Introduction

Successful interspecific mating can be infrequent because of reproductive barrier diversity1. Nevertheless, interspecific hybridization does occur in nature among both plants and animals, albeit at low frequencies2. Novel genomic data have revealed multiple incidences of multispecies hybridization3. Interspecific hybridization can be an important generator of genetic diversity, which drives adaptation and speciation under changing or fluctuating environments4,5,6. However, it can also negatively impact the maintenance of species boundaries7,8,9. In birds, various phenotypic traits—such as mating, foraging behavior, and migration—can act as pre-mating barriers or extrinsic post-zygotic barriers to successful interspecific mating9. In addition, low levels of social bonding and migration are associated with a lack of hybridization10. In recent years, climate change and anthropogenic habitat disturbance have been implicated in the dismantling of some barriers to hybridization11,12,13,14. As such, issues of hybridization in nature pose challenges to biological and phylogenetic species concepts and conservation practices.

Hybrid detection using genomic data has potential applications in the conservation of endangered species that exhibit hybridization and introgression. It can also be used to measure the rate of hybridization, differentiate natural vs. human-caused hybridization, and in the genetic management of species to improve their adaptability. In birds, interspecific hybrids have been identified in the natural populations of various species using different DNA markers, including Forbes’ parakeet, buntings, and chickadees (using mtDNA and microsatellites)15,16,17; partridges (using RAPD)18; warblers and owls (using AFLP)19,20,21; and flycatchers, woodpeckers, and eagles (using microsatellites and single nucleotide polymorphisms [SNPs])22,23,24. At present, third-generation DNA markers based on SNP and insertions/deletions (InDels) have gained popularity25,26. InDel markers have been widely used to differentiate hybrids25,27, as they can be genotyped with simple procedures based on size separation28.

Another advantage of InDel markers is the minuscule chance of two InDel mutations of exactly the same length occurring at the same genomic position. Therefore, shared InDels can be seen as representing identity by descent28. Recent advances in DNA sequencing technology have led to the reference genomes of various species becoming publicly accessible29.Consequently, InDel polymorphisms can be easily identified by mapping new re-sequencing data to reference genomes27. Recent studies on hybridization have advanced the identification and widespread availability of multiple nuclear markers. These markers can be used to directly link interspecific hybridization and its evolutionary significance with the adaptability of hybridizing species (e.g., their fitness or selection pressures), rather than with evolutionary noise30.

Two variants of rock pigeons—domestic (Columba livia) and feral (Columba livia var. domestica)—are among the most intensively studied birds worldwide. Wild rock pigeons (C. livia) are native to the Old World, whereas domestic pigeons with diverse forms have been produced through long-term artificial selection by humans31. Several domestic varieties have also been going feral worldwide32. In particular, racing breeds of domestic pigeons have made extensive contributions to feral populations29,33. The frequency of hybridization is known to be higher among sister species than among non-sister species34. Gene flow from feral pigeons to native populations has reduced the genetic originality of rock pigeons, leading to overlaps in morphology and behavioral ecology35. Furthermore, interspecific hybridization between two closely related species—feral pigeons and hill pigeons (Columba rupestris)—has been reported in some regions, including in Northern India, Southern Siberia, and South Korea36,37. This is a cause for concern, as the hybridization has been accompanied by a loss of C. rupestris population36,37,38. However, the patterns and processes of interspecific hybridization between the two species remain unclear because of the lack of contemporary genomic tools39.

In the present study, we developed agarose-resolvable InDel markers through whole-genome re-sequencing (WGS) to differentiate hill pigeons from feral pigeons and their F1 hybrid. We used these markers to verify the hybrid risk in wild populations as follows: (1) we identified InDel polymorphisms between hill and rock pigeons (i.e., domestic and feral pigeons) using WGS; (2) we conducted experimental validation of species-specific InDel polymorphisms through agarose gel electrophoresis using native populations in South Korea (i.e., hill and feral pigeons); and (3) we developed hybrid markers between hill and feral pigeons. Polymorphisms in these species-specific InDel regions were re-investigated (through agarose gel electrophoresis) using four pairs of hill and feral pigeons and their F1 hybrids. On the basis of our results, we discuss the conservation perspectives of endangered species that exhibit interspecific hybridization with target species.

