Animals and SARS‐CoV‐2: Species susceptibility and viral transmission in experimental and natural conditions, and the potential implications for community transmission

Abstract The current COVID‐19 global pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) of probable bat origin, has highlighted the ongoing need for a One Health response to emerging zoonotic disease events. Understanding the human–animal interface and its relevance to disease transmission remains a critical control point for many emerging zoonoses. Determination of the susceptibility of various animal species to infection with SARS‐CoV‐2 and the role of animals in the epidemiology of the disease will be critical to informing appropriate human and veterinary public health responses to this pandemic. A scoping literature review was conducted to collect, evaluate and present the available research evidence regarding SARS‐CoV‐2 infections in animals. Experimental studies have successfully demonstrated SARS‐CoV‐2 infection and transmission in cats, ferrets, hamsters, bats and non‐human primates under experimental settings. Dogs appear to have limited susceptibility to SARS‐CoV‐2, while other domestic species including pigs and poultry do not appear susceptible. Naturally occurring SARS‐CoV‐2 infections in animals appear uncommon, with 14 pets, 8 captive big cats and an unreported number of farmed mink testing positive at the time of writing (early July 2020). Infections typically appear asymptomatic in dogs, while clinical signs of respiratory and/or gastrointestinal disease tend to be mild to moderate in felines, and severe to fatal in mink. Most animals are presumed to have been infected by close contact with COVID‐19 patients. In domestic settings, viral transmission is self‐limiting; however in high‐density animal environments, there can be sustained between‐animal transmission. To date, two potential cases of animal‐to‐human transmission are being investigated, on infected mink farms. Given the millions of COVID‐19 cases worldwide and ongoing potential for further zoonotic and anthroponotic viral transmission, further research and surveillance activities are needed to definitively determine the role of animals in community transmission of SARS‐CoV‐2.


| INTRODUC TI ON
First detected in late 2019, the COVID-19 pandemic caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has since been declared a public health emergency of international concern, approaching eleven million human infections and causing over half a million human deaths by early July 2020 (Dong et al., 2020). The zoonotic virus was first identified in Wuhan, China, and the initial cluster of cases was initially believed to be linked to a live "wet" seafood and animal market, although results of back tracing activities have now called this into question (Letzter, 2020;Woodward, 2020). Phylogenetically related to SARS-CoV-1, which caused the disease SARS that emerged from China in 2003, SARS-CoV-2 is also believed to have emerged from a betacoronavirus circulating in rhinolophid (horseshoe) bats (Lau et al., 2020;Zhou et al., 2020b), which then made the species jump to humans, potentially by way of an intermediate animal host.
From the initial zoonotic spillover event in China, the virus quickly demonstrated efficient human-to-human transmission, with virus in respiratory droplets expelled by COVID-19 patients during coughing, sneezing and talking reportedly remaining viable for multiple hours in aerosols and up to days on some surfaces (van Doremalen et al., 2020;Meselson, 2020). With billions of people around the world harbouring the virus, and the presumptive animal-origin and high recombination rates of SARS-coronaviruses (Stavrinides & Guttman, 2004), reverse spillover of SARS-CoV-2 from humans to animals (anthroponotic transmission) was surely inevitable.
Incalculable numbers of human-animal interactions occur each day in domestic, agricultural, research, recreational, educational, therapeutic, entertainment and wild settings. Over 470 million dogs and 370 million cats were kept as pets worldwide in 2018 (Bedford, 2020), with the USA, China and Russia the top three dogand cat-owning nations (Walden, n.d.). Over half of 27,000 surveyed households owned at least one pet in 2016, and increases in pet ownership are reportedly highest in China, India and Latin America (GfK, 2016). Worldwide, an estimated 70 billion terrestrial animals were raised and slaughtered for human consumption in 2018, with farm animal production the fastest growing agricultural sub-sector and increases particularly expected to occur in developing countries (Ilea, 2009). Flourishing legal and illegal wildlife trades exist worldwide, with the global trafficking in illegal wildlife commodities including pangolin scales, elephant ivory, tiger bone and bear bile estimated at $7-23 billion USD in 2016 (TRAFFIC, 2020).

