In total, 435 specimens of the Southeast Asian freshwater leech species within the Hirudinidae family were collected from 17 locations of various types of aquatic habitats in northeastern Thailand. They were all morphologically placed within the genus Hirudinaria Whitman, 1886 and there were three distinct species: the common Hirudinaria manillensis, 78.2% of all collected specimens and at all 17 locations, Hirudinaria javanica at 20.3% of collected samples and from five locations and a rarer unidentified morphospecies (Hirudinaria sp.) with six samples from only two locations. The karyotypes of these three species were examined across their range in this study area for 38, 11 and 6 adult specimens of Hirudinaria manillensis, Hirudinaria javanica and Hirudinaria sp., respectively. This revealed different chromosome numbers among all three species, with Hirudinaria javanica having n = 13, 2n = 26, Hirudinaria manillensis lacked one small chromosome pair with n = 12, 2n = 24, and the unknown Hirudinaria sp. differed from any known Hirudinaria karyotypes in exhibiting a higher chromosome number (n = 14, 2n = 28) and a gradual change in size from large to small chromosomes. This suggests that the unknown Hirudinaria sp. is a new biological species. However, phylogenetic analysis based upon a 658 bp fragment of the cytochrome oxidase subunit I gene placed this unknown morphospecies within the Hirudinaria manillensis clade, perhaps then suggesting a recent sympatric speciation, although this requires further confirmation. Regardless, the chromosomes of all three species were asymmetric, most with telocentric elements. A distinct bi-armed chromosome marker was present on the first chromosome pair in Hirudinaria javanica, whilst it was on pairs 1, 2, 3 and 5 in Hirudinaria manillensis, and on pairs 3 and 5 for the unknown Hirudinaria sp.

Freshwater leeches, Hirudinea , karyotypes, morphology, COI, sanguivorous

The family Hirudinidae (Arhynchobdellida, Hirudiniformes) is comprised of mainly blood-sucking (sanguivorous) freshwater leeches, or medicinal leeches, although four terrestrial species are known. It includes approximately 60 hirudinids ranging across all continents, except for Antarctica, and from temperate to tropical regions (Elliott and Kutschera 2011). On the basis of the number of complete somites, the distance between the third and fourth pair of eyes, number of sensillae, position of the nephropore opening, and the presence or absence of auricular characters, the Hirudinidae family is divided into the two subfamilies of Hirudininae and Haemadipsinae. The Hirudininae, or buffalo leeches, contains 12 known species (Moore 1927) within six genera (Dinobdella, Hirudinaria, Hirudo, Limnatis, Myxobdella and Whitmania), and are distributed in temperate and tropical Asia, Africa and the Caribbean islands (Richardson 1969, Sket and Trontelj 2008, see Phillips and Siddall 2009 for alternative classification). The high species diversity and their wide geographic distribution make the hirudinid leeches attractive material for systematic and biogeographical studies. However, due to their conserved morphology, it is not easy to establish a reliable phylogenetic hypothesis for this group. There is only one recent published paper regarding their phylogenetic relationship that considered both morphological and molecular analyses and described a new species of the Asian buffalo leech, Hirudinaria bpling Phillips, 2012.

The genus Hirudinaria Whitman, 1886 consists of only three known species Hirudinaria javanica (Wahlberg, 1856), Hirudinaria manillensis (Lesson, 1842), and Hirudinaria bpling that are widely distributed over tropical South and Southeast Asia, being recorded from within Peninsular Malaysia, Thailand, Indo-China, Indonesia, Philippines, China, Myanmar, Bangladesh, India and Sri Lanka (Moore 1938, Lai and Chen 2010). Chromosomal data for hirudinid leeches have only been recorded for three species of Hirudo (Utevsky et al. 2009).

In this study, we examined the karyotypes of 38, 11 and 6 specimens of the three species (H. manillensis, H. javanica, and a third distinct and different morphospecies, Hirudinaria sp.) collected from across 17 locations in northeastern Thailand, representing 13.4%, 12% and 100% of the collected samples, respectively. Their systematic implications are then discussed in comparison with other previously reported hirudinid karyotypes. The phylogenetic analysis, based upon a 658 bp fragment of the cytochrome oxidase subunit I gene, was also conducted to clarify the systematics of all collected morphospecies.

Locality, co-ordination and sample size for all collected species are given in Table 1. Species identification of each specimen was made on the basis of Lesson (1842), Wahlberg (1856), Whitman (1886), Moore (1927), Richardson (1969), Klemm (1972) and Lai and Chen (2010). Voucher specimens were deposited in the Museum of Zoology, Chulalongkorn University, Bangkok, Thailand (CUMZ).

