Sci. 969 , 323-339 (2002).

28. Haislip, N. A., Gray, M. J., Hoverman, J. T. & Miller, D. L. Development and disease: how susceptibility to an emerging pathogen changes through anuran development. PLoS One 6, e22307 (2011).

29. Creemers, R. C. M. & van Delft, J. J. C. W. in De amfibieën en reptielen vanNederland – Nederlandse fauna 9. (Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey - Nederland, Leiden, 2009).

______Risk analysis CMTV-like virus April 2013 74 30. Mulder, J. & Creemers, R. C. M. in De amfibieën en reptielen van Nederland – Nederlandse fauna 9 (eds Creemers, R. C. M. & van Delft, J. J. C. W.) (Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey, Nederland, Leiden, 2009).

31. Dorenbosch, M. in De amfibieën en reptielen van Nederland – Nederlandse fauna 9 (eds Creemers, R. C. M. & van Delft, J. J. C. W.) (Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey, Leiden, Nederland., 2009).

32. Buggemum, H. J. M. in De amfibieën en reptielen van Nederland. – Nederlandse fauna 9. (eds Creemers, R. C. M. & van Delft, J. J. C. W.) (Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey – Nederland, Leiden, 2009).

33. Martens, G. & Snep, R. in De amfibieën en reptielen van Nederland – Nederlandse fauna 9. (eds Creemers, R. C. M. & van Delft, J. J. C. W.) (Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey, Leiden, Nederland, 2009).

34. Van Maanen, E. in De amfibieën en reptielen van Nederland. – Nederlandse fauna 9 (eds Creemers, R. C. M. & van Delft, J. J. C. W.) (Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey - Nederland, Leiden, 2009).

35. van Delft, J. J. C. W., Creemers, R. C. M. & Spitzen-van der Sluijs, A. S. Basisrapport Rode Lijsten Amfibieën en Reptielen volgens Nederlandse en IUCN-Criteria. Stichting RAVON, Nijmegen, in opdracht van Directie Kennis, Ministerie van LNV. (2007).

36. Crombaghs, B. H. J. M., van Eijk, J. & Creemers, R. C. M. in De amfibieën en reptielen van Nederland. – Nederlandse fauna 9. (eds Creemers, R. C. M. & van Delft, J. J. C. W.) (Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey, Leiden, Nederland, 2009).

37. Van den Broek, T. G. Y. & Frissen, D. in De amfibieën en reptielen van Nederland. – Nederlandse fauna 9. (eds Creemers, R. C. M. & van Delft, J. J. C. W.) (Nederlandse fauna 9. Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey – Nederland, Leiden, 2009).

38. Creemers, R. C. M. Brulkikkers in Baarlo 2010-2011. Stichting RAVON, Nijmegen. (2011).

39. Spitzen-van der Sluijs, A. S. & Zollinger, R. Risk Assessment on the American bullfrog and the fungus Batrachochytrium dendrobatidis . Stichting RAVON, Nijmegen. (2010).

40. Bauwens, D. & Claus, K. in Verspreiding van amfibieën en reptielen in Vlaanderen (De Wielewaal, Turnhout, 1996).

41. Günther, R. in Die Amphibien und Reptilien Deutschlands. (Gustav Fischer Verlag, Jena, 1996).

42. Juszczyk, W. The migrations of the aquatic frog Rana esculenta L. Bulletin of the Polish Academy of Sciences, Biological Sciences Letters, serie B 2, 341 – 369 (1951).

43. Tunner, H. G. in Proceedings of the Sixth Ordinary General Meeting of the Societas Europaea Herpetologica (eds Korsós, Z. & Kiss, I.) 449-452, Boedapest, 1992).

44. Günther, R. in Die Wasserfröschen Europas (Anura-Froschlurche). (A. Ziemsen Verlag, Wittenberg, Lutherstadt., 1990).

45. Bergers, P. J. M. & Luijten, L. in De amfibieën en reptielen van Nederland. (eds Creemers, R. C. M. & van Delft, J. J. C. W.), 2009).

