Helminth fauna of the invasive American red-eared slider Trachemys scripta in eastern Spain: potential implications for the conservation of native terrapins

Abstract In this study we report on the helminth fauna of the invasive American red-eared slider Trachemys scripta in five localities from eastern Spain where this species co-occurs with two native, endangered freshwater turtles, i.e. Emys orbicularis and Mauremys leprosa. In total, 46 individuals of T. scripta were analysed for parasites. Adult individuals of three helminth species were found: the monogenean Neopolystoma orbiculare, the digenean Telorchis solivagus and the nematode Serpinema microcephalus. Telorchis solivagus and S. microcephalus are trophically transmitted parasites of native turtles that probably infected T. scripta through shared infected prey. Neopolystoma orbiculare infects T. scripta in its native Nearctic range and probably survived the overseas shipping of hosts due to the combination of a direct life cycle, long lifespan in turtles and crowding conditions that allowed frequent (re)infections. These findings entail potentially significant conservation consequences that deserve further attention. First, there was a successful spill-back of Telorchis solivagus and S. microcephalus to large invasive populations of T. scripta in the study area, which could eventually increase infection levels in native turtles. Second, the possibility of spill-over of N. orbiculare to native turtles should not be underestimated as this phenomenon has recently been reported in France. Given the pathogenic potential of N. orbiculare, its transference to native turtles could be troublesome.


Introduction
Over recent decades, most of the biogeographic barriers that had isolated continental biotas for millions of years have been erased due to increased human global trade. As a consequence, exchange of alien species among biogeographic regions has accelerated greatly (Hulme 2009). Nonetheless, only a small fraction (1%) of alien species become but surveys in its areas of invasion are relatively scarce. In Japan, red-eared sliders have been reported to harbour native parasites of the freshwater turtles Mauremys japonica (Temminck and Schlegel, 1835) and Chinemys reevesii (Gray, 1831), but also exotic parasites typically found in North America (Asakawa 2005;Oi et al. 2012). In southwestern Spain, Hidalgo-Vila et al. (2009) reported four parasite species typically infecting the native freshwater turtles E. orbicularis and M. leprosa. Interest in the conservation consequences associated with parasite exchange between T. scripta and native turtles has been shown recently Verneau et al. 2011).
In this study we examine the parasitic fauna of introduced T. scripta in eastern Spain, an area where a large population of this species co-exists with two species of native freshwater turtles ) and information about parasite fauna of any of these species is scarce (Aparicio et al. 2008). Our study is particularly aimed at providing a reliable diagnosis of potentially exotic parasite taxa. We also provide a preliminary assessment on the conservation consequences of parasitological findings for native freshwater turtles.
The 95% confidence interval (95% CI) for prevalence was calculated with Sterne's exact method (Reiczigel 2003), and for mean values of intensity and abundance with the bias-corrected and accelerated bootstrap method using 20,000 replications (Rózsa et al. 2000). Calculations were performed using the free software Quantitative Parasitology v. 3 (Reiczigel and Rózsa 2005).

Results
Three helminth species were found ( Table 1). The digenean Telorchis solivagus (Odhner, 1902) ( Given that N. orbiculare is a putative exotic species, a detailed description of specimens follows. Morphometric measurements are based on five individuals and given as mean ± SD with range in parentheses (in µm unless otherwise stated).

