Demographic history and genetic diversity of wild African harlequin quail (Coturnix delegorguei delegorguei) populations of Kenya

Abstract Hunting wild African harlequin quails (Coturnix delegorguei delegorguei) using traditional methods in Western Kenya has been ongoing for generations, yet their genetic diversity and evolutionary history are largely unknown. In this study, the genetic variation and demographic history of wild African harlequin quails were assessed using a 347bp mitochondrial DNA (mtDNA) control region fragment and 119,339 single nucleotide polymorphisms (SNPs) from genotyping‐by‐sequencing (GBS) data. Genetic diversity analyses revealed that the genetic variation in wild African harlequin quails was predominantly among individuals than populations. Demographic analyses indicated a signal of rapid demographic expansion, and the estimated time since population expansion was found to be 150,000–350,000 years ago, corresponding to around the Pliocene–Pleistocene boundary. A gradual decline in their effective population size was also observed, which raised concerns about their conservation status. These results provide the first account of the genetic diversity of wild African harlequin quails of Siaya, thereby creating a helpful foundation in their biodiversity conservation.

African harlequin quails into farmlands encourages their incessant capture for sale and consumption. At the moment, control measures to check on illegal and uncontrolled hunting are nonexistent, and little to no effort has been made to monitor and study their conservation status.
The domestication of Japanese quail during the late 19th century and early 20th century has contributed immensely toward improved poultry meat and egg production (Lukanov & Pavlova, 2020). Currently, the domestic Japanese quail can be found in many countries where they are selectively bred into meat and egg types (Jeke et al., 2018). The introduction of domestic Japanese quails in Africa and other developing countries has provided entrepreneurial opportunities and an alternative way of alleviating animal protein deficiency in their populations (Nasar et al., 2016). However, this introduced new challenges such as the uncontrolled introduction of immense exotic Japanese quail breeds in Kenya and other sub-Saharan countries (Jeke et al., 2018). Considering that the domestic quail business bubble soon deflated, some of these exotic quail breeds were disposed of by frustrated farmers. Planned or accidental release of domestic quail individuals into the wild would bring adverse genetic changes to wild quail populations such as loss of genetic variation and altered population structures (Laikre et al., 2010). Admixture between the domestic and wild quail populations in some parts of the world has been observed where wild quails were found to be hybrids of the domestic Japanese quail (C. Japonica) and the wild common (C. Coturnix) quail (Amaral et al., 2007;Barilani et al., 2005;Chazara et al., 2010;Sanchez-Donoso et al., 2012).
The wild African harlequin quail is endemic to Africa, but its population genetics and demographic history are largely unknown. Information on its genetic integrity is critical to its utilization and conservation. Illegal and unsustainable harvesting, coupled with the change in land use, habitat fragmentation, and the introduction of exotic breeds, are factors that might impact wild African harlequin quail genetics. Here, mitochondrial and genome-wide variation patterns were used to assess wild African harlequin quail diversity and demographic history. In addition, tests were conducted to detect the presence of any wild African harlequin quail individuals admixed with the domestic Japanese quail.

| Sample collection and DNA extraction
One hundred quails were sampled for this study; 78 captured wild African harlequin quails (Siaya North = 24, Siaya Central = 28, Siaya South = 26) and 22 domestic Japanese quails from Siaya and Kajiado Counties of Kenya, respectively ( Figure 1). This comprised an equal number of male and female individuals for every visited farm. Wild African harlequin quails were captured around homesteads using traditional methods described by Wamuyu et al. (2017). This involves using previously captured quails that are placed in woven baskets then tied to long poles as bait to lure other quails which are captured by traps set on the ground (Figure 2). Siaya County was selected due to the significant seasonal presence of wild African harlequin quails in the region, compared to others, and their interaction with rural small-scale farmers. It lies in the Lake Victoria basin at an altitude of 1,134 meters above sea level, receiving an annual rainfall of 500-1000 mm with temperatures ranging 17.1-29.4°C (Moraa et al., 2015). In contrast, Kajiado County is a peri-urban area where purebred domestic Japanese quails are more commonly reared under intensive (caged) production system. Whole blood was drawn from the wing vein into 2-ml cryotubes containing 70% alcohol and preserved at −80°C. Genomic DNA was extracted from the whole blood samples using the phenol-chloroform method (Sambrook & Russell, 2006).