Freshwater leeches were collected from 17 localities in northeastern Thailand (Fig. 1 and Table 1) during invertebrate faunal surveys performed from April 2012 to February 2014. In total, 435 adult specimens were collected and examined. Specimens were photographed and kept alive in a glass aquarium in order to observe the body color pattern and other external morphological characteristics, plus any behavioral traits. Most specimens were relaxed in 10% (v/v) ethanol and then fixed and kept in 95% (v/v) ethanol for further external and internal morphological studies. Some specimens were brought back alive to the laboratory for karyotypic analysis.

Figure 1. 

Map showing the locality of the sampling sites (collection of specimens from the genus Hirudinaria) in northeastern Thailand. Further details of sample numbers and locations are given in Table 1.

Jaws of some specimens were examined by scanning electron microscopy (SEM). The dried specimens were sputter coated with 35 nm of gold/palladium before being examined using a LEO/Zeiss DSM982 Geminifield emission scanning electron microscope located in the Scientific and Technological Research Equipment Centre, Chulalongkorn University.

Chromosome preparations were made from the testisac using hypotonic, fixation and air-drying techniques modified from Patterson and Burch (1978) and Kongim et al. (2013). Live leeches were injected with 0.1 mL of 0.1% (v/v) colchicine, left for 3–4 h and then dissected to remove the testisacs into 0.07% (w/v) KCl solution (hypotonic) for 30 min. Samples were then fixed in fresh Carnoy’s fixative (3:1 (v/v) absolute ethanol: glacial acetic acid). The testisacs were cut into small pieces in fresh Carnoy’s fixative and the separated cells were collected by centrifugation at 1,500 rpm for 10 min. The supernatant was removed and the cell pellet resuspended in 0.5 mL of fresh Carnoy’s fixative. Cell suspensions were dropped onto clean pre-heated (60 °C) glass slides, air-dried and stained in 4% (v/v) Giemsa solution for 10 min. Photomicrographs of 10 to 15 well-spread metaphase cells were measured for their relative length and centromeric index. Mitotic karyotypes were arranged and numbered for chromosome pairs.

For the molecular analysis, the total genomic DNA was extracted from a part of the wall-body muscle to avoid contamination from the host DNA, following the standard protocol of the DNeasy Blood & Tissue Kit (Qiagen Inc., Valencia, CA, USA). A fragment of the mitochondrial cytochrome oxidase subunit I (COI) gene was amplified using the primers LCO1490 (5’-GGT CAA CAA ATC ATA AAG ATA TTG G-3’) and HCO2198 (5’-TAA ACT TCA GGG TGA CCA AAA AAT CA-3’), which is the region used in animal DNA barcoding (Folmer et al. 1994). Polymerase chain reaction (PCR) of a 50 µL final volume using 20 µM of 2×Illustra hot starts master mix (GE Healthcare), plus 10 µM of each primer and about 10 ng of DNA temfig was performed in an eppendorf Mastercycler® pro S PCR thermal cycler with the following thermal cycling conditions: 3 min at 94 °C followed by 35 cycles of 1 min at 94 °C, 1 min at 45 °C and 150 s at 72 °C, before a final extension at 72 °C for 5 min. The PCR products were purified with a QIAquick PCR purification Kit (QIAGEN Inc.) before being commercially direct cycle-sequenced at Macrogen, Inc, Korea.

Sequence alignment and editing were performed using MEGA 6.06 (Tamura et al. 2013). The best-fit models of nucleotide substitution, as judged by the Akaike information criterion (AIC: Akaike 1974), were estimated using Kakusan4 (Tanabe 2007; with maximum likelihoods calculated in Treefinder, Jobb et al. 2004). The best-fit evolution model obtained was GTR+G. Phylogenetic trees based on maximum likelihood (ML) and Bayesian inference (BI) were constructed. The ML analysis was performed with Treefinder (Jobb et al. 2004), using the likelihood-ratchet method with 1000 bootstrap replicates. The BI tree was constructed using MrBayes v3.2.2 (Ronquist et al. 2012), which employs a Metropolis-coupled, Markov chain Monte Carlo (MC-MCMC) sampling approach. The BI analysis was run twice in parallel for one million generations (with default heating values), starting with a random tree, and trees were sampled every 100 generations. The remaining trees, after discarding 25% of ‘‘burn-in’’ samples, were used for calculation of the bipartition posterior probability (Ronquist et al. 2012). Tree topologies with bootstrap values of 70% or greater for ML and/or a bipartition posterior probability of 0.95 or greater for the BI were regarded as sufficiently resolved (Huelsenbeck and Hillis 1993, Larget and Simon 1999). Pairwise (uncorrected-p) sequence distances were also calculated using MEGA 6.06 (Tamura et al. 2013).