Encroachment of humans into previously uninhabited areas for
logging, cropping and urbanization can create diverse wildlife-livestock-human interfaces that represent critical points for cross-species transmission and emergence of pathogens (Hassell et al., 2017).
Over 60% of recently identified emerging disease events have been zoonotic-the majority of which have had a wildlife origin-and such events have been increasing significantly in recent times (Jones et al., 2008;Wang & Crameri, 2014).
As recent outbreaks of zoonotic diseases with pandemic and panzootic potential-including highly pathogenic avian influenza, include primates-particularly of the catarrhine (Old world) order (Damas et al., 2020)-and most of the carnivores, even-and oddtoed ungulates, pangolins and scaly anteaters (Praharaj et al., 2020).
Medium to high probability was ascribed to pangolins, rabbits and cats; lower probability to chickens and turkeys; and rats, mice and ducks were rated as low to very low (Praharaj et al., 2020;Zhao et al., 2020).  noted that pigs and dogs, and even cattle and sheep appear to have ACE2 proteins capable of acting as a receptor for viral entry, and the apparent low susceptibility of infection in these species is likely related to the relatively low levels of ACE2 expressed in the respiratory tract. Notably, the Damas et al. (2020) model predicted domestic cats, tigers and golden hamsters to be only medium risk, and ferrets low risk. Contrary evidence of natural and experimental infection in these animals (see below) suggests that the utility of these models as predictive tools may be limited.
information regarding natural and experimental infections of the virus in animals, in order to assess the potential role of animals in community transmission of COVID-19.

| MATERIAL S AND ME THODS
This literature review was conducted using an exploratory scoping study approach, as its purpose was to rapidly map the broad state of knowledge regarding SARS-CoV-2 and animals from the heterogeneous types of available evidence, rather than to answer a clearly defined question (Arksey & O'Malley, 2005). We conducted successive daily scans of multiple electronic databases, preprint repositories, internet search engines, government and organizational websites, and media sources, and used different combinations of the following words and phrases to actively identify additional relevant publications: SARS-CoV-2, coronavirus, animal,pet,cat,dog,ferret,hamster,tiger,lion,monkey,primate,bat,pangolin, snake, mink, "reservoir host", "intermediate host", transmission, susceptibility, infection. The searches covered all years and any publications in English and Dutch. We also searched reference lists from key reviews and articles to identify additional publications or sources of interest. Study titles and abstracts were screened, with full articles obtained and re-evaluated for inclusion/exclusion under the following criteria: 1. Inclusion criteria: Epidemiological studies, case reports, commentaries, reviews, letters, editorials, preprints, government documents, media reports and blogs, published in English or Dutch between 31st December 2019 and 23rd June 2020, that were relevant under the following categories: (1) reports of natural SARS-CoV-2 infections in domestic or wild animal species; (2) experimental SARS-CoV-2 susceptibility, pathogenesis and/or transmission studies in animals; (3) discussions, predictions and/or modelling of animal reservoir and/or intermediate host species for SARS-CoV-2, and (4) general reviews or commentary about SARS-CoV-2 and animals.
2. Exclusion criteria: Any documents relating to SARS-CoV-2 vaccines or therapeutics, including nanobody and antibody therapy, and articles published in a language other than English or Dutch.
Based on the above criteria, a total of 351 documents were collated and evaluated for inclusion in this review: 78 journal articles; 79 preprints; 75 media articles (including newspaper and magazine articles); 113 official reports, documents and press releases; and 6 websites. Included references were sorted into the categories listed above. Data from experimental studies were entered into a Microsoft Word template by animal species. Case data on natural infections in animals were entered into a Microsoft Excel spreadsheet, from which summary tables were created.  Table 1 and presented below. soft palate and tonsils) from as early as two days post-infection (dpi).

