Endoparasites infecting exotic captive amphibian pet and zoo animals (Anura, Caudata) in Germany

Alongside exotic reptiles, amphibians, such as toads, frogs, salamanders, and newts, are nowadays considered popular pets worldwide. As reported for other exotic pet animals, amphibians are known to harbor numerous gastrointestinal parasites. Nonetheless, very little data are available on captive amphibian parasitic diseases. In this study, we applied direct saline fecal smears (DSFS) to examine in total 161 stool samples from 41 different amphibian species belonging to the orders Anura and Caudata. In addition, carbolfuchsin-smear (CFS) staining (n = 74 samples) was used to detect amphibian Cryptosporidium oocysts. Also, complete dissections of deceased amphibians (n = 107) were performed to specify parasite infections and to address parasite-associated pathogenicity. Overall, examined amphibian fecal samples contained 12 different parasite taxa. The order Rhabditida with the species Rhabdias spp. and Strongyloides spp. were the most prevalent nematode species (19.3%), followed by flagellated protozoans (8.7%), Amphibiocapillaria spp./Neocapillaria spp. (7.5%), Oswaldocruzia spp. (4.3%), Blastocystis spp. (3.1%), Cosmocerca spp. (3.1%), oxyurids (Pharyngonoidae) (3.1%), spirurids (1.2%), un-sporulated coccidian oocysts (0.6%), Tritrichomonas spp. (0.6%), Karotomorpha spp. (0.6%), and Cryptosporidium spp. (0.6%). One CFS-stained fecal sample (1.4%) was positive for Cryptosporidium oocysts. Within dissected amphibians, 31 (48.4%) of the anurans and 11 (26.2%) of the salamanders were infected with gastrointestinal parasites. One cutaneous Pseudocapillaroides xenopi infection was diagnosed in an adult African clawed frog (Xenopus laevis). Etiologically, 17 (15.9%) of them died due to severe parasitic and/or bacterial infections (e.g., Chryseobacterium indologenes, Citrobacter freudii, Sphingobacterium multivorum, Klebsiella pneumoniae). High prevalence and pathological findings of several clinical amphibian parasitoses call for more detailed investigation on gastrointestinal parasite-derived molecular mechanisms associated with detrimental lesions or even death.


Introduction
Amphibian species are ectothermic, tetrapod, and mainly carnivorous vertebrate species of the class Amphibia (Hill et al. 2015). Modern amphibians inhabit a wide variety of habitats, with most species living within terrestrial, fossorial, arboreal, or freshwater aquatic ecosystems. In the last decades, amphibians established themselves as domestic exotic pets/ zoo animals worldwide (Mutschmann 2010). Amphibian species have suffered a significant decline in the wild during the last decades, mainly due to anthropogenic pressure, such as environmental contamination, UV-B irradiation, introduction of alien/invasive species, direct mistreatment, habitat losses, c l i m a t e c h a n g e s , a n d e m e r g i n g d i s e a s e s ( e . g . , chytridiomycosis, Ranavirus) (Daszak et al. 1999;Stuart et al. 2004;Beebee and Griffiths 2005;Collins 2010;Bishop et al. 2012;Henle et al. 2012;Foden et al. 2013;Martel et al. 2014;Wren et al. 2015;Nguyen et al. 2017) and have attracted special media and public attention. This public attention among other factors has raised interest for these ectothermic animals resulting in an increased popularity of amphibians as private pets, also in Germany (Krautwald-Junghanns 2017).
International amphibian trade is becoming governmentally more and more restricted in many countries. Radical Section Editor: Elizabeth Marie Warburton * Malek J. Hallinger malekhallinger@web.de 1 restrictions seem to be impossible to implement and also require time, effort, and knowledge with no guarantee of success. Even non-regulated trade/black market activities happen as suggested elsewhere (Garner et al. 2009;Bishop et al. 2012). More importantly, diverse anuran (i.e., frogs and toads) and caudate (i.e., salamanders and newts) amphibians threatened by extinction are nowadays part of many zoological collections worldwide (Bishop et al. 2012;Ziegler 2016;Ziegler and Rauhaus 2019).