Remarks
Species of Neopolystoma (Price, 1939) are distinguished from those of allied genera, i.e.
Polystomoides (Ward, 1917) and Polystomoidella (Price, 1939), by the lack of hamuli (Morrison 2010). Other diagnostic traits are the presence of one testis, short uterus and ovary in the anterior part of the body. Also, ovarian follicles extend into the posterior part of the body and vaginas are present. The genus Neopolystoma currently contains 21 species that are distributed worldwide except in the Afro-tropical region, and usually occur in the urinary bladder, cloaca and, occasionally, nasal mucosa and conjunctival sac of freshwater turtles (Price 1939). Neopolystoma orbiculare was described in T. scripta and Chrysemys picta (Schneider, 1783) in North America (Morrison 2010). Diagnostic traits of N. orbiculare found in the present study agree with the measurements and other features provided for the holotype established by Stunkard (1916) and with re-descriptions by Price (1939) and Lamothe-Argumedo (1972). The species can be separated from the four other sympatric species of Neopolystoma from the Nearctic region as follows (Stunkard 1916;Caballero 1938). Neopolystoma domitilae (Caballero, 1938) infects the urinary bladder and cloaca of Trachemys ornata (Gray, 1831) and has 19-21 genital spines and a larger body size (4.04-4.06 by 1.32-1.72 mm). Neopolystoma elizabethae (Platt, 2000) infects the conjunctival sac of the eye of Chrysemys picta bellii (Gray, 1831) and has eight genital spines. Neopolystoma terrapenis (Harwood, 1932) infects the urinary bladder and cloaca of Terrapene carolina (Linnaeus, 1758); Price (1939) considered that this species was very similar to N. orbiculare, but Harwood (1932) stated that N. terrapenis has a smaller body size than N. orbiculare (1.90-2.50 by 0.72-0.82 mm), the vitellaria do not crowd into the intercaecal area posterior to the testis, and the pharynx and cirrus sac are also smaller. Neopolystoma chelodinae (MacCallum, 1918) infects the urinary bladder and cloaca of Chelodina longicollis (Shaw, 1794); this species is very similar to N. orbiculare but differs from all other species of Neopolystoma in that it has a testis with irregular margin, and vitelline follicles with irregular shape and variable size (Price 1939;Pichelin 1995) and more variability in the number of genital spines (12-16; Morrison 2010).
In the Palaearctic region, three species of Neopolystoma have been described. Neopolystoma exhamatum (Ozaki, 1935) is found in the urinary bladder and cloaca of Mauremys japonica from Japan, but it was recently reported from T. scripta in the same region (Oi et al. 2012). This species resembles N. orbiculare regarding haptor shape, but its testis has different morphology and size (1.20-1.60 by 0.70-1.30); N. exhamatum also has more genital spines (16-18; Morrison 2010). Neopolystoma palpebrae Strelkov, 1950, which occurs on the lower eyelid of Pelodiscus sinensis (Wiegmann, 1834), is very similar to N. orbiculare but differs in the shape and size of genital spines (Morrison 2010). Finally, Neopolystoma euzeti Combes and Katari, 1976 infects the urinary bladder and cloaca of the Spanish pond turtle, Mauremys leprosa in North Africa (Combes and Katari 1976). However, compared with the specimens examined in the present study, N. euzeti has more hooks in the genitalia (33-36) with a larger size (length: 48-57 µm); body (4.54 ± 0.75 × 1.51 ± 0.08) and haptor (1.20 ± 0.17 × 1.67 ± 0.32) are also larger, and vitelline follicles are more widely extended. Based on the above evidence, we conclude that our material is conspecific with N. orbiculare.

