915 J. Parasitol., 91(4), 2005, pp. 915–921  American Society of Parasitologists 2005 INTRASPECIFIC VARIATION OF HAEMATOLOECHUS FLOEDAE HARWOOD, 1932 (DIGENEA: PLAGIORCHIIDAE), FROM RANA SPP. IN NORTH AND CENTRAL AMERICA Virginia Leo ´ n-Re ` gagnon, Sergio Guille ´ n-Herna ´ ndez*, and Marı ´a Antonieta Arizmendi-Espinosa Departamento de Zoologı ´a, Instituto de Biologı ´a, UNAM. Apdo. Postal 70-153. C.P. 04510, Me ´xico, D.F. Me ´ xico. e-mail: vleon@ibiologia.unam.mx ABSTRACT: Haematoloechus floedae was originally described from the lungs of Rana catesbeiana in Texas, and later reported in Florida, in the United States. It was proposed to be synonymous with H. breviplexus, H. parviplexus, and H. varioplexus. We obtained specimens of H. floedae from Rana brownorum and R. vaillanti in Yucata ´n, Mexico; R. cf. forreri and R. taylori in Guanacaste, Costa Rica; and R. catesbeiana in Georgia and California. Some specimens were processed for morphological study; sequences of the 28S of the rDNA and the mitochondrial COI were obtained from several specimens of each population. Phylogenetic analysis of molecular data indicates studied populations constitute a single taxon, different from H. varioplexus, H. breviplexus, and H. parviplexus. Some morphological characters remain constant among populations of H. floedae, and are useful for differentiating this species. These include the oral sucker-pharynx ratio, oral sucker-ventral sucker ratio, ovary and testes shape, extension of uterine longitudinal loops, and extension of vitelline follicles (in fully developed worms). Sequence homo- geneity among populations of H. floedae suggests a recent spread, perhaps due to the introduction of R. catesbeiana for culture to the western United States, southeast Mexico, and Costa Rica. Morphological variability within species of Haematoloechus spp., the frog lung flukes, has been documented (Prokopic and Krivanec, 1974; Kennedy, 1980a, 1980b), and has caused con- troversy regarding taxonomy of the group (Odening, 1960; Pro- kopic and Krivanec, 1974; Kennedy, 1981). Kennedy (1981) considered that only 6 of the 15 previously known species from Canada and the United States were valid. This author concluded that morphological characters, such as the shape of the ovary and testes, arrangement of uterine loops, presence or absence of spines on the body surface, sucker ratio, and egg size, ex- hibited intraspecific variation as the result of several factors (i.e., developmental stage, host species, crowding effect), and, therefore, are not useful for differentiating species of Haema- toloechus. Species delimitation can be a controversial issue, especially in those cases in which morphological variability can confound the taxonomy of a group (Leo ´n-Re `gagnon et al., 1999; Stock et al., 2001), or in which morphological characters are so con- servative that independent lineages cannot be differentiated (Chilton et al., 1995; Verneau et al., 1997). In these cases, DNA sequence data provide an independent source of information for differentiating diverging lineages, and to evaluate morphologi- cal characters that have been traditionally used in the taxonomy of the groups (Nadler, 2002). In the present study, morphological variation of specimens of H. floedae from different localities and host species of North and Central America are documented; molecular data are used to corroborate whether differences among populations represent intraspecific or interspecific variation, and validate the use of some morphological characters in the differentiation of species of this genus. MATERIALS AND METHODS Collecting localities and hosts are shown on Table I. Worms collected from freshly killed amphibians initially were placed in saline (0.65%) for 5–10 min. For the morphological study, they were fixed by sudden immersion in hot 4% formalin and preserved in 70% alcohol. Speci- Received 13 July 2004; revised 13 October 2004; accepted 22 October 2004. * Universidad Auto ´noma de Yucata ´n. Facultad de Medicina Veterinaria, Km 105 carretera Me ´rida-Xmatkuil, Ap. Postal 14-6. C.P. 97000, It- zimna ´, Me ´rida, Yucata ´n, Me ´xico. mens were stained with Mayer paracarmine or Gomori trichrome, de- hydrated, cleared in methyl salicylate, and mounted in Canada balsam. Specimens were mounted permanently between coverslips and held in Cobb slides. Measurements are presented as the range with means in parentheses and expressed in millimeters, unless otherwise stated. Fig- ures were drawn with the aid of a drawing tube. Voucher specimen accession numbers are listed in Table I. The following abbreviations are used: CNHE, Coleccio ´n Nacional de Helmintos, Instituto de Biologı ´a, Universidad Nacional Auto ´noma de Mexico; USNPC, United States National Parasite Collection, Beltsville, Maryland; and HWML, Harold W. Manter Laboratory, Lincoln, Nebraska. Based on the phylogenetic hypothesis proposed by Leo ´n-Re `gagnon and Brooks (2003) for the genus, we chose the most closely related species of Haematoloechus for reference. The following specimens were examined for comparison: H. breviplexus Stafford, 1902 (USNPC 82015 and 84806; HWML 20293, 21957, and 22691); H. danbrooksi Leo ´n-Re `gagnon and Paredes-Caldero ´n, 2002 (CNHE 4112 and 4151; USNPC 92220); H. floedae (USNPC 30879, 84804, and 091507); H. parviplexus Irwin, 1929 (USNPC 75445 and 81467; HWML 20142– 43, 20753, and 21660; CNHE 4405); and H. varioplexus Stafford, 1902 (USNPC 75447 and 81915; HWML 20151–20160 and 38396). Two or 3 specimens from each locality and host species were used for molecular work. They were identified in vivo and preserved in 100% ethanol. Standard phenol extraction methods were used to recover DNA from individual specimens. Laboratory protocols followed those de- scribed by Palumbi (1996) and Hillis et al. (1996). Polymerase chain reaction (PCR) was used for amplifying the 5' end of the 28S ribosomal RNA gene and partial COI mitochondrial gene. Amplification and se- quencing were performed using the primers 28Sy 5'-cta acc agg att ccc tca gta acg gcg agt-3' (forward) and 28Sz 5'-aga ctc ctt ggt ccg tgt ttc aag ac-3' (reverse) for the 28S; and COIF 5'-ttt ttt ggg gat cct gag gtt tat-3' (forward) and COIR 5'-taa aga aag aac ata atg aaa atg-3' (re- verse). Amplification programs consisted of 1 min at 94 C followed by 35 cycles of 30 sec at 92 C, 30 sec at 50 C, and 1 min at 72 C; followed by 4 min at 72 C for final elongation for the 28S gene; and 5 min at 95 C followed by 35 cycles of 30 sec at 95 C, 45 sec at 50 C, and 1.5 min at 72 C, followed by 10 min at 72 C for final elongation for the COI gene. PCR products were sequenced directly on an ABI Prism 310 automated DNA sequencer using Thermo Sequenase radiolabeled ter- minator cycle sequencing kits (Amersham Life Science, Inc., Cleveland, Ohio). GenBank accession numbers are listed in Table I. We obtained more than 1 COI sequence from each population of H. floedae to evaluate genetic divergence among and within populations; only 1 sequence of each population was included in the combined par- simony analysis. Sequences were aligned visually using the computer program Bioedit (Hall, 1999). Uncorrected distance matrices were ob- tained for the pairs of examined sequences, and phylogenetic trees were constructed using PAUP (version 4.0b10 Swofford, 2002). Unweighted parsimony analyses using a branch-and-bound search were performed considering character states as unordered, and gaps as missing data for separate and combined matrices. In each case, a nonparametric boot-