Introduction

The co-evolutionary histories of hosts and parasites have long fascinated biologists. By studying hosts and their parasites jointly, we gain unique insights into their co-evolutionary interactions, adaptations, immune responses, and ecological sensitivities. For example, body lice co-evolved with human and non-human primate hosts, but also recently transferred hosts [1], and acquired specificity to different microhabitats on the human body [2]. The extent of congruence between host and parasite diversification can provide information to assess past geographical, evolutionary, climatic, and ecological drivers that shape faunal evolution [3–5]. In this way parasite epidemiology is relevant to understanding macroevolution. Documentation of the biological diversity and evolutionary history of parasite fauna is also important from the standpoint of biodiversity conservation, and parasite richness and diversity is often less well characterized as compared to their larger hosts.

African clawed frogs (genus Xenopus) are distributed across much of sub-Saharan Africa and are widely used in biomedical research and as food [6,7]. All but one Xenopus species is polyploid [8], and this genus contains 29 species in two subgenera: Silurana and Xenopus following the taxonomy of Evans et al. [8] and Furman et al. [9]. The chromosome number of the diploid ancestors of the subgenera Silurana and Xenopus are 20 and 18 respectively [10]. The subgenus Silurana comprises the only known diploid X. tropicalis and three tetraploid species: X. epitropicalis, X. mellotropicalis, and X. calcaratus [8]. In comparison, the subgenus Xenopus is represented by 25 described species, including allotetraploids, allooctoploids, and allododecaploids, and is subdivided into 3 groups: amieti, laevis, and muelleri [8]. In addition to parallel evolution of host and parasite, patterns of parasite infection are potentially influenced by allopolyploidy in Xenopus, wherein polyploid species are formed via the fusion of the genomes of two ancestral species. This is because allopolyploid species may acquire resistance or susceptibility to parasites that infect one or both of their lower ploidy ancestral species [11].

Along with other frogs in the family Pipidae, Xenopus species are primarily aquatic and live in slow moving or stagnant water as tadpoles and as adults [12]. Their aquatic lifestyle allows them to serve as definitive or intermediate hosts for several aquatic parasites [13]. The parasite fauna of Xenopus is extraordinarily rich with exceptional specialization to species and specific body parts, such as the cloaca, nostrils, and urinary bladder [14,15]. Within metazoan parasites, for instance, there are over 25 genera from seven invertebrate groups infecting Xenopus; a richer assemblage than in most other anurans [15,16]. This diversity reflects a dual origin of the parasites, which are derived from amphibian-specialist groups but also from fish-specialist groups that likely transferred to Xenopus due to overlap in habitat and diet with fish [15]. Among the parasites of Xenopus, helminths (principally nematodes and members of the major platyhelminth lineages, Monogenea, Digenea and Cestoda) are distinguished by morphology, the site of infection within the host, and the host species. Xenopus species have a complex web of interactions with other organisms [15] and may serve as intermediate hosts for helminth infections of aquatic predators. Some of the nematodes occurring in Xenopus have direct life cycles, whilst others (camallanids and cephalochlamydid cestodes) utilize copepods as intermediate hosts. Digeneans have two or more intermediate hosts, with a mollusk as the first intermediate host. There are no known intermediate hosts for monogeneans; they are transmitted by a swimming infective stage [15].

Being amphibians, Xenopus belongs to the most threatened vertebrate class, with over 40% of species in decline [17]. Infectious diseases are a major contributor to the decline of amphibian species [18] and the diversity and conservation status of anuran parasites is poorly characterized. A better understanding of host/parasite relationships in Xenopus is thus relevant to biodiversity conservation.

