Abstract

This study characterised mitochondrial genetic diversity in Vietnamese Ven dogs using a 582-bp fragment of the HV1 region from 21 individuals. A total of 32 polymorphic sites was identified, and sequence variation resolved 14 haplotypes assigned to three major haplogroups (A, B, and C). Haplogroup A was predominant, while haplotype C2 was the most frequent (19.06%). One novel haplotype (Cn) was detected at low frequency (4.67%). Genetic diversity indices indicate substantial maternal variation within the sampled population (Hd = 0.943; Pi = 0.01524), consistent with patterns reported in other Vietnamese and Southeast Asian village dogs. The haplotype network further supports the coexistence of both closely related and moderately divergent maternal lineages, although demographic inference remains limited by sample size. Overall, the findings suggest that Vện dogs represent a phenotypically defined subset embedded within a diverse regional gene pool rather than a distinct maternal lineage. This study provides a baseline HV1 dataset for Vện dogs and highlights the need for expanded sampling and genome-wide analyses to better resolve population structure, evolutionary history, and implications for conservation and breeding strategies.

Keywords:

Ven dog, HV1, D-loop, haplotype, haplogroup, genetic diversity, mitochondrial DNA

Highlights

1.0 Introduction

Village dogs – locally referred to as “Muc,” “Ven,” and “Co”- represent long-established indigenous canine populations of Vietnam, distinct from recently introduced cosmopolitan breeds. Archaeological and ethnobiological evidence suggests that these dogs have been embedded within human settlements across the Indochinese peninsula for at least ~6,000 years, reflecting a prolonged history of cohabitation, semi-managed breeding, and ecological adaptation (Corbett, 1995). Phenotypically, Vietnamese village dogs are typically medium-sized (10–25 kg; 45–65 cm at the withers), with short coats exhibiting extensive color polymorphism, including solid and patterned forms such as brindle. This phenotypic heterogeneity likely mirrors underlying genetic diversity shaped by weak artificial selection and strong local environmental filtering. Within this heterogeneous population, the brindled phenotype – commonly referred to as the “Ven” dog – constitutes a conspicuous and geographically widespread ecotype. While “Vện” is defined primarily by coat pattern rather than strict breed status, it represents a biologically meaningful subset for population genetic inference, as phenotypic clustering in village dogs may partially track lineage structure, historical gene flow, or localized selection regimes. Consequently, Ven dogs provide an informative model for interrogating the evolutionary history and standing genetic variation of Southeast Asian village dog populations.

To date, research on Ven dogs has been dominated by phenotypic and production-oriented descriptors, including morphology, physiology, and basic reproductive traits. These studies consistently depict Ven dogs as small-to-medium-sized, resilient, and well adapted to low-input management systems, with behavioral traits suited for guarding and hunting (Trieu et al., 2018, 2019, 2020). Population-level surveys further reveal substantial within-group variability in coat pattern, ear morphology, and minor anatomical traits such as dewclaw number and tongue pigmentation. While informative, such datasets remain largely descriptive and offer limited resolution for reconstructing population history or inferring evolutionary processes. At the molecular level, preliminary candidate gene analyses have identified polymorphisms of potential functional relevance, including adjacent SNPs in the HTR1D gene that may influence amino acid composition and, by extension, behavioral phenotypes such as aggression (Trieu et al., 2020). However, these locus-specific insights do not capture genome-wide or matrilineal patterns of diversity, which are critical for understanding demographic history, maternal lineage structure, and phylogeographic relationships.

In this context, mitochondrial DNA (mtDNA), particularly the hypervariable region I (HVI) of the control region, has emerged as a powerful marker for resolving maternal genetic structure in domestic dogs. Due to its high mutation rate, lack of recombination, and maternal inheritance, mtDNA HVI enables fine-scale discrimination of haplotypes and facilitates inference of population expansion, migration, and lineage diversification. Previous studies in Vietnamese dogs – including Phu Quoc, H’Mong, and populations from the Ma River basin and southern urban regions – have consistently revealed substantial haplotype diversity and the presence of globally distributed haplogroups (A, B, C), alongside the sporadic occurrence of the rarer haplogroup E (Quan et al., 2016a,b; Tran et al., 2016; Nguyen et al., 2019; Hải et al., 2021; Bùi et al., 2021). These findings position Vietnam within the broader Southeast Asian domestication landscape, a region increasingly recognized as a major reservoir of early canine genetic diversity.

Despite these advances, the maternal genetic architecture of Ven dogs remains poorly characterised. Specifically, it is unclear whether this phenotypically defined group represents a random subset of the broader village dog gene pool or retains distinct haplotypic signatures indicative of structured ancestry, localized selection, or demographic isolation. Addressing this gap requires a shift from descriptive phenotyping to sequence-based population genetic analysis. The present study therefore aims to characterize the genetic diversity of Vietnamese Ven dogs using mtDNA HVI sequences, with an emphasis on haplotype diversity, nucleotide variation, and phylogenetic affiliation. By integrating newly generated sequence data with published datasets from other Vietnamese dog populations, we seek to (i) resolve the maternal lineage composition of Ven dogs, (ii) assess their position within regional haplogroup structure, and (iii) provide a foundational dataset for subsequent genome-wide and coalescent-based analyses of village dog evolution in Vietnam.