Discussion
There are few studies on the helminth fauna of T. scripta outside its natural range, and they all indicate that this species can acquire parasites in the area of invasion, mainly from native turtles. In Japan, Asawaka (2005) reported the digeneans Telorchis clemmydis Yamaguti, 1933 and Telorchis geoclemmydis Yamaguti, 1933, and the nematodes S. microcephalus and Falcaustra sp. In other localities from Japan, Oi et al. (2012) found Telorchis clemmydis, S. microcephalus and the monogeneans Neopolystoma exhamatum and Polystomoides japonicum Ozaki, 1935 Oka, 1912 in Japan. Overall, these observations suggest that the specific composition of native elements of the parasite community in invasive individuals of T. scripta depends on both geographic factors and the local pool of available parasite species provided by native turtles.
In the present study, T. scripta was also observed to be infected with two species presumably acquired from native turtles. The digenean Telorchis solivagus is a common parasite infecting freshwater turtles in Europe, Africa and Asia (Roca et al. 2005 and references therein). Serpinema microcephalus naturally infects freshwater turtles throughout the Palaearctic region (Baker 1979); in Spain, the most common definitive host is M. leprosa (Roca et al. 2005;Villarán and Domínguez 2009). Species of Telorchis use turtles, snakes and salamanders as definitive hosts, aquatic snails as first intermediate hosts, and tadpoles and snails as second intermediate hosts (Font and Lotz 2008). The life cycle of species of Serpinema include copepods and aquatic snails as first intermediate hosts, fish, anurans and aquatic insects (damselflies) as paratenic hosts and turtles as definitive hosts (Moravec and Vargas-Vázquez 1998;Wiles and Bolek 2015). Therefore, we suspect that, in the study area, individuals of T. scripta became infected with both Telorchis solivagus and S. microcephalus through consumption of infected intermediate hosts. In support of this, the stomach contents of the turtles analysed in this study included mainly vegetal matter but also invertebrates (gastropods and crustaceans). There is an ontogenetic shift in diet from carnivory to herbivory in T. scripta (Bouchard and Bjorndal 2006), but adult turtles can still acquire parasites through accidental infection of infected intermediate or paratenic hosts (Hidalgo-Vila et al. 2009). Oi et al. (2012) first reported on exotic parasites from T. scripta in invaded areas; in particular, two vascular digeneans, Spirorchis artericola (Ward, 1921) and Spirorchis elegans Stunkard, 1923, and the nematode Falcaustra wardi (Mackin, 1936) in Japan. In our study area, a previous survey reported a digenean preliminarily identified as Telorchis attenuata Goldberger, 1911(Aparicio et al. 2008. If this identification were to be confirmed, this would represent the first report in Spain of an exotic species typical from Trachemys scripta. With regard to the monogenean N. orbiculare, Verneau et al. (2011) identified eggs of this species in individuals of T. scripta from a turtle farm in southern France using molecular methods. Recently, Meyer et al. (2015) also detected eggs of N. orbiculare in naturalized populations of T. scripta from the same area. In the present study, we report for the first time adults of N. orbiculare in wild individuals of T. scripta outside its natural range. The presence of this species in the region does not seem to be anecdotal as there are preliminary records of putative N. orbiculare in T. scripta from localities close to those sampled in the present study (F. Domènech unpub. data). The occurrence of N. orbiculare in non-native areas could have resulted from the interplay of three factors. First, N. orbiculare seems to be a common parasite of individuals of T. scripta sold as pets by some commercial suppliers in the native area. For instance, Henke et al. (1990) reported prevalences over 40% in commercial samples of T. scripta from Wisconsin (USA). Second, polystomatid monogeneans have a long lifespan (up to 3 years in extreme conditions, see Tinsley 1999), which could allow survival of the parasite during the overseas shipping of hosts. Third, N. orbiculare has a direct life cycle that includes a free-living oncomiracidium and T. scripta as the single host (Verneau 2004). Therefore, transport of overcrowded loads of T. scripta could have increased probabilities of host-parasite encounter (Verneau et al. 2011) allowing the completion of the life cycle of N. orbiculare first in captivity and later in the wild.
The finding of both native and exotic parasites in invasive populations of T. scripta from eastern Spain could entail potential conservation consequences that would deserve attention in the future. First, there is clear evidence of spill-back effects. Both Telorchis solivagus and S. microcephalus are generalist parasites of freshwater turtles that are able to reproduce in T. scripta, as shown in this study. In fact, the prevalence of Telorchis solivagus found in the present study (30.4%) is close to the upper range of prevalence of this species in M. leprosa, i.e. from 2.8 to 42.9% (López-Román 1974;Mishra and González 1978;Roca et al. 2005). In contrast, the prevalence of S. microcephalus in E. orbicularis and M. leprosa is generally higher (i.e. > 30%) than that found in the present study (Yildirimhan and Sahin 2004;Roca et al. 2005;Hidalgo-Vila et al. 2009). However, Kirin (2001) found S. microcephalus with a prevalence of just 1% in a sample of E. orbicularis from south Bulgaria (n = 69), suggesting that local factors strongly influence infection levels in native turtles. Overall, evidence suggests that T. scripta appears to be a competent host for at least some parasites of native turtles. However, the ecological impact of this phenomenon would depend on both the ability of T. scripta to produce infective stages, and the pathogenic potential of parasites. The red-eared slider has large populations in coastal wetlands from eastern Spain where both E. orbicularis and M. leprosa co-occur (LIFE-Trachemys 2011). Therefore, the possibility that T. scripta could help to significantly increase populations of native parasites cannot be ruled out. However, it is unclear whether this putative increase of parasite populations could have a pathogenic impact on native turtles, invasive turtles, or both. Interestingly, Hidalgo-Vila et al. (2011) suggested that S. microcephalus appears to be especially pathogenic to T. scripta, in contrast to the species of Serpinema that naturally infect this turtle in its native area. These authors hypothesized that the exotic population of T. scripta in southern Spain was immunosuppressed due to maladaptation to the newly colonized area and, therefore, they were prone to the pathogenic effects of native parasites.
On the other hand, the possibility of spill-over effects should not be underestimated. Verneau et al. (2011) and Meyer et al. (2015) demonstrated by DNA barcoding analysis of parasite eggs that host-switching of N. orbiculare from T. scripta to E. orbicularis and M. leprosa were successful under both captive and natural conditions. In the present study, the prevalence of N. orbiculare was lower (10.9%) than that reported in their natural range (20-74%; Everhart 1957). However, this monogenean exhibits low specificity among freshwater turtles (Harwood 1932;Price 1939;Zerecero 1948;Barus and Moravec 1966;Lamothe-Argumedo 1972) and, therefore, the possibility exists that the parasite can easily be transmitted in areas of high turtle density. Henke et al. (1990) reported that putative eggs of N. orbiculare provoked inflammatory lesions in the urinary bladder of T. scripta. If this is true, then this species could be of potential pathogenic significance, at least for invasive turtles.
In summary, whether or not potential spill-over and spill-back events have a population impact on native and non-native turtles is difficult to predict and cannot easily be assumed. However, parasites are an integral part of ecosystems, and this study highlights the need to monitor the complex effects of parasite exchanges between native and invasive species, as emphasized in recent reviews on the subject (e.g. Torchin et al. 2002;Heimpel et al. 2003;Keesing et al. 2006;Kelly et al. 2009).