An exceptionally rich complex of Sanguinicolidae von Graff, 1907 (Platyhelminthes : Trematoda) from Siganidae, Labridae and Mullidae (Teleostei : Perciformes) from the Indo-West Pacific Region

We describe an unprecedented radiation of sanguinicolid blood flukes ( Digenea: Sanguinicolidae) from two species of Labridae (Choerodon venustus and C. cauteroma), seven species of Mullidae (Mulloidichthys vanicolensis, Parupeneus barberinoides, P. barberinus, P. bifasciatus, P. cyclostomus, P. indicus and P. multifasciatus) and ten species of Siganidae (Siganus argenteus, S. corallinus, S. doliatus, S. fuscescens, S. lineatus, S. margaritiferus, S. puellus, S. punctatus, S. virgatus and S. vulpinus) from sites off Australia and Palau. The flukes were morphologically similar in having the combination of a long thread-like body, tegumental spines in lateral transverse rows, a vestigial oral sucker bearing concentric rows of fine spines, an H-shaped intestine, a cirrussac, a notch level with the male genital pore, a lateral or post-ovarian uterus, a uterine chamber and separate genital pores. These species are divided into two genera on the basis of testis number. Sanguinicolids from Siganus fuscescens have a single large testis between the intestinal bifurcation and the ovary and are placed in Ankistromeces Nolan & Cribb, 2004. Species from the remaining nine species of Siganidae, Labridae and Mullidae are placed in Phthinomita n. g.; these species have two testes, the anterior testis being large and between the intestinal bifurcation and the ovary whereas the small posterior testis is at the posterior end of the body and appears rudimentary or degenerate and probably non-functional. The second internal transcribed spacer (ITS2) of ribosomal DNA ( rDNA) from 29 host/parasite/location combinations (h/p/l) was sequenced together with that of Ankistromeces mariae Nolan & Cribb, 2004 for comparison. From 135 samples we found 19 distinct genotypes which were interpreted as representing at least that many species. Replicate sequences were obtained for 25 of 30 h/p/l combinations ( including A. mariae); there was no intraspecific variation between replicates sequences for any of these. Interspecific variation ranged from 1 - 41 base differences (0.3 - 12.7% sequence divergence). The 19 putative species were difficult to recognise by morphological examination. We describe 13 new species; we do not describe (= name) six species characterised solely by molecular sequences and three putative species for which morphological data is available but for which molecular data is not. We have neither morphological nor molecular data for sanguinicolids harboured in five hosts species ( Siganus margaritiferus, S. puellus, Choerodon cauteroma, Parupeneus indicus and P. multifasciatus) in which we have seen infections. Where host species were infected in different localities they almost always harboured distinct species. Some host species ( for example, S. argenteus and S. lineatus from Lizard Island) harboured two or three species in a single geographical location. This suggests that, for parts of this system, parasite speciation has outstripped host speciation. Distance analysis of ITS2 showed species from each host family ( Siganidae, Mullidae and Labridae) did not form monophyletic clades to the exclusion of species from other host families. However, a host defined clade was formed by the sequences from sanguinicolids from S. fuscescens.

Cardicola Short, 1953 and Braya n. gen. (Digenea : Sanguinicolidae) from five families of tropical Indo-Pacific fishes

A survey of Pacific coral reef fishes for sanguinicolids revealed that two species of Lutjanidae (Lutjanus argentimaculatus, L. bohar), six species of Siganidae (Siganus corallinus, S. fuscescens, S. lineatus, S. margaritiferus, S. punctatus, S. vulpinus), seven species of Chaetodontidae (Chaetodon aureofasciatus, C. citrinellus, C. flavirostris, C. lineolatus, C. reticulatus, C. ulietensis, C. unimaculatus), three species of Scombridae (Euthynnus affinis, Scomberomorus commerson, S. munroi) and three species of Scaridae (Chlorurus microrhinos, Scarus frenatus, S. ghobban) were infected with morphologically similar sanguinicolids. These flukes have a flat elliptical body, a vestigial oral sucker, a single testis, separate genital pores and a post-ovarian uterus. However, these species clearly belong in two genera based on the position of the testis and genital pores. Sanguinicolids from Lutjanidae, Siganidae, Chaetodontidae and Scombridae belong in Cardicola Short, 1953; the testis originates anteriorly to, or at the anterior end of, the intercaecal field and does not extend posteriorly to it, the male genital pore opens laterally to the sinistral lateral nerve chord and the female pore opens near the level of the ootype ( may be anterior, lateral or posterior to it) antero-dextral to the male pore. Those from Scaridae are placed in a new genus, Braya; the testis originates near the posterior end of the intercaecal field and extends posteriorly to it, the male pore opens medially at the posterior end of the body and the female pore opens posterior to the ootype, antero-sinistral to the male pore. The second internal transcribed spacer (ITS2) of ribosomal DNA from these sanguinicolids and a known species, Cardicola forsteri Cribb, Daintith & Munday, 2000, were sequenced, aligned and analysed to test the distinctness of the putative new species. Results from morphological comparisons and molecular analyses suggest the presence of 18 putative species; 11 are described on the basis of combined morphological and molecular data and seven are not because they are characterised solely by molecular sequences or to few morphological specimens (n= one). There was usually a correlation between levels of morphological and genetic distinction in that pairs of species with the greatest genetic separation were also the least morphologically similar. The exception in this regard was the combination of Cardicola tantabiddii n. sp. from S. fuscescens from Ningaloo Reef ( Western Australia) and Cardicola sp. 2 from the same host from Heron Island ( Great Barrier Reef). These two parasite/ host/location combinations had identical ITS2 sequences but appeared to differ morphologically ( however, this could simply be due to a lack of morphological material for Cardicola sp. 2). Only one putative species ( Cardicola sp. 1) was found in more than one location; most host species harboured distinct species in each geographical location surveyed ( for example, S. corallinus from Heron and Lizard Islands) and some ( for example, S. punctatus, S. fuscescens and Chlorurus microrhinos) harboured two species at a single location. Distance analysis of ITS2 showed that nine species from siganids, three from scombrids and five from scarids formed monophyletic clades to the exclusion of sanguinicolids from the other host families. Cardicola milleri n. sp. and C. chaetodontis Yamaguti, 1970 from lutjanids and chaetodontids, respectively, were the only representatives from those families that were sequenced. Within the clade formed by sanguinicolids from Siganidae there wasa further division of species; species from the morphologically similar S. fuscescens and S. margaritiferus formed a monophyletic group to the exclusion of sanguinicolids from all other siganid species.