Sequencing and phylogenetic analysis of the Pyrgilauda ruficollis (Aves,Passeridae) complete mitochondrial genome

Abstract

In this study, both long PCR and conserved primers walking sequencing methods were used to determine the complete sequence of the of Pyrgilauda ruficollis mitochondrial genome (KC836121). The results showed that the complete mitochondrial genome of P. ruficollis is 1 6909 bp in length with 55.0% A+T content, harboring the typical 37 genes. The mitogenome had the same gene order with that of Podoces hendersoni. All protein coding genes started with ATG codon, except ND3 with GTG. For the stop codon usage, most genes terminate with codons TAA or TAG, but ND5 terminated with AGA, while ND1 and COI genes with AGG, and both the genes COIII and ND4 have an incomplete termination codon (T). The secondary structures of 22 tRNA genes were also predicted, showing that all tRNAs can form typical clover-leaf secondary structures, except for the tRNA^Ser (AGN) which loses the DHU arm, while tRNA^Phe harbor an extra nucleotide inserted in the TψC arm. The predicted secondary structures of 12S rRNA and 16S rRNA exhibit 47 helices in 4 domains and 60 helices in 6 domains respectively. The control region of P. ruficollis with the length of 1 305 bp was located between tRNA^Glu and tRNA^Phe, and typical domains of which could be found as other bird groups. Using the data from 13 mitochondrial protein-coding genes, results of a final phylogenetic analysis strongly supports the traditional view that P. ruficollis is closely related with Passeridae and Fringillidae.

To date, there are seven known snow finches (including genus/Montifringilla and Geospiza), six of them distributed in the Qinghai-Tibet Plateau and adjacent areas, and four of these are endemic to the area. There are different views and arguments on the classification of snow finches-one is that all seven snow finches exist as an independent genus (Howard & Moore, 1980). Ivanitskii (1992) and Gebauer & Kaiser (1994) however proposed dividing them into Montifringilla and Pyrgilauda, according to their ecological habits. Montifringilla would then include Montifringilla nivalis and Montifringilla adamsi, while Pyrgilauda would contain Pyrgilauda taczanowskii, Pyrgilauda ruficollis, Pyrgilauda blanfordi, Pyrgilauda davidiana and Pyrgilauda theresae. A later clustering analysis by Lei et al (2000) reframed the issue somewhat, with P. taczanowskii, P. ruficollis, P. blanfordi and P. davidiana tentatively classified into subgenus Pyrgilauda or Pyrgilauda species group. A cladistic analysis by Lei et al (2001) also showed that P. davidiana, P. blanfordi and P. theresae have a close relationship and constitute a monophyletic group that forms a clade together with P. ruficollis, yielding the monophyletic group of Pyrgilauda. The two most recent studies by Zheng (2012) and Gill & Donsker (2013) likewise classified P. ruficollis as Pyrgilauda, and their analysis is generally accepted by most researchers in the fields.

The variety of findings from the studies on the snow finches and the evolution in the utilized study methodologies underscores both the continued debate on taxonomy and phylogeny. Recently, it has become increasingly popular to usewhole mitochondrial genome sequencing to help settle some of the ongoing disputes and discrepancies in classifying animals and in developing a more complete picture of their evolution. Genetic studies on the mitochondrial gene can be used to resolve the phylogenetic relationships of high-level category (Simon et al, 2006), and also has a better resolution on the genetic relationship of subfamily, genus, inter-species and even the genetic structure of populations (Zhou et al, 2010). Rather surprisingly, as useful as these studies are, to date there has been no research regarding the mitochondrial genome sequencing and phylogenetic studies of Pyrgilauda ruficollis-a bird living mainly in Palaearctic areas, belonging to Passeridae in Passeriformes-and instead most phylogenetic studies on this species are mainly focused on morphological evidence, while a small amount reports on RAPD-PCR classification (Qu, 2006).