| Ferrets
Highest viral titres were observed in nasal washes, peaking at 4 dpi before dropping below detection limits at 10 dpi (Kim et al., 2020).
Virus replication potentially also occurs in the digestive tract (viral RNA was detected in rectal swabs of infected ferrets) but does not appear to replicate in the lower respiratory tract, even in animals that were intratracheally inoculated. Antibodies were detectable in low quantities from 13 dpi, increasing substantially by 20 dpi . The virus is effectively transmitted via direct contact and to other in-contact ferrets via the respiratory droplet route (Kim et al., 2020;Richard et al., 2020;Shi et al., 2020).
Clinical signs in infected ferrets included elevated body temperatures, loss of appetite, reduced activity and occasional coughing (Kim et al., 2020;Shi et al., 2020). The similarity of the clinical disease picture in infected ferrets and humans, and the efficient replication of SARS-CoV-2 in the ferret upper respiratory tract makes them a highly suitable animal model for evaluating antiviral drugs or vaccine candidates against COVID-19.

| Cats
Cats are also susceptible to infection with SARS-CoV-2, with viral replication occurring in the upper respiratory tract following nasal inoculation . Peak oral and nasal viral shedding was recorded at 3 dpi in three directly inoculated cats, and on day 7 post-exposure in two in-contact cats (Bosco-Lauth et al., 2020).
Infectious virus is also excreted rectally .
Transmission occurs by both direct contact and indirectly via aerosols (Bosco-Lauth et al., 2020;Shi et al., 2020). Experimental exposure has resulted in both subclinical and symptomatic infections, with juvenile cats (70 to 100 days old) reportedly more susceptible to severe clinical disease and death .
Antibody titres have been recorded in both experimentally inoculated and in-contact cats (Bosco-Lauth et al., 2020;Shi et al., 2020). Bosco-Lauth et al. (2020) reported that infected cats (n = 3) developed significant neutralizing antibody titres that stayed stable or increased from 14 dpi up to 42 dpi. Following re-challenge with SARS-CoV-2 at 28 dpi, a moderate increase in antibody titre was observed, and no virus was shed by any of the cats in the subsequent 14-day period.
A study by Chen et al. (2020) suggests that cats have both a high frequency of cells expressing ACE2, and a high proportion of cells co-expressing ACE2 and TMPRSS2 (transmembrane serine protease 2), both of which are targets for SARS-CoV-2 entry. These target cells are also widely distributed among organs within the feline digestive, respiratory and urinary systems, suggesting that cats may be susceptible to infection and transmission via several routes.

| Dogs
Dogs appear to have low susceptibility to SARS-CoV-2 following experimental inoculation. One study  intranasally infected five beagles with SARS-CoV-2, and collected samples over the subsequent two week period. Viral RNA was detected in three of the dogs' rectal swabs between 2 and 6 dpi. No virus was detected in any oropharyngeal swabs taken from the dogs during the study period, nor in any organs or tissues on autopsy, and attempts at virus isolation were negative. Antibodies were detected in the sera of 2/4 experimentally infected dogs, but neither of two additional beagles that were housed in close proximity. Chen et al. (2020) reported very low levels of co-expression of ACE2 and TMPRSS2 target receptors in canine lung cells, as well as mutations in critical amino acid sequences in ACE2 receptors, which they suggested may be responsible for the low susceptibility of dogs to SARS-CoV-2 infection.

| Syrian golden hamsters
SARS-CoV-2 infections in Syrian golden hamsters appear to resemble features found in human patients with mild infections, with viral replication in the epithelial cells of the respiratory and gastrointestinal tracts following intranasal inoculation . Peak viral load 10 5 -10 7 was detected in the lungs at around 2 to 3 dpi and high copy numbers of viral RNA continued to be detectable beyond 7 dpi, although no infectious virus was detected from 7 dpi.
Syrian hamsters developed mild clinical signs including weight loss, rapid breathing, and postural changes, with older hamsters (32-24 weeks old) exhibiting more pronounced and consistent weight loss than younger hamsters (6 months old) (Osterrieder et al., 2020).
Other age-dependent observations included an earlier and stronger influx of immune cells in lung tissue, and rapid lung recovery at day 14 post-infection in younger hamsters. Older hamsters developed conspicuous alveolar and perivascular oedema indicative of vascular leakage (Osterrieder et al., 2020), similar to the more severe pathology observed in elderly COVID-19 patients.
Transmission studies reported efficient viral transmission from infected to naïve in-contact hamsters, resulting in similar pathology but not weight loss. All infected hamsters fully recovered and developed mean serum neutralizing antibody titre >1:427 by day 14 post-infection (Sia et al., 2020), with higher antibody titres seen in younger animals (Osterrieder et al., 2020). Immunoprophylaxis with early convalescent serum achieved significant decrease in lung viral load but not in lung pathology .

