Gyrodactylus jarocho sp. nov. and Gyrodactylus xalapensis sp. nov. (Platyhelminthes: Monogenea) from Mexican poeciliids (Teleostei: Cyprinodontiformes), with comments on the known gyrodactylid fauna infecting poeciliid fish Author Rubio-Godoy, Miguel Author Paladini, Giuseppe Author García-Vásquez, Adriana text Zootaxa 2010 2509 1 29 journal article 10.5281/zenodo.196033 a86ba3e8-b241-4166-a37a-5abb4b60edcc 1175-5326 196033 Gyrodactylus rasini Lucký, 1973 ( Figure 24 ; Tables 4 and 7) Type host: Xiphophorus hellerii Heckel (“green swordtail”, “cola de espada verde”). Site of infection: Gills. Type locality: Not known. Type material examined: Four paratypes ( BMNH 1994.11.24.13–18). Comments: Type locality not known as the original description of G. rasini was based on parasites collected from aquarium fish ( Lucký 1973 ). The distribution range of the host X. hellerii spans the Atlantic coast of México , Belize , Guatemala and Honduras ( Froese & Pauly 2009 ). Specimens identified as G. bullatarudis by Harris (1986) were reassigned to G. r a s i n i by Richards et al. (2000) , and their measurements are reviewed in Table 7. While the marginal hooks sickles of G. r a s i n i can be readily discriminated from those of G. j a ro c h o sp. nov. based principally on the shape of the sickle base ( Figure 25 ), their discrimination from G. xalapensis sp. nov. should be supported by information drawn from other haptoral elements and the MCO. If the dimensions of the two latter species are compared, then the ranges for both species also closely coincide, G. xalapensis sp. nov. being towards the upper end of the range of measurements reported for each structure. Only the length of the dorsal bar of G. xalapensis sp. nov. is apparently longer than that of G. r a s i n i ( av. 22.2 (19.4–25.7) c.f. 13– 19 ( Lucký 1973 ; Harris 1986 ; Richards et al . 2000 )) - although the four paratypes re-measured in this study had on average 22.7 µm (21.5–23.4) long dorsal bars ( Table 4 ). When the hamuli of the two species are compared, those of G. xalapensis sp. nov. are more robust, with longer ventral bar articulation regions and more pronounced dorsal bar attachment points. Although the ventral bars of both species are also similar, the membrane of G. rasini is more rhomboid in shape while that of G. xalapensis sp. nov. is lingulate. The armature of the MCO differs between the two species. The MCO of G. rasini is observed to possess a large spine facing 5–6 small spines (current study; 4–7 in the literature). The MCO of G. xalapensis sp. nov. , by comparison, possesses a large spine facing 2 medium-sized, upwardly pointing, terminal spines and 2 small, central spines. Given the morphological similarity of these two species, PCA was used to investigate whether they formed discrete clusters ( Figures 30 , 31 ). The first PCA ( Figure 30 ), which included all specimens considered in this study, suggested some overlap between the two gyrodactylids (labelled d and m). The subsequent removal of G. costaricensis , G. gambusiae , G. j a ro c h o sp. nov. , G. milleri and G. turnbulli and the re-analysis of the remaining specimens, indicated that G. r a s i n i and G. x a l a p e n s i s sp. nov. formed discrete groups ( Figure 31 ). The component loadings determined for each variable for the second PCA indicated that the hamulus point curve angle and the lengths of the marginal hook and its shaft are key to the separation of G. rasini from G. xalapaensis sp. nov. through Factor2 ( Table 5 ). This is confirmed when the average values for each of these variables is considered ( i.e. 13.5° (hamulus point curve angle), 22.8 µm (marginal hook total length) and 17.5 µm (marginal hook shaft length) for G. r a s i n i and 6.7°, 25.8 µm and 20.9 µm for G. xalapensis sp. nov. , respectively) (see Tables 3 and 4 ).