In this study, we sought to remedy this deficiency by sequencing the mitochondrial genome of P. ruficollis, describing in detail the starting position, composition content and structural characteristics of each gene. We also attempted to predict and analyze the secondary structures of 22 tRNAs and two rRNA genes. To provide some new insights into this specie’s systematic evolution and the mitochondrial phylogenomics of Passeriformes, we analyzed the phylogenetic position of P. ruficollis based on protein-coding genes (PCGs) datasets of 32 Passeriformes mitochondrial genomes published in NCBI using a molecular systematics method.

MATERIALS AND METHODS

Materials

Samples of P. ruficollis were collected in Hot Springs Village, Xinghai County, Hainan Prefecture, Qinghai Province on August 12, 2003, at an elevation of 3 960 m. Samples were immersed in absolute ethanol at -20 ℃, with replacement of alcohol several times. The voucher specimens (PJ917) were deposited in the Bird Herbarium at the Key Laboratory of Animal Systematic and Evolutionary, Institution of Zoology, Beijing, Chinese Academy of Sciences.

Methods DNA extraction

The total genomic DNA of P. ruficollis was extracted and purified using traditional phenol-chloroform-isoamyl alcohol method (Yang et al, 2010) and stored in refrigerator prior to await further analysis.

Primer design and PCR amplification

According to the mitochondrial genome sequence of <italic>Podoces hendersoni</italic> (NC_014879) alongside the 39 mitochondrial genomes of Passeriformes available from GenBank, we used ClustalX to align and search the highly conserved sequences, and then employed Primer Premier 5.0 to design 29 pairs of PCR amplification primers. The design process based on the principle of primer design, and referred to the primer location identified in <xref ref-type="bibr" rid="b41-ZoolRes-35-2-81">Sorenson et al (1999)</xref> and <xref ref-type="bibr" rid="b-notfound">Sorenson (2003)</xref>. Oligo was then used to evaluate and modify each primer. Of these, six pairs of primers were selected for L-PCR, while the rest were used as Sub-PCR primers (primers synthesized by Shanghai Sangon Biological Engineering Technology Co.). The approximate location and approximate length of amplified fragment in the whole genome is listed in Supplementary Tables 1 and 2 (Supporting information of <a href="http://www.zoores.ac.cn/">http://www.zoores.ac.cn/</a>).

Using 6 pairs of L-PCR primers, the whole mitochondrial genome of P. ruficollis was amplified with six overlapped large fragments (L1-L6), then Sub-PCR was used to amplify small fragments with length of 500−1300 bp. PCR products were detected using 0.8% agarose gel electrophoresis. Good bands were cut and purified using DNA purification kit (U Gene), then the target fragments were sequenced. To assist with Sub-PCR, some samples of the large fragments were stored.

The total volume of the reaction system was 50 μL, with DreamTapTM Green PCR Master Mix (2×)25 μL, Water, nuclease-free 22 μL, upstream and downstream primers (10 μmol/L)1 μL for each, DNA template 1 μL. The reaction program of L-PCR was as follows; 93 ℃ denaturation for 2 min; (92 ℃for 10 s, 58−53 ℃ for 30 s, 68 ℃ 10 min)× 20 cycles; (92 ℃ 10 s, 53 ℃30 s, 68 ℃10 min, and each cycle increases 20 s)× 20 cycles; 68 ℃ 7 min; 4 ℃ incubation. The reaction program of Sub-PCR were as follows: 95 ℃ denaturation 4 min (95 ℃ 45 s, 53−58 ℃ 60 s, 72 ℃60 s)× 30 cycles; 72 ℃ 4 min, 4 ℃ incubation.

Sequencing

Using six pairs of L-PCR amplification primers, the whole mitochondrial genome of P. ruficollis was amplified with six large overlapping fragments (L1−L6), the regions of walking sequencing ineffective or poor coverage between long fragments were amplified using Sub-PCR and then sequenced. Sequencing was completed by the Shanghai Sangon Biological Engineering Technology Company.