| Non-human primates
Experimental infection studies have been reported in a range of non-human primate species, including rhesus macaques (Macaca mulatta), cynomolgus or crab-eating macaques (Macaca fascicularis), common marmosets (Callithrix jacchus) and African green or vervet monkeys (Chlorocebus aethiops).
An experimental study by Lu et al. (2020) determined that that rhesus macaques (Macaca mulatta) were the most susceptible to SARS-CoV-2 infection, followed by cynomolgus or crab-eating macaques (Macaca fascicularis) and lastly common marmosets (Callithrix jacchus). SARS-CoV-2 was detected in nasal, throat and anal swabs as well as blood from all three monkey species, with viral shedding from the upper respiratory tract peaking between 6 and 8 dpi (Denis et al., 2020).
Infections in all tested monkey species ranged from asymptomatic, to mild serous nasal discharge and transient or intermittent increases in body temperatures (Hartman et al., 2020;Rockx et al., 2020), to moderate respiratory disease with coughing, anorexia, postural changes and weight loss, and some radiographic chest abnormalities similar to those from human COVID-19 patients observed (Lu et al., 2020;Munster et al., 2020). Hartman et al.
(2020) also reported a transient decrease in lung tidal volume at 7 dpi. Post-mortem examination of severely affected animals showed severe gross pathology in the heart, stomach and lower respiratory tract including diffuse interstitial pneumonia Shan et al., 2020).
Viral replication was highest in lung tissue, but was also detected in ileum and tracheobronchial lymph nodes (Rockx et al., 2020). Viral shedding in oropharyngeal swabs peaked between 1 and 5 dpi, decreasing below detectable limits by day 9, with one study recording a second recrudescent period of viral shedding from respiratory and gastrointestinal tracts in all infected macaques including subclinical animals between 14-21 dpi (Hartman et al., 2020). Anal and rectal swabs tested positive in some animals up to 11 and 20 dpi, respectively Shan et al., 2020). Higher levels of viral RNA were detected in nasal swabs of aged cynomolgus macaques compared to younger animals (Rockx et al., 2020). Viral RNA was detectable by RT-PCR in peripheral blood between 2 and 10 dpi (Lu et al., 2020), and antibodies against the S1 domain and nucleocapsid proteins of SARS-CoV-2 were detectable from 14 dpi (Rockx et al., 2020).
Study authors concluded that the experimental non-human primate species tested are permissive to SARS-CoV-2 infection, shed virus for pronged periods of time and display COVID-19-like disease, thereby providing suitable animal models for COVID-19 studies Totura et al., 2020).

| Tree shrews
Tree shrews (Tupaia belangeris) were considered as potential animal models for SARS-CoV-2 infection, as they are genetically similar to primates and have been used in biomedical research for animal models of viral infections including hepatitis B, influenza and Zika viruses (Zhou et al., 2020b). However, the one available study by Zhou et al. (2020b) reported limited viral replication in tree shrews (n = 24) intranasally inoculated with SARS-CoV-2. SARS-CoV-2 RNA was detectable in nasal, throat and anal swabs from 6 dpi; viral shedding was highest in younger animals during the early stage of viral infection, and of longest duration in the adult and old animals, particularly in the males of these groups. Organs appeared grossly normal and histopathological changes were generally mild on necropsy, except for one adult animal that had severe lung pathology.

| Poultry
Chickens, turkeys, ducks, quail, geese and pigeons were not able to be experimentally infected Shi et al., 2020;Suarez et al., 2020) and did not seroconvert (Suarez et al., 2020), and are presumed to be not susceptible to SARS-CoV-2. The virus did not replicate in embryonated chicken eggs (Suarez et al., 2020).
A study by Chen et al. (2020) suggests a lack of co-expression of ACE2 and the entry activator TMPRSS2 target receptors in lung cells of poultry, as well as mutations in critical amino acid sequences in ACE2 receptors, are responsible for the low susceptibility of these species to SARS-CoV-2.