Since there is a considerable overlap within described diseases for captive amphibians/reptiles/fishes and free-ranging amphibians, caretakers must be mindful on endoparasites and diseases while co-housing amphibians with wild ones or other zoological taxa, such as fishes (Densmore and Green 2007). Parasitic diseases of amphibians are closely related to parasitoses affecting other ectothermic vertebrates (Densmore and Green 2007). Therefore, parasite species can be transmitted from other ectothermic vertebrates to amphibians (i.e., fish ectoparasites, such as the protozoan species, Trichodina spp. and Ichthyobodo spp.) (Densmore and Green 2007;Mutschmann 2010). Most amphibian pets or other lower vertebrates, such as reptiles maintained in captivity, are often associated to inadequate husbandry and mismanagement conditions (Beck and Pantchev 2013;Wolf et al. 2014). In addition, specific intrinsic associated risk factors (e.g., age, sex, species, host immune status) and extrinsic risk factors (e.g., poor hygiene housing conditions, temperature, humidity, animal density, nutrition) might lead to relevant parasitic burdens (Mutschmann 2010;Beck and Pantchev 2013;Hallinger et al. 2019;2020). Reinfection with resistant reproductive stages of certain endoparasites (e.g., oxyurid eggs) can lead to heavy parasitism and/ or even death of pet reptiles/amphibians (Frank 1981;Pasmans 2008;Beck and Pantchev 2013;Wolf et al. 2014;Hallinger et al. 2018).
Amphibian hosts can be infected with different gastrointestinal parasites, such as protozoans, nematodes, cestodes, trematodes, acanthocephalans, and pentastomids (Frank 1984;Vaucher 1990; Al-Sorakhy and Amr 2003; Barton and Riley 2004;Densmore and Green 2007;Mutschmann 2010). Some of them bear zoonotic potential, such as helminths (Spirometra spp., Gnathostoma spp., Diphyllobothrium spp., Alaria spp., and Echinostoma spp.) and pentastomids (Pentastoma), since most of them are food-borne diseases (Graczyk und Fried 1998;de Górgolas et al. 2003;Dorny et al. 2009;Pantchev und Tappe 2011;Warwick et al. 2012). Up to date, most scientific research have focused on free-ranging amphibians or laboratory animals (Coggins and Sajdak 1982;Cunningham et al. 1996;Hamann et al. 2012;Kuzmin et al. 2003;Loras et al. 2010;Mohammad et al. 2010;Rizvi and Bhutia 2010;Amin et al. 2012;Yildirimhan et al. 2012), but seldom on captivity kept anuran and caudate amphibians. Thus, this comprehensive investigation on German captive amphibian pets of private households and German zoological collections aims to provide current data on the occurrence of gastrointestinal endoparasites, to assess differences in parasite occurrence between privately kept animals and zoo animals considering host species, keeping facility, sex, and/or order/taxon. In addition, we aim to assess presence of zoonotic parasites circulating in amphibian pets and further to gain a better understanding of parasite-derived pathogenicity in these exotic herpetic pets.

Fecal samples
Examined fecal samples originated either from animals owned privately, submitted by owners attending veterinarians, or by different German zoos entities, which had been referred to exomed® laboratory in Berlin, Germany. In order to identify both protozoan and helminth stages, we performed direct saline fecal smears (DSFS) for general parasitological diagnosis according to Barnard and Upton (1994). Clients were also asked to provide a printed form containing individual animal's signalement (i.e., species, sex, age), husbandry circumstances (i.e., origin, animal density, time in owner's possession), previous parasitological examinations, and anthelminthic treatments. At exomed® laboratory, all stool samples were labeled with corresponding forms and reference numbers and finally conserved at 4°C for up to 2 h in a lab refrigerator until further parasitological examination.
For DSFS, a uniform solution was created by mixing 1 g of amphibian feces at a ratio of 1:1 with 0.9% saline solution, carefully placed on glass cover slides (Nunc) with pipette (Nunc) and finally covered with cover slips (22 × 22 mm; Nunc). Both a 100× and/or 400× magnification for light microscopy examination (Axio Imager M1®, Zeiss, Jena) equipped with a digital camera were used here.

