Parasitological surveys of domestic livestock and children were carried out in March 2009 and 2010 in four areas across Senegal; the Senegal River Basin (SRB) in the North, Vallée du Ferlo which is central, Tambacounda in the South East and Kolda in the South. (Figure 1). These sites were specifically selected as they were known foci for human and livestock schistosomiasis (personal communication). As these were mainly pilot surveys aimed to positively identify the different Schistosoma species and any hybrids infecting livestock and children in each area only human urine samples and animal intestines were sampled.

gDNA from individual miracidia was extracted as described in Webster [33]. Individual worm pairs were washed in TE buffer to remove any residual ethanol and allowed time to relax in the TE buffer so that they could be separated. The male and female worm from each pair was recorded. gDNA from individual worms was extracted using the DNeasy blood and tissue kit (Qiagen) according the manufacture's protocol and DNA was eluted in a total of 100 µl.

To enable high throughput processing of the samples the majority of the gDNA extractions were carried out using the DNeasy blood and tissue kit 96 well plate spin protocol (Qiagen) according the manufacturer's protocol and DNA was eluted in a total of 100 µl.

Hybrid detection of schistosomes requires a multi-locus approach, analyzing both mitochondrial and nuclear DNA simultaneously from individual specimens [32]. A partial region of the mitochondrial cox1 gene and the complete ITS1+2 rDNA were analyzed from each individual as described below.

The cox1 and ITS reference sequences of S. haematobium, S. bovis and S. curassoni from Huyse and colleagues [32], were used to identify and compare the sequences from the samples in this study.

During the collection of the parasite material at the sampling sites, eggs were also collected to establish laboratory isolates of S. haematobium, S. bovis and S. curassoni.

Ethical approval for these studies was obtained from the Imperial College Research Ethics Committee (ICREC), Imperial College London and the Ministry of Health Dakar, Senegal. Before conducting the study, the MoH-approved plan of action had been presented and adopted by regional and local administrative and health authorities. Meetings were held in each village to inform the village leader, heads of the families, local health authority, teachers, parents and children about the study, its purpose and to invite them to participate voluntarily. According to common practice and with approval from the Imperial College Research Ethics Committee (ICREC), due to low levels of literacy all village leaders, teachers, parents and study participants gave oral consent for the studies to take place. Informed consent for the urine examinations was obtained from each study participant and their parents or guardians. Oral consent for each participant was documented by inscription at school committees comprising of parents, teachers and community leaders. All the data were analysed anonymously and all schistosomiasis positive participants were treated with PZQ (40 mg/kg). In schools or classes where the percentages of infections were more than 50%, mass treatment of all children was carried out at the end of the study.

Laboratory animal use was within a designated facility regulated under the terms of the UK Animals (Scientific Procedures) Act, 1986, complying with all requirements therein, including an internal ethical review process at the NHM and regular independent Home Office inspection. Work was carried out under the Home Office project license number 70/6834.

The prevalence of human urogenital schistosomiasis was high in all areas sampled, ranging from 57–100%, and visible haematuria was obvious in urines samples from all study areas. A total of 823 individual miracidia were collected and genetically analysed from 52 urine samples. 79% of the miracidia were molecularly identified as pure S. haematobium, however, 21% presented a mixed mt cox1 + nuclear ITS genotype suggesting that these miracidia had a hybrid origin. Hybrid miracidia were found in 88% of the urine samples analysed and in all areas surveyed. The numbers and type of hybrids varied between urine samples and areas (Table S1). Hybrids between S. bovis and S. haematobium were found in children from all areas except Barkedji in the Vallée du Ferlo and hybrids between S. curassoni and S. haematobium were only isolated in Tambacounda and the Vallée du Ferlo.

In total, 1004 schistosome worms (502 pairs) were dissected and genetically analysed from the mesenteric vessels of 33 animal intestines and 109 miracidia were hatched and analysed from three infected liver samples (Table S2). Only S. bovis was found in animals from Kolda and Richard Toll but S. bovis and S. curassoni were found in animals from Tambacounda and the Vallée du Ferlo. The number of worm pairs found in each individual animal varied considerably from 1 to over 100. S. bovis/S. curassoni hybrid worms with different genetic profiles were found in small numbers in the intestines of five animals from Tambacounda and all these hybrids were paired with either S. bovis or S. curassoni worms. No worms with pure or hybrid S. haematobium genetic profiles were found in any domestic animal.

All three species were successfully isolated from the field samples and maintained in laboratory Bulinus wrighti snails. The molecular ITS and cox1 identification of the cercariae resulting from the three groups of laboratory snail infections showed that each group was infected with one species (Group 1 S. curassoni, Group 2 S. bovis and Group 3 S. haematobium). No hybrid profiles were observed in these isolates. The mixed species infections in all the laboratory rodents (animal crosses 1–3, Table 1) produced heterospecific pairs between male and female worms of all three species (Table 1) and also homospecific pairs. The numbers of pairs varied between animals with the most abundant heterospecific parings occurring between S. bovis and S. curassoni worms. Miracidia were hatched from the infected livers of animals with mixed infections. Molecular typing of these miracidia detected hybrid offspring (Table 1) and confirmed viable heterospecific pairings between all three species. Homospecific miracidia of each species were also identified. The numbers of each type of miracidia correlated to the number of homospecific or heterospecific worm pairs present.

