|Other Abstract||Sinipercids are a group of freshwater percoid fish endemic to East Asia. Dollfustrema vaneyi and D. hefeiensis are two common digenean parasites in the intestines of sinipercid fish and even other non-sinipercid fish. Despite several taxonomical attempts and preliminary mtDNA-based molecular phylogeny among the sinipercids, their phylogenetic relationship and systematic position remained unsolved. D. vaneyi and D. hefeiensis are regarded as sibling species, which share morphological similarities. Thus, the molecular phylogeny of sinipercids and the two digeneans were investigated in the present study, with a preliminary discussion on the co-evolutionary relationship between the two digeneans and siniprcids.
The complete cytochrome b gene sequences from nine sinipercid species and four non-sinipercid fish species were cloned, and a total of 12 complete cyt b gene sequences from 10 species of sinipercids and 11 sequences from 10 species of non-sinipercid fish also in Perciformes were included in the phylogenetic analysis including neighbor-joining, maximum parsimony, maximum likelihood and Bayesian analyses. The competing sinipercid topologies were tested by using the Shimodaria-Hasegawa test (SH test), approximately unbiased test (AU test) and parametric bootstrap test. As expected, the two genera Siniperca and Coreoperca within sinipercids are recovered as monophyletic. However, nine species representing Moronidae, Serranidae, Centropomidae, Acropomatidae, Emmelichtyidae, Siganidae and Centrarchidae included in the present study are all nested between Coreoperca and Siniperca, which provides marked evidence for a nonmonophyly of sinipercid fishes. Coreoperca appears to be closest to Centrachus representing the family Centrarchidae. Coreoperca whiteheadi and C. herzi are sibling species, which together are closely related to C. kawamebari. In the Siniperca, the node between S. roulei and the remaining species is the most ancestral, followed by that of S. fortis. S. chuatsi and S. kneri are sibling species, sister to S. obscura. However, the sinipercids do not seem to have a very clear phylogenetic history, for different methods of phylogenetic reconstruction result in different tree topologies, and the only conclusive result in favor of a paraphyletic origin of the two sinipercid genera is the parametric bootstrap test. The paraphyly of Sinipercidae may suggest that the ‘‘synapomorphs’’ such as cycloid scales, upon which this family is based, were independently derived at least twice within sinipercid fishes, and further study should be carried out to include the other two Siniperca species and to incorporate other genes. By using a Bayesian relaxed molecular clock approach for inferring dating information from molecular phylogeny, the first molecular clock timescale was presented for sinipercids evolution. The genus Siniperca appears to have emerged approximately 22.71 ± 4.84 Mya, and radiated from Oligocene–Miocene boundary to Pleistocene, along with several rapid speciation events in Miocene; while genus Coreoperca have emerged most possibly approximately 22.90 ± 3.78 Mya.
Nuclear introns are commonly used as phylogenetic markers. Based on different alignment parameters, the rpS7 intron 1 sequences are aligned with Clustal x（GO = 20, 10, 5, 2.5 and 1, respectively; GE = 10, 5, 1 and 0.5, respectively）and POY 3.0.11 (gap costs = 1, 2, 4 and 8, respectively). The gaps generated in the process of alignment are treated as missing data, generating a “simple indel coding” with the program GapCoder, and excluding the areas of ambiguous alignment. The matrixes generated after the gap treatment were analyzed using maximum parsimony (MP), maximum likelihood (ML) and Bayesian (BI) methods. The different analytical strategies are assessed based on the totality of congruent nodes (MP: BP≥75%; ML: BP≥ 75%; BI: PP ≥ 0.95). It is suggested that there are more congruent nodes of phylogenetic trees reconstructed on the matrix generated from POY (gap cost = 1 and 2, respectively) than from Clustal x. The topology of the trees based on the matrix containing new information generated from GapCoder are prevalently better than the other two-gap treatments. It is also shown that excluding the areas of ambiguous alignment will decrease the phylogenetic information. In the tree reconstructed on the matrix aligned by POY (gap cost = 2), the monophyly of sinipercids was disputed by the injection of N. spinosus. The two genera Siniperca and Coreoperca were recovered with high posterior probability. In the Coreoperca, C. whiteheadi formed a sister species with C. kawamebari, and C. herzi was a basal species. The Siniperca consisted of two main clades. One was formed by S. fortis and S. undulata, and the other clade was represented by the rest members of the Siniperca. S. scherzeri 1 branched basal to this clade, and (S. roulei, S. scherzeri 2) was the secondly basal species. S. chuatsi and S. kneri was the most derived species, and formed a sister species that was sister to S. obscura.
An innate immunity gene, the virus-induced protein (viperin) gene was for the first time employed in the present study to illustrate the phylogeny of sinipercids. The A total of 7307 bp data from three mitochondrial partitions (cytochrome b, CO 1 and 16S rRNA) and viperin gene plus the first two introns of nuclear S7 ribosomal protein gene was analyzed in order to estimate the relationship among sinipercids and to assess the phylogenetic utility of viperin gene. The phylogenetic utility and homogeneity of all partitions were estimated via a combination of homogeneity partition tests, congruent nodes, and partitioned Bremer support. Despite the detection of significant heterogeneity of phylogenetic signal between the mitochondrial and nuclear partitions, the Bayesian tree by viperin gene represented the best-supported topology of all the data by using the approximately unbiased test and the Shimodaira-Hasegawa test, with the following clades: ((C. whiteheadi, (C. kawamebari, C. herzi)), ((S. obscura, S. scherzeri), ((S. undulata, S. fortis), (S. roulei, (S. kneri, S. chuatsi))))). While all Bayesian tree from other individual and simultaneous data were significantly rejected. Topological appraisals of the six gene partitions suggested that the viperin gene is a better estimator of phylogenetic relationship of sinipercids than the cyt b, COI, 16S rRNA partitions and the first two introns of rpS7 at the taxonomic levels under consideration, and it would be interesting to employ this gene for the phylogenetic analysis of other fish groups.
The complete internal transcribed spacer 1(ITS1), 5.8S rDNA, and ITS2 region of the ribosomal DNA from 60 specimens belonging to two closely-related bucephalid digeneans (Dollfustrema vaneyi and Dollfustrema hefeiensis) from different localities, hosts and microhabitat sites were cloned to examine the level of sequence variation and the taxonomic levels to show utility in species identification and phylogeny estimation. It is obvious that these molecular markers may enable the discrimination of the two species. A total of 21 haplotypes was found with 44 polymorphic positions in 38 individuals of D. vaneyi, and 16 haplotypes found with 43 polymorphic positions in 22 individuals of D. hefeiensis. There is no shared haplotypes between the two species. Haplotype rather than nucleotide diversity is similar between the two species. For ITS1, the largest intraspecific genetic distance is 0.0421, much lower than the average interspecific distance, which has a value of 0.0536, and still much lower than the shortest interspecific distance, 0.0441. For ITS2, the largest intraspecific genetic distance is 0.0435, much lower than the average interspecific distance, which has a value of 0.0576, and still much lower than the shortest interspecific distance, 0.0513. It is thus suggested that ITS1 and ITS2 are useful genetic markers for the identification of bucephalid digeneans. Phylogenetic analyses revealed two robustly supported clades, one corresponding to D. vaneyi and the another to D. hefeiensis. However, the population structures between the two species seem to be incongruent, and show no geographic and host-specific structure among them, further indicating that the two species may have had a much more complex evolutionary history than expected.|