2.0 Materials and methods

2.1. Sample collection and total DNA extraction
A total of 21 putatively unrelated Vietnamese Vện dogs were sampled across multiple provinces of the Mekong Delta, including Ca Mau (n = 4), Sóc Trang (n = 3), Hau Giang (n = 3), An Giang (n = 4), Can Tha (n = 3), and Vinh Long (n = 4) (Trieu, 2024). Unrelatedness was approximated through owner interviews to minimize sampling of closely related individuals, a common constraint in village dog populations lacking formal pedigree records. The sampled dogs were distributed across communities representing diverse ethnic backgrounds (Kinh, Khmer, and Hoa), thereby capturing a degree of socio-ecological heterogeneity that may influence gene flow and breeding structure.

All individuals were phenotypically classified as Ven dogs based on the presence of brindle coat patterns, typically expressed as combinations of black, yellow, and white pigmentation. Each sample was assigned a unique identifier (ST1–ST21). To reduce the inclusion of recently introgressed or non-native lineages, a morphological screening protocol was applied during sampling, excluding individuals exhibiting diagnostic traits of recognized foreign breeds (Figure 1). While morphology-based filtering cannot fully eliminate cryptic admixture, it provides a pragmatic first-pass control in field-based population genetic studies of free-breeding dogs.

Genomic DNA was extracted from hair samples using a standardized protocol adapted from Quan et al., (2016c) Quan et al. (2016c), optimized for low-input keratinized tissues. DNA quality and concentration were assessed via spectrophotometric analysis (NanoDrop™ One/OneC, Thermo Scientific™), ensuring adequate purity for downstream applications. All DNA extracts were subsequently normalized to a working concentration of 20 ng/µL to standardize template input for amplification of the mitochondrial hypervariable region I (HVI).

Figure 1. Some primary dog coat colours with the brindle phenotype (top) and characteristics of the brindle ridge dog (bottom)

2.2. Sequencing
The HV1 region gene sequence (582 bp) was amplified by PCR using a specific primer pair (15412F: CCACTATCAGCACCCAAAG, and 16625R: AGACTACGAGACCAAATGC) (Gundry et al., 2007). This locus was selected due to its elevated mutation rate and established utility in resolving fine-scale maternal lineage structure in domestic dogs. PCR amplifications were performed in a total reaction volume of 25 µL, containing 1× MyTaq™ Red Mix (Bioline), 0.2 µM of each primer, approximately 20 ng of template DNA, and nuclease-free water. Thermal cycling was conducted under the following conditions: an initial denaturation at 95°C for 5 min; followed by 35 cycles of denaturation at 95°C for 1 min, annealing at 52°C for 30 s, and extension at 72°C for 1 min; with a final extension at 72°C for 5 min. These parameters were optimized to ensure robust amplification of mitochondrial templates while minimizing nonspecific products. Amplification success was verified by electrophoresis on 1.5% agarose gels stained with GelRed®, with fragment size assessed against a standard DNA ladder. PCR products exhibiting single, clear bands of the expected size were purified to remove residual primers and dNTPs prior to sequencing. Purified amplicons were then subjected to bidirectional Sanger sequencing at a commercial facility (1st BASE, Malaysia). This targeted sequencing approach enables high-confidence base calling across the HVI region, providing the resolution necessary for downstream haplotype identification, polymorphism detection, and phylogenetic inference of maternal lineages.

2.3. Data analysis
The chromatograms obtained from sequencing were analyzed using FinchTV 1.4.0. Any inconsistencies observed between the forward and reverse sequences were resolved manually. Subsequently, all sequences were aligned with a reference sequence (GenBank no: U96639.2) using MEGA 11 (Tamura et al., 2021) and trimmed to generate sequences of 582 bp. Nucleotide positions were annotated according to Pereira’s scheme (Pereira et al., 2004). These 582 bp sequences were then subjected to Haplotype Identifier for haplotyping (Thai et al., 2017). Any sequence exhibiting new mutations was designated with the suffix “n” to indicate novelty, for instance, A1n, C2n. For a comprehensive understanding of haplotype relationships, a haplotype medium-joining network was constructed using Network 10.2.0.0 (Bandelt et al., 1999) and DnaSP version 6.11 (Rozas et al., 2017). Genetic diversity indices such as haplotype diversity (Hd), nucleotide diversity (Pi), and average number of distinct nucleotides (K) were computed using Arlequin (Excoffier et al., 2010).