Nigeria is situated in West Africa and is home to an extraordinary diversity of species including components of the Guinean Forest biodiversity hotspots [19], and several other diverse ecoregions [20]. Nigeria is the most populous country in Africa (~240 million people as of 2025), whose population size is forecasted to increase at a much higher rate than the global average over this century [21]. Consequently, study of Nigerian fauna is pressing from the standpoint of understanding the influences of humans (e.g., climate change, habitat alteration and loss, pollution). In this study, we present results from seven of the nine ecological zones in Nigeria: the Sahel, Sudan, Guinea, and Derived Savanna, the Jos Plateau, Montane, and Lowland Rainforest. The four types of savanna zones are distinguished by several factors including temperature, rainfall, and anthropogenic activity, and the vegetation varies from tall grasses and sparse trees (Guinea Savanna) to sparse grass and thorny bushes (Sudan Savanna). Compared to the savanna ecological zones, the montane ecological zone is characterized by lower temperatures, higher rainfall, and a high diversity of endemic fauna. The Jos Plateau ecological zone is a mosaic of montane woodland, scrublands and grasslands. The Lowland Rainforest ecological zone is characterized by tall dense and diverse trees.

Previous studies of Xenopus species and their parasites in Nigeria have focused on morphology [22,23]. This study aims to build on these studies by using molecular data to characterize distributions of Xenopus hosts and their parasites in Nigeria. We further aim to evaluate the null hypotheses of strict host/parasite co-evolution in this incompletely studied system, with the predictions that parasite species have species-specific infections of hosts, and that phylogenetic relationships among the hosts and their parasites are congruent. We additionally aimed to qualitatively characterize the extent of habitat specificity of parasites of Nigerian Xenopus in terms of the number of host species and tissue types they infect, and the number of distinct ecoregions in which they occur.

Results

Distributions of Xenopus species in Nigeria

A total of 336 Xenopus individuals from five species were collected across seven ecological zones of Nigeria (Fig 1, S1 Table). All of the frogs we collected were adults. Xenopus fraseri (225/336 individuals; 67% of the collection) was the most abundant and widely distributed, and was collected in the Sahel, Sudan, Guinea and Derived Savanna ecological zones. Xenopus fischbergi (7/336 individuals; 2%) was the least abundant and was collected from two locations within the Derived Savanna ecological zone. Xenopus tropicalis and X. calcaratus (14/336; 4%) were collected from one locality in the Lowland Forest ecological zone. We sequenced seven of 14 samples; all of these had X. tropicalis mitochondrial sequences, but this sample included at least one admixed individual (EUA0334) described elsewhere [24], and for this reason we are unable to conclusively assign specimens from this locality to X. tropicalis or X. calcaratus. Xenopus poweri (90/336 individuals; 27%) was collected in the Jos Plateau and Montane Forest ecological zones. Slightly more females were collected than males (across all species, 183 females, 153 males; S2 Table). We did not observe a significant difference between the sexes in the mean number of parasites per individual (the mean of the female-male difference in abundance across all parasite species in all hosts:–0.35, 95% confidence interval of this mean difference: –3.10–1.29).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Distribution of ecological zones in Nigeria following [20] (top left), host Xenopus species (top right), and four major groups of Xenopus parasites (bottom row). Numbered ecological zones are: (1) Sahel Savanna, (2) Sudan Savanna, (3) Guinea Savanna, (4) Jos Plateau, (5) Montane Forest, (6) Derived Savanna, (7) Lowland Rainforest, (8) Freshwater Swamp Forest, and (9) Coastal Vegetation.

A small red box on the inset in the top left panel shows the location of Nigeria; the large red box on this panel demarcates the area depicted in other panels, which illustrate sampling localities of frogs (top right) and parasites (bottom row). Parasite species are labeled except the nematodes where species are represented by the following symbols: Camallanus kaapstaadi: blue diamonds; Chabaudus leberrei: green circles; Batrachocamallanus occidentalis: red triangles; Batrachocamallanus xenopodis: orange star.

https://doi.org/10.1371/journal.pone.0348516.g001

Distributions and host specificities of Xenopus parasites in Nigeria

In total, thirteen species of helminths from two phyla and five major lineages were isolated: camallanoid and seuratoid nematodes (roundworms), and cestode, digenean, and monogenean platyhelminths (flatworms). These were present in three host species collected from seven ecological zones (Fig 1, Table 1, S3 Table).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Observed distributions of parasites in Xenopus host species, Nigeria ecological zones, and Xenopus host organs.