| Mitochondrial DNA amplification and sequencing
A 760bp mitochondrial DNA control region fragment was amplified via polymerase chain reaction (PCR) using primers AV1F2 and CR1b (Mwacharo et al., 2011) in 56 wild African harlequin and 17 Japanese quails. The amplicons were then sequenced using Applied Biosystems 3730xl DNA Analyzer (Life Technologies) following the manufacturer's protocol. Forward and reverse sequences of each sample were edited using SeqMan Pro version 7.1.0 (DNASTAR, Inc.) and used to generate a consensus sequence. Sequence alignment was done using Clustal X version 2.1 (Thompson et al., 1997) and MUSCLE version 3.6 (Edgar, 2004). Highly conserved regions of the sequences were trimmed, and a variable 347bp fragment was retained for further analysis.

| GBS library preparation and sequencing
A GBS predesign experiment approach was used to evaluate the enzymes and sizes of restriction fragments using training data. The following criteria were applied: 1. the number of tags must be suitable for the specific needs of the research project; 2. the enzymatic tags must be evenly distributed through the sequences to be examined; 3. repeated tags should be avoided.
These considerations improved the effectiveness of GBS for the dataset at hand. A strict length range was selected to maintain the sequence depth uniformity of different fragments (~50 bp).
We then constructed the GBS library in accordance using the predesigned scheme. Genomic DNA was incubated at 37℃ with MseI Polymerase chain reaction (PCR) was conducted using the purified samples, the Phusion Master Mix (NEB, United States) universal and index primers, and i5 and i7 sequences. The PCR products were purified using Agencourt AMPure XP, pooled, and electrophoresed on a 2% agarose gel. A Gel Extraction Kit (Qiagen, Germany) was used to isolate 375-400 basepair fragments (with indexes and adaptors).
These fragments were purified using Agencourt AMPure XP, and the resulting products were diluted for sequencing. Then, paired-end sequencing of the selected tags was performed using the Illumina NovaSeq high-throughput sequencing platform at the Novogene Bioinformatics Technology Company, China.

| Sequence data processing and SNP calling
Raw reads were processed through a series of quality control procedures which involved removing reads with ≥10% unidentified nucleotides (N), with >50% bases with a Phred quality <5, with >10 F I G U R E 1 Sampling locations in Kenya nucleotides alignment to the adapter, with >10% mismatches, and that contain the HaeIII enzyme sequence.
Burrows-Wheeler Aligner (BWA) version 0.7.17 was used to align the retained reads against the Coturnix Japonica 2.0 genome (Assembly accession number GCF_001577835.1) with the parameters "mem -t 4 -k 32 -M" ). Variant calling was performed using SAMtools version 1.11 mpileup command ) in conjunction with BCFtools version 1.11 call command (Li, 2011). Variant filtering was performed by restricting the dataset to biallelic SNPs found in at least 80% of samples, with a minimum depth of 2 reads, minimum Phred score of 30, and minimum minor allele frequency (MAF) of 0.05 using VCFtools version 0.1.13 (Danecek et al., 2011).

| Genetic diversity and population structure
Mitochondrial genome haplotype diversity was assessed using DnaSP version 5.10.1 (Librado & Rozas, 2009). The phylogenetic relationship between the study mtDNA haplotypes and GenBank reference mtDNA haplotypes (C. japonica and C. coturnix) was estimated using a median-joining network and a maximum likelihood tree generated with network version 5.0 (Bandelt et al., 1999) and

| Demographic analysis
MtDNA neutrality tests (Tajima's D and Fu's Fs) and mismatch analysis (sum of squares deviation (SSD), raggedness index (rag.), theta0 (θ 0 ), theta1 (θ 1 ), and Tau (τ)) were calculated using Arlequin version 3.5.2.2 (Excoffier & Lischer, 2010) to check for signals of demographic expansion through investigating mismatch distributions of pairwise differences among mtDNA haplotypes. Time since population expansion (t) was calculated using the equation τ = 2 μLt, where τ is the tau value, μ is the substitution rate, and L is the length of the sequence (Aoki et al., 2018). Like most other avian species, the mitochondrial genome mutation rate of wild African harlequin quail or other quail species, especially for the control region, is unknown. Therefore, the commonly used mitochondrial genome mutation rate of 2% per million years (Myr) was applied in this study (Nabholz et al., 2016). The divergence times between African harlequin quails and related quail species such as the Japanese and common quails were estimated using the TimeTree database (Kumar et al., 2017).
To examine the recent demographic history of the wild African harlequin quail using the SNP dataset, all populations were combined, and easySFS (Gutenkunst et al., 2009) was used to generate folded site frequency spectrum (SFS) information. The historical population size changes were then estimated using Stairway plot version 2.1 (Liu & Fu, 2015). The stairway plot has proved applicable to low-depth sequencing and reduced-representation sequence data (Liu & Fu, 2020). The mutation rate of the wild African harlequin quail is unknown, but since the mean mutation rate of most avian genomes ranges between 1.23 and 2.21 × 10 −9 site −1 year −1 , an intermediate rate observed in chicken (1.91e-9) and a generation time of 1 year was assumed (Nam et al., 2010).

