Open Access This article is
- freely available
Genes 2019, 10(1), 54; https://doi.org/10.3390/genes10010054
Genomes of Three Closely Related Caribbean Amazons Provide Insight for Species History and Conservation
Department of Biology, University of Puerto Rico at Mayaguez, Mayaguez, PR 00680, USA
Department of Biology, University of Konstanz, 78464 Konstanz, Germany
Theodosius Dobzhansky Center for Genome Bioinformatics, St. Petersburg State University, 199034 St. Petersburg, Russia
Department of Genetic Medicine, Weill Cornell Medical College, New York, NY 10021, USA
Department of Biological Sciences, Oakland University, 118 Library Drive, Rochester, MI 48309, USA
Department of Biological Sciences, Uzhhorod National University, 88000 Uzhhorod, Ukraine
Beaumont BioBank, William Beaumont Hospital, Royal Oak, MI 48073, USA
Conservation Program of the Puerto Rican Parrot, U.S. Fish and Wildlife Service, Rio Grande, PR 00745, USA
The Recovery Program of the Puerto Rican Parrot at the Rio Abajo State Forest, Departamento de Recursos Naturales y Ambientales de Puerto Rico, Arecibo, PR 00613, USA
Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California Davis, Davis, CA 95616, USA
Program in Individualized Medicine (PrIMe), Pharmacogenomics Laboratory, Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Washington State University, 100 Grimes Way, Pullman, WA 99164, USA
Authors to whom correspondence should be addressed.
Received: 15 October 2018 / Accepted: 8 January 2019 / Published: 16 January 2019
Islands have been used as model systems for studies of speciation and extinction since Darwin published his observations about finches found on the Galapagos. Amazon parrots inhabiting the Greater Antillean Islands represent a fascinating model of species diversification. Unfortunately, many of these birds are threatened as a result of human activity and some, like the Puerto Rican parrot, are now critically endangered. In this study we used a combination of de novo and reference-assisted assembly methods, integrating it with information obtained from related genomes to perform genome reconstruction of three amazon species. First, we used whole genome sequencing data to generate a new de novo genome assembly for the Puerto Rican parrot (Amazona vittata). We then improved the obtained assembly using transcriptome data from Amazona ventralis and used the resulting sequences as a reference to assemble the genomes Hispaniolan (A. ventralis) and Cuban (Amazona leucocephala) parrots. Finally, we, annotated genes and repetitive elements, estimated genome sizes and current levels of heterozygosity, built models of demographic history and provided interpretation of our findings in the context of parrot evolution in the Caribbean.
Keywords:genomics; parrots; birds; heterozygosity; Puerto Rican parrot; Cuba; Hispaniola; conservation; demography
The Bird 10,000 Genomes (B10K) Project resulted in a large number of genomic sequences that are being quickly assembled and incorporated into studies on evolution, ecology, population genetics, neurobiology, development and conservation [1,2,3]. Genome-wide sequencing and assembly has expanded to the point that it allows for completion of the genome-based phylogeny of all birds. Attention has recently started to shift away from representation of the overall bird phylogeny to instead filling in gaps and resolving specific lineages. A narrowed focus on speciation and adaptation processes on the species level can allow for decoding of the links between genotypes and phenotypes; determining genetic, evolutionary, biogeographical and biodiversity relationships across species; and evaluation of how various ecological factors affect avian evolution . Finally, by focusing on groups that include endangered species, genome studies provide the means to elucidate the conservation issues that would help in our efforts to preserve biodiversity. Neotropical parrots represent a fascinating group that includes many species with endangered conservation status that have not yet been represented in whole-genome phylogenetic analyses [2,3].
Islands became an important source of evolutionary ideas [5,6,7]. Since Darwin’s Voyage on the Beagle, they provided valuable model systems for fundamental studies of migration, diversification and extinction [5,8]. Amazon parrots (Amazona sp.) that inhabit the Greater Antillean Islands are a fascinating example of speciation on islands, in many ways similar to that of Darwin’s finches in the Galapagos . Several attempts have been made in the past to shed light on evolution and speciation of these birds based on morphological [10,11] and molecular data  but the picture still falls shy of full resolution. Considering the significance of parrots to these islands’ history and ecology [13,14,15,16], we believe it is important to understand how these species came to be and how they adapted to specific island environments. In this article, we focus on the clade of amazon parrots that originated in Central America and spread across the Caribbean islands of Cuba, Hispaniola and Puerto Rico [10,11].
Parrots have long been thought to have first originated in and diversified from, Gondwana, based on current distribution across the southern continents that formerly composed this giant ancient supercontinent . Initial biogeographic analyses, based on multi-loci phylogenies, extensive taxa sampling and different analytical approaches, support a hypothesis of origin and initial diversification in Gondwana during the Cretaceous . Consequently, separation of Arinae (New World parrots that include amazons, macaws, conures and parakeets) from other groups of parrots was associated with drift of major Gondwana plates around 35 Mya . Accordingly, Amazona parrots were thought to have split from all the other neotropical parrots around 23 Mya. Other, more recent and robustly supported independent phylogenomic analyses [3,19,20], as well as fossil evidence [21,22], support post-Gondwana divergence of stem Psittaciformes from Psittacopasseres between 55–60 Mya. These studies estimate the earliest divergence of crown group Psittaciformes (Nestor-Psittacidae) to have occurred between 42–32 Mya and the divergence of New World parrots from others as recently as 14 Mya [3,19,20,21,22,23].
The parrots on the Greater Antilles appear closely related to the small A. albifrons of Central America (Figure 1) and there may have been two separate dispersal events to these islands, one directly to Jamaica and one to Cuba, followed by the stepping stone dispersal from island to island, as far as to Puerto Rico [10,11,24]. Unfortunately, the sequence and timing of many of these significant evolutionary events are inferred from limited molecular or geological data. With the arrival of genome sequencing and new fossil data, the history of speciation is being refined using the additional evidence. Evolutionary history of the Amazona clade was previously assessed using a small set of genetic markers (cytochrome B gene, COI gene etc.) producing the first molecular phylogeny [11,23]. This analysis did not provide speciation times and left other unresolved issues due to the insufficient amount of information: at least two contradicting colonization scenarios for the speciation order among the four major Caribbean islands (Cuba, Hispaniola, Jamaica and Puerto Rico) have been proposed [10,11]. Even if the full mitogenomic sequences were used, an analysis based solely on mtDNA may not have been sufficient due to the incomplete lineage sorting and subsequent gene flow between the islands, which can interfere with proper interpretation of phylogenetic trees . Nuclear genomes or at least multiple nuclear genes, along with mitochondrial data, should be incorporated into the analysis to rule out misinterpretations and to reconstruct events leading to parrot speciation on islands.
So far, the only publicly available genome from the Caribbean amazon clade (Figure 1) has been from the Puerto Rican parrot (Amazona vittata): a short-read assembly with only 76% coverage that probably included a number of mis-assemblies . The critically endangered A. vittata is the only surviving indigenous parrot species anywhere in the U.S. . Once abundant throughout the island of Puerto Rico, its drastic population decline followed the decimation of the old-growth forest . Despite early DNA fingerprinting efforts [29,30], the genetic consequences of the severe population bottleneck, as well as the population expansion associated with the recent recovery, have not been fully evaluated and a more comprehensive analysis of the genome on the population and species levels is needed. Further detailed research on A. vittata conservation genomics is necessary to provide data and better tools to study inbreeding depression, mutation and adaptation to captivity [31,32].
In the current study, we used additional genome wide and transcriptome data to improve that assembly, as well as to assemble and annotate the genomic sequences of two additional amazon species from the Caribbean: the Cuban amazon (A. leucocephala) and the Hispaniolan amazon (A. ventralis). Using both genome and transcriptome data we have generated the improved de novo assembly of the A. vittata genome, performed reference-assisted assembly for two other closely related Caribbean amazon species, annotated protein-coding genes and repeats, analyzed demographic history and genomic levels of heterozygosity and discussed our results in the context of conservation biology of these species.