https://doi.org/10.1371/journal.pone.0348516.t001

In our sample, nematodes were the most abundant and widely distributed helminth and were isolated from three host species within six ecological zones. Camallanoid nematodes sampled included: (i) Camallanus kaapstaadi (Fig 2a) which was isolated from the esophagus and stomach of X. fraseri and X. poweri, (ii) Batrachocamallanus xenopodis (Fig 2c) which was isolated from the esophagus and stomach of X. fraseri and X. poweri, and (iii) Batrachocamallanus occidentalis (Fig 2d) which was isolated from the esophagus and stomach of X. fraseri and X. fischbergi. Thus, the relatively distantly related host species X. fraseri and X. poweri were both host to the camallanoid nematodes C. kaapstaadi and B. xenopodis. These two species plus X. fischbergi were all host to the another morphologically-identified seuratoid nematode species Chabaudus leberrei (Fig 2e), even though these host species occupy distinctive ecological zones: the savannas – Sahel, Sudan, Guinea, and Derived – and montane grassland of the Jos Plateau, respectively (though C. leberrei was not detected in the Sahel).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Photographs of (a) a female Camallanus kaapstaadi; (b) a male Batrachocamallanus sp.; (c) a female Batrachocamallanus xenopodis; (d, e) a female and male Batrachocamallanus occidentalis, respectively; (f,g,h,i) anterior & posterior views of a male and femaleChabaudus leberrei, respectively; (j,k,l) scolex, immature proglottid, and gravid proglottid of Cephalochlamys compactus,, respectively; (m,n,o) scolex, immature proglottid, and gravid proglottid of Cephalochlamys namaquensis, respectively; (p) Protopolystoma xenopodis; (q) Protopolystoma microsclera; (r) Protopolystoma occidentalis; (s) Diplodiscus fischthalicus; (t) Progonimodiscus doyeri; (u) Oligolecithus elianae; (v) Strigeatoid metacercaria (a larvae of an unidentified digenetic trematode).

The scale bar of 5 mm for all panels is in (a).

https://doi.org/10.1371/journal.pone.0348516.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Phylogenetic inferences among Sanger sequences of Nigerian Xenopus species (top left insert) and their helminth parasite (Nematoda, Platyhelminthes: Cestoda, Trematoda (Digenea) Monogenea); black circles over nodes indicate bootstrap support of at least 90%, except for terminal nodes where this is omitted for clarity.

For the host phylogeny, two divergence time estimates are provided based on mitochondrial (mtDNA) and nuclear DNA (nDNA) data, as described in [8].

https://doi.org/10.1371/journal.pone.0348516.g003

Two species of platyhelminth cestode (genus Cephalochlamys) were isolated. Cephalochlamys namaquensis (Fig 2h) and C. compactus (Fig 2f) were isolated from intestines of X. poweri and from X. fraseri + X. fischbergi, respectively. That X. fraseri and X. fischbergi were host to the same species of cestode (C. compactus) and nematode (B. occidentalis; see above), is consistent with their close phylogenetic affinities, and their co-occurrence in the Derived Savanna ecological zone.

In platyhelminth parasites, several other examples of host-specificity were detected in our sample. For instance, the trematode parasites Progonimodiscus doyeri (Fig 2m) and Oligolecithus elianae (Fig 2n), the monogenean parasites Protopolysoma xenopodis (Fig 2i) and P. microsclera (Fig 2j), and an unidentified digenean parasite at the strigeatoid metacercariae developmental stage (Fig 2o) were detected only in X. poweri. Likewise, the monogenean parasite Protopolysoma occidentalis (Fig 2k) and the trematode parasite Diplodiscus fishthalicus (Fig 2l) were detected only in X. fraseri. But it remains unclear whether these parasites are host-specific or whether this is a consequence of incomplete sampling.

Phylogenetic relationships inferred among Sanger sequences of Xenopus host species and their helminth parasites are presented in Fig 3. As discussed above, nematodes (camallanoids and seuratoids) exhibited host generalism by infecting several host species (e.g., C. kaapstaadi and B. occidentalis infect X. fraseri and X. poweri, C. leberrei infects X. fraseri, X. fischbergi and X. poweri). In contrast (and based on our sample), platyhelminthes (cestodes, digeneans, monogeneans) tended to exhibit higher host specificity (e.g., in our sample C. namaquensis was found only in X. poweri, O. elianae and P. doyeri were found only in X. poweri, P. occidentalis was found only in X. fraseri).