| SNP identification and MtDNA haplotype analyses
More than 95% of the wild African harlequin quail sequenced reads in all individuals were mapped to the Japanese quail reference genome, thereby ascertaining confidence in the SNP identification process. A total of 8,237,001 variants were called in 96 individuals, and after variant filtering, 119,339 SNPs were retained for further analysis.
The 56 mtDNA control region sequences of the wild African harlequin quails revealed 32 haplotypes in total (haplotype 1-32), which translated to a haplotype diversity of 0.949±0.016, whereas only one haplotype (haplotype 33 farm-reared) was observed among the 17 domestic Japanese quails mtDNA control region sequences.
These generated haplotypes were then used in subsequent analysis.

| Genetic diversity and population structure analysis
Out of the 32 wild African harlequin quail mtDNA haplotypes observed, haplotype 1 was the most common and present in 9 individuals. In contrast, the majority of the haplotypes were only present in one individual each (singletons). None of the other reference quail mtDNA haplotypes (Table S1) clustered together with the wild African harlequin mtDNA haplotypes; therefore, no haplotype sharing was observed ( Figure 3). The median-joining network of our study haplotypes together with reference quail mtDNA haplotypes (C. japonica and C. coturnix) from GenBank formed three haplogroups ( Figure 4). All the wild African harlequin quail mtDNA haplotypes (yellow) clustered under haplogroup 1, whereas haplogroup 2 contained the single domestic Japanese quail mtDNA haplotype from this study (blue), reference Japanese quail mtDNA haplotype that was similar to the Japanese quail mtDNA haplotype of this study (orange) and other reference Japanese quail mtDNA haplotypes (purple) from GenBank (Table S2). Haplogroup 3 contrarywise contained reference common quail mtDNA haplotypes (green). The king quail (C. chinensis) mtDNA haplotype was the chosen outgroup.
Principal component analysis showed no signs of population structuring in wild African harlequin quails of Siaya County as most individuals were grouped in a single cluster ( Figure 5). The first and second eigenvectors (EV) accounted for 6.6% and 2.1% of the total observed variation, respectively. This observation was also supported by a low average pairwise F ST value of 0.000075, signifying slight genetic variation among the three sampled populations.
Overall, the wild African harlequin and domestic Japanese quails were separated into two distinct clusters.
Admixture results were similar to PCA as the wild African harlequin quails were separated from domestic Japanese quails at K = 2, which had the least cross-validation error ( Figure S1). However, one wild African harlequin quail sample exhibited low levels of shared genetic ancestry with Japanese quail at K = 2, an observation that was not supported by other K values ( Figure 6).

| Demographic history
The estimated mean Tajima However, genetic information on the wild African harlequin quail is lacking. In this study, we managed to assess the genetic diversity and demographic history of wild African harlequin quail populations of Siaya County using mtDNA d-loop data and SNP information generated through genotyping by sequencing approach.
There has been an uncontrolled introduction of domestic Japanese quail breeds into Kenya and a noticeable population size reduction of wild African harlequin quail numbers in parts of Western Kenya over the years. There have also been suspected cases of domestic Japanese quail release into the wild but so far, no credible evidence has been provided yet. Before any reasonable management practices are adopted to allow for proper utilization of the wild African harlequin quail as a resource and increase the effectiveness of conservation activities, more information with respect to the genetic diversity, demographic history, and population structure is required.

| Genetic diversity and population structure
The seasonal migration of wild birds such as the wild African harlequin quails affects their population size and structure (Newton, 2006 (Jetz et al., 2012;Seabrook-Davison et al., 2009;Stein et al., 2015).
The mtDNA haplotype and phylogenetic relationship analyses were also able to show the genetic variation among the wild African harlequin quail individuals, their similar ancestry, and their distinction from other quail species. In contrast, only a single mtDNA haplotype was detected among sampled Japanese quail individuals, which was consistent with low genetic diversity within this domesticated population. MtDNA haplotype sharing between wild and domestic quails has previously been used to detect hybridization (Barilani et al., 2005). However, there was no mtDNA haplotype sharing between wild African harlequin quails of Siaya County and Japanese quail individuals, separating the two species into different haplogroups.
Even though no mtDNA haplotype sharing was observed, the wild Detecting hybridization is crucial as it has been linked to reduced or absent migratory behavior in wild quail populations (Barilani et al., 2005). Currently, there is no reason to suspect the artificial release of domestic Japanese quail breeds into the wild as was the concern.