Nucleotide sequences obtained in this study have been deposited in the GenBank database under the GenBank ID: KJ551848–KJ551855.

All 435 examined specimens in this study were assigned as belonging to the genus Hirudinaria by the following distinct characters; male pore and female pore separated by 5–7 annuli, sensillae large and elongated, salivary papillae present, and without vaginal stalk. From these identified characters, the specimens were determined to be three species: as Hirudinaria javanica, Hirudinaria manillensis, and an unidentified morphotype, Hirudinaria sp. (Figs 2 and 3).

Figure 2. 

The color pattern of A Hirudinaria javanica CUMZ 3424 from Mukdahan B Hirudinaria manillensis CUMZ 3403 from Nakhon Phanom, and C Hirudinaria sp. CUMZ 3405 from Nakhon Phanom.

Figure 3. 

The dorsal and ventral sides of A Hirudinaria javanica CUMZ 3404 from Nakhon Phanom B Hirudinaria manillensis CUMZ 3403 from Nakhon Phanom, and C Hirudinaria sp. CUMZ 3406 from Nakhon Phanom.

The chromosomes were typically indistinct because of their small size. Nevertheless, all cleared metaphase arrangements could be observed and the spermatogonial meiotic and mitotic chromosome numbers could be confirmed for all the examined species (Fig. 6). Haploid and diploid numbers of the three species of Hirudinaria were found to differ, ranging from n = 12, 2n = 24 for Hirudinaria manillensis, n = 13, 2n = 26 for Hirudinaria javanica, and n = 14, 2n = 28 for Hirudinaria sp. (Figs 6 and 7) and did not differ within each species across their respective geographic populations (Table 3). Chromosomal data of the three investigated Hirudinaria species obtained in the present study are summarized in Table 3 along with that for three other hirudinid species (all from the genus Hirudo) from the literature for comparison.

The karyotypes of all three species were asymmetric, and mostly telocentric, chromosomes. The distinct bi-armed chromosome marker varied among the three species, being found on the first pair in Hirudinaria javanica, on pairs 1, 2, 3 and 5 for Hirudinaria manillensis and on pairs 3 and 5 for Hirudinaria sp.

Figure 7. 

Karyotypes of A Hirudinaria javanica B Hirudinaria manillensis, and C Hirudinaria sp.

The samples used for phylogenetic analysis and their collection locations are summarized in Table 4. A total of 3, 4 and 2 adult specimens of Hirudinaria manillensis, Hirudinaria javanica and Hirudinaria sp., respectively, were included. Fragments of the mitochondrial COI gene (DNA barcode region) containing 658 base pairs (bp) were used for the phylogenetic tree estimation. The final alignment data metric contained a total of 224 variable sites, 162 sites of which were parsimony informative. The nucleotide compositions of the gene fragments were A (28.32%), C (15.78%), G (15.51%) and T (40.39%). The phylogenetic tree showing the evolutionary relationships among Hirudinaria species and related taxa is shown in Fig. 8. Tree topology estimated by ML and BI analyses gave identical topologies with a high support for all major nodes (ML bootstrap values of 99.3–100% and a BI bipartition posterior probability of 1). The phylogenetic tree strongly supported the monophyly of the genus Hirudinaria. Hirudinaria bpling was basal to the Hirudinaria javanica and Hirudinaria manillensis clades. Hirudinaria sp. came out within the Hirudinaria manillensis clade.

Figure 8. 

Phylogenetic relationships of the genus Hirudinaria and their related species, with chromosome number data. Tree topology was obtained from ML analysis based on a 658 bp fragment of the mitochondrial COI gene (DNA barcode region). Nodes with a 0.95 or higher bipartition posterior probability for BI and/or 70% or higher bootstrap value for ML were regarded as sufficiently resolved nodes, and are shown for the major clades (ML/BI). Numbers in parentheses refer to sampling localities in Figure 1 and the list in Table 1. Chromosome data of the related species were taken from Vitturi et al. (2002) and Utevsky et al. (2009).

The uncorrected p-distances between the members of the genus Hirudinaria are shown in Table 5. The highest value of 0.132 was between Hirudinaria bpling and Hirudinaria sp. (2n = 28) and the lowest value of 0.014 was between Hirudinaria manillensis (2n = 24) and Hirudinaria sp. (2n = 28).