There were also examples where phylogenetically diverged parasite species co-infected the same host species. For example, species from at least three different trematode genera and at least two different nematode genera can infect X. poweri (Fig 3). Similarly, at least three diverged species of nematode in two genera can infect X. fraseri (Fig 3). Interestingly, however, the diversity of monogenean parasites in the genus Protopolystoma was comparatively modest with different but only modestly diverged species infecting X. fraseri and X. poweri (Fig 3). Attempts to amplify the DNA sequences of the monogenean P. microsclera and the nematodes C. leberrei and Batrachocamallanus sp. were unsuccessful, and these taxa thus were not included in the molecular analyses.

Higher niche generalism of nematodes as compared to platyhelminths is also evidenced by differences in diversity of ecological zones they occupy (S3 Table). Nematode species were sampled in five or six ecological zones, whereas platyhelminth species were detected in only one, two, or three ecological zones.

All parasite species detected in this study infected only one tissue, with the exception of three nematode species that infected the esophagus and stomach and one that infected the lower intestine and rectum (Table 1). Both of these pairs of tissues are closely associated portions of the digestive tract, though differences do exist, for example in acidity and rugosity of the lining. Overall these observations highlight the distinctive ecological conditions of different tissue types – and the commensurate specialization of the parasites that infect them.

Discussion

This study is currently the most comprehensive, molecularly confirmed perspective on the distributions of Xenopus species and their helminth parasites in Nigeria. Xenopus fischbergi is a recently described species [8] and the taxonomic status of X. fraseri was recently clarified [25]; both of these species formerly were referred to as X. muelleri (e.g., Kobel et al. 1996). It is therefore unclear whether earlier reports of the earlier use of species names “X. muelleri” in the forest ecological zones of Nigeria [26] and “X. fischbergi” in the Guinea Savanna ecozone of Nigeria [22] actually are X. fraseri or X. fischbergi because molecular tools were not employed for these earlier species identifications. This study provides additional information on the distribution of X. fraseri in the Sahel savanna and three other vegetation zones, and expands its geographical range to include Nigeria, which is perhaps unsurprising because Nigeria lies between previously known localities on either side of Nigeria (in northern Ghana and Cameroon) [25]. The high abundance of Xenopus fraseri in four savanna ecological zones of Nigeria (Sahel, Sudan, Guinea and Derived Savanna) indicates that this frog species has a wide ecological tolerance. Xenopus fraseri was collected in the Derived Savanna zone in the same location as X. fischbergi, its close relative. Derived Savanna forms as a result of intense anthropogenic activities on the Lowland Forest ecological zone [20], which suggests X. fischbergi also tolerates disturbed habitats. The holotype of Xenopus fischbergi was collected on the Jos Plateau in Nigeria, even though surveys in this study only detected X. poweri in this locality. Xenopus fischbergi also has a wide range in western Africa [8,25]. Xenopus poweri was previously identified as a subspecies of X. laevis (Xenopus laevis sudanensis) [27], or within X. laevis sensu lato [9]. The presence of X. poweri on the Jos Plateau and Montane Forest Ecozone (both >1200 m) [20] suggests that this species occurs in areas with high altitude in multiple ecological zones.

Xenopus tropicalis and X. calcaratus were collected from the Lowland Forest ecological zone. Several studies documented X. tropicalis in forest habitat [26,28], which is consistent with our findings, but this could also refer to X. calcaratus. Anele et al. [22] collected X. tropicalis from a relatively undisturbed location within the northern Guinea Savanna ecological zone, indicating occupancy of this species in both savanna and forest ecozones.

Generalist species are those with broad ecological tolerances and/or diets whereas specialist species have more narrow ecological niches, dietary needs, or other requirements [29]. It is sometimes assumed that specialists do better in their optimal habitat whereas generalists often do better in variable habitats [30]. One way to characterize parasite ecology is to define species that infect multiple distantly related species as host generalists and those that infect one (or few) closely related species as host specialists. According to these definitions – and based on our sample – the generalists include nematodes - C. kaapstaadi, B. xenopodis and C. leberrei and the specialists include all of the trematodes, cestodes and monogeneans isolated in this study. The camallanid nematode B. occidentalis has intermediate host specificity in that it infects two closely related species (X. fraseri, X. fischbergi).