Sequence assembly, annotation and analysis

The Staden Package 1.7 was used for the sequencing assembly and annotation, tRNAScan-SE 1.21 (<a href="http://lowelab.ucsc.edu/tRNAscan-SE/">http://lowelab.ucsc.edu/tRNAscan-SE/</a>) was then applied to predict the secondary structure of the tRNA. The mitochondrial DNA sequences of <italic>P. hendersoni</italic> was used as the template during annotation. Using MEGA 5.0 software (<xref ref-type="bibr" rid="b43-ZoolRes-35-2-81">Tamura <italic>et al</italic>, 2011</xref>), basic group composition, codon usage frequency of protein genes and other information in mitochondrial genome were calculated.

Phylogenetic analysis

Using the protein-coding genes (PCGs) of 32 Passeriformes species and 4 Anseriformes species published in NCBI (see Supplementary <xref ref-type="table" rid="T3-ZoolRes-35-2-81">Table 3</xref>, Supporting information of <a href="http://www.zoores.ac.cn/">http://www.zoores.ac.cn/</a>), as well as those obtained from <italic>P. ruficollis</italic>, we compiled a dataset to construct maximum likelihood (ML), Bayesian inference (BI) and maximum parsimony (MP) trees. The maximum likelihood method was implemented by RAxML 7.0.4 (<xref ref-type="bibr" rid="b-notfound">Stamatakis, 2006</xref>). The most suitable model (GTR+G) was selected by Modeltest3.7 (<xref ref-type="bibr" rid="b30-ZoolRes-35-2-81">Posada & Crandall, 1998</xref>).The process performed rapid bootstrap analysis, and searched for ML tree with best-scoring during a separate program. A random seed number started rapid bootstrap analysis, using model GTRGAMMA. Nodal support was evaluated using 1 000 replications of rapid bootstrapping implemented in RaxML.

MrBayes 3.1 (Ronquist & Huelsenbeck, 2003) was used to reconstruct the BI tree with four Anseriformes species acting as an outgroup. GTR + I + G modelwas selected as the optimal evolutionary model using Modeltest 2.2 (Nylander, 2004). Four Markov chains (one cold chain and three hot chains) runs at one million generations, sampling every 100 generations, were used. The first 1000 samples were discarded as burn-in, and consensus tree was then acquired with the posterior probability (PP) in each branches.

PAUP * 4.0b10 was used to construct MP tree with a 1 000 bootstrap resampling test.

RESULTS

Composition and gene arrangement

The mitochondrial genome of P. ruficollis contains 13 protein-coding genes (PCGs), 22 tRNA genes, 2 rRNA genes (lrRNA and srRNA) and a non-coding control region (D-Loop region), with length of 16 909 bp (GenBank accession number KC836121). The position of the 37 other genes (see Table 1) were identical to those other published Passeriformes species. In general, gene arrangement is quite tight, and adjacent genes have little or no non-coding nucleotides except for in the control region. There are a total of 30 bp overlapping sequences between 7 genes with length of 1−10 bp. There are 105bp interval sequences between 18 pairs of genes with length of 1−22 bp. The largest intergenic region was 22 bp, located between the tRNA^{leu (UUR)} and ND1 gene. Finally, 12 pairs of genes had neither overlap nor interval.

Table 1.

