The snail specimens studied were collected in the field, from lymnaeid populations present in geographical areas of the Departamento de Cajamarca in the northern Andean part of Peru, with human infection and/or animal fascioliasis endemicity. The majority of the snail populations studied were found close to schools whose children proved to be infected [13] or around neighbouring villages.

For an estimation of lymnaeid population densities, the method of counting the number of snails in a marked area in a unit of time was used. Marked areas were mainly a segment of a canal or, less frequently, a portion of the edge of a pool. In each locality, snails were collected by the same four people during 1 hour, with the purpose of collecting as many specimens as possible, so as to increase the probability of detecting infected snails.

Localities and their altitudes furnishing the lymnaeid specimens sequenced are noted in Table 1 and Figure 1. Snail specimens for molecular analyses, as well as trematode larval stages found in several of them, were fixed in 70% ethanol for DNA extraction procedures.

Populations of lymnaeid species found are noted in Table 1, including coordinates, altitudes, number of lymnaeid specimens collected and population densities. Within a high altitude range between 2,080 and 3,130 m, the very high population densities, in localities such as Santa Rosa de Chaquil, Tauripampa and Yanamarca, are worth mentioning.

Nuclear rDNA ITS-2 and ITS-1 and mtDNA 16S and cox1 nucleotide sequence data reported in this study are available in the GenBank™, EMBL and DDBJ databases under the accession numbers noted in Table 1. Initial snail species classifications, based on previous complete morphological decriptions [42,50], were subsequently verified by sequence comparisons, as noted in the following.

Four DNA sequence alignments, one for each marker used, where made to show the total nucleotide differences between the respective haplotypes of the three lymnaeid species present in the fascioliasis hyperendemic area of Cajamarca.

When comparing the ITS-2 sequences, a 487-bp-long alignment was obtained, in which a total of 172 variable positions (35.3%) including mutations and indels appeared (Figure 6A). In the analysis of the ITS-1 sequences, the alignment obtained had a length of 562 bp, with a total of 187 variable positions (33.2%) including mutations and indels (Figure 6B).

In the comparison of the mtDNA 16S sequences, the alignment obtained was 425 bp long, including a total of only 31 variable positions (7.2%), with very few gaps (Figure 7A). In the analysis of the mtDNA cox1 sequences, a total of 96 variable positions (14.2%) including only mutations were found in the 672-bp-long alignment. Most of these mutations were however silent, as indeed only two variable positions (0.8%) appeared in the 224-aa-long alignment of the corresponding partial protein sequences (Figure 7B).

The complete sequence of the rDNA ITS-1 obtained from the trematode rediae found in two specimens of G. truncatula from Santa Rosa de Chaquil and another two specimens of the same lymnaeid species from Tauripampa is 432 bp long and with a 51.85% GC content. It showed no nucleotide difference when compared with the sequence of that spacer in F. hepatica from the Northern Bolivian Altiplano and Spain [EMBL: AJ243016], and thus corresponds to the haplotype code Fh ITS1-HA.

DNA marker sequences prove that there are three different lymnaeid species present in the fascioliasis endemic area of Cajamarca: G. truncatulaL. neotropica and L. schirazensis. However, previous studies only refered to one lymnaeid species as being responsible for the disease in that area, namely L. viatrix[36,37,59-61]. This lymnaeid species appears repeatedly in the Peruvian literature as the main responsible for fascioliasis transmission throughout Peru [62-69]. Indeed, Peruvian L. viatrix refer to L. viator variety B elongata described long time ago from the surroundings of Lima [70,71] that recently proved, by multiple DNA marker sequencing, to belong to a new species to which the name L. neotropica was ascribed [50]. Thus, the existence of true L. viatrix, which corresponds to L. viator variety A ventricosa [70,71], has so far only been molecularly verified in Argentina [50] and Chile [21]. Additionally, it should be considered that the correct species name for this species is Lymnaea viator D’Orbigny, 1835, because the feminine spelling viatrix proposed time ago [72] does not fit articles 31.2.1 and 34.2.1 of the International Code of Zoological Nomenclature, as it has been recently highlighted [21].