Another approach would be to classify parasites based on the number of ecological zones they were found in (e.g., specialists occupy one whereas generalists occupy more than one), irrespective of their host specificity. Under this approach the generalists include all nematodes, cestodes, monogeneans and trematodes (digeneans), except P. doyeri and O. elianae which are specialists to the Jos Plateau. More realistically, niche specialization could be considered a continuous trait with cestodes, monogeneans and trematodes being more specialized because they occupy only 1–3 ecological zones and nematodes being more generalized because they occupy 5–6.

Specimens of X. tropicalis and X. calcaratus (both are in subgenus Silurana) that were already preserved in ethanol were examined for helminths, but none were detected. It is possible that this is a consequence of the small sample size (14 individuals) or because these specimens were handled differently from the others. Other surveys have recorded infections in subgenus Silurana of several parasite species that also infect the sister subgenus Xenopus [28]. In general, species in subgenus Silurana tend to have low abundance and diversity of helminths (JAJ and RCT, personal observation).

Nematodes had the highest overall prevalence in the Xenopus species we examined. The prevalences of members of the genera Camallanus and Batrachocamallanus in this study are summarized in Supplemental Table 2, and have been reported from other Xenopus species and places as well [22,23]. Batrachocamallanus occidentalis and B. xenopodis are known to infect X. muelleri, which is distributed in East Africa [31]. Hence their presence in X. fraseri and X. fischbergi – both close relatives of X. muelleri – was not unexpected. However, X. fraseri seems to harbour several of the same parasites as X. poweri, indicating some degree of host generalism for B. xenopodis and C. kaapstaadi. Camallanus kaapstaadi has a wide distribution in sub-Saharan Africa and infects X. laevis other species in the laevis group, as well as X. fraseri [31]. The presence of the nematode B. occidentalis in the esophagus and stomach of X. fischbergi and X. fraseri but not X. poweri could indicate a degree of host specificity. The nematode Chabaudus leberrei was isolated from the lower intestine and rectum of X. fraseri, X. fischbergi, and X. poweri, which indicates that this parasite is a Xenopus generalist. Consistent with this, C. leberrei is also known to infect other species in the muelleri and laevis groups [32]. Cephalochlamys compactus was detected in the closely related host species X. fraseri and X. fischbergi, whereas C. namaquensis showed host specificity for X. poweri. Cephalochlamys namaquensis is also known to infect the laevis group, and C. compactus occurs in X. muelleri in addition to X. fischbergi and X. fraseri [33]. The presence of Oligolecithus elianae in the intestine of X. poweri from the Jos Plateau but not in X. poweri collected from the Montane ecozone could indicate habitat specificity or alternatively incomplete sampling. Previous studies have recorded this parasite from X. l. laevis and X. l. victorianus from diverse areas in Africa (the Democratic Republic of the Congo, South Africa, Uganda, Rwanda) [34].

The presence of the digeneans P. doyeri and D. fischthalicus in the rectum of X. poweri and X. fraseri respectively could be facilitated by sympatry of these host species. This could indicate a degree of generalism for both parasites. Progonimodiscus doyeri is known to infect species in the Xenopus subgenus Xenopus, including hosts in the laevis and muelleri groups [11], but it has not been found in the subgenus Silurana. Diplodiscus fischthalicus is known to infect the rectum of X. poweri and has been isolated from the rectum of Hoplobatrachus occipitalis, which is a ranid frog [22]. Using morphology, three species of Protopolystoma were identified from the urinary bladder of Xenopus species. These were confirmed with DNA isolated from P. xenopodis and P. occidentalis, though DNA extraction was not achieved from P. microsclera. The occurrence of P. xenopodis and P. microsclera in a single X. poweri individual is unusual, as is the high abundance of Protopolystoma per individual host (as high as 7 and 9 parasites in X. fraseri and X. poweri, respectively). More typically, only one Protopolystoma individual is found per host individual [3,35,36]. In our sample, Protopolystoma species are highly host specific, with P. xenopodis and P. microsclera infecting X. poweri and P. occidentalis restricted to X. fraseri.