Results

Identification of InDel polymorphisms between hill pigeons and rock pigeons

A total of 112,638,908 and 841,269,524 clean reads were generated from seven hill pigeons and seven rock pigeons (Table 1). The clean reads were mapped to the reference genome assembly (Clive_1.0) using Burrows–Wheeler Aligner (BWA), and 88–94.45% of the reads from hill pigeons and rock pigeons were mapped with a depth of 8.61–45.67, respectively (Table 1). The overall genome coverage from hill pigeons and rock pigeons was 98.61–99.63% (Table 1), and genome-wide analysis revealed 226,030–814,877 InDels (Table 2). InDels occurred most frequently in intron regions (43.83%), followed by intergenic regions (24.13%) (Table 3). A total of 67 primers with an InDel length of > 34 bp (HM 1–HM 67) (Table S1) were used for experimental validation of species-specific InDel polymorphisms.

Table 1 Summary of the original sequencing data of seven hill pigeons and seven feral/domestic pigeons.
Full size table
Table 2 InDel polymorphisms identified in the genome sequences of seven hill pigeons and seven feral/domestic pigeons.
Full size table
Table 3 Types of InDel polymorphisms identified in the genome sequences of seven hill pigeons and seven feral/domestic pigeons.
Full size table

Experimental validation of species-specific InDel markers for hill pigeon and feral pigeon

To identify the species-specific InDel regions between the two species, we used 16 hill pigeons and 14 feral pigeons. All hill pigeons used in the experiment were confirmed as hill pigeons based on their feather phenotype (Fig. S1). Using 67 primers, we performed PCR and agarose gel electrophoresis with randomly selected individuals: hill pigeon 4 (or 2) and feral pigeon 5 (or 2). Primer pairs that amplified different base pairs between the two species (Fig. S2) were selected. As a result, species-specific InDels between the two species were identified in 17 out of 67 primer pairs (Fig. S2). PCR and agarose gel electrophoresis were re-performed using the 17 selected primers for 16 hill pigeons and 14 feral pigeons (Fig. S3). Finally, 11 species-specific InDel markers were identified (Fig. S3).

Experimental validation of InDel markers for F1 hybrids between hill pigeons and feral pigeons

Eleven of the species-specific InDel markers were used to develop the hybrid markers. A total of eight F1 hybrids were produced using four individuals of hill pigeon and feral pigeon each (Fig. S4). Five out of eight hybrid F1 individuals died in the brooding stage. Therefore, the phenotype of only three individuals could be confirmed. Using the 16 individuals, 11 species-specific InDel markers were tested. For 10 (HM30, HM37, HM41, HM42, HM43, HM45, HM51, HM55, HM56, and HM59) out of the 11 markers, two DNA fragments (hill and feral pigeon) were simultaneously amplified from all F1 individuals (Fig. 1). In contrast, HM60 amplified 304-bp or 338-bp fragments in some individuals of hybrid F1 (Fig. 1). Ten of the selected hybrid InDel markers were sequenced using the PCR products of two randomly selected individuals of each species. The sequencing results of HM30, HM37, HM41, HM43, HM45, HM55, HM56, and HM59 matched (Fig. 2) with the InDel regions identified during primer design (Table S1). However, in HM42, a mutation occurred in the feral pigeon (Fig. 2), and the sequencing results did not match those obtained during primer design (Table S1). Furthermore, in HM51, the position of the InDels did not match that of the InDel regions identified during primer design (Fig. 2; Table S1).

Figure 1
figure 1

Testing of 11 InDel regions to develop the markers of hybridization between two wild populations (Columba rupestris and C. livia var. domestica) and their hybrid using agarose gel electrophoresis. The names of accepted markers are shown in bold. Black arrows indicate 300 bp. Representative phenotypes (i.e., white band in tail feathers) of pigeons. Phenotypes are shown for (a) a hill pigeon (C. rupestris), (b) a feral pigeon (C. livia var. domestica), and (c) their F1 hybrid. Groups of gels are cropped from different gel images and original images are included in Fig. S5.