46. Arntzen, J. W. in De amfibieën en reptielen van Nederland, België en Luxenburg. (ed Sparreboom, M.) 57-117 (A.A. Balkema, Rotterdam.

______Risk analysis CMTV-like virus April 2013 75 47. Blab, J. Biologie, Ökologie und Schutz von Amphibien. Schriftenreihe für Landschaftspflege und Naturschutz 18, Kilda Verlag, Bonn (1986).

48. Jehle, R. & Sinsch, U. Wanderleistung und Orientierung von Amphibien:eine Übersicht. Zeitschrift für Feldherpetologie 14 , 137-152 (2007).

49. Schlüpmann, M. & Günther, R. in Die Amphibien und Reptilien Deutschlands. (ed Günther, R.) 143-174 (Gustav Fischer Verlag, Jena.

50. Aarts, H. & Staal, H. J. Onderzoek naar de voortplantingsmigratie van Bufo bufo bufo (L.) m.b.v. radiotelemetrie. Report 206 (1981).

51. Henriksen, K. Colonization of newly constructed ponds by amphibians. Flora og Fauna 106 , 41-44 (2000).

52. van Delft, J. J. C. W. in De amfibieën en reptielen van Nederland. – Nederlandse fauna 9. (eds Creemers, R. C. M. & van Delft, J. J. C. W.) (Nationaal Natuurhistorisch Museum Naturalis, European Invertebrate Survey - Nederland, Leiden, 2009).

53. Blab, J. Untersuchungen zur Ökologie, Raum-Zeit-Einbindung und Funktion von Amphibienpopulationen. Dissertation Friedrich-Alexander Universität Erlangen-Nürnberg. (1978).

54. Kaufmann, R. Intensivteichbewirtschaftung und Amfibienvorkommen am Beispeil des Aischgrundes. Staatsexamarbeit Erlangen. (1976).

55. Feldman, R. in Die Amphibien und Reptilien Westfalens (Westfälischen Landesmuseums für Naturkunde, Münster, 1981).

56. Willis, Y. L., Moyle, P. B. & Baskett, T. S.Emergence, breeding, hibernation, movements, and transformation of bullfrog, Rana catesbeiana , in Missouri. Copeia , 30-35 (1956).

57. Gahl, M. K. & Calhoun, A. J. K. Landscape setting and risk of ranavrius mortality events. Biological Conservation 141 (11) , 2679-2689 (2008).

58. Bang Jensen, B., Reschova, S., Cinkova, K., Ariel, E. & Vesely, T. Common carp ( Cyprinus carpio ) and goldfish (Carassius auratus ) were not susceptible to challenge with ranavirus under certain challenge conditions. Bull. Eur. Ass. Fish Pathol. 31 (3) , 112-117 (2011).

59. Gahl, M. K. & Calhoun, A. J. K. The role of multiple stressors in ranavirus-caused amphibian mortalities in Acadia National Park wetlands. Canadian Journal of Zoology 88 , 1-108-121 (2010).

60. Hoverman, J. T., Gray, M. J., Miller, D. L. & Haislip, N. A. Widespread occurrence of ranavirus in pond-breeding amphibian populations. Ecohealth 9, 36-48 (2012).

61. Rojas, S., Richards, K., Jancovich, J. K. & Davidson, E. W. Influence of temperature on Ranavirus infection in larval salamanders Ambystoma tigrinum . Dis. Aquat. Organ. 63 , 95-100 (2005).

62. Gray, M. J., Miller, D. L., Schmutzer, A. C. & Baldwin, C. A. Frog Virus 3 prevalence in tadpole populations at cattle-access and non-access wetlands in Tennessee, U.S.A. Diseases of aquatic organisms 77 , 97-103 (2007).