2. Materials and Methods
Blood samples for DNA sequencing from female Puerto Rican (A. vittata) and Hispaniolan parrots (A. ventralis) were obtained during routine veterinary procedures from birds housed at the US Fish and Wildlife Service “Iguaca” Aviary, a captive-breeding facility for the Puerto Rican parrot near El Yunque National Rainforest in Puerto Rico. All procedures were approved by the University of Puerto Rico at Mayagüez Institutional Animal Care and Use Committee (IACUC#201109.1) and were in accordance with the guidance for the Endangered Species Act. The Cuban parrot (A. leucocephala) DNA sample was extracted from the living cell cultures of the Frozen Zoo® collection, at the Institute for Conservation Research at the San Diego Zoo.
Samples for the RNA sequencing were obtained from five different A. ventralis individuals (4 females, 1 male) with an average age of 17 years (±7.4 years) and weight ranging from 234–410 g (median of 262 g) at the School of Veterinary Medicine, University of California, Davis. One liver (sample 336) and four blood samples (sample 335, 140, 341 and 13) were obtained. All birds were part of a research flock and were housed individually in wire cages (61 × 58 × 66 cm) in a room maintained at 23 °C (73.4 °F) with a photoperiod of 12 h. They were fed a pelleted diet (ZuPreem FruitBlend, Premium Nutritional Products, Shawnee, KS, USA) ad libitum and had constant access to water. This study was approved by the University of California, Davis, Institutional Animal Care and Use Committee.
2.2. DNA and RNA Extraction
DNA was extracted from whole blood using the Qiagen QIAmp Mini Kit following manufacturer’s protocol. Total RNA was extracted from blood cells and liver tissue using a column method (RNeasy Kit; Qiagen, Hilden, Germany). RNA quality concentration was determined by a fluorometric technique (Qubit, Thermo Fisher Scientific, Waltham, Massachusetts, U.S.A.) and quality was verified by a small fragment analyzer (Bioanalyzer 2100, Agilent, Santa Clara, California, U.S.A.). The globin mRNA content of blood RNA samples (only) was first reduced using a commercial kit (Globin-Zero, Illumina, San Diego, California, U.S.A).
2.3. Genome and Transcriptome Sequencing
Genomes of all three species included in this study (A. vittata, A. ventralis and A. leucocephala) were sequenced using the Illumina HiSeq2000 platform. For A. vittata one PE (paired-ends) library with 300 bp target IS (insert size) and 2 MP (mate-pairs) libraries with 3 kbp and 8 kbp target IS, respectively, were generated using TruSeq DNA PCR-Free Library Prep Kits. For A. ventralis and A. leucocephala only one PE library with 300 bp target IS was sequenced (Table S1).
RNA sequencing was performed by the Genomics Core facility at Washington State University (Spokane, WA, USA) using one liver and four blood samples from five different A. ventralis individuals. Then, RNA libraries were generated using Illumina TruSeq RNA Library Prep Kit v2 for each sample using 100 ng of input total RNA. Obtained libraries were sequenced using Illumina’s HiSeq2500 machine.
2.4. Data QC and Filtering
The initial QC of NGS data was performed using FastQC . Both genomic and transcriptomic reads were filtered in a two-stage process. First, long fragments of Illumina adapters were trimmed using Cookiecutter . Then Trimmomatic v0.36  was used to remove short adapter fragments and perform filtering by quality (Trimmomatic options: ILLUMINACLIP: TRIMMOMATIC_ADAPTERS:2:30:10:1 SLIDINGWINDOW:20:20 MINLEN:50). The resulting output is shown in the Table S1.
2.5. Genome and Transcriptome Assembly
The new A. vittata genome was assembled from 1 PE library and 2 MP libraries (3 kbp and 8 kbp, see Table S1) followed by scaffolding using A. ventralis de novo transcriptome assembly (see Supplementary Figure S1 for the assembly pipeline) and post-assembly filtration. De novo transcriptome assembly for A. ventralis was performed using Trinity v2.8.2  from filtered reads for each RNAseq library independently and for merged library including reads from all five libraries generated.
Initial genome contigs were generated from the PE library using Fermi v 1.1 assembler . Then reads from all libraries were aligned to the initial contigs using BWA  to estimate actual insert sizes. Only alignments to contigs whose length was equal to or greater than, 3× target IS were used for estimation in order to avoid bias introduced by alignment artifacts. For actual IS see Table S1. At the following step, initial contigs were scaffolded by SSPACE  using all read libraries, followed by gap closing with GapCloser  using only the PE library. Next, all scaffolds with length of less than 100 bp (i.e., less than read length from PE library) were removed as assembly artifacts. Finally, transcripts from A. ventralis transcriptome assembly were aligned to the scaffolds by BLAT  and the obtained alignment was used for scaffolding with L_RNA_scaffolder . Finally, all scaffolds of length lower than 1000 bp were discarded. Genome size and actual coverage of PE libraries were estimated using the Jellyfish 2  and KrATER  for each species. Assembly integrity was verified using BUSCO v3 and aves_odb9 gene set  (Table S2).
2.6. Repeat Masking in the A. vittata Genome
Repeat identification in the A. vitatta genome was performed de novo from the PE library and the repeat library generated. It was then combined with aves repeats from the RepBase  and the combined library was used to annotate repeats with RepeatMasker [47,48,49]. Finally, repeats in the A. vittata genome were soft-masked using BEDtools  for prediction of protein-coding genes.
2.7. Annotation of Protein-Coding Genes in the A. vittata Genome
The annotation of protein-coding genes was performed using a combined approach that unifies homology-based, transcriptome-based and de novo predictions. However, de novo predictions were used only to fill gaps and to extend homology- and transcriptome-based predictions. Proteins of three reference species: Gallus gallus (Gallus_gallus-5.0 (GCA_000002315.3)), Melopsittacus undulatus (melUnd1) and Taeniopygia guttata (taeGut3.2.4) were aligned to the A. vittata assembly by Exonerate  using the protein2genome model with a maximum of five hits per protein. The obtained alignments were divided into the top (primary) and secondary hits; the coding sequence (CDS) fragments were cut from each side by 3 bp for the top hits and by 9 bp for the secondary hits. Then, A. ventralis RNAseq reads from all libraries were aligned to A. vittata genome by STAR  and the obtained splice junctions alongside with CDS segments from protein alignments were clustered and supplied as hints to the AUGUSTUS software package . The CDS segments of genes were predicted in a soft-masked A. vitatta assembly using chicken gene models. Proteins were extracted from the predicted genes and aligned by HMMER v3.1  and BLAST  to the Pfam  and Swiss-Prot  databases, respectively. Only genes supported by both hints and hits to one of the protein databases were retained; the rest were discarded.
2.8. Genome Read Alignment and Variant Calling
Filtered reads of A. vittata, A. ventralis and A. leucocephala were aligned to the assembled A. vittata genome using BWA mem with default options, followed by duplicate marking using Picard  MarkDuplicates. Next, a mask track was created for each genome using deduplicated alignments and based on coverage. Only regions with coverage of 50–250% (10–50× for A. ventralis, 8–40× for A. leucocephala, 6–34× for A. vittata) of mean coverage were retained unmasked. Then, HaplotypeCaller from GATK pipeline  was used to call variants. Only the SNPs (Single Nucleotide Polymorphisms) and indels passing hard filters from GATK Best Practice were retained (QD > 2.0, FS < 20.0, MQ > 40.0, MQRankSum > −12.5, ReadPosRankSum > −8.0 for SNPs and QD > 2.0, FS < 20.0, ReadPosRankSum > −20.0 for indels, respectively).