Amphibian autopsies
Amphibian corpses of deceased pet animals were necropsied at exomed® laboratory (Table 2, Online supplement). Clients were also asked to provide a printed form containing individual animal's signalement (i.e., species, sex, age) and husbandry circumstances (i.e., origin, animal density, time in owner's possession). Additionally, amphibian hosts were morphologically identified using corresponding published literature (Hofrichter 2000). In addition, pathohistological examinations were performed using standard hematoxylin and eosin (H&E) batch staining (Buesa 2007) of the following amphibian organs: liver, lungs, intestine, and kidneys. First, a visual inspection of the whole digestive tract was conducted to unveil presence of macroscopic helminth endoparasites. Afterwards, intestinal contents were examined by DSFS method. Morphological identification of endoparasites was performed under a light microscope equipped with a digital camera (Axio Vision M1®, Zeiss, Jena). External examinations and necropsies were performed as described previously by Mutschmann (2010).
Concerning non-infectious death causes, we identified in a Betsileo Madagascar frog (Mantidactylus betsileanus) a severe egg-related stasis most probably due to reproductive disorders. Moreover, in a Dyeing dart frog (Dendroabtes tinctorius), we found an advanced renal carcinoma, and in an African clawed frog (Xenopus laevis), a cutaneous lymphadenoma. In 14 (13.1%) animals, the final cause of death remained unknown. Besides, one Batrachochytrium dendrobatidis-infected axolotl salamander (Ambystoma mexicanum) was co-infected with Ichthyobodo spp.
ectoparasites. An overview on dissected animals, the diagnosed endoparasites and combined microbiological, pathological, and pathohistological findings is provided in Table 6.

Discussion
Parasite infections in free-ranging amphibians seem to appear obligatory worldwide, and thus, very high prevalences of up to 90% have previously been described (Coggins and Sajdak 1982;Al-Sorakhy and Amr 2003;Amin et al. 2012). For instance, Rizvi et al. (2011) sampled free-ranging amphibians in an Indian Wildlife Sanctuary (Haryana) and found that endemic common dicroglossid frogs (Euphlyctis cyanophlyctis) were frequently infected (52.9%) by nematodes. In contrast to this wildlife study, there is very little knowledge on parasitic infections of dicroglossid frogs (E. cyanophlyctis) kept in captivity. While comparing our prevalence data with previous published studies, it should be considered that most of these surveys were conducted in wild animals, and this fact might explain prevalence differences. Most likely to dicroglossid frogs, other freeranging amphibians are also showing higher parasitic prevalences when compared with those kept in captivity (Coggins and Sajdak 1982;Amin et al. 2012). Moreover, sensitivity and specificity of applied DSFS to detect helminth and protozoan stages might have influenced observed prevalence as different diagnostic methods in former wildlife studies have been used (Rizvi et al. 2011;Amin et al. 2012 quality can directly affect parasitic burdens and prevalences in free-ranging amphibians (Vaucher 1990;Kehr and Hamann 2003;Marcogliese and Pietrock 2011;Thiemann and Wassersug 2000), very little is still known whether these factors might also influence the outcome of parasitic burdens in pet amphibians kept in households or zoos (Mutschmann 2010). In this study, helminth infections occurred frequently in investigated animals (Table 4). All nematode species found in this survey have been reported to possess pathogenic significance for amphibians (Mutschmann 2010;Amin et al. 2012;Langford and Janovy 2009;Langford 2010;Yildirimhan et al. 2012 (Densmore and Green 2007). In free-ranging amphibians, Rhabdias infections seem to occur frequently and sometimes result in pneumonia (Kuzmin et al. 2003;Mohammad et al. 2010;Fernández Loras et al. 2011). Consistently, Rhabdias spp. infection rates for captive German amphibians were rather high in this study (19.3%) and resulted in the most prevalent parasites. Rhabdias/Strongyloides infection rates varied significantly within taxon, i.e., caudates were less frequently infected (3.13%) than anurans (22.83%). Nonetheless, it is well known from literature that Rhabdias is more frequently parasitizing frogs/toads (Langford and Janovy 2009;Langford 2010). In line, Rhabdias ranae seems not capable to infect caudates and to be restricted to frogs/toads as suitable hosts, but in the past two decades, first Rhabdias infections in caudates have been reported (Kuzmin et al. 2001;Kuzmin et al. 2003;Eisenberg and Pantchev 2009). Therefore, it seems assumable that anurans might be more often infected with Rhabdias than caudates, especially because the correlation was rather high (r = 0.44) when comparing these two amphibian groups (Cohen 1988). Clinical relevance of rhabdiosis was also underlined in dissections, since in one adult male Australian green tree frog (L. caerulea), a Rhabdias spp.