| Pigs
Pigs were not able to be experimentally infected and are presumed to be not susceptible to SARS-CoV-2 . A study by Chen et al. (2020) reported that cells that co-express ACE2 and

| Natural SARS-CoV-2 infection events in animals
The World Organisation for Animal Health (OIE) defined the following criteria for confirmed cases of SARS-CoV-2 in animals: isolation of SARS-CoV-2 from a sample taken directly from an animal, or identification of SARS-CoV-2 viral RNA in a sample taken directly from an animal that either targets at least two specific genomic regions at a level indicating the presence of infectious virus, or targets a single genomic region followed by sequencing of a secondary target (OIE, 2020e).
To date (July 2020), confirmed cases of SARS-CoV-2 in animals worldwide remain limited and sporadic, with global totals of fourteen domestic pets, eight captive big cats and an unreported number of mink (Neovison vison) on nineteen fur farms having definitively tested positive to SARS-CoV-2 (see Table 2). Most cases are strongly linked to close contact with confirmed or suspected human COVID-19 patients, indicating that human-to-animal (anthroponotic) transmission is the primary mode of spread in domestic settings (Denis et al., 2020). In high-density animal environments, including zoos and intensive breeding farms, subsequent between-animal transmission has been presumed to occur, and at present there are two cases of probable animal-to-human transmission being investigated on Dutch mink farms. If confirmed, these cases would represent the first known cases of zoonotic transmission of the virus since the initial spillover event/s from bats to humans in China.
Summary case data are presented below.
Of the pet cats to have tested positive for SARS-CoV-2, nearly all have had known (n = 8) or suspected (n = 2) close contact with human COVID-19 cases. Infected cats have shown clinical signs in 8/11 cases, from mild upper respiratory disease including oral lesions and tongue ulceration (n = 1), fever (n = 2), sneezing and ocular discharge (n = 4), to moderate respiratory and/or gastrointestinal disease (n = 3). The Spanish cat was euthanized due to severe and deteriorating respiratory distress; however, its condition was attributed to underlying hypertrophic cardiomyopathy and the positive SARS-CoV-2 test finding was considered incidental (MAPA, 2020a).
Of the remaining infected cats, nine have fully recovered either naturally or following supportive treatment; follow-up data on the Russian cat have not been provided to date.
While most of the infected cats were from single-cat households, the German case provided evidence of variable susceptibility to infection with and low transmission of SARS-CoV-2 between cats: three cats were resident in a retirement home that was experiencing a COVID-19 outbreak, and only one tested positive, even after all three cats were subsequently quarantined together (ProMED-mail, 2020j). reported to have tested positive for SARS-CoV-2 during a university study, but negative results were obtained in subsequent confirmatory testing and the animal was eventually classified as negative (Fisher, 2020;Hauser & Gross, 2020).

| Farmed mink
The Netherlands

SARS-CoV-2 infections in farmed American mink (Neovison vison)
were first reported on 26th April 2020, after "a few" mink on two Dutch fur farms showed respiratory and/or gastrointestinal disease, and increased mortality (Schouten, 2020b). A national reporting requirement for any sign of respiratory disease and/or increased mortalities in farmed mink was introduced from 26th April (Schouten, 2020b). Protective measures were implemented on and around all infected farms, including visitor bans, transport bans for mink and mink manure, and PPE for farm employees. Public roads within a 400m radius of each premises were initially closed, but later reopened after testing of dust and air samples revealed that no virus was present outside the animal facilities (Schouten, 2020a). Additional surveillance and testing activities are described in the Surveillance section. Larsen & Zuferov, 2020).

| Testing of pets of human COVID-19 patients
The 5th meeting of the OIE Advisory Group on COVID-19 (OIE, 2020a) reported that routine testing of companion animals is being discouraged to "avoid concerned owners making unnecessary visits to their private veterinarian, which may increase public health risks through person to person contact". It was also noted that "there have also been instances of humans submitting their own samples disguised as their pet's sample to veterinarians in a desperate attempt to be tested". The possibility of including a species probe in the diagnostic testing was discussed as a way of detecting this activity.