63. Greer, A. L. & Collins, J. P. Habitat fragmentation as a result of biotic and abiotic factors controls pathogen transmission throughout a host population. Journal of Animal Ecology 77 , 364-369 (2008).

64. St-Amour, V., Wong, W. M., Garner, T. W. & Lesbarreres, D. Anthropogenic influence on prevalence of 2 amphibian pathogens. Emerg. Infect. Dis. 14 , 1175-1176 (2008).

______Risk analysis CMTV-like virus April 2013 76 65. IUCN/SSC Re-introduction specialist group. Guidelines for re-introductions. IUCN, Gland, Switzerland and Cambridge, UK. (1998).

66. Speare, R. et al . Hygiene protocol for handling amphibians in the field. Amphibian Diseases Group James Cook University Townsville Australia School of Public Health and Tropical Medicine .

67. Janssen, I. De vroedmeesterpad in Nederland: leven in een dierentuin. RAVON werkgroep monitoring, nieuwsbrief Meetnet Amfibieën 19 , 14-15 (2006).

68. Van Laar, V. De verspreiding van amfibieën op Texel in relatie tot de saliniteit van binnenlandse wateren. RAVON 7 (2) , 25-30 (2005).

69. Boere, G. C. & Stroud, D. A. in Waterbirds around the world (eds Boere, G. C., Galbraith, C. A. & Stroud, D. A.) 40-47 (The Stationary Office, Edinburgh, UK, 2006).

70. Branquart, E. et al . in ISEIA guidelines, Harmonia information system (version 2.5) 1-4 (Biodiversity.be, 2009).

71. Duffus, A. L. J. & Cunningham, A. A. Major disease threats to European amphibians. Herpetological Journal 20 , 117--127 (2010).

72. Pasmans, F., Blahak, S., Martel, A., Pantchev, N. & Zwart, P. Ranavirus-associated mass mortality in imported red tailed knobby newts (Tylototriton kweichowensis): a case report. Vet. J. 176 , 257-259 (2008).

73. Duggan, I. C. The freshwater aquarium trade as a vector for incidental invertebrate fauna. Biological invasions 12 (11) , 3757-3770 (2010).

74. Whittington, R. J., Kearns, C., Hyatt, A. D., Hengstberger, S. & Rutzou, T. Spread of epizootic hematopoietic necrosis virus (EHNV) in redfin perch ( Perca fluviatilis ) in southerne Australia. Australian Veterinary Journal 73 , 112-- 114 (1996).

75. Garmyn, A. et al . Waterfowl: potential envrionmental reservoirs of the chytrid fungus Batrachochytrium dendrobatidis . PloS one 7(4) , e35038. doi:10.1371/journal.pone.0035038 (2012).

76. Majji, S. et al . Rana catesbeiana virus Z (RCV-Z): a novel pathogenic ranavirus. Dis. Aquat. Organ. 73 , 1-11 (2006).

77. Storfer, A. et al . Phylogenetic concordance analysis shows an emerging pathogen is novel and endemic. Ecol. Lett. 10 , 1075-1083 (2007).

78. Jancovich, J. K. et al . Evidence for emergence of an amphibian iridoviral disease because of human-enhanced spread. Mol. Ecol. 14 , 213-224 (2005).

79. Williams, T., Barbosa-Solomieu, V. & Chinchar, V. G. A decade of advances in iridovirus research. Adv. Virus Res. 65 , 173-248 (2005).

80. Blaustein, A. R. et al . Ecophysiology meets conservation: understanding the role of disease in amphibian population declines. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 367 , 1688-1707 (2012).

81. Kerby, J. L., Hart, A. J. & Storfer, A. Combined effects of virus, pesticide, and predator cue on the larval tiger salamander ( Ambystoma tigrinum ). Ecohealth 8, 46-54 (2011).

82. Daszak, P. et al . Emerging infectious diseases and amphibian population declines. Emerg. Infect. Dis. 5, 735-748 (1999).