In this study, host specificity was observed in cestodes (C. namaquensis), digeneans (O. elianae) and monogeneans (P. xenopodis and P. microsclera) infecting X. poweri, as well as P. occidentalis infecting X. fraseri. However, nematodes showed a wider range of host selection, with B. xenopodis and C. kaapstadi infecting both X. fraseri and X. poweri and C. leberrei infecting X. fraseri, X. fischbergi and X. poweri. However, one nematode species (B. occidentalis) and one cestode species (C. compactus) were found to infect only the closely related host species X. fraseri and X. fischbergi.

In several cases, we observed infection of multiple, physically connected tissue types (C. kaapstaadi, B. xenopodis and B. occidentalis from the esophagus and stomach of three different Xenopus species; Chabaudus leberrei from the lower intestine and rectum of three different Xenopus species). Food regurgitation in Xenopus [37,38] may facilitate exchange of parasites between the stomach and esophagus.

There are several caveats to our conclusions. For instance, sampling of parasites was performed over two years, and seasonal variation could have differently influenced parasite abundance. We also note that a relatively small sample of X. fischbergi was available for analysis and we were unable to isolate parasites from X. tropicalis or X. calcaratus (possibly as a consequence of preservation conditions); clearly a larger sample would provide a more comprehensive perspective on parasite abundance and specificity in these host species and across the ecological zones that they inhabit. Additionally, the sizes and developmental stages of some parasites hampered efforts to infer taxonomy based on morphology or DNA sequences; this limited our ability use these specimens to inform our understanding of parasite distributions and ecological tolerances. As well, our sampling efforts were restricted to Nigeria and patterns observed in this country may not be broadly representative – particularly in widespread species. Moreover, Cephalochlamys species, C. leberrei, and D. fischthalicus can occur in non-Xenopus hosts [15], so at some level are quite generalistic.

Conclusions

African clawed frogs (Xenopus) and their parasites are a complex and fascinating system with which to explore host-parasite relationships, specialization to different habitat types, tissues, and host species, and to better understand biodiversity. In this study we have explored diversity and ecological specialization of this system with a focus on representatives from Nigeria. Our analyses increase understanding of the distributions of Xenopus and several parasites in Nigeria, identify several examples of host specialization where parasites occupy only one host species (O. elianae, P. doyeri, C. namaquensis, in X. poweri and, D. fischthalicus in X. fraseri), one habitat type but several host species (C. compactus from X. fraseri and X. fischbergi within the Derived Savanna), and one tissue type but several species or habitat types (C. leberrei in the intestine of X. fraseri, X. fischbergi and X. poweri in all the sampled ecological zones except the Sahel Savanna and the Lowland Forest ecozone). As well, we identified several instances of host generalism, where parasites infect several habitat types (all the helminths isolated in this study except P. doyeri which was present in the Jos Plateau only), or multiple distantly related Xenopus species (C. leberrei, C. kaapstaadi and B. xenopodis from X. fraseri and X. poweri). Our efforts are accompanied by voucher specimens, genetic samples, and Sanger sequences that are publicly available at a museum collection (Museum of Comparative Zoology) and a database (GenBank) that will assist with future studies of this dynamic system.

Materials and methods

Supporting information

Acknowledgments

We wish to acknowledge the efforts Mr. Sunday Baba, Department of Biology, Mr. Michael Matthias and Mr. D.A. Adebote, Department of Zoology, Ahmadu Bello University, Zaria, Mr. Linus, Sarkin Ruwa Bagauda, Prof. MSO Aisien University of Benin, Edo State, Mr. Lawan Ibrahim (Hadejia wetlands), Lt. Col O.A. Anele and Lt. Col. C.E. Anele for assistance with the collection of samples. Our warmest appreciation go to the Anele’s family especially late Ex WO Cyriacus Anele, late Mrs. Appolonia MG Anele, Maj. Gen. E. Okoro, Cdre. C. Chinaka for financial and emotional support especially during field work. We wish to thank Lindsey Kukoly, Jianlong Zhu, Jay Patel for their hospitality when EUA visited the Evans lab and Jonathan Dushoff for statistical advice. We thank the staff at the Museum of Comparative Zoology for their support with archiving the specimens from this study and three anonymous reviewers for helpful comments on an earlier version of this manuscript.