Full size image
Figure 2
figure 2

Sequences of 10 selected InDel markers to identify the F1 hybrids between hill and feral pigeons. Data are shown for two randomly selected individuals in the two species.

Full size image

Discussion

The rock pigeon (C. livia) has a long history of hybridization with its sister taxa37. Taxonomically, rock pigeons include domestic pigeons and free-living feral pigeons40. Feral pigeons (C. livia var. domestica) are not a distinct species, but a variety of rock pigeons40. The wild ancestors of feral pigeons were first domesticated approximately 1000 years ago, and hybridization has occurred among domestic pigeons throughout their history of domestication41. Subsequently, domesticated pigeons began escaping captivity and were termed “feral”40. Feral pigeons have been introduced worldwide by humans (The World Bird Database: https://avibase.bsc-eoc.org/avibase.jsp?lang=EN). Such artificial introductions can cause overlaps in the distribution areas of species, and pre-zygotic barriers (e.g., life cycle, behavior, and distribution) can be disrupted among closely related species. The two study species, hill and feral pigeons, have similar characteristics of inhabiting man-made structures. However, feral pigeons mainly inhabit urban areas, and hill pigeons inhabit forest areas. Occasionally, feral pigeons invade man-made structures in forest areas where hill pigeons live, resulting in hybridization. Hybrids between feral pigeons and native populations in the Columbidae family have recently been reported in North America42 and South Korea36. The hybrid markers that evolved throughout this complicated history were evaluated in this study.

Prior to experimental validation, we confirmed the phenotype of all individuals used in the experiment. All hill pigeons used in this study were collected from Gurey-gun and Goheung-gun in South Korea, where interspecific hybridization was not detected using mtDNA36, and the phenotype also reflected the characteristics of hill pigeons (Figs. S1, S4). In contrast, the feral pigeons showed various phenotypes, and it was not possible to confirm their species-specific phenotype. However, the white tail-band across the black tail was clearly distinguished in the phenotype between the two species (Fig. S4). The tail pattern of F1 hybrids showed a clear tendency toward mixing the phenotypes of the two species, with the gray background of the feral pigeon’s traits and a rare white band of hill pigeon’s traits. This was consistent with the characteristics of individuals suspected to be hybrids using mtDNA in a previous study36.

To develop interspecific markers, species-specific InDel regions were identified in 11 of the 67 primers (Fig. S3). We tested 11 species-specific InDel markers and performed PCR, agarose gel electrophoresis, and sequencing analysis using these eight F1 hybrids and their eight parents (i.e., four pairs). The results of agarose gel electrophoresis from the F1 hybrid were expected to show two DNA fragments from each parent. Although HM60 clearly distinguished the two species in the parental generation, some offspring did not inherit all of the parental characteristics (Fig. 1). We assumed that this was either because of allelic dropouts or null alleles. This phenomenon is more likely to occur when primers do not perfectly match the flanking sequences and fail to amplify one or both alleles of a diploid individual43. The 10 selected InDel markers (except HM60) were sequenced using two randomly selected individuals from the two species (Fig. 2). Through this process, we identified whether the expected InDels matched the sequences obtained during experimental validation (PCR and gel-loading). We confirmed that HM42 was unstable because of the occurrence of mutations in the InDel region. Moreover, HM51 was excluded from the final marker list because InDels did not occur in the expected region. In total, we developed eight InDel markers (HM30, HM37, HM41, HM43, HM45, HM55, HM56, and HM59) for F1 hybrids between hill and feral pigeons.