83. Miller, D. L. et al . Pathologic findings in larval and juvenile anurans inhabiting farm ponds in Tennessee, USA. J. Wildl. Dis. 45 , 314-324 (2009).

______Risk analysis CMTV-like virus April 2013 77 84. Chen, G. & Robert, J. Antiviral immunity in amphibians. Viruses 3, 2065-2086 (2011).

85. Fox, S. F., Greer, A. L., Torres-Cervantes, R. & Collins, J. P. First case of ranavirus-associated morbidity and mortality in natural populations of the South American frog Atelognathus patagonicus . Dis. Aquat. Organ. 72 , 87-92 (2006).

86. Brunner, J. L., Richards, K. & Collins, J. P. Dose and host characteristics influence virulence of ranavirus infections. Oecologia 144 , 399-406 (2005).

87. Van der Lugt, G., Prudon, B. & Crombaghs, B. De knoflookpad langs de Vecht, Regge en Dinkel in Overijssel. Inventarisatie in 2000 & een overlevingspan voor de periode 2000-2010. Natuurbalans/Limes Divergens i. s;m. Stichting RAVON (2000).

88. Lenders, A. J. W. Beschermingsplan vroedmeesterpad en geelbuikvuurpad 2000-2004. Ministerie van Landbouw, Natuurbeheer en Vissierij. 's Gravenhage (2000).

89. Crombaghs, B. & Bosman, W. Beschermingsplan geelbuikvuurpad & vroedmeesterpad 2006-2010 -Platform geelbuikvuurpad en vroedmeesterpad. Natuurbalans/Limes Divergens & Stichting RAVON,Nijmegen (2006).

90. Ariel, E. et al . Ranavirus in wild edible frogs Pelophylax kl. esculentus in Denmark. Dis. Aquat. Organ. 85 , 7-14 (2009).

91. Drury, S. E., Gough, R. E. & Cunningham, A. A. Isolation of an iridovirus-like agent from common frogs ( Rana temporaria ). Vet. Rec. 137 , 72-73 (1995).

92. Teacher, A. G. F., Cunningham, A. A. & Garner, T. W. J. Assessing the long-term impact of Ranavirus infection in wild common frog populations. Anim Conserv. 13 , 514--522 (2010).

93. Chinchar, V. G., Hyatt, A., Miyazaki, T. & Williams, T. Family Iridoviridae: poor viral relations no longer. Curr. Top. Microbiol. Immunol. 328 , 123-170 (2009).

94. Smith, K. F., Acevedo-Whitehouse, K. & Pedersen, A. B. The role of infectious diseases in biological conservation. Animal Conservation ISSN 1367-9430 (2009).

95. Rubbo, M. J., Belden, L. K., Storrs-Mendez, S. I., Cole, J. J. & Kiesecker, J. M. Species loss in the brown world: are heterotrophic systems inherently stable? Aquatic Sciences 74(3) , 397-404 (2012).

96. Whiles, M. R. et al . The effects of amphibian population declines on the structure and function of neotropical stream ecosystems. Frontiers in Ecology and Environment 4(1) , 27-34 (2006).

97. Mao, J., Wang, J., Chinchar, G. D. & Chinchar, V. G. Molecular characterization of a ranavirus isolated from largemouth bass Micropterus salmoides. Dis. Aquat. Organ. 37 , 107-114 (1999).

98. Moody, N. J. G. & Owens, L. Experimental demonstration of the pathogenicity of a frog virus, Bohle iridovirus, for a fish species, barramundi Lates calcerifer. . Dis. Aquat. Organ. 18 , 95--102 (1994).

99. Rubbo, M. J., Lanterman, J. L., Falco, R. C. & Daniels, T. J. The influence of amphibians on mosquitoes in seasonal pools: Can wetlands protection help to minimize disease risk? Wetlands 31 , 799-804 (2011).