2.9. Reference-Assisted Assembly of A.ventralis and A. leucocephala Genomes
In all three species, filtered reads from PE libraries of all three amazons were aligned by the BWA to the previously assembled A. vittata reference genome, followed by variant calling using GATK Haplotype caller with extensive filtration in accordance with the GATK best practices. In the case of a heterozygous position being encountered during reference assisted-assembly, the algorithm had to choose between two possible nucleotides. Therefore, several options would be available. First, if both alleles were different from the reference, the algorithm could choose one randomly. Second, if one allele is identical to the one in the reference genome, it would be reasonable to choose the reference allele in the new genome as well. However, the tool for generation of genome sequence in the GATK pipeline  chooses the alternative allele by default. Therefore, we had to remove SNPs with reference alleles from the vcf file prior to the reference assisted assembly.
2.10. Phylogeny Reconstruction and Divergence Time Estimation
Ortholog identification for the longest proteins corresponding to each predicted gene of A.vittata, A. ventralis, A. leucocephala and other species from songbird, parrot and falcon groups was performed using Emapper V 1.0.1  and veNOG subset (dataset for vertebrate orthologs) from the eggNOG database of orthologous groups . Other species included: Serinus canaria , Ficedula albicollis , Parus major , Zonotrichia albicollis , Manacus vitellinus , Cyanistes caeruleus , Melopsittacus undulatus , Geospiza fortis , Taeniopygia guttata , Aquila chrysaetos  and Falco peregrinus .
Single-copy orthologs were extracted from the obtained groups and corresponding CDSs (coding sequences) were aligned by codon using PRANK , followed by removal of hypervariable regions with Gblocks [71,72]. Obtained alignments were concatenated, being treated as a single partition and used to reconstruct a maximum likelihood tree with RAxML v8.2  under the GTRGAMMA with 1000 bootstrap replications. The reconstructed tree was rooted with Falconiformes species (Falco cherrug, Falco peregrinus) as an outgroup. The resulting tree was drawn using FigTree software .
2.11. Demographic History Inference
Based on the variation data from the genomes, we estimated population dynamics using the pairwise sequentially Markovian coalescent (PSMC) model . The PSMC approach uses the coalescent model to estimate changes in population size, which allowed us to create a TMRCA (Time to the Most Recent Common Ancestor) distribution across the genome and estimate the effective population size (Ne) in recent evolutionary history (e.g., from 10,000 to 1 million years). Demographic history was inferred separately for each species using a generation time of six years calculated by the captive breeding program for Puerto Rican parrot  and mutation rates recently estimated from bird pedigrees available in the literature .
2.12. Amazona Genome Browser Hub
To provide convenient access to our data, we organized the UCSC genome browser hub  containing annotated genomic features of A. vitatta, A. ventralis and A. leucocephala genomes. The features available on the hub include protein-coding genes and RepeatMasker-detected repeats of A. vitatta, as well as genomic variants (SNVs and indels) of the three related to A. vitatta species as a reference. Also shown are the BigWig tracks  that visualize coverage of the reference genome by aligned reads of the three genomes. The track hub file is publicly available online at: http://public.dobzhanskycenter.ru/AmazonaHub/hub.txt. To view the Amazona hub in the UCSC Genome Browser, the user must add the track hub file to the Track Hubs web page: http://genome.ucsc.edu/cgi-bin/hgHubConnect.
3.1. Assembly and Annotation
In this study we used a combination of short read paired-end (PE) read sequences, mate pairs (MP) and transcriptomes to assemble genomes of three closely related amazon species from the Caribbean: Amazona vittata, A. leucocephala and A. ventralis. Among these, A. vittata was chosen as reference species for genome assembly. Recent demographic history inadvertently shaped the genome of A. vittata into an ideal candidate for a de novo sequencing project: its relatively small (1.58 Gb, less than half of the human) genome  was expected to be highly invariable due to the recent population bottleneck [26,27]. In addition, this was the only assembly with the long reads: one PE and two MP libraries available for this species. For the two other parrots, only one PE library per species was generated (Table S1).
Genome size estimations, that were based on distributions of 23-mers extracted from the PE libraries, demonstrated similarity between the species with less than 10% difference (Table 1; Figure S2). Moreover, these values were in relative concordance with the publicly available haploid DNA content estimates evaluated using flow cytometry method (C-values) from the Animal Genome Size Database . The observed discrepancy of about 10% (Table 1) is common among all three estimates and may be attributable to the genome regions that have not been covered by sequencing.
Unfortunately, the transcriptomes of the highly endangered status of A. vittata were difficult to obtain due to the restrictions at breeding facilities at the Conservation Program of the Puerto Rican Parrot, U.S. Fish and Wildlife Service and the Recovery Program of the Puerto Rican Parrot at the Rıo Abajo State Forest, Departamento de Recursos Naturales y Ambientales de Puerto Rico. Transcriptome sequencing is crucially important for annotation of protein-coding genes, however the de novo assembled transcripts can also be used for the additional scaffolding stage of the genome assembly. In this study, we sequenced transcriptomes for five tissue samples from the Hispaniolan parrot (A. ventralis; 4 blood samples and 1 liver sample) using Illumina HiSeq platform. Transcripts were assembled from both independent libraries and merged into one. The characteristics of the assembled transcriptome are presented in Table 2.
The assembly of A. vittata genome followed the multistage pipeline described in Section 2.4. of Materials and Methods (Figure S1). To achieve the best results, PE and MP from A. vittata were complemented by the transcriptome reads from A. ventralis during the transcriptome based scaffolding stage (Figure S1). This approach worked well, probably due to the close evolutionary proximity between the two species and resulted in the total assembly length of 1.45 Gbp (Table 3), which is slightly longer than the 23-mer based estimation, that used PE libraries of the single species (1.42 Gbp; Figure S2). This small increase in the size of genome assembly is likely to be due to the imprecise gap length estimates during scaffolding, in particular at the transcriptome scaffolding stage.
Quality assessment of our assembly was performed using BUSCOv3 with the aves_odb9 gene set . Out of the 4915 Benchmarking Universal Single-Copy Orthologs (a conservative gene set or BUSCOs), 87.4% were found as a single-copies, 6.4% as fragmented and 6.2% were not found. Therefore, despite the relatively low PE library coverage common to all three parrot datasets (Table 1), the contig N50 of 101 kbp and the relatively high BUSCO score point to the high integrity of this assembly, sufficient for protein-coding gene prediction. The total number of scaffolds in our assembly, even after filtration of very short scaffolds (shorter than 1000 bp), was relatively high, at more than 62 k (Table 3). However, this is a known issue for the short-read based assemblies with few mate-pair libraries sequenced and it was observed in the earlier assembly as well .
Repeat masking is strictly necessary for prediction of protein-coding genes, as interspersed repeats often include mobile elements with ORFs. Unfortunately, the most commonly used database of repetitive elements RepBase still includes a very small number of avian repeats (less than 500). To address this problem, we assembled repeats de novo from the PE library and combined the results with the RepBase aves library [46,47,48,49]. Subsequently, 7.57% of the genomic sequences in this study were identified as repeats (Table 4). The most common repeat class appears to be L3/CR1 LTRs, which comprises almost 1/3 of all repeats (2.47% of the genome, Table 4). At the same time, more than half of all repeats still fall into the unclassified category (3.92%, Table 4).
Prediction of protein-coding genes was performed in accordance with the hybrid pipeline (described in Materials and Methods 2.7) using homology (Pfam  and Swiss-Prot [57,58] databases), transcriptome (A. ventralis RNAseq reads) and de novo predictions. As a reference for homology-based transfer, we have chosen protein sequences of three species: Gallus gallus (chicken), Taeniopygia guttata (zebra finch) and Melopsittacus undulatus (budgerigar). Chicken and zebra finch were selected as avian species, with the best available chromosome-level assembly and annotation, based on extensive usage of RNAseq. Meanwhile, the budgerigar genome was included because it is the highest quality parrot genome available, which makes it the best currently available closely related species with sufficient assembly and annotation data.