Unfortunately, it is at present impossible to ascertain whether the populations from Cajamarca classified as L. viatrix were in fact L. neotropica. These two species are very similar and may be easily confused when only relying on traditional malacological methods [50]. Moreover, both species belong to the Galba/Fossaria group, which is well known due to the pronounced morphological similarity of the numerous species it includes and which makes specimen classification and species differentiation extremely difficult [42]. Consequently, reports on L. viatrix in the Cajamarca area may be result of confusion, not only with L. neotropica, but also with G. truncatula and L. schirazensis.

The pronouncedly less numerous nucleotide differences found in the two mtDNA markers 16S and cox1 between the three lymnaeid species present in Cajamarca (Figure 7A,B) when compared to their two rDNA markers ITS-2 and ITS-1 (Figure 6A,B) should be highlighted. This result, together with the high number of silent mutations in cox1 (Figure 7B), suggests a saturation in both mtDNA markers, a phenomenon already described in lymnaeids [42] and well known in several invertebrate groups [49].

Despite the numerous nucleotide differences in their DNA sequences in both rDNA and mtDNA (Figure 6 and Figure 7), the three lymnaeid species found in Cajamarca may easily be confused one another both in the field and also in the laboratory. Nevertheless, there are several details that may help for at least a preliminary classification (Table 2). However, although a morphological shell trend may help in distinguishing a given species [73], confirmation of the species to which a population belongs can so far only be obtained by DNA sequencing. Moreover, it should be taken into account that mixed populations of G. truncatula and L. schirazensis have already been described in the field [42]. Hence, care should even be taken when working experimentally in the laboratory.

The apparent monomorphic genetic characteristics of each one of the three lymnaeid species in Cajamarca should be highlighted. Each lymnaeid species appear to present only one haplotype for each molecular marker sequenced, although of course additional studies on more populations of these species throughout the endemic area in question are needed to verify this assumption. Anyway, the lack or pronouncedly reduced genetic variability of populations living at very high altitude is known in different organisms and has already been also observed in lymnaeids such as G. truncatula in the Northern Bolivian Altiplano [51,74], an area located between 3800 and 4100 m high altitude [2]. This lack of genetic variability of G. truncatula in the Northern Bolivian Altiplano is related to the usual selfing trend followed by this lymnaeid [75]. Autofecundation has been verified to be the normal fecundation process in L. schirazensis and has also been observed in other species of the Galba/Fossaria group [42].

The aforementioned monomorphic genetic characteristics suggest an introduction of each one of the three lymnaeid species from only one source. In the case of G. truncatula, although its European origin appears evident [1,51], the unexpected new haplotypes of ITS-1 (G.tru-HE), 16S (G.tru-16S-B) and cox1 (G.tru-cox1d) found in this species in Cajamarca pose a problem when attempting to elucidate the direct geographical source. In South America, the presence of G. truncatula has already been molecularly verified in the Northern Altiplano of both Bolivia [51] and Peru [10], Argentina [76,77], Chile [21] and Venezuela [22]. In Colombia, the only report of G. truncatula published so far [78] appears to be a misclassification [42].

Lymnaea neotropica, originally described from Lima and surroundings [50], appears to be a species restricted to South America but with a very broad geographical distribution covering from Argentina in the Southern Cone [52] up to Venezuela [22]. The combined rDNA and mtDNA haplotype of this species in Cajamarca is identical to the one from the type locality. The presence of this combined haplotype in both Peruvian areas may probably be related to the livestock trade between Lima and Cajamarca which was very intense along the old rural Inca routes at a given period of the early Spanish colonisation [79].

The combined rDNA and mtDNA haplotype of L. schirazensis present in Cajamarca appears to be a mix. It shares the ITS-2 haplotype L.schi-H1 with Spain, the Dominican Republic and Venezuela, and is different from the one in Rio Lurin, close to Lima. Its ITS-1 haplotype L. schir-HB has also been found in Mexico and Ecuador and in Peru it is present in both Cajamarca and Rio Lurin. The ribosomal 16S gene of the mtDNA does unfortunately not furnish any biogeographical information, as the haplotype L.schir-16S-HA found in Cajamarca is the same as everywhere. And finally, the mtDNA cox1 haplotype L.schir-cox1d appears to be unique [42]. Although a trans-Andean livestock introduction route was launched between western Andean Venezuela and the Colombian Bogota and also further southward [80], the introduction of L. schirazensis with livestock and humans should have most probably occurred from northern Peru, probably by livestock transported between Quito and Lima through Cajamarca along the old rural Inca routes followed by the Spanish conquerors [42].