100. KIWA Water Research, E. Knelpunten- en kansenanalyse Natura 2000-gebied 30 – Dwingelderveld. KIWA Water Research / EGG, Nieuwegein / Groningen. (2007).

101. Anonymous. Gebiedendocument – werkdocument Natura 2000: aanwijzingsbesluit 1 Natura 2000 gebied 30 - Dwingelderveld SynBioSys. www. synbiosys. alterra. nl/natura2000/ (2012).

102. Beebee, T. Impact of Ranavirus on garden amphibian populations. Herpetological Bulletin 120 , 1-3 (2012).

______Risk analysis CMTV-like virus April 2013 78 103. Van den Hurk, B. et al . KNMI climate change scenarios 2006 for the Netherlands. Scientific Report WR 2006-01. KNMI, De Bilt (2006).

104. OIE. in Manual of Diagnostic tests for aquatic animals 2012 (ed OIE) 256-276 (OIE, Paris, 2012).

105. Whittington, R. J., Becker, J. A. & Dennis, M. M. Iridovirus infections in finfish - critical review with emphasis on ranaviruses. J. Fish Dis. 33 , 95-122 (2010).

106. Peel, A. J., Hartley, M. & Cunningham, A. A. Qualitative risk analysis of introducing Batrachochytrium dendrobatidis to the UK through the importation of live amphibians. Dis. Aquat. Organ. 98 , 95-112 (2012).

107. Garner, T. W. J., Stephen, I., Wombwell, E. & Fisher, M. C. The amphibian trade: bans or best practice? EcoHealth 6, 148-151 (2009).

108. Gray, M. J., Miller, D. L. & Hoverman, J. T. Reliability of non-lethal surveillance methods for detecting ranavirus infection. Dis. Aquat. Organ. 99 , 1-6 (2012).

109. Pavlin, B. I., Schloegel, L. M. & Daszak, P. Risk of importing zoonotic diseases through wildlife trade, United States. Emerg. Infect. Dis. 15 , 1721-1726 (2009).

110. Driskell, E. A., Miller, D. L., Swist, S. L. & Gyimesi, Z. S. PCR detection of ranavirus in adult anurans from the Louisville Zoological Garden. J. Zoo Wildl. Med. 40 , 559-563 (2009).

111. Greer, A. L. & Collins, J. P. Sensitivity of a diagnostic test for amphibian ranavirus varies with sampling protocol. Journal of wildlife diseases 43(3) , 525-532 (2007).

112. Phillott, A. D. et al . Minimising exposure of amphibians to pathogens during field studies. Dis. Aquat. Organ. 92 , 175-185 (2010).

113. OIE. Terrestrial animal health code. 1 & 2 (2012).

______Risk analysis CMTV-like virus April 2013 79 Contributors

The consortium conducting this risk analysis consisted of four partners, three from the Netherlands and one from Belgium. A list of contact details per organization is provided in Annex 1.