As a result, a total of 19,669 genes have been predicted for the current version of the A. vittata genome. This number is somewhat higher than the comparable numbers in genomes of other birds (Table 5). However, the number of genes with the longest protein containing less than 100 amino-acids is 4–9× higher in our annotation than in annotations of other bird genomes (Table 5). This probably reflects the higher fragmentation level of the assembly. All of the assembly data from this study is available in the form of the browser hub. The track hub file is publicly available by the following link: http://public.dobzhanskycenter.ru/AmazonaHub/hub.txt.
3.2. Genome-Wide Heterozygosity
Given the fragmented nature (N50 101.028 kbp, Table 3) of our assembly we were unable to use the most common window-based approach for heterozygosity-level assessment. Instead we calculated a simple metric of whole genome heterozygosity: counting and dividing the number of heterozygous SNP by the unmasked genome length. As expected, the A. vittata genome had the lowest level of heterozygosity, with a mean density of 0.96 heterozygous SNPs/kbp. A. ventralis and A. lecucocephala showed higher heterozygosity 1.6–1.7 SNPs/kbp (Table S6).
3.3. Phylogenetic and Demographic Analysis
In order to resolve evolutionary relationships between the species in this study, we performed in the wider context of bird phylogeny. In addition to the genome data for the three amazon parrots from this study, we chose 11 additional avian genomes. We included two species of falcons (Falconiformes), eight species of passerine birds (Passeriformes) and one additional parrot (Psittaciformes) with assembly and annotation based on extensive usage of RNAseq (see Materials and Methods). The tree was calculated using filtered alignment of 4135 single-copy orthologs in RaxML v.8 with 1000 bootstrap replicates. The two falcon species (Falco cherrug and F. peregrinus) were placed in an outgroup. All nodes in the resulting tree had the highest 100% bootstrap support suggesting high stability of the tree to the noise in input data. According to our reconstruction A. vittata and A. ventralis form a monophyletic group with A. leucocephala as a sister species (Figure 2).
We also attempted to estimate population dynamics using the pairwise sequentially Markovian coalescent (PSMC) model. The resulting estimate is based on the assumption of a generation time of six years calculated by the captive breeding program for the Puerto Rican parrot  and mutation rates recently estimated from bird pedigrees available in the literature  and provides the assessment of these species’ effective population sizes (Ne) in recent evolutionary history (Figure 3). This analysis indicates that the first split between the three species of amazons occurred at least 2 MYA. This date may be suggesting the time of the initial dispersal of the ancestral population of parrots from Central America (Figure 3).
Publicly available genome assemblies and gene annotations for the three Caribbean parrots are the major result of this study. In a combination approach that used both genome and transcriptome sequences, we were able to obtain enough coverage to allow for the identification of the major types of repeats, as well as locations of the protein coding genes (Table 4 and Table 5), with high confidence: 90% of BUSCO genes were found as complete, single copy or duplicated and only 6.4% as fragmented; Table S2. Still, the number of genes (19,669) is somewhat higher than the respective numbers estimated for other birds (Table 5). While BUSCO assessments elevate the confidence that this increase can be correlated with the increase in genome size, we cannot rule out the possibility that the higher fragmentation level of our assemblies has contributed as well. Specifically, the number of genes with longest protein shorter than 100 amino-acids is 4–9 times higher in our annotation than in annotations of other bird genomes (Table 5). Nevertheless, these assemblies were sufficient to produce the first estimates of genome wide heterozygosity (Table S4) and allowed inferences of the phylogeny based on the genome wide data (Figure 2), as well as estimates of demographic histories, in these three island species for the first time (Figure 3).
This study reinforces the observation that the Amazona parrots have slightly larger genomes than other parrots on average. The Animal Genome Size Database  features haploid DNA content (C-values) for 56 genomes of Psittaciformes, of which 16 belong to the genus Amazona. The average genome size for the Amazon parrots is 1.58 pg (±0.09), while the rest of the parrots have significantly (p < 0.0001) smaller genomes of 1.35 pg (±0.11). Our estimates were based on distributions of 23-mers extracted from PE libraries (Figure S2, Table 1). Therefore, current genome size estimations for the three Caribbean parrots in this study provide an independent confirmation and demonstrated remarkable similarity with the publicly available C-values. The 10% discrepancy in genome size estimates between the two methods (Table 1) could be attributed to the genomic regions which have not been covered by sequencing.
Only 7.57% of the parrot genomic sequences in this study were identified as repeats (Table 4). Such a small fraction is not unusual for birds. In fact, birds have the least number of repeated elements in their genomes compared to any other group of tetrapods, comprising only 4–10% of the total genome size (compared to the 34–52% in mammals) [4,80] and an up to two-fold genome contraction had occurred before the divergence of birds from a theropod ancestor [82,83,84]. However, since repeats are more difficult to assemble and a higher proportion of them would be omitted in comparison with the rest of the sequence, the reduced number of repeats we found could also be explained by the high percentage of gaps (~24%) in the present genome assembly.
There is a possibility that the large genome sizes in Amazons could be attributed to expansion of one of the transposable element classes. Unfortunately, because of the high gap proportion in the current genomes, it is difficult to determine which of these elements could be the culprit with high confidence. The most common of repeats we have identified are the Chicken repeat 1 (CR1) elements that make up 2.47% of the Amazona genome (Table 4). CR1 elements belong to the non-long repeat class of retrotransposons and are subdivided into at least six distinct subfamilies, comprising sequences of about 300 bp long, all of which share substantial sequence similarity. CR1-like elements were found in various genomes from invertebrates to mammals, suggesting their importance for genome structure and/or function . However, it is too early to arrive at any conclusion, since the majority of the repetitive content is either not represented in the assembly (because repeats are more difficult to assemble and place) or has been labelled as “unclassified” (3.92% of the genome; Table 4).
There were two main reasons why we could not calculate the timing of speciation using the divergence time analysis given our data. First, we noted a significant difference in cumulative branch length (especially for Passeriformes) and suspected existence of substantial differences in the mutation rates that would bias our estimates. Second, the current paleontological record is lacking fossil calibrations inside Psittaciformes. The only calibration point within the parrot clade for a split between Melopsittacus undulatus and Amazons (22.5 MYA) is in fact a derived second level molecular calibration point. Fossil-based calibration points within the passerine clade (split between Manacus/Taeniopygia/Geospiza at 13.6–16.3 MYA and Taeniopygia/Geospiza at 7.2–11.6 ) are not useful for timing parrot phylogenies because of the mentioned issue involving mutation rate. The split between Psittaciformes/Passeriformes (53.5–65 MYA)  is too old to be helpful for dating of the relatively recent split between Amazona species. Hopefully, paleontological discoveries of new fossils of ancient parrots combined with the rapid advances in genome sequencing and analysis will soon bridge this gap. All mentioned fossil calibrations are listed in a supplementary table (Table S3).
In this study we present an early attempt to estimate genome wide heterozygosity given our data. This early estimate can be further evaluated and discussed in conjunction with the effects for life-history, morphological and physiological traits [87,88,89]. Species that are endangered and/or threatened taxa generally display lower heterozygosity than related unthreatened taxa . However, we can already use these preliminary values to address certain theoretical questions in conservation genetics. For instance, we used the heterozygote estimates in context with the hypothesis that genetic diversity should be positively correlated among islands . This hypothesis is based on two of the most prominent theories of island diversity, MacArthur and Wilson’s  theory of island biogeography and Sewall Wright’s  island model of population genetics . While we have observed a connection between island size and heterozygosity (Figure S3; r2 > 0.99), more island species are needed for a better evaluation of this hypothesis (Figure S3). The heteroygosity estimates for SNPs based on the genome sequences indicate that the Puerto Rican parrot has the lowest heterozygosity (0.96 SNPs/kbp), with almost half of that in the other two species and is similar to the number reported for another endangered parrot: kea (Nestor notabilis; 0.91 SNPs/kbp). At the same time, two earlier investigated critically endangered/vulnerable avian species, the white-tailed eagle (0.4 SNPs/kbp) and the dalmatian pelican (0.6 SNPs/kbp), have lower heterozygosity values  (Table S4).
