Multispecies reservoir of Spirometra erinaceieuropaei (Cestoda: Diphyllobothridae) in carnivore communities in north-eastern Poland

Background Spirometra erinaceieuropaei is a diphylobothriid tapeworm with a complex life-cycle including definitive, intermediate and paratenic (transport) hosts. Multiple routes of parasite transmission often make it impossible to determine what type of host a specific infected animal is considered to be. Spargana larvae cause sparganosis, a severe food- and water-borne disease mainly found in Asia. In Poland, Spirometra sp. was reported in large carnivores in Białowieża Primeval Forest for the first time in the 1940s and was recently confirmed as S. erinaceieuropaei in several mammals and snakes using molecular methods. Methods In total, 583 carcasses of 9 carnivore species were necropsied between 2013 and 2019 in north-eastern (NE) Poland. The larvae of S. erinaceieuropaei (spargana) were isolated from subcutaneous tissue, counted, and preserved for genetic analyses. We calculated the prevalence and intensity of infection. To assess spatial variation in S. erinaceieuropaei infection probability in NE Poland, we applied a generalized additive model (GAM) with binomial error distribution. To confirm the species affiliation of isolated larvae, we amplified a partial fragment of the 18S rRNA gene (240 bp in length). Results Spirometra larvae were found in the subcutaneous tissue of 172 animals of 7 species and confirmed genetically as S. erinaceieuropaei. The overall prevalence in all studied hosts was 29.5% with a mean infection intensity of 14.1 ± 33.8 larvae per individual. Native European badgers and invasive raccoon dogs were characterized by the highest prevalence. An analysis of parasite spread showed a spatially diversified probability of infection with the highest values occurring in the biodiversity hot spot, Białowieża Primeval Forest. Conclusions Our study revealed that various mammal species (both native and non-native) can serve as S. erinaceieuropaei reservoirs. The frequency and level of infection may differ between selected hosts and likely depend on host diversity and habitat structure in a given area. Further studies are needed to assess the distribution of the parasite throughout Europe and the environmental and biological factors influencing infection severity in wild mammals.

new environments [2]. Parasite transmission is common among members of foraging guilds [3]. For most of helminth parasites, host specificity appears less defined at the intermediate host stage than at the definitive host stage, with non-adult intermediate (larval) stages able to infect different organs and tissues of diverse intermediate hosts. The range of specificity of the parasite is a crucial determinant of its invasive capacity and the likelihood of new host-parasite combinations occurring [4]. In addition, several parasites use paratenic (transport) hosts, where parasite larvae show no development [5]. The role of paratenic hosts is important; they facilitate contact between parasite larvae and the definitive host, contributing to an increase of prevalence in a specific host population [6].
The life-cycle of Spirometra spp. may also include paratenic hosts. Second intermediate hosts infected with plerocercoid larvae can be preyed upon and thus reach a wide variety of tetrapods such as birds or mammals (including humans) that may serve as paratenic hosts for this parasite [21,22]. Multiple routes of parasite transmission often make it impossible to determine what type of host a specific infected animal is considered to be. After passing through the intestinal wall, spargana settle in host tissues and cause sparganosis [23,24]. Paratenic hosts, however, are unnecessary for the completion of life-cycle, thus the larval Spirometra can infect and subsist in numerous species of paratenic hosts until finally consumed by felids or canids, serving as definitive hosts [25], though they are also reservoirs of spargana for other carnivores. It has also been found that some canids, such as the red fox (Vulpes vulpes Linnaeus, 1758) and the raccoon dog (Nyctereutes procyonoides Gray, 1834), may serve as both definitive and paratenic hosts [7,26,27].
The role of wild carnivores as transmission vectors for zoonotic diseases has been widely described [41][42][43][44] and includes both native as well as non-native invasive species. Invasive species are important for spreading and transmitting of diseases because they can carry their own parasites and acquire new ones during the colonization of new territories [45,46]. Nothing is known about the role of invasive species in the spread of sparganosis, and very little about the contribution of native carnivores to the maintenance of this disease in European wildlife.
The main goal of this paper was to investigate the spread of S. erinaceieuropaei in well-preserved communities of wild carnivores in north-eastern (NE) Poland. We aimed to: (i) investigate which carnivore species (native and/or non-native invasive) may serve as S. erinaceieuropaei hosts in wildlife and whether the parasite's presence is only local or perhaps more widespread; (ii) calculate the infection parameters, prevalence and intensity; (iii) assess spatial patterns of infection probability in the study area; and (iv) reveal whether wild canids and felids, the typical definitive hosts, may also serve as paratenic hosts carrying spargana.

