Genome‐wide variant analyses reveal new patterns of admixture and population structure in Australian dingoes

Admixture between species is a cause for concern in wildlife management. Canids are particularly vulnerable to interspecific hybridisation, and genetic admixture has shaped their evolutionary history. Microsatellite DNA testing, relying on a small number of genetic markers and geographically restricted reference populations, has identified extensive domestic dog admixture in Australian dingoes and driven conservation management policy. But there exists a concern that geographic variation in dingo genotypes could confound ancestry analyses that use a small number of genetic markers. Here, we apply genome‐wide single‐nucleotide polymorphism (SNP) genotyping to a set of 402 wild and captive dingoes collected from across Australia and then carry out comparisons to domestic dogs. We then perform ancestry modelling and biogeographic analyses to characterise population structure in dingoes and investigate the extent of admixture between dingoes and dogs in different regions of the continent. We show that there are at least five distinct dingo populations across Australia. We observed limited evidence of dog admixture in wild dingoes. Our work challenges previous reports regarding the occurrence and extent of dog admixture in dingoes, as our ancestry analyses show that previous assessments severely overestimate the degree of domestic dog admixture in dingo populations, particularly in south‐eastern Australia. These findings strongly support the use of genome‐wide SNP genotyping as a refined method for wildlife managers and policymakers to assess and inform dingo management policy and legislation moving forwards.