In conclusion, new data generated on three Caribbean amazons has contributed to the body of knowledge on parrot genomics and conservation genetics and in combination with other genomes it will allow for future analyses that will provide valuable insights into the evolution of functional elements in the genomes of these parrot species.
All the data has been uploaded to NCBI: Bioproject accession PRJNA496322; Genome submission ID: SUB4629817. The following are available online at https://www.mdpi.com/2073-4425/10/1/54/s1, Figure S1. Pipeline used to assemble A. vitatta genome. One PE and two MP A. vitatta genome libraries were used to generate preliminary scaffolds complemented by additional scaffolding step using five A. ventralis RNAseq libraries. Figure S2. Distributions of 23-mers from K-mer distributions PE libraries of three parrot species. Corresponding genome size estimation present on figure and also in Table 1. Figure S3. Connection between heterozygosity and areal size for three Amazona species. Table S1. The sequencing outputs for each genome used in the current assemblies. Abbreviations: paired-end (PE) mate pairs (MP); Table S2. BUSCO scores for all steps of assembly of A. vittata genome. Assembly evaluation was performed using BUSCO v3 and Avian dataset (Simão et al., 2015). Table S3. Available fossil-based calibrations for speciation time dating within Psittaciformes and Passeriformes clades. Table S4. Mean heterozygosity in the genomes of three Amazona species compared with the values reported earlier for other avian species.
Conceptualization, S.K. (Sofiia Kolchanova), S.K. (Sergei Kliver), M.C., J.L.R.-F., J.C.M.-C. and T.K.O.; Data curation, S.K. (Sofiia Kolchanova), S.K. (Sergei Kliver), G.T., K.G. and W.W.; Formal analysis, S.K. (Sergei Kliver), A.K., P.D. and K.G.; Funding acquisition, A.J.M. and T.K.O.; Investigation, S.K. (Sofiia Kolchanova), S.K. (Sergei Kliver), A.J.M., J.R.P.-M., D.G., R.V.d.l.R., J.V. and T.K.O.; Methodology, S.K. (Sergei Kliver), M.C., J.C.M.-C. and T.K.O.; Project administration, T.K.O.; Resources, A.J.M., J.R.P.-M., D.G., M.C., R.V.d.l.R., J.V. and T.K.O.; Software, S.K. (Sergei Kliver), G.T., A.K., P.D. and W.W.; Supervision, J.L.R.-F. and T.K.O.; Visualization, S.K. (Sergei Kliver); Writing—original draft, S.K. (Sofiia Kolchanova), S.K. (Sergei Kliver), and T.K.O.; Writing—review & editing, S.K. (Sergei Kliver), J.L.R.-F., J.C.M.-C. and T.K.O.
This research was funded in part by US National Science Foundation, grant number NSF DUE 1044714; US FWS Grant (2012) and Toyota Community Awards (2012–2013) for Puerto Rican Parrot Project; Russian Foundation for Basic Research (RFBR) 17-00-00144; St. Petersburg State University grant 1.52.1647.2016.
We would like to thank personnel of the Conservation Program of the Puerto Rican Parrot, U.S. Fish and Wildlife Service and the Recovery Program of the Puerto Rican Parrot at the Rio Abajo State Forest, Departamento de Recursos Naturales y Ambientales de Puerto Rico for help and assistance in this project. We thank Oliver Ryder at the Institute for Conservation Research at the San Diego Zoo for providing valuable DNA samples from the Frozen Zoo® collection. A special thanks to Klaus-Peter Koepfli from the Smithsonian Conservation Biology Institute for helpful suggestions. We thank Kara Fore from the UPRM Research & Development Center for help with editing and submitting this manuscript. We thank the two reviewers for their criticism and useful comments that helped to significantly improve our paper.
Conflicts of Interest
The funders had no role in the design of the study; in the collection, analyses or interpretation of data; in the writing of the manuscript or in the decision to publish the results.
- Zhang, G.; Jarvis, E.D.; Gilbert, M.T.P. A flock of Genomes. Science 2014, 346, 1308–1309. [Google Scholar] [CrossRef]
- Zhang, G. Genomics: Bird sequencing project takes off. Nature 2015, 522, 34. [Google Scholar] [CrossRef] [PubMed]
- Jarvis, E.D.; Mirarab, S.; Aberer, A.J.; Li, B.; Houde, P.; Li, C.; Ho, S.Y.W.; Faircloth, B.C.; Nabholz, B.; Howard, J.T.; et al. Whole-genome analyses resolve early branches in the tree of life of modern birds. Science 2014, 346, 1320–1331. [Google Scholar] [CrossRef][Green Version]
- Zhang, G.; Li, C.; Li, Q.; Li, B.; Larkin, D.M.; Lee, C.; Storz, J.F.; Antunes, A.; Greenwold, M.J.; Meredith, R.W.; et al. Comparative genomics reveals insights into avian genome evolution and adaptation. Science 2014, 346, 1311–1320. [Google Scholar] [CrossRef] [PubMed][Green Version]
- Grant, P.R.; Grant, B.R. Adaptive radiation of Darwin’s finches. Am. Sci. 2002. [Google Scholar] [CrossRef]
- MacArthur, R.H.; Wilson, E.O. The Theory of Island Biogeography, 1st ed.; Princeton University Press: Princeton, NJ, USA, 1967; ISBN 0691088365. [Google Scholar]
- Darwin, C. Journal of Researches into the Natural History and Geology of the Countries Visited during the Voyage of H.M.S. Beagle Round the World, under the Command of Capt. Fitz Roy, R.N. (Voyage on the Beagle), 2nd ed.; John Murray: London, UK, 1845. [Google Scholar]
- Whittaker, R.J.; Fernández-Palacios, J.M.; Matthews, T.J.; Borregaard, M.K.; Triantis, K.A. Island biogeography: Taking the long view of nature’s laboratories. Science 2017, 357, 6354. [Google Scholar] [CrossRef]
- O’Brien, S.J. Genome empowerment for the Puerto Rican parrot—Amazona vittata. Gigascience 2012, 1, 13. [Google Scholar] [CrossRef] [PubMed]
- Snyder, N.; Wiley, J.W.; Kepler, C.B. The Parrots of Luquillo: Natural History and Conservation of the Puerto Rican Parrot; Western Foundation of Vertebrate Zoology: Los Angeles, CA, USA, 1987. [Google Scholar]
- Lack, D. Island Biology Illustrated by the Land Birds of Jamaica; Blackwell: Oxford, UK, 1976. [Google Scholar]
- Russello, M.A.; Amato, G. A molecular phylogeny of Amazona: Implications for Neotropical parrot biogeography, taxonomy, and conservation. Mol. Phylogenet. Evol. 2004, 30, 421–437. [Google Scholar] [CrossRef]
- Blanco, G.; Hiraldo, F.; Rojas, A.; Dénes, F.V.; Tella, J.L. Parrots as key multilinkers in ecosystem structure and functioning. Ecol. Evol. 2015, 5, 4141–4160. [Google Scholar] [CrossRef] [PubMed][Green Version]