Hong Kong
Early in the pandemic, Hong Kong authorities took the unique step of routinely quarantining and testing companion animals owned by COVID-19 patients (Denis et al., 2020). News reports suggested that as of 26th April, only the three known cases (two dogs and one cat as described in the case summaries) from 52 sampled pets were positive for SARS-CoV-2 (2/32 dogs, 1/18 cats and 0/2 hamsters) following PCR testing of nasal, oral and faecal samples (Cheng, 2020;Deacon, 2020). While the sample size is small, the infection prevalence of 5.8% among the tested animals is nevertheless of interest.

Europe
Some laboratories in Italy are reportedly preparing to test pets (OIE, 2020b). The universities of Padua and Venice are reportedly planning serosurveys for SARS-CoV-2 in the small town of Vo, which was the epicentre of the initial outbreak in Italy, in order to assess the potential role of cats in the epidemiology of the disease (ProMED-mail, 2020b;ProMED-mail, 2020c). No routine testing of pets appears to be occurring in other European countries, although in Germany any animal that shows respiratory disease symptoms and has had contact with a COVID-19 positive human is eligible for testing (Denis et al., 2020;OIE, 2020b). In the Netherlands, some companion animals of COVID-19 patients with clinical disease appear to be being tested for SARS-CoV-2, as evidenced by a report of a positive test result in eight-year-old American bulldog that was euthanized due to severe deteriorating respiratory disease (Schouten, 2020a).
Molecular diagnostic testing did not detect SARS-CoV-2; however, serological testing identified antibodies to SARS-CoV-2 in the dog's blood. It was not reported whether the dog's condition was attributed to SARS-CoV-2 or to another cause. The Dutch Agriculture Minister subsequently implemented requirements for veterinarians to report suspected SARS-CoV-2 infections in animals, and for laboratories to report positive test results, to the Netherlands Food and Consumer Product Safety Authority (Schouten, 2020a).
A French study by Temmam et al. (2020)
Specimens were also tested in parallel with three assays from the CDC and all samples tested negative (Brulliard, 2020;IDEXX, 2020d;ProMED-mail, 2020a bears, yellow-throated martens, weasels, red pandas and wild boar) using a double-antigen sandwich ELISA, with no positive results (Deng et al., 2020). Of most note is the lack of positive ELISA results among 15 pet and 99 street dogs from Wuhan (Deng et al., 2020).
Otherwise, there is very limited information presented regarding the animal populations from which the samples were derived, and so very little inference can be made from this paper (Denis et al., 2020;Weese, 2020).  ProMED-mail, 2020l).

Great apes and other wild animal populations
While no natural or experimental cases of SARS-CoV-2 in great apes have been reported to date, their known susceptibility to common human respiratory viruses including rhinovirus C (the common cold) (Negrey et al., 2019) and the similarity of the ape form of the ACE2 receptor used by SARS-CoV-2 to infect human cells (Melin et al., 2020) has caused concern among primatologists and conservationists worldwide. Wildlife preserves across Africa and Asia have reportedly closed to the public to minimize human contact.
Research teams in Tanzania, Cote d'Ivoire and Uganda, among others, are reportedly collecting faecal samples from wild gorillas and chimpanzees for virus testing, and training wildlife rangers to identify signs of respiratory distress in apes while also introducing PPE, quarantine and social distancing measures for in-contact researchers (Gibbons, 2020). Use of PPE including face masks, gloves and coveralls was also recommended by the US National Wildlife Health Center for researchers conducting any wildlife handling or investigations into wildlife mortality events (Sleeman, 2020).
One predictive modelling study identified 291 bat species that are likely to be undetected hosts of betacoronaviruses, including 30 Rhinolophus species in addition to the 16 known hosts in this genus, suggesting that the potential bat reservoir for SARS-CoV-2 may be two-thirds higher than currently described . A rapid risk assessment conducted by the US Geological Survey states a non-negligible risk of human-to-bat transmission of SARS-CoV-2 in north America at present, but estimates that implementation of appropriate PPE including N95 respirator masks, dedicated outer clothing, and gloves, would reduce the exposure risk by 94%-96% (Runge et al., 2020). Should SARS-CoV-2 be introduced into the north American bat population, there would be an estimated 33.3% likelihood that the virus could spread and become established, thereby creating a potential reservoir of infection for humans and other animal species (Runge et al., 2020).
The OIE "EBO-SURSY" project is reportedly proposing to test 3,000 samples previously collected from bats for haemorrhagic fever virus surveillance in West Africa for the presence of coronavirus, to investigate potential circulation of precursor viruses to SARS-CoV-2 (OIE, 2020b). American research institutes are also reportedly conducting infection studies in brown bats, to investigate the possibility of whether these wide-ranging animals could become a reservoir for SARS-CoV-2 following human-to-bat transmission (OIE, 2020b), and surveillance of native north American wildlife generally to investigate their potential as reservoirs for SARS-CoV-2 (Sleeman, 2020).