Hybridisation and genetic admixture are of growing concern to conservation scientists particularly as hybridisation occurs more frequently between species or populations experiencing climate change, urbanisation and/or culling (Bohling & Waits, 2015;Canestrelli et al., 2017;Rutledge et al., 2012). Hybridisation between species can lead to a hybrid swarm or unimodal hybrid zone when first-generation hybrids are fertile and there are no barriers to reproduction or mating preferences (Allendorf et al., 2001;Fitzpatrick et al., 2015). Hybrid swarms consist of predominately intermediate hybrids rather than backcross progeny, or pure parental populations.
When hybridisation occurs, but there are pre-mating barriers or assortative mating preferences, a bimodal hybrid zone may be formed, characterised by a population where parental forms predominate and intermediate hybrids are rare (Jiggins & Mallet, 2000;McFarlane & Pemberton, 2019). Identifying both the occurrences and repercussions of admixture can inform wildlife management and conservation policies for at-risk or keystone species (Allendorf et al., 2001;Fitzpatrick et al., 2015;vonHoldt et al., 2018).
To accurately detect hybridisation and subsequent admixture, the genetic identity and diversity of each contributing species must be well described (McFarlane & Pemberton, 2019). Canis species (ssp.) are particularly vulnerable to hybridisation, particularly with domestic dogs, due to conserved genetic compatibility and reproductive behaviours, overlapping habitat and the population density of domestic dogs (Bohling, 2016;Bohling & Waits, 2015;Lord et al., 2013;Macdonald et al., 2019;vonHoldt & Aardema, 2020).
Indeed, hybridisation with free-ranging domestic dogs is a threat to wild canid populations in many geographic regions (Doherty et al., 2017;Hughes & Macdonald, 2013;Rahaman, 2017;Sykes et al., 2020). In this study, we sought to understand population structure and admixture patterns in Australian wild canids using genome-wide genetic markers. We hypothesise that undescribed population structure has biased previous admixture assessments in Australian wild canids, yielding significant implications for management and conservation .
The Canis (spp.) present in Australia includes dingoes, as well as pet and feral domestic dogs. Dingoes are an early lineage of dog, divergent from both modern domestic dogs and the free-ranging domestic dogs observed in Asia (Bergström et al., 2020;Cairns, 2021;Field et al., 2022;Surbakti et al., 2020). Phylogeny and ancient DNA studies estimate that the dingo lineage diverged from other early dogs 8000-11,000 years before present (YBP), prior to the rise of human agriculture and intensive artificial selection (Bergström et al., 2020;Cairns & Wilton, 2016;Zhang et al., 2020).
The management, conservation, nomenclature and taxonomy of dingoes are controversial, in large part due to a lack of information regarding a uniform definition of dingoes and the extent of dingo × dog hybridisation (Cairns, 2021;Crowther et al., 2014;Jackson et al., 2017;Smith et al., 2019). At a basic level, there are still key knowledge gaps concerning the genomic, morphological, physiological and ecological differences between dingoes and domestic dogs (Field et al., 2022;Smith et al., 2019;Eeden et al., 2018).
In Australia, the term wild dog is often used to refer to dingoes, free-ranging domestic dogs and their hybrids (Crowther et al., 2014;Fleming et al., 2001). We hereafter use the term dingo to refer to pure dingoes; dingo × dog hybrids to refer to the first-generation offspring of a dingo and dog; and domestic dog to refer to free-ranging or pet domestic dogs. We use the terms dingo backcross and dog backcross to refer to admixed animals with predominately dingo or domestic dog ancestry. Conservation policies aim to keep dingoes free of domestic dog admixture and minimise the impact of dingoes on livestock producers. While the dingo is legislatively classified as native wildlife, they are only protected within National Parks of some states and there are efforts to suppress or eradicate their populations across most of the Australian mainland (Bulte et al., 2003;Corbett, 2001;Fleming et al., 2001;Philip, 2019;Smith, 2015).
Contemporary and historical domestic dog admixture, particularly in south-eastern Australia, has raised significant concern that dingoes will become extinct through genetic dilution Claridge et al., 2014;Major, 2009;Stephens et al., 2015). The broad geographic range of dingoes, historical time period during which they and dogs have been in contact and intrusive management practices that may increase dog × dingo hybridisation are all factors thought to affect the genetic integrity of dingo populations (Cairns, 2021;Stephens et al., 2015).
Historically, skull morphology and pelage were used to discriminate dingoes from dogs and identify dingo × dog hybrids (Newsome et al., 1980;Newsome & Corbett, 1985). However in 1999, this changed with the first application of DNA-based testing (Wilton et al., 1999), and a major continental survey (n ≥ 3000) of dingo × dog hybridisation across Australia was carried out using a microsatellite assay comprised of 23 markers (Stephens et al., 2015). Stephens et al. (2015) concluded that in south-eastern Australia, ≤1% of the wild canid population were pure dingoes (Stephens et al., 2015).
However, subsequent microsatellite DNA surveys, incorporating the data set of Stephens et al. (2015) and an additional 1364 samples from south-eastern Australia, found that most dingoes in northern, western and central Australia were comparatively pure (68%-98%) while in south-eastern Australia, only 18%-41% of animals were pure dingoes . Importantly, both Stephens et al. (2015) and  showed that the prevalence of free-ranging domestic dogs in Australia was very low with less than 1% of samples identified as feral dogs.
The reported low occurrence of pure dingoes in southern and eastern Australia has been a key factor driving government policy and practice (Allen et al., 2017;Bird & Bowman, 2016;DEPI, 2013;Major, 2009;NWDAP, 2020). Government policy and communications widely use the term 'wild dog' rather than dingo to emphasise the mixed ancestry of wild dingo populations and the presence of feral dogs (Letnic et al., 2012). Many Australians are unaware that dingoes are referred to as 'wild dogs' and as such targeted for killing during pest management programmes on public lands (Eeden et al., 2018. The purportedly low occurrence of pure dingoes in south-eastern Australia led to Victoria listing dingoes as a threatened species, and New South Wales listing 'hybridisation by feral domestic dogs' as a key threatening process to dingoes. However, those and similar decisions were made based on limited resolution microsatellite DNA testing. More advanced approaches using higher density genomic data have been used extensively to address similar questions in other species, but had not been applied to the problem of hybridisation between dingoes and dogs (McFarlane & Pemberton, 2019;Oliveira et al., 2015;Pilot et al., 2021;Stroupe et al., 2022;Szatmári et al., 2021), as we do here.
Assessments of dingo × dog hybridisation are confounded by a reliance on (1) predefined and geographically restricted reference populations and (2) use of low numbers of variable markers. Recent studies utilising predominately Y chromosome or mitochondrial markers and small sample sizes (n < 130) observed a broad pattern of biogeographic subdivision between dingoes in south-eastern Australia from those in the rest of Australia (Cairns et al., 2017(Cairns et al., , 2018Cairns & Wilton, 2016;Greig et al., 2018;Zhang et al., 2020). Similar patterns of west to east subdivision have been observed in dingoes using three-dimensional cranial morphology (Koungoulos, 2020).
However, persistent sampling gaps and extensive variation in dingo pelage, size and body shape suggest that additional uncharacterised biogeographic subdivisions exist.
In this study, we sought to better characterise population structure in Australian dingoes, as well as examine admixture between the dingo and domestic dog using genome-wide genotyping approaches. Using over 195,474 SNPs on a commercial microarray, we provide compelling evidence that domestic dog admixture is uncommon in contemporary dingo populations. We further demonstrate that SNP-based technologies are a comparatively cheaper and more rigorous method for determining the composition of dingo populations across Australia. In doing so, we provide the first detailed study of geographic population structure in dingoes across all regions of Australia.