All 435 examined specimens in this study were found by morphological analysis to belong to three distinct species within the genus Hirudinaria, and were identified as Hirudinaria javanica, Hirudinaria manillensis and an unidentified morphospecies (Hirudinaria sp.). They all shared various diagnostic characters reported in other studies, such as: a medium to large body size; five pairs of large eyes with the third and fourth pairs separated by one annulus, and the fourth and fifth pairs separated by two annuli; a large jaw; the presence of salivary papillae; gonopores separated by 5–7 annuli, and the absence of a vaginal stalk (Whitman 1886, Moore 1927, Klemm 1972, Nesemann and Sharma 2001, Lai and Chen 2010).

The unidentified species (Hirudinaria sp.) was different from the other two (Hirudinaria javanica and Hirudinaria manillensis) in both its morphology and also in its chromosome number and karyotype. Morphologically, Hirudinaria sp. had fewer salivary papillae (25) than the other two species (43 and 30 for Hirudinaria javanica and Hirudinaria manillensis, respectively) and a higher estimated number of teeth per jaw (167 versus 134 and 148 for Hirudinaria javanica and Hirudinaria manillensis, respectively) (Fig. 5). Although previous studies have reported a higher number of teeth for Hirudinaria javanica and Hirudinaria manillensis at 150 and 145, respectively (Moore 1927, Phillips 2012), than found in this study, these were still lower than that found for Hirudinaria sp. in this study. Comparison of all the taxonomic characters (Table 2) revealed that Hirudinaria sp. was quite similar to Hirudinaria manillensis in terms of having the gonopores separated by five annuli, but it differed in color pattern (Figs 2 and 3). However, the phylogenetic analysis, based upon the 658 bp of the mitochondrial COI gene sequence, placed Hirudinaria sp. in the same clade as Hirudinaria manillensis. Thus, they may represent recently sympatrically separated species. Hirudinaria manillensis was the most abundant and frequently found species in this study (346/435 or 79.5% of the collected specimens and found in all 17 sampled localities), compared to 83 (19%) specimens from five locations for Hirudinaria javanica and the seemingly rarer 6 samples (1.4%) from only two locations for unidentified Hirudinaria sp. Surprisingly, the northeastern Thailand population of Hirudinaria manillensis examined in this study showed a distinctly different internal morphology from that previously reported elsewhere. It contained a nerve cord running along on the left side of the atrium, instead of the right side as previously reported (Lai and Chen 2010), and also as found in Hirudinaria javanica and Hirudinaria sp. in this study.

With respect to the karyotypic analysis, the haploid and diploid chromosome numbers were similar to those reported previously in other genera of Hirudinidae (n = 14 in Hirudo medicinalis, n = 12 in Hirudo orientalis and n = 13 in Hirudo verbena) (Utevsky et al. 2009), but differed in chromosome structure and morphology. Moreover, distinctive karyotypic chromosome markers were presented, such as a distinct bi-arm chromosome that was only found on the first pair in Hirudinaria javanica, on pairs 1, 2, 3 and 5 in Hirudinaria manillensis, and on pairs 3 and 5 in Hirudinaria sp. That Hirudinaria manillensis showed the lowest chromosome number and the widest distribution across northeastern Thailand of the sampled species is of interest since, in general, it is believed that the original or ancestor species has the lowest chromosome number and is often the most common species (Sumner 2003). The unidentified species (Hirudinaria sp.) in this study had the same haploid and diploid chromosome numbers as Hirudo medicinalis, but their karyotypes were different (Utevsky et al. 2009) and their phylogenetic placement was markedly different, being placed in well-supported distinct clades, confirming that they are indeed separate biological species.

Our current identification of these 435 samples to three morphospecies (two nominal species and one unidentified morphospecies) was quite clear because of the distinct appearance of their external and internal organs, and was supported by the distinct chromosome numbers and karyotypes of the analyzed samples of each species. However, given the apparent variation between that reported here for, for example, Hirudinaria manillensis and that reported for the same nominal species elsewhere, indicates a need for further comparative studies utilizing type specimens and additional molecular analysis of these and congener species, for species confirmation and prior to any further systematic discussion and taxonomic re-classification. In particular, the potential recent sympatric speciation of Hirudinaria manillensis and Hirudinaria sp. requires further confirmation.

This project was funded partly by a grant from the 90^th Anniversary of Chulalongkorn University Fund (Ratchadaphiseksomphot Endowment Fund). The main funding was from The TRF Senior Research Scholar, The Thailand Research Fund (RTA 5580001) to SP. We appreciated the critical reading the manuscript by Dr. Robert Butcher from the PCU Unit, Faculty of Science, Chulalongkorn University. We thank all members of the Animal Systematics Research Unit, Chulalongkorn University, and Kridsada Deein, Pramook Ruekaewma and Parinda Ratanadang from Fisheries Department, Ministry of Agriculture for assistance in collecting some material.