Study area
Our study was performed in Podlasie and the eastern part of the Masurian Lakes region, north-eastern (NE) Poland. Both regions are characterized by well-preserved forest ecosystems (with four national parks and a large area covered by Natura 2000, a network of nature protection areas in the territory of the European Union) and greater forest cover (32%) in comparison to the rest of Poland [47]. Our main sampling sites were large forested areas: Augustów Forest (AF); Białowieża Primeval Forest (BPF); Biebrza Valley (BV); Knyszyn Forest (KF); and the Masurian Lake District (MLD). AF, KF and MLD are dominated by coniferous forests [48]. In BPF over 60% of the area is covered by deciduous and mixed forests with high species richness [49]. Wet habitats cover over 40% of the area [48], and BV is the largest marshland in central Europe with concentrated patches of woodlands [50]. Moreover, the northern part of the study area in MLD and AF is interspersed by numerous lakes.
The region is characterized by a low degree of urbanization and the lowest human densities in Poland (59 individuals/km 2 ) [47]. NE Poland is located within a zone of temperate transitional climate with marked continental influences. The study area is inhabited by well-preserved animal communities with as many as 12 local species of carnivores [51].

Carcass collection
A total of 583 carcasses of 9 mammal species were collected in BPF and AF between 2013 and 2019, in BV, KF and MLD between 2016 and 2019, and in the surroundings of studied forests between 2015 and 2019 ( Fig. 1, Table 1). Mammal carcasses originated from road kills or   legal culling from predator control for gallinaceous bird, i.e. the black grouse (Lyrurus tetrix Linnaeus, 1758) and the western cappercaillie (Tetrao urogallus Linnaeus, 1758), conservation projects. No animals were killed specifically for this study.

Necropsies and genetic identification of larvae
The collected carcasses were kept frozen at − 80 °C for a minimum of one week to minimize the risk of Echinococcus multilocularis (Leuckart, 1863) infection [52]. During necropsies the animals were weighed, measured and sexed. Spargana were isolated from subcutaneous tissue, counted, and preserved in 99% ethanol for molecular identification as described below. Prevalence (the proportion of infected individuals in %) was calculated as the ratio of infected individuals to all collected study animals.
The mean intensity of infection in each host species was calculated as the average number of spargana per one infected individual of a particular species. We also calculated a standard deviation of the mean intensity of infection as the square root of its variance.
To confirm larval species affiliation, we molecularly confirmed 25 larvae isolated from different carnivore species (see details in Additional file 1: Table S1, Additional file 2: Figure S1, Additional file 3: Figure S2). We used primers and procedures described by Liu et al. [53] to achieve a sequence of an evolutionarily conserved nuclear 18S rRNA gene of over 240 bp in length. The sequences were aligned with previously analysed ones from wild boars, European badgers, and grass snakes from BPF [9,32,40] (Additional file 1: Table S1), as well as with sequences available on GenBank subsequently trimmed to 222 bp in length and analysed using the ClustalW multiple alignment test (BioEdit sequence alignment editor [54]) and Basic Local Alignment Search Tool (BLAST) [55]. A Tamura-Neil model (TrN) was selected for maximum likelihood analysis using MEGA v6 [56]. The following reference sequences for other Diphyllobothriidae were retrieved from GenBank: Spirometra erinacei (D64072.1); S. erinaceieuropaei (KX528090 and KX552801); and Dibothriocephalus latus (Linnaeus, 1758) (KF218247, KF218246.1 and DQ925309). Taenia pisiformis (Bloch, 1780) (JX317675.1) and Taenia krabbei (Moniez, 1879) (MH843684.1) were used as the outgroup in analyses (Additional file 2: Figure S1 and Additional file 3: Figure S2). A list of all newly generated sequences of 18S rRNA is presented in Additional file 1: Table S1.

Statistical analysis
To assess the spatial variation of S. erinaceieuropaei infection probability in NE Poland, we applied a generalized additive model (GAM) with binomial error distribution in the mgcv package implemented in R [57]. We added the presence/absence of larvae in the sampled individuals as a binomial dependent variable, while the interaction of longitude and latitude of sample locations (explanatory variables) was fitted as a non-parametric spline. We limited analyses to samples collected in BPF, KF, AF and BV due to larger numbers of carcasses collected during the study period.