| Canid sampling
We collected tissue, blood or buccal samples from 307 wild and 84 captive dingoes from locations across Australia (Figure 1, Dryad dataset: [https://doi.org/10.5061/dryad.vq83b k3ww]). A majority of the wild dingoes were sampled as part of programmes to suppress dingo populations. Additional samples were provided by private landholders, government officers, found as roadkill or during carcass disposal or sampled from puppies found in the wild. Unrelated captive dingo samples were contributed by dingo sanctuaries, zoos and private citizens. We included captive dingo samples to serve as controls in a comparison to wild dingoes, due to their known parentage. A subset of the wild and captive samples were from existing scientific collections and had been previously tested using a 23-marker microsatellite-based test developed for distinguishing dingoes from dingo × dog hybrids Stephens et al., 2015). We also collected blood or buccal samples from 36 Australian domestic pet dogs and 111 North American domestic pet dogs. We chose dog samples from working, herding or mixed breeds as they represent the populations most likely to have hybridised with Australian dingoes. A set of five unrelated Vietnamese putative primitive breed samples were sourced from a private owner in Hồ Chí Minh City, Vietnam. DNA was extracted from blood, tissue or buccal samples using either (1) Qiagen DNeasy Blood and Tissue (Qiagen Sciences) or (2) Monarch Genomic DNA Purification kits (New England Biolabs) or (3) a modified proteinase K, phenol-chloroform extraction method (Bell et al., 1981). quality control (QC) filtering was conducted using Plink v1.9 (Purcell et al., 2007). We trimmed the North American domestic dog data to include only the SNPs from the Axiom Canine HD Genotyping array (Thermo Fisher Scientific Inc.), as the Axiom Canine Genotyping Set A and Set B arrays (Thermo Fisher Scientific Inc.) contain probes for a larger set of SNP markers and were used in the development of the Axiom Canine HD Genotyping array (Thermo Fisher Scientific Inc.).

| Axiom canine genotyping
We excluded individuals or sites with ≥10% missing data using the commands --geno 0.1 and --missing. We did not perform minor allele frequency filtering because it has a negligible impact on ascertainment bias and may negatively impact genetic diversity or population structure assessments. Instead, we employed linkage disequilibrium (LD) filtering (Malomane et al., 2018), by removing regions in high LD using the command --indep-pairwise 50 5 0.6 in Plink v1.9.