The genomic evaluation of interspecific hybridization within sister taxa is a conservation management tool for rare species that are close to extinction because of low genetic diversity, inbreeding depression, or poorly adapting phenotypes in rapidly changing environments44. The occurrence and function of hybridizations in nature may be underestimated, and the effects of anthropogenic habitat disturbance should also be considered from the perspective of conservation biology12,45,46. Among birds, hybrids exhibit higher fitness than parental species in fluctuating environments4. Hybridization between species can also assist species range expansion47. Nevertheless, the survival probability of fertile hybrids seems to be low under strong pre-zygotic selection in birds, although hybridizing species may not be rare in the wild7. Ecosystem-scale alterations in the physical habitat (e.g., clear-cutting, road and city construction and urbanization, eutrophication, and irrigation) and the introduction of invasive species have been on-going in the wild, resulting in a breakdown of reproductive barriers (i.e., increasing hybridization or introgression)12. The overall proportion of feral pigeons cohabiting with hill pigeons was approximately 20% within the three known colonies (i.e., 140 individuals) in South Korea, where the small numbers of interspecific hybrids have been reported to invade and compete with feral pigeons.

In the present study, we developed eight InDel markers (HM30, HM37, HM41, HM43, HM45, HM55, HM56, and HM59) using WGS. Our findings can be applied to the conservation of hill pigeons, which have recently shown a rapid decline in South Korea36. In addition, the method used in this study (i.e., selecting InDel markers using WGS) may be used as a precedent for the discrimination of hybrids occurring in other animal taxa. Genomic tools for hybrid detection are expected to play a crucial role in the assessment of hybridization frequency in the wild. Therefore, our method can be a convenient approach for hybrid detection; however, it is expected that better results can be obtained if the structure analysis method using SNP48 is additionally employed to confirm and evaluate the hybrid ratio of individuals other than F1. Future studies are needed to elucidate the process of interspecific hybridization in the presence of reproductive barriers. In addition, the systematic captive propagation of the hybrids of hill pigeons and feral pigeons can help mitigate the population decline of hill pigeons.

Methods

Study system

The hill pigeon is closely related to the rock pigeon and snow pigeon (C. leuconota). The two subspecies of the study species are recognized as C. r. turkestanica (western form) and C. r. rupestris (eastern form)37. The hill pigeon is similar in appearance to the rock pigeon in terms of size and plumage but is differentiated by its tail pattern. Specifically, hill pigeons have a broad, white tail-band across the black tail and a white patch on the back (Fig. 1, Fig. S1). This white-banded tail pattern is similar to that of the snow pigeon. Recently, researchers have reported a decline in the populations of hill pigeons in India and South Korea, primarily because of invasion, nest-site competition, and interspecific hybridization by feral pigeons. Nevertheless, hill pigeons have been designated a non-endangered species on the basis of their overall distribution36,37. The frequency of hybridization between the two sister taxa is stable (3–10% of the population), and there are some ethological barriers to potential random mating (e.g., breeding asynchrony and habitat isolation)37. Historical records from 1948 indicate that hill pigeons (but not rock pigeons or feral pigeons) were commonly found in palace grounds and coastal river cliffs in the Korean Peninsula49. However, recent reports indicate that there are only three breeding populations with less than 100 individuals in South Korea36. Accordingly, the taxon has been recommended for priority conservation with interspecific hybridization (with ca. 20% cohabiting feral pigeons) in South Korea.

DNA sampling

All samples were collected in South Korea between 2015 and 2022. Hill pigeons were captured in Gurye-gun and Goheung-gun, which are their known habitats36. Feral pigeons were captured from among individuals that had infiltrated the habitat of hill pigeons and caused hybridization. The captured individuals were transferred to the Research Center for Endangered Species of the National Institute of Ecology (NIE) and National Institute of Biological Resources (NIBR), South Korea. The phenotypes of the birds were identified, and blood samples were collected. Following this, the birds were either used for captive propagation or released back to the collection site36. Feathers or blood samples of six hill pigeons and one feral pigeon were used to extract DNA for WGS (Table 4)50. In a previous study that used the genome sequences of the samples in Table 4, all samples were clearly divided into hill and rock pigeon groups through both principal component analysis and non-synonymous SNP phylogenetic analysis, and it was confirmed that previous introgression did not occur between the two sample groups51. Furthermore, additional blood samples were collected from 20 hill pigeons, 18 feral pigeons, and 8 F1 hybrids (Figs. S1, S4). From these samples, DNA was extracted from these samples and analyzed by agarose gel electrophoresis. These hill pigeons were captured in Gurye and Goheung-gun, South Korea, and a previous study confirmed their distinction from rock pigeons through a haplotype network analysis using mtDNA36. The F1 hybrids of the two species consisted of eight chicks obtained from four pairs of parents. In order to produce F1 hybrids, one pair of hill and feral pigeon were placed in a total of four cages using each two female and two male of each species. And natural breeding was induced by providing a sufficient amount of grains. The brooding process after egg hatching also induced a natural feeding process by the parents. The study design was approved by the Research Experimental Ethics Committee of NIE (NIEIACUC-R-2021-019). And all experiment were performed in accordance with the ARRIVE guidelines, and carried out Accordance with relevant guidelines and regulations.