Composition of the Pyrgilauda ruficollis mitochondrial genome

Gene	Coding strand	Start position	End position	Length	Intergenic length	Initiation codons	Termination condones
tRNAPhe	H	1	68	68
SrRNA	H	69	1046	978	0
tRNAVal	H	1047	1116	70	0
LrRNA	H	1117	2713	1597	0
tRNALeu (UUR)	H	2714	2788	75	0
ND1	H	2811	3788	978	22	ATG	AGG
tRNAIle	H	3794	3865	72	5
tRNAGln	L	3872	3942	71	6
tRNAMet	H	3942	4010	69	-1
ND2	H	4011	5051	1041	0	ATG	TAA
tRNATrp	H	5051	5120	70	-1
tRNAAla	L	5122	5190	69	1
tRNAAsn	L	5202	5276	75	11
tRNACys	L	5277	5343	67	0
tRNATyr	L	5343	5413	71	-1
COI	H	5415	6965	1551	1	ATG	AGG
tRNASer (UCN)	L	6957	7029	73	-9
tRNAAsp	H	7033	7101	69	3
COII	H	7108	7791	684	6	ATG	TAA
tRNALys	H	7793	7862	70	1
ATP8	H	7864	8031	168	1	ATG	TAA
ATP6	H	8022	8705	684	-10	ATG	TAA
COIII	H	8713	9496	784	7	ATG	T
tRNAGly	H	9497	9565	69	0
ND3	H	9566	9916	351	0	GTG	TAG
tRNAArg	H	9918	9987	70	1
ND4L	H	9989	10285	297	1	ATG	TAA
ND4	H	10279	11656	1378	-7	ATG	T
tRNAHis	H	11657	11726	70	0
tRNASer (AGN)	H	11727	11792	66	0
tRNALeu (CUN)	H	11792	11862	71	-1
ND5	H	11863	13680	1818	0	ATG	AGA
CytB	H	13689	14831	1143	8	ATG	TAG
tRNAThr	H	14846	14926	81	14
tRNAPro	L	14928	14997	70	1
ND6	L	15013	15531	519	15	ATG	TAG
tRNAGlu	L	15533	15604	72	1
A+T rich		15605	16909	1305	0

Nucleotide composition

The base content of P. ruficollis mtDNA was C> A > T > G, with the content of G + C% is slightly greater than A + T%, but we observed no significant difference. The content of the four bases are very close, which is consistent with mitochondrial genome base content known for the vast majority of reported birds. Base content in protein-coding genes is most similar to the entire mitochondrial genome. One issue in the base content of different codons was that the third codon of P. ruficollis exhibited some fluctuations, which is in line with the findings reported by Gao et al (2009) which noted the third base content of different codons in 74 mitochondrial genomes of birds, suggesting that the base changes in the mitochondrial genome were mainly caused by the changes of the third codon.

Protein-coding genes and codon usage

The length of 13 protein-coding genes in P. ruficollis were equal to those of the other published Passeriformes species. All protein-coding genes used ATG as an initiation codon, except the COI gene which was GTG. TAA is also the most commonly termination codon, but ND1 and COI genes used AGG as termination codons, while ND5 gene used AGA, and COIII and ND4 genes had an incomplete termination codon T. The remaining genes used typical TAA or TAG. Relative synonymous codon usage in P. ruficollis was also similar to that found in other birds. Leu, Thr, Ala, Ser and Ile had the highest frequency in 3 787 encoded amino acids, accounting for 49.41% of the total observed content.

Secondary structure of tRNA

Totally, there are 22 tRNAs in the <italic>P. ruficollis</italic> mitochondrial genome, including 14 located on the heavy chain and 8on light chain. Anticodons used in <italic>P. ruficollis</italic> are the same as those in <italic>Pseudopodoces humilis</italic>. All tRNA secondary structure except for <italic>tRNA^{Ser (AGN)}</italic> were predicted by tRNAScan-SE 1.21, and all of them form a typical clover structure (Supplementary <xref ref-type="fig" rid="F1-ZoolRes-35-2-81">Figure 1</xref>, Supporting information of <a href="http://www.zoores.ac.cn/">http://www.zoores.ac.cn/</a>). There were 36 mismatches in the 20 predicted tRNA secondary structures, though the remaining <italic>tRNA^Lys</italic> and <italic>tRNA^Arg</italic> had no mismatch. Among these, there are 27 pairs of GU mismatches, and the remaining are as follow: 2 pairs AC mismatch occurs in anticodon arm of <italic>tRNA^Asp</italic> and amino acid acceptor arm of <italic>tRNA^His</italic>; 2 pairs AA mismatch located in TψC arm of <italic>tRNA^Asp</italic> and the DHU arm of <italic>tRNA^Trp</italic>; 2 pairs of UU mismatches in the anticodon arm of <italic>tRNA^Gly</italic> and the TψC arm in <italic>tRNA^Met</italic>; 3 pairs of CC mismatches in the anticodon arm of <italic>tRNA^Gly</italic> and amino acid acceptor arm of <italic>tRNA^{Leu (UUR)}</italic> and the TψC arm of <italic>tRNA^Phe</italic>. The <italic>tRNA^{Ser (AGN)}</italic> lacks a DHU arm, and a CC mismatch is present in the TψC arm of <italic>tRNA^Phe</italic>.