Results
Plerocercoid larvae of Spirometra were found in 172 of 583 mammals belonging to 7 out of 9 studied species (Table 1, Fig. 2). The overall prevalence of spargana infection in carnivore hosts was 29.5%. The highest overall prevalence was estimated for European badgers (36.4%) and raccoon dogs (30.9%). Twenty-seven stone martens (Martes foina Erxleben, 1777) and 3 Eurasian lynxes were uninfected. The overall mean infection intensity was 14.1 ± 33.8 with the highest values up to 276 found in European badgers (38.2 ± 56.0) ( Table 1).
All 25 newly generated 18S rRNA sequences of larvae isolated from autopsied animals of 7 different species showed 99% identity with the 3 sequences for S. erinaceieuropaei deposited on GenBank (Additional file 2: Figure S1, Additional file 3: Figure S2). Comparisons with previously published 18S rRNA sequences from isolates in European badgers, wild boars, and grass snakes, revealed that the new sequences all showed 99.4-100% similarity (BLAST). The only variation was observed in the DNA of 2 out of 4 analysed larvae originating from European polecats and was, in both cases, an insertion of a T nucleotide at position 207 of sequences (see Additional file 2: Figure S1). The occurrence of S.  13:560 erinaceieuropaei was confirmed for the first time using molecular methods in the pine marten (Martes martes Linnaeus, 1758), red fox, and river otter (Lutra lutra Linnaeus, 1758) (Additional file 1: Table S1; Additional file 2: Figure S1, Additional file 3: Figure S2).
The GAM showed a significant spatial pattern in the probability of S. erinaceieuropaei infection in the study area (χ 2 = 47.8, P < 0.001). The highest probability of infection was observed in BPF and was observed to decrease northward (Fig. 3).