Dutch Wildlife Health Centre DWHC Expertise - The purpose of the Dutch Wildlife Health Centre (DWHC) is to enhance knowledge and expertise in wildlife health in the Netherlands. This serves to provide scientifically based information for political and practical decisions concerning public health, wild and domestic animal health, and nature conservation issues. The first outbreak of ranavirus in free-living amphibians was detected through the general wildlife disease surveillance that is one of the main activities of the DWHC. Subsequently, and together with RAVON, DWHC conducted the 2011 ranavirus surveillance project. Nature of contribution - Within this study, DWHC coordinated the activities of the partners including final reporting. Focus was on CMTV-like virus infection occurring in amphibians. DWHC performed a literature review in order to obtain an overview of CMTV-like virus, potential hosts, and potential environmental factors. In addition, DWHC used data obtained from earlier and ongoing CMTV-like virus cases and studies. DWHC obtained information from official sources, to try to be able to quantify risks. This included import data from Schiphol airport with regards to amphibian species known to be susceptible to CMTV-like virus infection and information on officially permitted translocations of potential hosts. DWHC participated in the risk assessment score workshop. DWHC assessed possible detection methods with regard to amphibian hosts, and defined possible risk management options. DWHC had contact with the pathologist experts in the UK. The experts involved : ∗ Prof. Dr. Andrea Gröne, DVM, Dipl. ECVP, Dipl. ACVP is a veterinary pathologist, the Director of the DWHC and the Head of the Division Pathology in the Department of Pathobiology at the University of Utrecht. She supervises DWHC activities and provided input in the final reporting. ∗ Dr. Marja Kik DVM, Dipl. Vet Path RNVA, Dipl. ECZM (Herpetology) is a staff veterinary pathologist working in pathology of exotic animals and wildlife and is an expert in ranavirus-induced disease. She was involved as pathologist in the 2011 ranavirus surveillance project. ∗ Dr. Jolianne Rijks , DVM, works at DWHC as an epidemiologist. She studied veterinary medicine and obtained her PhD degree with research on the 2002 phocine distemper virus epidemic in seals. She has (co- )authored 7 peer-reviewed papers on wildlife related topics. She has prepared and managed the 2011 ranavirus surveillance project on behalf of DWHC, and coordinated the present study.

RAVON Expertise - RAVON stands for Reptile, Amphibian and Fish Conservation Netherlands. The RAVON Foundation is a non governmental organization (NGO) with 1800 volunteers, more than 1000 contributors and 28 professional staff members with offices in Nijmegen and Amsterdam. Organization members inform and educate, conserve species and their habitats, organize two national monitoring networks and collect distribution data for the conservation of reptiles, amphibians and fish in the Netherlands. RAVON staff also gives advice and carries out research for a wide range of clients. RAVON and DWHC have worked together on the ranavirus. Nature of contribution - RAVON contributed in providing the host species related data relevant to natural pathways, establishment possibilities and likelihood of dispersal for CMTV-like virus. RAVON provided information on the host species, vector species and affected species necessary to define risk areas. RAVON provided data collected in 2011 in the National Park Dwingelderveld, supplemented by data collected during two field visits in 2012 (August and September), to assess the impact of CMTV-like virus on host populations and to define the ecological characteristics of the described outbreaks, with special emphasis on defining the future scenarios that may occur. In this RAVON and RU collaborated closely. RAVON contributed to define the ecological, economic and social consequences of ranavirus outbreaks and participated in the risk assessment score workshop. The experts involved : ∗ Ir. A. M. Spitzen – van der Sluijs is project leader and professional staff member at RAVON. She is also a PhD student at Ghent University under the supervision of professor Frank Pasmans and professor An Martel, where she’s studying the impact of B. dendrobatidis on native amphibian species in relation to environmental conditions. She has (co-)authored 9 peer-reviewed papers on diverse topics, and her main interests are the effects of invasive pathogens on amphibian communities, the resistance of landscape matrices on dispersal possibilities of reptiles and amphibians, and the auto-ecology of the legless lizard, Anguis fragilis . She has conceived managed the 2011 ranavirus surveillance project on behalf of RAVON. ∗ Drs. R. Zollinger , team leader and manager/researcher at Research and Conservation department of RAVON. He is a biologist (terrestrial ecology, population dynamics, nature conservation), published several

______Risk analysis CMTV-like virus April 2013 80 papers on biology of amphibians, birds and nature conservation topics. Involved in several interdisciplinary projects and is active member of the European Herpetological Society.