| Dingo × dog ancestry and admixture analysis
We performed genetic ancestry analyses on the complete data set of 307 wild dingoes, 84 captive dingoes and 152 domestic breed dogs using FastStructure (Raj et al., 2014). We employed FastStructure, a Bayesian population clustering method similar to STRUCTURE (Pritchard et al., 2000), due to its capacity to handle large genomic and simple prior. A simple prior was suitable for our study because we are primarily interested in strong population structure between dingoes and domestic dogs rather than substructure between domestic dog breeds. FastStructure reports the optimal K values to maximise marginal likelihood and explain structure in the data (Raj et al., 2014).
Ten replicates were completed for each modelling scenario that best fit the data (K = 8-10), calculating the average proportion of dingo ancestry for each individual along with the standard error.  FastStructure ancestry component q-values were used to categorise each animal as a dingo, historical or recent dingo backcross, dingo × dog hybrid, recent or historical dog backcross or domestic dog (Table 1). Rather than using a strict threshold, we assigned animals to a specific category using the calculated 95% confidence intervals (McFarlane & Pemberton, 2019). We considered that the occurrence of different backcross classes in wild dingo populations may inform conservation and management policy directions, with some previous studies considering animals with >93% dingo ancestry as pure (Allen et al., 2017). Therefore, we categorised backcrosses as being either: (1) historical backcrosses (>93% parental ancestry) that were likely to be third generation or higher or (2) recent backcrosses (55%-93% parental ancestry) which described animals that were likely to have had a dingo × dog hybrid ancestor in the last one to three generations (Dziech, 2021;Harmoinen et al., 2021;McFarlane & Pemberton, 2019).
A map depicting the FastStructure ancestry assignment of each wild origin sample was created using the QGIS v3.12 (QGIS, 2020).
Principal coordinates analyses (PCoA) were carried out using the R package SambaR function 'find_structure' . In the SambaR analyses, populations were defined according to the FastStructure results with ancestry categorised as in Table 1. However, Vietnamese dogs were allocated to their own category, given that dogs from South Asia and dingoes may share ancestry (Cairns, 2021). F ST values for animals categorised as dingoes, (historical and recent) dingo backcrosses, dingo × dog hybrids, (historical and recent) dog backcrosses, domestic dogs and Vietnamese dogs were calculated in SambaR using the 'calcdistance' function.

| Dingo population structure and biogeography
To explore patterns of biogeographic variation in dingoes across Australia, the SambaR 'find_structure' function was used to run PCoA using only animals with an ancestry category of dingo or historical dingo backcross, excluding all animals with greater than 7% ancestry deriving from domestic dogs. Analyses were run with and without dingoes of captive origin. A map of wild origin dingo samples and their primary FastStructure population cluster ancestry identity was plotted using QGIS v3.12 (QGIS, 2020).
Inbreeding coefficients (F het ) were calculated in Plink v 1.9 using the command --het based on the equation O is the number of observed homozygous markers of the individual, E is the expected number of homozygous markers under the Hardy-Weinberg equilibrium calculated from the allele frequencies estimated on the dingo data set and N is the total number of markers (Gazal et al., 2014;Purcell et al., 2007).

| Comparisons between microsatellite and SNP ancestry estimates
We compared the modelled ancestry q-values for a set of 113 animals that were tested using both the 195,434-marker Axiom SNP array method and the 23-marker microsatellite method of  and Stephens et al. (2015). We calculated the difference between our modelled SNP ancestry q-values and previously reported values based on microsatellite DNA testing Stephens et al., 2015). We then calculated the average difference between SNP and microsatellite q-values for individuals within the five dingo subpopulations. We also assessed the number of individuals which were categorised by microsatellite DNA testing as recent backcrosses or dingo × dog hybrids, but were pure or historical backcrosses based on SNP analysis. TA B L E 1 Genetic thresholds used for the categorisation of dingo ancestry during microsatellite a or SNP (this paper) DNA testing methods.  Each sample was assigned using the Table 1 SNP testing criteria to an ancestry category after assessing calculated 95% confidence interval dingo q-values ( Figure 3). In circumstances where the 95% confidence interval spanned a category threshold, we used the lower bound of the 95% confidence interval to classify individuals.