| D ISCUSS I ON
While the initial SARS-CoV-2 spillover event is believed to have been a zoonotic transfer from bats to humans, possibly by way of an unidentified intermediate animal host, the driving force behind the pandemic has undeniably been human-to-human transmission (WHO, 2020 There is no evidence that people are at risk of contracting infection from their pets, and the OIE has advised that there is no justification in taking measures directed at companion animals that may compromise their welfare (OIE, 2020f). During the early stages of the pandemic, there were reports of stray dogs and cats rounded up in Russia (Balmforth, 2020), companion animals owned by confirmed COVID-19 patients routinely euthanized by authorities in China, of large-scale abandonment of pet dogs and cats and even of pets being thrown to their deaths from high-rise buildings by fearful owners following early reports of pets testing positive for SARS-CoV-2 (Campbell, 2020). This occurred despite no evidence of animal-to-human transmission of the virus in domestic settings. In fact, the results of one modelling study predicted that abandoning cats could actually increase incidence of COVID-19 in urban settings .
While SARS-CoV-2 infections in humans and animals within domestic settings are unlikely to contribute to community transmission, in high-density animal environments, such as on farms with susceptible species, the risk of anthroponotic and zoonotic transmission must increase substantially. The lack of susceptibility of poultry and pigs to SARS-CoV-2 infection is reassuring, particularly given the heavy losses to these highly intensive commercial industries following recent and ongoing outbreaks of avian influenza and African swine fever, respectively (FAO, 2020b;Tian & Cramon-Taubadel, 2020).
Commercial farming of mink and ferrets, which are known to be susceptible to SARS-CoV-2 infection and replication (Kim et al., 2020;Richard et al., 2020;Schlottau et al., 2020;Shi et al., 2020), however, provides ideal conditions for viral transmission and spillover. China produced 26.2 million mink pelts or nearly 40% of the global mink pelt harvest in 2016 (Lung & Lin, 2019), while other major mink producers include Denmark (17.8 million), Poland (8.5 million), the Netherlands (5.5 million), the USA (3.5 million) and Canada (2.1 million) (Bale, 2016;Ingman, 2015;Lung & Lin, 2019). The high numbers of confirmed human COVID-19 cases in these countries (Dong et al., 2020), coupled with particularly high densities of mink and other small mammals at this time of year due to seasonal breeding patterns (Amstislavsky et al., 2008), means it is possible that unre- Of additional concern is the potential for anthroponotic transmission of SARS-CoV-2 to vulnerable wild animal populations.
While no natural or experimental cases of SARS-CoV-2 in great ape species have been reported to date, their known susceptibility to common human respiratory viruses including rhinovirus C (the common cold) (Negrey et al., 2019), the similarity of their ACE2 receptors to the human form that SARS-CoV-2 uses to infect human cells (Melin et al., 2020), and the experimental studies that have demonstrated SARS-CoV-2 infections in other primate species (Lu et al., 2020;Rockx et al., 2020;Woolsey et al., 2020), suggest that SARS-CoV-2 is a tangible threat to great ape populations.
Concern for highly vulnerable wild populations of great apes including chimpanzees and gorillas has already seen implementation of protective measures in wildlife preserves across Africa and Asia. Visitors have been banned from many such parks since early March 2020 to limit human contact with wild apes, and PPE, quarantine and social distancing measures for in-contact researchers have also been implemented (Gibbons, 2020 (Runge et al., 2020).
The authors of this risk assessment identified several critical uncertainties that could affect their estimates of SARS-CoV-2 entering bat populations and associated sequelae, including likelihoods of human-to-bat, bat-to-bat, bat-to-animal and bat-to-human viral transmission dynamics; pathogenesis and replication of the virus in bat tissues; and seasonal impacts including bat breeding and hibernation patterns on virus replication and transmission (Runge et al., 2020).