- Blanco, G.; Hiraldo, F.; Tella, J.L. Ecological functions of parrots: An integrative perspective from plant life cycle to ecosystem functioning. Emu Aust. Ornithol. 2018, 118, 36–49. [Google Scholar] [CrossRef]
- Aslan, C.E.; Zavaleta, E.S.; Croll, D.; Tershy, B. Effects of Native and Non-Native Vertebrate Mutualists on Plants. Conserv. Biol. 2012, 26, 778–789. [Google Scholar] [CrossRef]
- Anderson, S.H.; Kelly, D.; Ladley, J.J.; Molloy, S.; Terry, J. Cascading effects of bird functional extinction reduce pollination and plant density. Science 2011, 331, 1068–1071. [Google Scholar] [CrossRef]
- Cracraft, J. Avian evolution, Gondwana biogeography and the Cretaceous-Tertiary mass extinction event. Proc. R. Soc. B Biol. Sci. 2001, 268, 459–469. [Google Scholar] [CrossRef]
- Wright, T.F.; Schirtzinger, E.E.; Matsumoto, T.; Eberhard, J.R.; Graves, G.R.; Sanchez, J.J.; Capelli, S.; Müller, H.; Scharpegge, J.; Chambers, G.K.; et al. A multilocus molecular phylogeny of the parrots (Psittaciformes): Support for a gondwanan origin during the cretaceous. Mol. Biol. Evol. 2008, 25, 2141–2156. [Google Scholar] [CrossRef] [PubMed]
- Rheindt, F.E.; Christidis, L.; Kuhn, S.; de Kloet, S.; Norman, J.A.; Fidler, A. The timing of diversification within the most divergent parrot clade. J. Avian Biol. 2014, 45, 140–148. [Google Scholar] [CrossRef]
- Prum, R.O.; Berv, J.S.; Dornburg, A.; Field, D.J.; Townsend, J.P.; Moriarty Lemmon, E.; Lemmon, A.R. A Comprehensive Phylogeny of Birds (Aves) using Targeted Next Generation DNA Sequencing Online Data and Software Archive. Nature 2015, 526, 569–573. [Google Scholar] [CrossRef] [PubMed]
- Claramunt, S.; Cracraft, J. A new time tree reveals Earth history’s imprint on the evolution of modern birds. Sci. Adv. 2015, 564, 136–141. [Google Scholar] [CrossRef]
- Mayr, G. The origins of crown group birds: Molecules and fossils. Palaeontology 2014, 57, 231–242. [Google Scholar] [CrossRef]
- Mayr, G. Paleogene Fossil Birds; Springer-Verlag: Berlin/Heidelberg, Germany, 2009. [Google Scholar]
- Ottens-Wainright, P.; Halanych, K.M.; Eberhard, J.R.; Burke, R.I.; Wiley, J.W.; Gnam, R.S.; Aquilera, X.G. Independent geographic origin of the genus Amazona in the West Indies. J. Caribb. Ornithol. 2004, 17, 23–49. [Google Scholar]
- Avise, J.C. Phylogeography: The History and Formation of Species; Harvard University Press: Boston, MA, USA, 2000. [Google Scholar]
- Kolchanova, S. Molecular Phylogeny and Evolution of Amazon Parrots in the Greater Antilles; Univeristy of Puerto Rico at Mayaguez: Mayagüez, Puerto Rico, 2018. [Google Scholar]
- Oleksyk, T.K.; Pombert, J.-F.; Siu, D.; Mazo-Vargas, A.; Ramos, B.; Guiblet, W.; Afanador, Y.; Ruiz-Rodriguez, C.T.; Nickerson, M.L.; Logue, D.M.; et al. A locally funded Puerto Rican parrot (Amazona vittata) genome sequencing project increases avian data and advances young researcher education. Gigascience 2012, 1, 14. [Google Scholar] [CrossRef][Green Version]
- Brinkley, D. The Wilderness Warrior: Theodore Roosevelt and the Crusade for America; Harper Collins: New York, NY, USA, 2009. [Google Scholar]
- Afanador, Y.; Velez-Valentín, J.; Valentín de la Rosa, R.; Martínez-Cruzado, J.C.; vonHoldt, B.; Oleksyk, K.T. Isolation and characterization of microsatellite loci in the critically endangered Puerto Rican parrot (Amazona vittata). Conserv. Genet. Resour. 2014, 6, 885–889. [Google Scholar] [CrossRef]
- Brock, M.K.; White, B.N. Application of DNA fingerprinting to the recovery program of the endangered Puerto Rican parrot. Proc. Natl. Acad. Sci. USA 1992, 89, 11121–11125. [Google Scholar] [CrossRef] [PubMed]
- Allendorf, F.W.; Hohenlohe, P.A.; Luikart, G. Genomics and the future of conservation genetics. Nat. Rev. Genet. 2010, 11, 697–709. [Google Scholar] [CrossRef] [PubMed]
- Ouborg, N.J.; Pertoldi, C.; Loeschcke, V.; Bijlsma, R.K.; Hedrick, P.W. Conservation genetics in transition to conservation genomics. Trends Genet. 2010, 26, 177–187. [Google Scholar] [CrossRef] [PubMed]
- Andrews, S. FASTQC. A Quality Control Tool for High Throughput Sequence Data. 2010. Available online: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ (accessed on 15 July 2019).
- Starostina, E.; Tamazian, G.; Dobrynin, P.; O’Brien, S.; Komissarov, A. Cookiecutter: A tool for kmer-based read filtering and extraction. bioRxiv 2015, 2015, 24679. [Google Scholar] [CrossRef]
- Bolger, A.M.; Lohse, M.; Usadel, B. Trimmomatic: A flexible trimmer for Illumina sequence data. Bioinformatics 2014, 30, 2114–2120. [Google Scholar] [CrossRef] [PubMed]
- Grabherr, M.G.; Haas, B.J.; Yassour, M.; Levin, J.Z.; Thompson, D.A.; Amit, I.; Adiconis, X.; Fan, L.; Raychowdhury, R.; Zeng, Q.; et al. Trinity: Recontructing a full-length transcriptome assembly without a genome from RNA-Seq data. Nat. Biotechnol. 2011, 29, 644–652. [Google Scholar] [CrossRef]
- Grigorev, K.; Kliver, S.; Dobrynin, P.; Komissarov, A.; Wolfsberger, W.; Krasheninnikova, K.; Afanador-Hernández, Y.M.; Paulino, L.A.; Carreras, R.; Rodríguez, L.E.; et al. Innovative assembly strategy contributes to the understanding of evolution and conservation genetics of the critically endangered Solenodon paradoxus from the island of Hispaniola. GigaScience 2018, 7, giy025. [Google Scholar] [CrossRef]
- Li, H.; Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 2009, 25, 1754–1760. [Google Scholar] [CrossRef][Green Version]
- Boetzer, M.; Henkel, C.V.; Jansen, H.J.; Butler, D.; Pirovano, W. Scaffolding pre-assembled contigs using SSPACE. Bioinformatics 2011, 27, 578–579. [Google Scholar] [CrossRef]
- Luo, R.; Liu, B.; Xie, Y.; Li, Z.; Huang, W.; Yuan, J.; He, G.; Chen, Y.; Pan, Q.; Liu, Y.; et al. SOAPdenovo2: An empirically improved memory-efficient short-read de novo assembler. Gigascience 2012, 1, 18. [Google Scholar] [CrossRef]
- Kent, W.J. BLAT—The BLAST-like alignment tool. Genome Res. 2002, 12, 656–664. [Google Scholar] [CrossRef] [PubMed]
- Xue, W.; Li, J.T.; Zhu, Y.P.; Hou, G.Y.; Kong, X.F.; Kuang, Y.Y.; Sun, X.W. L_RNA_scaffolder: Scaffolding genomes with transcripts. BMC Genom. 2013, 14, 604. [Google Scholar] [CrossRef] [PubMed]
- Marçais, G.; Kingsford, C. A fast, lock-free approach for efficient parallel counting of occurrences of k-mers. Bioinformatics 2011, 27, 764–770. [Google Scholar] [CrossRef] [PubMed][Green Version]
- Kliver, S. KRATeR. Available online: https://github.com/mahajrod/KrATER (accessed on 15 July 2019).