Secondary structure of rRNA

The srRNA gene of P. ruficollis is located between tRNA^Phe and tRNA^Val with a length of 978bp. The predicted srRNA gene secondary structure (Supplementary Figure 2) is roughly the same as other published bird srRNA structures, containing four domains with 47 stem-loop structures. The lrRNA gene is located between tRNA^Val and tRNA^{Leu (UUR)} with a length of 1 597bp. The predicted lrRNA gene secondary structure of the P. ruficollis lrRNA gene contains six domains, containing 60 stem-loop structures (Supplementary Figure 3).

Figure 2.

ML tree (a), BI tree (b), and MP tree (c) reconstructed using complete mitochondrial PCG sequence data sets

Figure 3.

Comparison on the 174 site of the mitochondrial ND3 gene sequence among five representative avian species

Control region

The control region of P. ruficollis is located between tRNA^Glu and tRNA^Phe with a length of 1 305 bp. Similar to the structure in P. hendersoni and Larus brunnicephalus (Yang et al, 2012), the control region of P. ruficollis contains an F, E, D, C, B domain, and conserved sequence blocks (CSB) (Figure 1).

Figure 1.

Structure of Pyrgilauda ruficollis mitochondrial control region

Phylogenetic analysis

The maximum likelihood (ML), Bayesian inference (BI) and maximum parsimony (MP) trees based on PCG dataset of 33 Passeriformes and 4 Anseriformes species are shown in Figure 2.

The tree topology obtained via three different methods are generally similar. The main difference is that ML and MP trees support the partition of Oscines and Suboscines, while the BI tree supports that Acanthisitta chloris is the sister group with Oscines and Suboscines. In Passerii, Menuridae is the sister group with Corvida and Passerida, and the monophyly of Passerida and Corvida except for Menuridae is supported. In the 3 superfamily of Passerida, Sylvioidea is located at the base of the trees, and the evolutionary relationship of Passeroidea and Musicicapoidea is closer. Several families of Sylvioidea formed a clade, except P. humilis gathered together with Musicicapoidea. Pycnonotidae, Acrocephalidae, Sylviidae and Timaliidae are mutually sister group. Among them, the relationship of Pycnonotidae and Acrocephalidae is closer, while Sylviidae has a close relationship with Timaliidae. InPasseroidea, P. ruficollis is close to Fringillidae, but has certain distant with Estrildidae. Musicicapoidea is divided into Sturnidae and Turdidae, and P. humilis get together with Turdidae.

DISCUSSION

The cytosine insertion phenomenon in ND3 gene

According to Gao et al (2009), cytosine insertion phenomenon in ND3 gene was found in almost all of the Palaeognathae species while most Neognathae species had no such phenomenon. The ND3 sequence of P. ruficollis and four Passeriformes published in NCBI shows no cytosine insertion phenomenon at position 174 (Figure 3).

Secondary structure of tRNA

The TψC arm of tRNA^Phe in the P. ruficollis mitochondrial genome has no “C”insertion but forms a CC mismatch, which is different from other Passeriformes species (Figure 1, Box 1). This findings is differs from that in Harrison et al (2004), in which three matching forms of TψC arm in tRNA^Phe in the avian mitochondrial genome is normally present (Figure 4a-c). This result also differs from that reported by Yang et al (2012) regarding the Larus brunnicephalus mitochondrial genome (Figure 4-d). It is generally accepted that form one (Figure 4-a) is the ancestor form of modern birds, and commonly exist in four Model populations and owls. Form two meanwhile (Figure 4-b) in prevalent in modern birds, and is more primitive than form three, which is itself a derviative of form two that has lost guanine (G). According to statistics, TψC arm of tRNA^Phe has a “C” insertion in the majority of published Passeriformes species (Figure 4-b), while the TψC arm of tRNAPhe in P. ruficollis formed a CC mismatch (Figure 4-e). The DHU arm is also missing in the tRNA^{Ser (AGN)} gene, which is common in many Vertebrate mitochondria including avians (Wolstenholme, 1992a, b). The missing length of the DHU arm of tRNA^{Ser (AGN)} gene gets shorter from invertebrates to higher vertebrates. An earlier study about mammaltRNA^{Ser (AGN)} showed that they can form potential L-shaped structure to maintain the distance between the anticodon arm and CCA acceptor arm (Hanada et al, 2000). The function of tRNA transferring specific amino acids and tertiary structure of L shape are maintained by the amino acid acceptor arm and anticodon arm, while the TΨC loop and DHU loop as well as the corresponding arms seems to not inhibit the function of tRNA (Dirheimer et al, 1995; Wolstenholme et al, 1994).