Radboud University Nijmegen Expertise - The research focus of the Institute for Water and Wetland Research at the Radboud University Nijmegen (RU) is the natural environment, in particular aquatic ecosystems and wetlands. The Department of Environmental Science was founded in 1991. It focuses on understanding and predicting biological responses to physical and chemical pressures. Nature of contribution - If there were specific environment related factors (e.g., climate, water quality parameters) affecting the probability of establishment and spread, RU analyzed whether and in which water systems these occur, and how these match now and in the future with species. RU assisted RAVON with defining the ecological, economic and social consequences of CMTV-like virus outbreaks, supervising and warranting the quality of the impact assessment. RU prepared, presided and wrote down the results of the risk assessment score workshop. The experts involved : ∗ Dr. Rob S.E.W. Leuven (1957) studied biology (aquatic ecology, animal physiology, fisheries & aquaculture). His PhD-thesis concerned the impacts of acidification on the biodiversity and functioning of aquatic ecosystems. He was project leader of several large research projects commissioned by the World Bank, European Commission, the Netherlands Organization for Scientific Research and several Dutch governmental and non-governmental organizations. He successfully supervised 10 PhD-projects and (co)authored more than 210 papers (among which 90 ISI-WOS articles), 18 scientific books and 50 professional reports. His main research interest is understanding the impact of multiple stressors on biodiversity and functioning of aquatic ecosystems (rivers, floodplains, moorland pools and peat bogs). Recent research focuses on effects of exotic invasive species on aquatic ecosystems. Special attention is paid to the invasion process (including dispersal patterns, causes of establishment success, physiological tolerances and biological traits of invaders), risk assessment of aquatic invasions, effects of climate change on global redistribution of species, ecological and socio-economical impacts and invasion management. ∗ Laura N.H. Verbrugge (MSc) studied Environmental Sciences at Utrecht University (BSc) and Radboud University Nijmegen (MSc - science communication track). At present she is working as a researcher on a joined PhD project of the Institute for Science, Innovation and Society (ISIS) and the Institute of Water and Wetland Research (IWWR), both stationed at the Radboud University. Her research activities include a broad spectrum of topics concerning non-native species including (ecological) risks, public perception and preventive measures and communication campaigns. She had published several papers and reports on risk assessment of non-native species and has gained experience in moderating risk assessment workshops held in the Netherlands. Currently she is also involved in the evaluation of the Dutch code of conduct for invasive aquatic plants, a voluntary agreement between the government and horticulture sector with the objective prevent further introductions of invasive species by increasing public awareness.

6.4. Faculty of Veterinary Medicine, Ghent University, Belgium Expertise - The research group “Amphibian and reptile diseases” is hosted by the Division of Avian and Exotic Animals (An Martel) and the Laboratory of Veterinary Bacteriology and Mycology (Frank Pasmans) at the Department of Pathology, Bacteriology and Avian Diseases, Faculty of Veterinary Medicine, Ghent University, Belgium (UGent). This research group focuses on amphibian diseases in the wild and captivity, has conducted several studies on ranavirus infection (see reference list) and was involved in the diagnosis of ranavirus in the two ranavirus associated mortality events in the Netherlands . Nature of contribution - The main contribution of UGent to this study was to make an inventory of trade in amphibians in Belgium and from Belgium to the Netherlands, and an inventory of privately kept amphibians in Belgium and their possible exchange to the Netherlands. The experts involved : ∗ Prof. Dr. An Martel , DVM, is the Head of the Division of Avian and Exotic Animals at the Department of Pathology, Bacteriology and Avian Diseases at the Faculty. Her research focus is infectious diseases in birds, reptiles and amphibians. She currently supervises 5 PhD students working on infectious diseases in amphibians. She has been promoter of 4 PhD theses and (co-)authored more than 80 ISI listed publications. She recently described a novel pathogen (the genus Amphibiichlamydia ) in amphibians. ∗ Prof. Dr. Frank Pasmans , DVM, is the Director of the Laboratory of Veterinary Bacteriology and Mycology. He has a lifelong interest in reptiles and amphibians, one of his major research topics being reptile and amphibian diseases, with emphasis on host pathogen interactions. He has been promoter of 19 defended PhD theses and (co-)authored 200 ISI listed publications.

______Risk analysis CMTV-like virus April 2013 81

______Risk analysis CMTV-like virus April 2013 82