Until further research results are available to fill these knowledge gaps, a precautionary use of PPE and a conservative approach to wildlife handling and interaction would be encouraged.
Other adverse effects of SARS-CoV-2 on animals may include the "panic slaughter" of species that are mistakenly blamed for transmitting disease. The reputation of bats as the probable source of COVID-19 has reportedly led to instances of fearful citizens setting fire to roosting bats , and prompted calls for their mass slaughter to protect public health (Zhao, 2020). Even less extreme requests for the removal of hibernating bats from human residential areas could disturb their delicate physiological balances, lead to high bat mortality and potentially increase the spread of other viruses (Zhao, 2020). In addition, many bat species are endangered, and those that have become habituated to urban life may not survive in the wild, placing already fragile populations at further risk (Zhao, 2020). Masked palm civets (Paguma larvata) are widely accepted to have been the intermediate host species for SARS-CoV-1 (Gong & Bao, 2018), which led to the slaughter of thousands of wild civets in southern China using a variety of inhumane methods, including clubbing and drowning in disinfectant, following outbreaks in the region (Parry, 2004). Pangolins have been put forward as potential intermediate hosts for the current COVID-19 pandemic (Andersen et al., 2020;Liu et al., 2020;Zhang, et al., 2020b), leading to concern from conservationists that pangolins, already endangered and reportedly the world's most trafficked mammal, may be placed at further risk by "ecological killing" as part of disease control efforts (Standaert, 2020). On the other hand, the zoonotic origins of the COVID-19 pandemic have increased global calls to end the trading of wildlife for all purposes, including "wet" animal markets and wild game hunting for human consumption, and the traditional medicine, tourism, wild pet and fur farming sectors still existing across the globe. In March 2020, China introduced a ban on hunting, trading, transporting and eating the meat of wild animals (NPC, 2020) and in early June upgraded the protection of all pangolin species to the highest level, banning all trade in pangolins and their products and removing pangolin scales from their approved list of traditional medicines (Yueming, 2020). While this is to be applauded, China's reported proposal to reclassify mink, raccoon dogs, and silver and blue foxes from wild animals to domestic livestock (Dalton, 2020) may cause additional concerns, both for animal activists, and public health advisors concerned about extended SARS-CoV-2 transmission and the potential for future zoonotic pandemics.

ACK N OWLED G EM ENTS
The authors would like to thank Trevor Drew, Debbie Eagles, Dwane O'Brien, Kim Halpin, James Watson and Mark Ford for their inputs throughout the manuscript writing process. We are also grateful to Peter Durr and Kerryne Graham for their support with this study.
Funding for the publication was provided by the Commonwealth Scientific and Industrial Research Organisation (CSIRO)'s Australian Centre for Disease Preparedness (ACDP).

D I S C LO S U R E S TAT E M E N T
Please note that similar sections of this paper have been previously published by the authors in the "Covipendium", a non-peer-reviewed living paper curated by the Rega Instituut, KU Leuven, Belgium (Denis et al., 2020).

CO N FLI C T O F I NTE R E S T
The authors have no conflicts of interest to declare.

E TH I C A L A PPROVA L
The authors confirm that the ethical policies of the journal, as noted on the journal's author guidelines page, have been adhered to. No ethical approval was required as this is a review article with no original research data.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analysed in this study.