Table 4 Whole-genome re-sequencing sources of seven hill pigeons (Columba rupestris) and seven feral/domestic pigeons (C. livia var. domestica).
Full size table

DNA extraction and high-throughput DNA sequencing

The collected samples were stored at − 30 °C before DNA extraction, and DNA was extracted using the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany), following the manufacturer’s protocol. The quality of extracted DNA was evaluated by electrophoresis, and the DNA concentration measured using a Nanodrop system (Thermo Fisher Scientific, Waltham, MA, USA). Genomic DNA (200 ng) was sheared into ~ 500-bp fragments using the Covaris system (Covaris, Woburn, MA, USA), and libraries were constructed using a library kit (Illumina, San Diego, CA, USA). High-throughput paired-end sequencing was performed on an Illumina NovaSeq platform.

Mining of InDel polymorphisms from the genomes of hill and rock pigeons

To identify InDel polymorphisms between hill pigeons and rock pigeons (domestic and feral pigeons), we explored the reference genome of the rock pigeon29 downloaded from the National Center for Biotechnology Information (NCBI) (accession no: Assembly Clive_1.0). The genome sequences of five breeds of domestic pigeon and one hill pigeon29 were also obtained from NCBI (Table 4). Genomic DNA isolated from the six hill pigeons and one feral pigeon was also used for analysis, and InDel polymorphisms between two groups of 14 individuals (Table 4) were identified using the reference genome (assembly Clive_1.0). Clean reads were obtained by trimming the ends of low-quality reads (< Q30) (Table 1). The Sickle software was used to trim the sequences (https://github.com/najoshi/sickle). The cleaned reads were aligned to the reference genome with the Burrows–Wheeler Aligner (BWA 1.7.10-r789). Quality control of FastQ files was performed using the FastQC software (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). InDels were filtered and called using the Genome Analysis Tool Kit (GATK) version 3.1 (https://gatk.broadinstitute.org/hc/en-us). InDel polymorphisms were extracted under the following conditions: homozygous within the groups (hill pigeon or rock pigeon) and heterozygous between the two groups (hill pigeon and rock pigeon).

Primer design, PCR amplification, and sequencing

The online Primer3-Plus tool (http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) was used to design PCR primers for the locations of InDel polymorphisms between hill and rock pigeons. The parameters were as follows: amplicon size, 150–400 bp; primer melting temperature (Tm), 48–60 °C; and 40–60% primer GC content. Based on these parameters, the primers that included a total of 1703 InDel polymorphisms > 20 bp in length were identified. To select primers that can be visually confirmed as conveniently as possible after gel loading, the InDel polymorphisms were listed in order of the longest. Subsequently, a total of 67 primers were selected, that included 18 primers with InDels > 50 bp, 19 primers with InDels > 40 bp, and 30 primers with InDels > 34 bp (HM 1–HM 67) (Table S1), which were used to investigate species-specific InDel polymorphisms between hill and feral pigeons through agarose gel electrophoresis. PCR amplifications for all samples were carried out (PCR volume, 30 µL) using 2 × Lamp Taq PCR Pre-Mix (Biofact, Daejeon, South Korea). The PCR thermal profile of each primer has been described in Table S2. All PCR products were checked on 2% agarose gels and visualized using ultraviolet light at 300 nm on Gel Doc XR plus (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The sequencing was performed by a commercial sequencing service (Macrogen Inc., Seoul, South Korea) on two randomly selected individuals from two species (hill pigeon and feral pigeon).