Discussion
Most European cases of sparganosis in wildlife represent exclusively single findings. Thus far, the only comprehensive research has been carried out in Belarusian Polesie, which is closely related to BPF where the parasite has been found in amphibians, reptiles and mammals (including 7 species of carnivores) [26,27,33,37,[58][59][60][61]. Our study focused on S. erinaceieuropaei spread in well-preserved carnivore communities of NE Poland. We revealed high species richness of intermediate/paratenic mammalian hosts for S. erinaceieuropaei, including seven out of nine studied species, both native (European badger, European polecat, river otter, pine marten and red fox) and invasive (American mink and raccoon dog).
Previously, the parasite was also confirmed in Poland in the wild boar (larvae) as well as in the Eurasian lynx and wolf (adult parasites) [7,8,32,39,40]. We molecularly confirmed pine marten and river otter as hosts of S. erinaceieuropaei for the first time and provided the first comprehensive epidemiological data for seven mammal species in NE Poland. Previous research from Europe has also shown six other mammal species as hosts of S. erinaceieuropaei, including the brown rat (Mus decumanus Berkenhout, 1769), European hedgehog (Erinaceus europeaus Linnaeus, 1758), European mink (Mustela lutreola Linnaeus, 1761), European mole (Talpa europaea Linnaeus, 1758), stoat (Mustela erminea Linnaeus, 1758) and weasel (Mustela nivalis Linnaeus, 1766) [34,35,[60][61][62]. The species richness of infected carnivores indicates a wider than previously expected spread and complex circulation of the parasite in wildlife. The lack of other comprehensive data about S. erinaceieuropaei occurrence in Europe has previously not given us a reason to suppose that the parasite is so widespread and prevalent in European wildlife.
The highest infection rates were revealed in the invasive raccoon dog and the native European badger. Therefore, the role of invasive species may be particularly significant in disease spreading; by settling into new areas, they can either bring alien parasite species into colonized areas [44,63], or become new hosts for native parasites and facilitate their spread [44]. High infection rates in both species may result from consuming prey (primarily amphibians and reptiles) in addition to carrion (wild boar and other mesocarnivores) for the raccoon dog that serve as the source of Spirometra infection [51]. Moreover, European badgers and raccoon dogs are preyed upon and consumed by large predators (wolf and Eurasian lynx) [51,64]; predation by wolves and dogs is one of the main sources of raccoon dog mortality [64]. This allows Spirometra to complete its life-cycle and continue spreading. The parasite can be effectively spread over large areas by its medium-sized and large carnivores hosts (European badger: maximal daily movement distance 17.5 km; Eurasian lynx: 24.8 km; wolf: 64 km) [65][66][67], or even to human settlements through dogs. Further studies are needed to reveal how diet composition influences infection severity in Spirometra-infected hosts.
We did not find S. erinaceieuropaei plerocercoids in stone marten and Eurasian lynx specimens. This could be due to stone marten habitat selection as well as the diet of both species. The stone martens' diet does not consist of amphibians and reptiles; they mainly prey on small mammals and birds [68], which can be a potential source of infection. In NE Europe, stone martens usually occur in urban and rural areas and avoid large, continuous forest complexes [69], which limit their potential to come into contact with possible wild hosts of S. erinaceieuropaei. So far, stone martens have been considered a host for Spirometra sp. accidentally in Italy [70]. However, the Eurasian lynx is a known definitive host of S. erinaceieuropaei [7,8,71]. Its diet only sporadically consists of amphibians, and it mostly preys on ungulates, which do not serve as hosts for Spirometra [51]. Moreover, carnivores like the red fox, raccoon dog and raccoon (Procyon lotor Linnaeus, 1758) may serve as both intermediate/ paratenic and definitive host of Spirometra sp. [26,27,72]. We did not find any spargana in the three Eurasian lynx carcasses, but the number of studied animals was very low. Raccoons have not yet been reported as either intermediate/paratenic or definitive host for Spirometra in European wildlife; they have only been confirmed experimentally [72]. However, our study has confirmed that the red fox and raccoon dog, the typical definitive hosts, can also act as intermediate/paratenic hosts for S. erinaceieuropaei. This may be important for sparganosis spread in wildlife, since it highlights the diversity of the disease transmission routes, which range from parasite eggs present in the faeces of definitive hosts, to intermediate/paratenic hosts infected with spargana and preyed upon by large predators [46,64]. We focused on intermediate/paratenic hosts for S. erinaceieuropaei in which larvae may be found in subcutaneous tissue and other organs. Morphological identification of spargana to the species level is impossible and such analyses are justified only in the case of adult tapeworms that complete their life-cycle in a definitive host [24]. Thus, it is highly recommended to use molecular methods for reliable identification of larval Spirometra. So far, notwithstanding our studies in NE Poland over the last few years [8,9,32,39,40], no genetic analyses of Spirometra larvae in Europe have been carried out; therefore, it is not clear which species have caused previously reported infections. Additionally, the adult tapeworms of S. janickii described by Furmaga in the 1950s [7], have not yet been found again, and thus cannot be analysed in more detail. Our study revealed only minor genetic divergence in the studied 18S rRNA gene sequences, which can be interpreted as inter-individual differences. All studied parasite specimens isolated from different hosts and locations undoubtedly belong to the same species, S. erinaceieuropaei. This may be explained by the fact that the sites are separated by no more than 200 km and hosts such as the Eurasian lynx, wolf, red fox, raccoon dog, and European badger can move over significant distances, dispersing parasites between sites [73][74][75][76]. Our study showed that the probability of S. erinaceieuropaei infection varies spatially. The highest probability of infection and prevalence (54.8%) occurred in BPF, indicating beneficial conditions for the parasite, likely because this site had the highest species richness of mammal species among all studied sites [49,51,77]. Additionally, S. erinaceieuropaei was found in grass snakes from BPF [9], which may promote the transmission of sparganosis in wildlife in this area. It was found that a decrease in the number of intermediate/paratenic hosts results in a lower prevalence in definitive hosts [78]; thus, host diversity may be an important factor responsible for the spread of sparganosis. Mammalian species richness was lower in other locations [48,51], although we observed that the probability of infection was highest inside forest complexes and decreased as the distance from the forest complex increased. BPF was also characterized by various forest and non-forest habitats, including wetlands [48], which are crucial to the life-cycle of Spirometra [10,24]. Wet habitats in BPF are the optimal environment for the development of freshwater copepods, which are the first intermediate hosts for the parasite. In S. erinaceieuropaei hosts inhabiting AF and BV, the probability of infection was lower than in BPF, but higher than in KF. This may be due to lower habitat heterogeneity, but with a relatively high percentage of wetland habitats [48]. KF, characterized by the lowest probability of infection and prevalence (13.7%), is generally less diversified in terms of species and habitat structure, with a lower proportion of wetlands [48,49].

Conclusions
Our study revealed that the spectrum of S. erinaceieuropaei intermediate/paratenic hosts in Europe is broad, and therefore, that the trophic dependencies that enable S. erinaceieuropaei to spread in the environment are complex. Sparganosis is likely more widespread in European wildlife than expected and may be transmitted by both native and not-native mammals. Presumably, infection rates vary spatially and depend on numerous factors, including habitat structure, species richness, and density of potential hosts. Further research is required to confirm which environmental and biological factors have the most significant impact on shaping the level of S. erinaceieuropaei infection in European wild mammalian hosts.
Additional file 1: Table S1. Spirometra erinaceieuropaei larvae isolated from different mammalian host species which were analysed genetically with accession numbers of obtained sequences. Abbreviations: BPF, Białowieża Primeval Forest; KF, Knyszyn Forest.