-infected lung was found and which might have caused severe pneumonia, hepatitis, and nephritis. Nonetheless, other pathogens could not be ruled out as the same animal showed secondary bacterial infections with Chryseobacterium indologenes (+++) and Pseudomonas fluorescens (+) isolated from the frog's coeloma. Alongside Rhabdias, other nematode genus, i.e., Oswaldocruzia, was frequently diagnosed (2.8%) in domestic kept amphibian pets. Oswaldocruzia nematodes infect amphibian hosts exclusively by the oral uptake of exogenous infective L3 (Hendrikx 1983). Noteworthy, a cutaneous Pseudocapillaroides xenopi infection was diagnosed in an adult African clawed frog (X. laevis). This X. laevis-infected animal suffered not only of a severe verminous dermatitis but also of secondary Gram-negative P. fluorescens (+++), Aeromonas hydrophila (++), and Citrobacter braakii (++) dermal infections. The amphibian nematode P. xenopi infects the epidermis and can cause clinically symptoms, such as erythematous/erosive dermatitis, with characteristic roughness of affected skin, petechiae, and dermal ulcera (Cunningham et al. 1996;Mutschmann 2010  P. xenopi can complete its direct life cycle within epidermis of frogs/toads in which burrowing activities of subdermal nematodes can lead to the damage of parasitized skin. Therefore, P. xenopi-infected animals are more susceptible for bacterial and/or fungal secondary dermal infections (Cunningham et al. 1996), as confirmed in our investigation. According to protozoan enteric infections, in 14 cases (8.7%), potentially pathogenic, flagellated protozoan genera, such as Proteromonadida, Reteromonadida, Diplomonadida, and Trichomonadida, were additionally diagnosed. Nonetheless, the literature considers many of these enteric flagellates as commensals within intestinal tract of amphibians (Densmore and Green 2007;Mutschmann 2010). Conversely, some genera of diplomonadids (Giardia, Hexamita, Spironucleus) and trichomonadids (Monocercomonas, Hexamastix, Tritrichomonas) can cause weight loss, general edema, and enteritis in severely infected animals.
The clinical relevance of flagellated protozoan infections was demonstrated during conducted dissections: Out of all dissected animals, four (3.7%) died because of severe Tritrichomonas spp.-and/or Spironucleus spp.-derived enteritis. These animals showed severe catarrhalic-to hemorrhagicnecrotic enteritis combined with secondary bacterial infections (e.g., Pseudomonas spp./Sphingobacterium spp.) of liver and gut mucosa.
Only five animals (3.7%) were positive for Blastocystis spp. infections. Conversely to our findings, Yoshikawa et al. (2004) found anurans and newts from distinct locations in Japan to be infected with Blastocystis showing very high prevalences (47.8-100%) by using in vitro culture diagnostic methods. Our observed Blastocystis prevalence might have been higher if this in vitro cultivation method would have been applied, but it cannot be excluded that this parasite is simply less frequently found in German pet amphibians. Unfortunately, there is still very little knowledge on amphibian-related blastocystiosis. The same holds true for its possible impact on animal health kept in captivity (Mutschmann 2010). Nevertheless, Blastocystis should be considered as potentially pathogenetic protozoan species and infections should be considered according to clinical symptoms. Moreover, during dissections we here diagnosed Entamoeba spp. cysts in three (2.8%) animals.
Several studies have focused on gastrointestinal apicomplexan coccidian parasites in amphibians (Duszynski et al. 2007). So far, monoxenous coccidian genera Eimeria, Goussia, Hyaloklossia, and Cystoisospora (former Isospora according to new nomenclature) have been described in diverse amphibian host species (Duszynski et al. 2007), and for further review a disposed online version (http://biology.unm. edu/coccidia/anura.html) is recommended. In accordance with these reports, we also diagnosed un-sporulated coccidian oocysts in one animal (0.6%), but amphibian oocysts were not fully identifiable to species level. Furthermore, non-sporulated Eimeria spp. oocysts were found within gut lumen of one dissected fire salamander, (Salamandra salamandra) but coccidian-derived death was ruled out as this animal was also co-infected with Aplectana spp., Spironucleus spp., and Tritrichomonas sp. and showed a manifested mycotic dermatitis.