Children were found excreting S. haematobium/S. curassoni hybrids in Tambacounda and the Vallée du Ferlo and S. haematobium/S. bovis hybrids in Tambacounda, Kolda and the SRB. A low number of S. bovis and S. curassoni hybrid worms were found in cows slaughtered in the Tambacounda abattoir but no S. haematobium worms or hybrids thereof were found in ruminants, however this could be due to the fact that only the intestines of the animals were sampled. . These data do suggest that at some point host switching has been able to take place between these three sister species enabling the two species involved to interact, hybridize and produce viable offspring.

The laboratory mixed infection experiments further indicate that males and females of each species readily pair and produce viable hybrid offspring and in these experimental crosses there does not appear to be any competition, exclusion or mating preference between the species and each cross produced viable hybrid offspring.

In each of the three types of hybrids two hybrid lines were observed, resulting from bidirectional introgressive hybridization with mtDNA from both the parental species introgressing into the other species involved in the hybridization (see Table S1+S2). Regarding the nuclear DNA profiles, initial hybrid generations usually display both parental nuclear rDNA ITS copies, resulting in double chromatogram peaks at the species-specific mutation sites [32], [47]–[49]. In subsequent hybrid generations or backcrossing of hybrids with parental species, biased homogenisation towards one of the parental species may result in nuclear rDNA ITS sequences that can appear as just one species or the other [49]. As the dynamics of this homogenisation and silencing of the genetic signal from one species over the other is unknown it is impossible to decipher which generation the natural hybrids are from by looking at their genetic profiles. Nevertheless, the observation of both pure and mixed nuclear rDNA ITS sequences within our hybrid populations strongly suggest that there are different generations of hybrids and/or hybrid backcrosses persisting in nature.

The hybridization of S. bovis and S. curassoni with S. haematobium is of particular interest, as for this to occur, host switching must have taken place of either, S. bovis and S. curassoni into humans or S. haematobium into domestic livestock. The oviposition site of a schistosome pair is generally assumed to primarily be dependent on the species of the male worm [37], [43], with S. bovis and S. curassoni males carrying their females to the intestinal tract and S. haematobium males carrying their females to the urinary tract. It is unknown how hybridization will affect this phenotypic/behavioural characteristic, however it would be expected that in the initial parental cross the worm pair would migrate to the oviposition site determined by the species of the male worm. Hybrid miracidia analysed from the human urine samples that present S. curassoni cox1 and mixed ITS sequences or S. bovis cox1 and mixed ITS sequences could be first generation hybrids resulting from pure parental crosses between S. haematobium males and S. curassoni or S. bovis females, however, it is also possible that second generation hybrids or hybrid backcrosses could also present these genetic profiles. It would be expected that first generation hybrids of the reciprocal cross (S. haematobium females paired with males of S. bovis or S. curassoni) would be excreted in stool samples, therefore miracidia from any eggs present in stool samples also need to be analysed to identify further hybrids and also any possible homospecific infections of S. bovis and or S. curassoni in humans.

No S. haematobium/S. bovis or S. haematobium/S. curassoni hybrids or S. haematobium worms were found in the domestic livestock, possibly suggesting that S. haematobium and indeed S. haematobium hybrids may lack the ability to penetrate and or develop in ruminants. It is also necessary to consider whether major differences in the vasculature between ruminants and humans would restrict the migration of S. haematobium to the blood vessels of the bladder of ruminants. However, as only the intestinal tracts of the slaughtered animals were routinely available for inspection at the abattoirs, and as S. haematobium is a parasite of the urinary tract, the presence of S. haematobium or the hybrids may have been missed. A much earlier study in Zambia did report the possible finding of S. haematobium/S. mattheei (another closely related ruminant schistosome) hybrids in the mesenteric veins of cattle [50] however these findings remain unconfirmed. It is clear that more extensive and detailed dissections off and egg collections from the ruminants is warranted.

The infectivity of a schistosome to a mammalian host depends on several physical and immunological factors, however, the variety of host use by different species of the Schistosoma genus suggests that host switching has occurred at several time points in the evolution on this genus [38]. It is possible that the close phylogenetic ancestral position of S. haematobium to its sister species S. curassoni and S. bovis, together with physical and anatomical host differences may enable the latter two species to retain an ability to infect humans. Another possibility is that the initial pairing between these human and domestic livestock parasites occurred first in another susceptible host, such as a rodent. Rodents have proved extremely efficient for passaging a variety of schistosome species in the laboratory [17] and are utilised as reservoir hosts by other species of schistosome [51]–[54].

The hybridization between S. curassoni/S. bovis is not that surprising given their neighbouring phylogenetic position and their overlap in definitive host associations [38] and the molecular data presented here conclusively confirm that these parasites are able to hybridize.