- Simão, F.A.; Waterhouse, R.M.; Ioannidis, P.; Kriventseva, E.V.; Zdobnov, E.M. BUSCO: Assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics 2015, 31, 3210–3212. [Google Scholar] [CrossRef] [PubMed]
- Bao, W.; Kojima, K.K.; Kohany, O. Repbase Update, a database of repetitive elements in eukaryotic genomes. Mob. DNA 2015, 6, 11. [Google Scholar] [CrossRef]
- Smit, A.; Hubley, R.; Green, P. RepeatMasker Open-4.0. 2013–2018. Available online: http://www.repeatmasker.org (accessed on 10 June 2018).
- Smit, A.; Hubley, R. RepeatModeler Open-1.0. Available online: https://github.com/rmhubley/RepeatModeler/blob/master/README (accessed on 14 July 2018).
- Tarailo-Graovac, M.; Chen, N. Using RepeatMasker to identify repetitive elements in genomic sequences. Curr. Protoc. Bioinform. 2009, 5, 4–10. [Google Scholar] [CrossRef]
- Quinlan, A.R.; Hall, I.M. BEDTools: A flexible suite of utilities for comparing genomic features. Bioinformatics 2010, 26, 841–842. [Google Scholar] [CrossRef]
- Slater, G.S.C.; Birney, E. Automated generation of heuristics for biological sequence comparison. BMC Bioinformatics 2005, 6, 31. [Google Scholar] [CrossRef]
- Dobin, A.; Davis, C.A.; Schlesinger, F.; Drenkow, J.; Zaleski, C.; Jha, S.; Batut, P.; Chaisson, M.; Gingeras, T.R. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics 2013, 29, 15–21. [Google Scholar] [CrossRef]
- Stanke, M.; Keller, O.; Gunduz, I.; Hayes, A.; Waack, S.; Morgenstern, B. AUGUSTUS: Ab initio prediction of alternative transcripts. Nucleic Acids Res. 2006, 34, W435–W439. [Google Scholar] [CrossRef] [PubMed]
- Johnson, L.S.; Eddy, S.R.; Portugaly, E. Hidden Markov model speed heuristic and iterative HMM search procedure. BMC Bioinform. 2010, 11, 431. [Google Scholar] [CrossRef] [PubMed]
- Altschul, S.F.; Gish, W.; Miller, W.; Myers, E.W.; Lipman, D.J. Basic local alignment search tool. J. Mol. Biol. 1990, 215, 403–410. [Google Scholar] [CrossRef]
- Bateman, A.; Coin, L.; Durbin, R.; Finn, R.D.; Hollich, V.; Griffiths-Jones, S.; Khanna, A.; Marshall, M.; Moxon, S.; Sonnhammer, E.L.L.; et al. The Pfam protein families database. Nucleic Acids Res. 2004, 32, D138–D141. [Google Scholar] [CrossRef] [PubMed]
- Bateman, A.; Martin, M.J.; O’Donovan, C.; Magrane, M.; Alpi, E.; Antunes, R.; Bely, B.; Bingley, M.; Bonilla, C.; Britto, R.; et al. UniProt: The universal protein knowledgebase. Nucleic Acids Res. 2017, 45, D158–D169. [Google Scholar] [CrossRef]
- Van der Auwera, G.A.; Carneiro, M.O.; Hartl, C.; Poplin, R.; del Angel, G.; Levy-Moonshine, A.; Jordan, T.; Shakir, K.; Roazen, D.; Thibault, J.; et al. From fastQ data to high-confidence variant calls: The genome analysis toolkit best practices pipeline. Curr. Protoc. Bioinform. 2013, 43, 11.10.1–11.10.33. [Google Scholar] [CrossRef]
- Huerta-Cepas, J.; Szklarczyk, D.; Forslund, K.; Cook, H.; Heller, D.; Walter, M.C.; Rattei, T.; Mende, D.R.; Sunagawa, S.; Kuhn, M.; et al. eggNOG 4.5: A hierarchical orthology framework with improved functional annotations for eukaryotic, prokaryotic and viral sequences. Nucleic Acids Res. 2015, 44, gkv1248. [Google Scholar] [CrossRef] [PubMed]
- Frankl-Vilches, C.; Kuhl, H.; Werber, M.; Klages, S.; Kerick, M.; Bakker, A.; de Oliveira, E.H.C.; Reusch, C.; Capuano, F.; Vowinckel, J.; et al. Using the canary genome to decipher the evolution of hormone-sensitive gene regulation in seasonal singing birds. Genome Biol. 2015, 16, 19. [Google Scholar] [CrossRef]
- Ellegren, H.; Smeds, L.; Burri, R.; Olason, P.I.; Backström, N.; Kawakami, T.; Künstner, A.; Mäkinen, H.; Nadachowska-Brzyska, K.; Qvarnström, A.; et al. The genomic landscape of species divergence in Ficedula flycatchers. Nature 2012, 491, 756. [Google Scholar] [CrossRef]
- Laine, V.N.; Gossmann, T.I.; Schachtschneider, K.M.; Garroway, C.J.; Madsen, O.; Verhoeven, K.J.F.; de Jager, V.; Megens, H.-J.; Warren, W.C.; Minx, P.; et al. Evolutionary signals of selection on cognition from the great tit genome and methylome. Nat. Commun. 2016, 7, 10474. [Google Scholar] [CrossRef][Green Version]
- Balakrishnan, C.N.; Mukai, M.; Gonser, R.A.; Wingfield, J.C.; London, S.E.; Tuttle, E.M.; Clayton, D.F. Brain transcriptome sequencing and assembly of three songbird model systems for the study of social behavior. PeerJ 2014, 2, e396. [Google Scholar] [CrossRef] [PubMed]
- Mueller, J.C.; Kuhl, H.; Timmermann, B.; Kempenaers, B. Characterization of the genome and transcriptome of the blue tit Cyanistes caeruleus: Polymorphisms, sex-biased expression and selection signals. Mol. Ecol. Resour. 2016, 16, 549–561. [Google Scholar] [CrossRef]
- Ganapathy, G.; Howard, J.T.; Ward, J.M.; Li, J.; Li, B.; Li, Y.; Xiong, Y.; Zhang, Y.; Zhou, S.; Schwartz, D.C.; et al. High-coverage sequencing and annotated assemblies of the budgerigar genome. Gigascience 2014, 3, 2011–2047. [Google Scholar] [CrossRef] [PubMed]
- Zhang, G.; Parker, P.; Li, B.; Li, H.; Wang, J.; Parker, P.; Li, B.; Li, H.; Wang, J. The genome of Darwin’s Finch (Geospiza fortis). GigaDB. 2012. Available online: http://dx.doi.org/10.5524/100040 (accessed on 10 June 2018).
- Warren, W.C.; Clayton, D.F.; Ellegren, H.; Arnold, A.P.; Hillier, L.W.; Künstner, A.; Searle, S.; White, S.; Vilella, A.J.; Fairley, S.; et al. The genome of a songbird. Nature 2010, 464, 757–762. [Google Scholar] [CrossRef] [PubMed]
- Doyle, J.M.; Katzner, T.E.; Bloom, P.H.; Ji, Y.; Wijayawardena, B.K.; DeWoody, J.A. The Genome Sequence of a Widespread Apex Predator, the Golden Eagle (Aquila chrysaetos). PLoS ONE 2014, 9, e95599. [Google Scholar] [CrossRef] [PubMed]
- Zhan, X.; Pan, S.; Wang, J.; Dixon, A.; He, J.; Muller, M.G.; Ni, P.; Hu, L.; Liu, Y.; Hou, H.; et al. Peregrine and saker falcon genome sequences provide insights into evolution of a predatory lifestyle. Nat. Genet. 2013, 45, 563. [Google Scholar] [CrossRef]
- Löytynoja, A.; Goldman, N. webPRANK: A phylogeny-aware multiple sequence aligner with interactive alignment browser. BMC Bioinform. 2010, 11, 579. [Google Scholar] [CrossRef]
- Talavera, G.; Castresana, J. Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments. Syst. Biol. 2007, 56, 564–577. [Google Scholar] [CrossRef] [PubMed]
- Castresana, J. Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol. Biol. Evol. 2000, 17, 540–552. [Google Scholar] [CrossRef]
- Stamatakis, A. RAxML version 8: A tool for phylogenetic analysis and post-analysis of large phylogenies. Bioinformatics 2014, 30, 1312–1313. [Google Scholar] [CrossRef]
- Rambaut, A. FigTree. 2016. Available online: http://tree.bio.ed.ac.uk/software/figtree/ (accessed on 10 June 2018).