Figure 4.

Comparison of the mitochondrial tRNA^Phe TψC arm structure among five representative avian species

Secondary structure of rRNA

The single-stranded loop region and double-stranded stem region are alternately arranged in the rRNA gene, being subject to different selection pressures. In general, the loop region has very fast evolution, with high variability between species, while stem region is relatively conserved (Noller, 1984; Woese et al, 1980). However, there are some stem regions containing the same variable length as the loop region, and some loci with a higher variability, while some loop regions do have very conserved sequences (Simon et al, 1994; Vawter & Brown, 1993). Predicting the secondary structure of rRNA can help explain the function and evolutionary information of different sections observed in the P. ruficollis rRNA genes, which can in turn help answer some issues in phylogeny.

For srRNA, the secondary structure of Passeriformes mitochondrial genome is very similar on the whole. The secondary structure of srRNA is the same in P. Ruficollis and Remiz consobrinus (Gao et al, 2013), and zone one in domain I has a extra base and a cytosine insertion compared to Gallus gallus (Eberhard et al, 2001) and P. hendersoni (Ke et al, 2010) (Figure 5-a). A similar phenomenon is also appearea in Pharomachrus pavoninus (Espinosa De los Monteros A, 2003). There are 5 base pairs and 2 base pairs in zone 24 of Domain II of P. ruficollis and R. consobrinus respectively, while G. gallus and P. hendersoni have not formed into paired stems, and 3 base pairs were formed in Anas platyrhynchos and Coscoroba coscoroba, which has a prominent cytosine insertion between the first and the second base pairs of the left stem. Moreover, the loop region has more differences among species, indicating that this region has great variability in different species (Figure 5-b). Zone 39 in the stem region of P. ruficollis has a significantly reduced stem and relative increased loop regions compared to other species (Figure 5-c). Most Stem variations tend to arise from compensatory mutation, which can guarantee the conservation of secondary structure not interfered by primary sequence variation.

Figure 5.

Comparison of variable regions in rRNA secondary structure

For lrRNA, the predicted secondary structure of P. ruficollis is identical with R. consobrinus, and is basically the same with P. hendersoni and Bos taurus (Burk et al, 2002). There are some differences: four species have formed a 15 bp base pairing stem in zone 19 of domain II, and formed a hairpin structure stem-loop, while the length of this stem is 19 bp and 11 bp in P. hendersoni and B. taurus respectively, but the loop region in B. taurusis much larger than the other species (Figure 5-d). P. ruficollis, R. consobrinus and P. hendersoni formed a 2−4 bp matching in zone 24 of domain II, while B. taurus has matching bases in the same position (Figure 5-e).

Control region

Like vast majority of birds, one control region presented in P. ruficollis’ mitochondrial genome is located between tRNA^Glu and tRNA^Phe with a length of 1305 bp. The control region contains regulatory mixeds of mitochondrial DNA transcription and replication. It suffered high variability according to under less evolutionary pressures, causing it as the fastest evolution region in mitochondrial DNA. However, control region also includes some conservative fragment. According to the structure information of control region in other species, we make a analysis and prediction to control region in P. ruficollis and determines the domains of F, E, D, C, B, CSB and Bird similarity (Figure 1).

The control region in P. ruficollis mitochondrial genome has three domains, that is domain I, domain II and domain III, which is similar to other vertebrate. Domain II is more conservative, while domain I and domain III have heterogeneity in base composition and length (Baker & Marshall, 1997).