Enteropathogenic apicomplexan Cryptosporidium is known to infect also the microvillus border of amphibian gastrointestinal epithelial cells (Jirků et al. 2008). Consistently, we diagnosed Cryptosporidium oocysts in an Australian frog (L. caerulea) via CFS analysis. If here identified Cryptosporidium oocysts were shed during a patent infection or whether they were passed because of Cryptosporidium spp.-infected prey animal consumption (e.g., feeding of baby mice) remains unclear. Since Cryptosporidium can be transmitted by ingestion of infected food animals, poorly treated water as well as direct contact with infective oocysts, it is possible to assume that human infections might occur through ingestion of under-cooked frog (Rana spp.) meat and/or handling and processing of Cryptosporidium-infected frogs as recently demonstrated in Africa . Former study revealed a high prevalence of Cryptosporidium spp. (35.9%) in the intestine of 117 frogs (Rana spp.) sold at the Hanwa frog market Zaria, Kaduna State, Nigeria, for human consumption . Therefore, further public health studies on different transmission routes of this neglected anthropozoonotic parasite should be conducted, including amphibians designated for human consumption .
Aside from protozoans, nematodes, cestodes, and trematodes, no acanthocephalan infections were here detected. Nonetheless, during necropsies, also cestode-parasitized animals were found. As such, in three dissected animals (2.8%), various long cestode specimens containing mature proglottids were diagnosed. Noteworthy was a heavily Nematotaenia-infected male Dyeing dart frog (Dendrobates tinctorius), which showed obstipation and congestion of ground substrate in the gut lumen. Amphibians are known to be infected by different cestode genera, i.e., Proteocephalus, Ophiotaenia, Cephalochlamys, Bothriocephalus, Nematotaenia, Distoichometra, Cylindrotaenia, and Baerietta. Clinical symptoms of nematotaeniosis manifest in affected animals during stress and/or in case of heavy infections (Mutschmann 2010). Then, ileus with obstipation, blood loss, necrosis of intestinal mucosa, edema or even death may also occur if untreated (Mutschmann 2010). Interestingly, a digenean trematode infection was found in a deceased yellowbellied toad (Bombina variegata), showing clinical symptoms, including hydrocoeloma, generalized edema, and pathohistological findings, such as hepatitis, enteritis, nephritis, and a bacterial co-infection (Citrobacter spp. +). Amphibians represent not only intermediate hosts for various digenean trematode orders (e.g., Amphistomida, Echinostomatida, Gasterostomida, Hemiurida, Holostomida, Plagiorchida) but also second or even final hosts. Nevertheless, trematode-driven pathological effects are mostly unknown for amphibians (Mutschmann 2010). Since many of examined amphibians in this study are considered as threatened endemic species of neotropical regions, e.g., Adelphobates galactonotus, Phyllomedusa bicolor, and Trachycephalus resinifictrix, and thus being kept as zoo animals for conversation reasons, detected parasites in these animals might represent imported parasites from their natural tropical habitats. Therefore, it seems noteworthy to mention that new wild amphibians introduced into zoological gardens should undergo a mandatory quarantine regime in order to avoid further spread of neozoan parasites as suggested elsewhere (Hallinger et al. 2019;2020).
Interestingly, parasitic infection rates in investigated anurans (51.12%) were significantly higher than the ones observed in caudate species (12.88%). As proposed for zoo animals, it is also recommended that newly purchased frogs, newts, and toads by private owners should be submitted to parasitological examination in order to detect presence of gastrointestinal parasites during quarantine as a routine health screening.

Conclusions
Our representative epidemiological survey on endogenous parasites of captive amphibians in Germany found several pathogenic parasite species resulting in clinically manifested disease. If correctly diagnosed, identified parasitoses should be medicated taking into account commensalism, metabolic features of amphibians, clinical signs, and more importantly prophylactic approaches in order to prevent future infections. Applied DSFS technique on scat samples proved to be valid for detection of many relevant parasitic stages, including tiny protozoan oocysts. Since biology, epidemiology, as well as pathogenesis of most amphibian parasitoses are not well understood, further investigations in these directions are needed. Similarly, the current lack of suitable therapy options for many of these amphibian parasitoses calls for more research in new drug development within the field of neglected herpetology medicine.