Observations from the different survey sites clearly demonstrate the focality of the transmission of the different species and their hybrids. Hybridization between the human and domestic livestock schistosomes appears common with hybrid genetic profiles recovered from human urine samples from several areas where the different species are transmitted sympatrically. Also, S. haematobium/S. bovis hybrids have previously been sampled from humans from several widely dispersed villages along the SRB [32], and the possibility of the hybridization between S. haematobium/S. bovis or S. curassoni was reported by [42] in Niger. Due to the wide distribution of the intermediate snail hosts B. globosus and B. truncatus, S. haematobium and S. bovis are common infections across much of Senegal, providing the opportunity for these species to interact, however the distribution of S. curassoni, transmitted through B. umbilicatus, is more restricted. The data presented here confirm an intense S. curassoni focus in the Vallée du Ferlo, increasing the known distribution of this species and the hybrids thereof. S. haematobium/S. curassoni hybrids were found in the urines of children sampled from the village of Barkedji, where only S. curassoni was found in the slaughtered sheep. This, together with the data from Tambacounda, provides the first confirmation of this species or hybrids thereof infecting children in Senegal.

Only small numbers of S. curassoni/S. bovis hybrids were found in cattle at Tambacounda abattoir, with most worms being identified as S. curassoni in agreement with an earlier report [41]. There are no obvious isolating barriers other than intermediate snail host preferences that prevent S. curassoni and S. bovis from interacting and hybridizing as both species readily infect ruminants. However, transmission of these two species does appear to be localised, with the majority of the animals sampled in one particular place being infected with either S. bovis or S. curassoni. The origins of the animals sampled at the small abattoirs in this study are usually unknown and would be almost impossible to trace due to changes of ownership during an animal's lifetime. The worm burden reflects past exposure: S. bovis transmission being associated with B. truncatus habitats while S. curassoni is associated with B. umbilicatus [54].

With regard to the role of intermediate snail hosts, hybridization between Schistosoma species potentially enables the schistosomes to increase their host range with the hybrids being able to utilize both the intermediate snail hosts of the parental species, which will have important implications for schistosomiasis epidemiology by increasing transmission and distribution [19]–[20], [27], [32]. Snail surveys were not conducted during this study, however, the study of Huyse and colleagues [32] did provide evidence for the transmission of S. haematobium/S. bovis hybrids through both B. globosus and B. truncatus the intermediate snail hosts of S. haematobium and S. bovis respectively. Snail surveys and molecular screening of cercariae from the hybrid zones are needed to clarify what role each snail species plays in the transmission of the parental species and their hybrids.

Some degree of sympatric transmission enabling interbreeding between these species may have always occurred but remained undetected due to the limitations of previously available sampling and analysis tools. However, hybridisation between species can be further facilitated by the loss of ecological barriers existing between species due to natural and or man-made changes. [35]–[37]. In Northern Senegal, it was speculated that the hybridization between S. haematobium and S. bovis was facilitated by the creation of water bodies for agriculture, through dam construction, which led to an increased prevalence and distribution of the intermediate snail hosts and movements of humans and livestock to these resources, creating sympatric transmission of these species, resulting in hybridization. In the additional hybrid zones confirmed in this study, while no such ecological change can be attributed to enabling hybridization to have taken place, the natural progression in farming, population (both human and livestock) movements and expansion will result in areas of increased close associations between humans and their domestic livestock, increasing the chances of interspecific interactions between the schistosomes they carry. Other behavioural factors could also have an important impact, for instance during the sampling of cattle in the abattoir in Richard Toll in the SRB where S. bovis was highly prevalent, it became apparent that due to the lack of running water the intestines of slaughtered animals were routinely washed in the local river, which was also frequented by the local people for their everyday activities, thus creating a potential sympatric transmission site for S. haematobium and S. bovis.

Introgressive hybridization may lead to phenotypic changes that can dramatically influence disease dynamics and evolution of the parasites. Although treatment with praziquantel, the drug routinely used to control human schistosomiasis across Africa, is successful against S. haematobium, S. bovis and S. curassoni [55]–[56], hybridization between different Schistosoma species have been reported to affect the success of drug treatment in cattle [57], cause severe disease outbreaks and competitive exclusion of one species by the other [35]–[37] and laboratory hybrids have been observed to acquire enhanced characteristics such as infectivity, fecundity and growth rates [19], [20], [54].

S. haematobium has long been recognised as a parasite with few, if any, reservoir hosts [38]. However, it seems that in Senegal and possibly in other areas of West Africa, new genotypes may emerge that may pass from people to domestic livestock and vice versa. Furthermore, if the hybridization events reported here result in phenotypic characteristics that influence drug sensitivity, pathology and transmission, it will be highly important to re-evaluate control strategies in these hybrid zones. The increased host range of the hybrid parasites and changes in host distribution may have a direct impact on transmission of these schistosomes. Human and veterinary schistosomiasis in Senegal and neighbouring countries needs to be further monitored to clarify further the epidemiology and dynamic interactions of these closely related schistosome species.