The domain I in the control region contains an interrupted poly-C structure at 25−42 sites. In Galliformes and Anseriformes, this region and a short sequences below can form hairpin structures Ruokonen & Kvist (2002), while this phenomenon is not found in P.ruficollis, and the the causes is not clear. Refers to the termination associated sequence (TAS) of control region in L.brunnicephalus (Yang et al, 2012) and Chrysolophus pictus (Huang & Liao, 2011), we confirmed the TAS in P.ruficollis.

We described the conservative box B, C, D, E and F in domain II of P. ruficollis based on P.hendersoni and L.brunnicephalus, and a more conservative Bird similarity-box was founded before B-box. The specific function of these conserved boxes is not entirely clear, it may related to the formation of D loop and the regulation of H chain synthesis (Figure 1).

For domain III of control region, CSB1-3 boxes in vertebrates domain III are closely related to the replication initiation of mitochondrial genome (Yokobori & Pääbo, 1995). CSB1 box in P. ruficollis is located 892−920 bp of domain III and is extremely conservative compared with other birds. According to CSB2-3 boxes description of C. pictus and A. chukar, we determined the position of CSB2/3 boxes at 952 − 991 bp, but there is no obvious conservative box features founded in the sequence analysis.

Phylogeny

One of the major reasons for pursuing this study of the mitochondrial genome was to help answer some long-standing questions regarding how several of these avian species fit together. According to the phylogenetic treewe constructed in this study, the position of categories under Passeriformes are basically the same as those found in the other currently available studies. In general though, the mitochondrial genome data presented in this study supports the division of Passeriformes by Sibley & Ahlqulst (1990). In the aspect of Suboscines, the position of Acanthisitta chloris is different in different phylogenetic tree. One is that A. chloris is clustered with Cnemotriccus fuscatus and Smithornis sharpie, and another is located in the base of Passeriformes, which is the sister group to Oscines and Suboscines. The former is consistent with the conclusion of traditional morphological classification (Mayr & Greenway, 1956) and DNA hybridization studies (Sibley & Ahlqulst, 1990), while the latter is supported by the study based on nuclear genes and mitochondrial genes (Pacheco et al, 2011; Ericson et al, 2002). From this perspective, the specific location of A. chloris needs to be verified using more molecular markers.

For Suboscines, there were many conclusions drawn from difference taxa being different from traditional conclusions reached by taxonomy and DNA hybridization. Menuridae was placed in Suboscines according to traditional taxonomy (Mayr & Greenway, 1956), while it was served as a superfamily under Corvida of Oscines based on DNA hybridization (Sibley & Ahlqulst, 1990). Our results, meanwhile, are consistent with the majority study conclusions based on DNA sequences (Barker et al, 2002; Ericson et al, 2002), supporting the contention that Menuridae is the sister group with Corvida and Passerida. Unfortunately this is largely a preliminary investigation the reliability of this phenomenon has yet to be tested, and the mainly reason is inadequate taxon sampling. Perhaps further studies can shed some light on this issue.

For Passerida, the partition of its three general superfamilies are basically consistent with the results of our study. That said, for P. humilis, which was moved from Corvidae to Paridae over the last debate, (James et al, 2003), the taxonomic status is still in doubt and its specific location in phylogenetic tree needs more research. In Passeroidea, Fringillidae first clustered into a branch, and then get together with P. ruficollis to form a large branch, which constituted a sister group. Finally, they gather together with the clade containing Taeniopygia guttata and Vidua chalybeate. According to this, we would speculate that T. guttata and V. chalybeate were first separated from Fringillidae, and P. ruficollis is the evolutional intermediate form. Though an interesting possibility, due to a dearth of data on Passeridae, we can make no definitive conclusions on the evolutionary position of P. ruficollis. We do however hope that this data can be paired with other datasets that will allow researchers to provide a more satisfactory answer.

Funding Statement

The National Science Foundation for Outstanding Youth (30925008); Open topic of the Chinese Academy of Sciences Key Laboratory of the Animal Evolution and Systematic (O529YX5105)