| Demographic history
The demographic history of the wild African harlequin quails of Siaya was outlined on a stairway plot from 560,000 years ago. The population bottleneck event that was observed ca. 180 ka is consistent with climate change during the Pliocene-Pleistocene boundary where climate oscillations resulted in range contraction of species during dry arid climate periods (deMenocal, 1995;Hewitt, 2004). The observed bottleneck is also consistent with the emergence of Homo sapiens in East Africa some 200,000 years ago (McDougall et al., 2005;White et al., 2003) which was associated with updated and improved technology (Bradfield et al., 2020;Hallett et al., 2021). African hominins were known for hunting terrestrial prey and the development of advanced tools facilitated the incorporation of animal protein into their diets (Faith, 2014). However, the time period when early hominins became skilled hunters is still debatable (Ferraro et al., 2013;Marean & Assefa, 1999).
Both mtDNA analysis and the stairway plot suggested that wild African harlequin quail populations of Siaya experienced a population expansion soon after the bottleneck. The negative mtDNA Tajima's D and Fu's Fs test values, and the demographic expansion estimates all signaled a large-scale demographic expansion of the wild African harlequin quail populations of Siaya County. The unimodal mismatch distribution that was also detected when the observed nucleotidesite differences between pairs of mtDNA haplotypes were computed is characteristic of a population that has experienced demographic expansion (Rogers & Henry, 1992). This could be through a range expansion with high levels of migration between neighboring demes that is commonly observed in wild quails that migrate seasonally (Excoffier, 2004;Ray et al., 2003). In addition, the demographic expansion parameter τ and stairway plot analysis were both used to infer time since population expansion and that estimate was placed around the quaternary glaciation period. Glacial changes during this period promoted the general cooling of African climates, and East African vegetation shifted from closed canopy to open savannah and grassland vegetation starting in the mid-Pliocene (deMenocal, 1995(deMenocal, , 2004. Grasslands, bushy areas, and croplands serve as habitats for many wild quail species, including the African harlequin quail, and this may explain their expansion during that period. In East Africa, vegetation changes and the development of periodical cooler and drier climatic conditions after 2.8 Ma greatly influenced species distribution and genetic isolation that further intensified after 1.7 and 1.0 Ma (deMenocal, 1995). Oscillations between wetter and drier climates have been found to repeatedly promote cycles of range contraction and expansion (Lorenzen et al., 2010). This may also explain the seasonal presence of wild African harlequin quails following the long and short rainy seasons (Wamuyu et al., 2017). Their sudden availability and increased numbers are also encouraged by food availability, especially after the rainy season, where grain farms create suitable habitats.
The effective population size of wild African harlequin quails of Siaya has been declining over the last 20,000 years, which coincides with climate change in East Africa, the massive late Pleistocene changes in range and abundance of smaller vertebrates (Steele, 2007) and the extinction of megafauna (50 ka −10 ka) (Van Der Kaars et al., 2017;Stuart, 1999). Climate change during the last glacial maximum (~20 ka) has been linked to a decline in the effective population size of several species in Southern and tropical Africa (Sithaldeen et al., 2015). Changes in temperature and precipitation were found to have a profound effect on food supply, reproduction, and the migration patterns of birds (Carey, 2009). In addition, this decline began after hydroclimatic changes that favored Homo sapiens demographic expansion and dispersal to other parts of the world approximately 70-60 ka (Rito et al., 2019;Schaebitz et al., 2021). It is hypothesized that during this period, human hunters were responsible for the massive extinction events and habitat destruction as they spread from Africa to other continents (Faith, 2014;Martin, 1967). However, the role played by humans in the late Pleistocene extinction events is still debated (Nagaoka et al., 2018).
Over the past 200,000 years, human occupation in East Africa has contributed to regional vegetation and habitat change (Schaebitz et al., 2021). This change was further facilitated by the transition from hunting and gathering to agriculture where rainforests and woody vegetation in the savanna were slowly converted into farms and pasturelands (Patin et al., 2014). Wild quails survive away from protected areas and end up in farms and pasturelands, where they are commonly hunted, thus also contributing to their population changes over time (Gatesire et al., 2014). Traditional hunting of wild African harlequin quails involves using captured females to attract males (Wamuyu et al., 2017). This leads to the capture of more males than females, thereby affecting the breeding sex ratio, which impacts the effective population size. This could also explain the dis-

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
Stephen