Fascioliasis in Cajamarca has been the focus of several studies on both livestock [36,37,61,81] and humans [12,13,82-84]. According to results from these studies, Cajamarca is a typical representative of the "valley transmission pattern" of fascioliasis [1,7,81] and the area presenting the highest prevalences in children, both in average (24.4%) and local maximum (47.7%), among all human endemic areas in Peru [13]. With regard to human prevalences, Cajamarca is only surpassed by the Northern Bolivian Altiplano [2,19]. Interestingly, only one lymnaeid species, G. truncatula, is responsible for such a human hyperendemic situation in the Northern Altiplano [51], whereas DNA sequence results here obtained demonstrate that there are three different lymnaeid species present in the hyperendemic area of Cajamarca.

Among the three lymnaeid species present in Cajamarca, Galba truncatula is considered the best F. hepatica transmitter known, with such a parasite/vector interaction as to support that this lymnaeid may be considered the original vector of this fasciolid [1]. Moreover, G. truncatula is known to give rise to very high human infection rates, both prevalences and intensities, at the very high altitude of the Northern Altiplano of both Bolivia [2,18,19] and Peru (10). Such high fascioliasis transmission rates have been proved to be the consequence of life cycle modifications in both F. hepatica and G. truncatula as an adaptation response to the extreme conditions of the very high altitude [51]. Several aspects suggest that G. truncatula may behave in the 2600–3100 m high altitudes of Cajamarca similarly to in the 3800–4100 m altitudes of the Northern Altiplano, namely:

a) the results of the field surveys carried out showing that populations of this species are the most abundant in the endemic area;

b) the typical anthropophylic characteristics of this lymnaeid appeared evident when proving to be the usual snail species in the neighbourhood of human communities such as Encañada, Santa Rosa de Chaquil, Tauripampa and Yanamarca (Figure 8), close to localities where schoolchildren appeared infected [13];

c) the high population densities of thois species found in the high altitude localities studied (Table 1);

d) the finding of infected specimens in two populations of this lymnaeid by means of the DNA sequencing methods (Table 1), despite the relatively few specimens analysed for infection detection (prevalences in lymnaeids are usually low).

In front of the aforementioned similarities, the genetic differences between G. truncatula populations from two different environments such as Cajamarca and the Northern Bolivian Altiplano should be emphasized (Figure 2). Whereas in Cajamarca the disease transmission follows a seasonality mainly related to the cyclic temperatures [37], in the Northern Altiplano fascioliasis transmission takes place throughout the year due to (i) the scarcely varying temperatures [8] and (ii) the link of G. truncatula populations to permanent water bodies as the consequence of the absence of sufficiently long-standing temporal water bodies due to the high evapotranspiration rates [2,8].

Lymnaea neotropica has also been found to be linked to human infection. This species was originally described from the surroundings and other areas near Lima, Peru [50], where human infection has repeatedly been detected [14,85]. Although none of the L. neotropica specimens sequenced in Cajamarca was infected by F. hepatica, Peruvian L. viatrix (= L. viator) variety B elongata (= L. neotropica according to [50]) has been shown to transmit fascioliasis both experimentally and in nature [62,67,86]. The transmission capacity of this lymnaeid vector has also been molecularly confirmed in Argentina, where it appeared linked to animal infection [52], as in Venezuela [22].

With regard to L. schirazensis, experimental infection assays of different geographical strains of this snail species from different continents with F. hepatica and F. gigantica have proved that fasciolid larval stages are not able to fully develop within this lymnaeid, which does therefore not participate in disease transmission [42]. A large multidisciplinar study has demonstrated that L. schirazensis has an outstandingly broad distribution from lowlands to highlands, including from below sea level (−23 m) in Iran up to very high altitude areas in Ecuador (3158 m). Throughout this wide altitudinal range, it has always been confused with G. truncatula in the Old World and with G. truncatula and other similar Galba/Fossaria vector species in the Americas. This hitherto overlooked species has been masking not only the geographical distribution of fascioliasis, but also fasciolid population specificity/susceptibility analyses. It shall be considered that L. schirazensis has been shown molecularly to be able to cohabit with other Galba/Fossaria group species in such proximity that a natural population of specimens of an apparently unique lymnaeid species may in fact involve specimens from two different species unnoticeably mixed. Such mixed populations have already been described in the case of G. truncatula and L. schirazensis[42].