- Li, H.; Durbin, R. Inference of human population history from individual whole-genome sequences. Nature 2011, 475, 493–496. [Google Scholar] [CrossRef][Green Version]
- Earnhardt, J.; Vélez-Valentín, J.; Valentin, R.; Long, S.; Lynch, C.; Schowe, K. The Puerto Rican parrot reintroduction program: Sustainable management of the aviary population. Zoo Biol. 2014, 33, 89–98. [Google Scholar] [CrossRef] [PubMed]
- Smeds, L.; Qvarnström, A.; Ellegren, H. Direct estimate of the rate of germline mutation in a bird. Genome Res. 2016, 6, 1211–1218. [Google Scholar] [CrossRef] [PubMed]
- Raney, B.J.; Dreszer, T.R.; Barber, G.P.; Clawson, H.; Fujita, P.A.; Wang, T.; Nguyen, N.; Paten, B.; Zweig, A.S.; Karolchik, D.; et al. Track data hubs enable visualization of user-defined genome-wide annotations on the UCSC Genome Browser. Bioinformatics 2014, 30, 1003–1005. [Google Scholar] [CrossRef] [PubMed]
- Kent, W.J.; Zweig, A.S.; Barber, G.; Hinrichs, A.S.; Karolchik, D. BigWig and BigBed: Enabling browsing of large distributed datasets. Bioinformatics 2010, 26, 2204–2207. [Google Scholar] [CrossRef] [PubMed]
- Tiersch, T.R.; Wachtel, S.S. On the evolution of genome size of birds. J. Hered. 1991, 82, 363–368. [Google Scholar] [CrossRef]
- Gregory, T. Animal Genome Size Database. Available online: http://www.genomesize.com (accessed on 15 July 2018).
- Volfovsky, N.; Oleksyk, T.K.; Cruz, K.C.; Truelove, A.L.; Stephens, R.M.; Smith, M.W. Chimpanzee chromosome 23 vs. human 22: Genomic insertion, deletion and ancestral indel polymorphisms. BMC Genom. 2009, 10. [Google Scholar] [CrossRef]
- Organ, C.L.; Shedlock, A.M.; Meade, A.; Pagel, M.; Edwards, S.V. Origin of avian genome size and structure in non-avian dinosaurs. Nature 2007, 446, 180. [Google Scholar] [CrossRef]
- Kapusta, A.; Suh, A.; Feschotte, C. Dynamics of genome size evolution in birds and mammals. Proc. Natl. Acad. Sci. USA 2017. [Google Scholar] [CrossRef]
- Coullin, P.; Bed’Hom, B.; Candelier, J.J.; Vettese, D.; Maucolin, S.; Moulin, S.; Galkina, S.A.; Bernheim, A.; Volobouev, V. Cytogenetic repartition of chicken CR1 sequences evidenced by PRINS in Galliformes and some other birds. Chromosom. Res. 2005, 13, 665–673. [Google Scholar] [CrossRef]
- Ksepka, D.; Clarke, J. Phylogenetically vetted and stratigraphically constrained fossil calibrations within Aves. Palaeontol. Electron. 2015, 18, 1–25. [Google Scholar] [CrossRef]
- DeWoody, Y.D.; DeWoody, J.A. On the estimation of genome-wide heterozygosity using molecular markers. J. Hered. 2005, 96, 85–88. [Google Scholar] [CrossRef] [PubMed]
- Coltman, D.W.; Slate, J. Microsatellite measures of inbreeding: A meta-analysis. Evolution 2003, 57, 971–983. [Google Scholar] [CrossRef]
- Chapman, J.R.; Nakagawa, S.; Coltman, D.W.; Slate, J.; Sheldon, B.C. A quantitative review of heterozygosity-fitness correlations in animal populations. Mol. Ecol. 2009. [Google Scholar] [CrossRef]
- Spielman, D.; Brook, B.W.; Frankham, R. Most species are not driven to extinction before genetic factors impact them. Proc. Natl. Acad. Sci. USA 2004, 101, 15261–15264. [Google Scholar] [CrossRef] [PubMed][Green Version]
- Vellend, M. Island Biogeography of Genes and Species. Am. Nat. 2003, 162, 358–365. [Google Scholar] [CrossRef] [PubMed]
- Wright, S. Breeding Structure of Populations in Relation to Speciation. Am. Nat. 1940, 74, 232–248. [Google Scholar] [CrossRef]
- Fleming, T.H. The theory of island biogeography at age 40. Evolution 2010. [Google Scholar] [CrossRef]
- Li, S.; Li, B.; Cheng, C.; Xiong, Z.; Liu, Q.; Lai, J.; Carey, H.V.; Zhang, Q.; Zheng, H.; Wei, S.; et al. Genomic signatures of near-extinction and rebirth of the crested ibis and other endangered bird species. Genome Biol. 2014, 15, 557. [Google Scholar] [CrossRef][Green Version]
Figure 1. Amazon parrots included in this study (Amazona leucocephala, A. ventralis and A. vittata) may all have originated from Central America, where the white-fronted amazon (A. albifrons) can be found today (modified from Kolchanova (2018) ).
Figure 2. Reconstructed phylogenetic tree for 14 species. Reconstruction was performed using RAxML 8  with falcons (Falco cherrug and F. peregrinus) as an outgroup. A. vittata and A. ventralis form a monophyletic group, with A. leucocephala as their sister taxon.
Figure 3. Population history of the three Amazona species. For all of them, A. vittata genome was used as a reference. Generation times were calculated by the captive breeding program for Puerto Rican parrot  and mutation rates recently estimated from bird pedigrees available in the literature . Trajectories of all three species suggest an initial founder effect that may be attributed to parrot dispersal from Central America between 2 and 3 MYA.
Table 1. Genome size estimates for the three Amazon parrot species in this study.
|Parrot Species||PE Library Coverage||Genome Size (Gbp)||C-Value (pg)|
23-mer based estimate based on sequencing data in this study; C-values are from the Animal Genome Size Database .
Table 2. Statistics for RNAseq libraries and assembled transcripts.
|Library ID||Tissue||Read Pairs (Millions)||Bases (Gbp)||Assembled Transcripts|
Table 3. Metrics for the Amazona vittata de novo genome assembly.
|N50 (kbp)||L50 (kbp)||Longest Contig (Mb)||Number of Ns Mb||Number of Scaffolds||Assembled Genome Length (Gbp)|
|Class||Number of Repeats||Total Length (bp)||Percentage of the Genome (%)|
Table 5. Statistics for protein-coding genes of species used in the current analysis. Among the three Amazona species used, only the A. vittata is listed, since as for A. leucocephala and A. ventralis only reference-assisted assemblies were performed, so that identical (or almost identical) gene counts are reported for these species.
|Species||N of Genes||N of Genes with Longest Protein <100 aa||N Genes Assigned to the EggNOG Clusters||Genome Size (pg) *|
* C-values are from the Genome Size Database .
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).