| Dingo × dog ancestry and admixture analysis
This allowed us to conservatively assign animals and minimise the number of Type II errors where a backcross animal is assigned as a parental type. Within our data set of 307 putative dingoes from across Australia, 70.0% were pure dingoes; 15.0% were historical dingo backcrosses; and 15.0% were recent dingo backcrosses (   (Table 4). which is consistent with whole genome sequencing data that also reported higher levels of homozygosity in dingoes compared to domestic dogs (Kumar et al., 2023;Zhang et al., 2020). The Big

| Dingo population structure and biogeography
Desert population has the highest median F het values, nearly 30% higher than the other dingo populations, indicating the population is extremely homozygous and likely inbred. We did not compare homozygosity between dingoes and domestic dogs due to potential ascertainment bias associated with the design of the Axiom Canine HD Genotyping array (Thermo Fisher Scientific Inc.).

| DISCUSS ION
Ancestry modelling, conducted using high-density genome-wide SNPs, reveals two key findings: (1) there are at least four wild dingo populations and a separate captive dingo population in Australia; (2) the presence of dog ancestry in wild dingoes is much less common than previously hypothesised by microsatellite DNA testing Stephens et al., 2015). These data provide much needed clarity regarding the identity of dingoes across Australia, which can inform policy debate regarding how these animals are managed. Furthermore, these findings indicate that genetic surveys of wild dingo populations using genome-wide technologies are important for understanding dingo × dog ancestry across Australia. Future use of genome-wide technology, particularly restriction-site associated DNA sequencing (RADseq) or whole genome sequencing, will permit simultaneous analysis of domestic dog admixture, demography, homozygosity and genomic diversity in dingo populations while also enabling comparisons to domestic dogs or wolves. Finally, these data argue for significant changes in current genetic testing methods.

| Biogeography and regional variation in the dingo
We observed the presence of four distinct geographically distributed dingo populations: West, East, South, Big Desert (Figures 2, 6 and 7).
Captive dingoes formed their own population, distinct from wild dingo populations (Figures 1 and 6). The Big Desert dingo population was the most strongly differentiated, with F ST values twice that of other populations (Table 4). While there are several phenotypically described varieties of dingo, it is unclear if these align with the genetic populations (Corbett, 2001;Walters, 1995). Our biogeographic findings corroborate and extend the observation of geographic subdivision in dingoes from skull morphometrics, mitochondrial DNA, Y chromosome and nuclear DNA data (Cairns et al., 2017(Cairns et al., , 2018Cairns & Wilton, 2016;Greig et al., 2018;Koungoulos, 2020;Zhang et al., 2020). We hypothesise that there is additional regional substructure within the four geographic dingo populations, with preliminary evidence of a regionally distinct subtype of the East dingo population around the NSW Southern Highlands, southwest of Sydney ( Figure 2). Singing Dog (Surbakti et al., 2020).

F I G U R E 6
Dingoes from all four of the wild populations were observed to have increased levels of homozygosity (median F het > 0.3) relative to recent dingo backcrosses and dingo × domestic dog hybrids ( Figure 5). We observed elevated homozygosity levels (median F het > 0.7) in the Big Desert compared to other dingo populations, suggesting the population is experiencing inbreeding (Figure 5). A recent microsatellite-based study found the Big Desert dingo population to have genetic diversity levels up to 50% lower than other dingo populations (Stephens et al., 2022). While the absolute value of homozygosity measures calculated for dingoes in this study could be inflated due to ascertainment bias, whole genome data from a small number of captive and wild dingoes also reveal elevated homozygosity levels relative to domestic dogs (Field et al., 2022;Kumar et al., 2023;Zhang et al., 2020). Further research is needed to assess whether lethal management programmes have had an impact on homozygosity and genetic diversity levels in regional dingo populations, such as the Big Desert population (Hohenlohe et al., 2021;Quevedo et al., 2019;Weeks et al., 2016). The increased levels of homozygosity (median F het > 0.3) found in dingoes may also reflect ancestral bottlenecks, which could be investigated in the future with runs of homozygosity (RoH) analysis of genomic data.

| Domestic dog admixture in Australia
We herein provide the most accurate data regarding the prevalence of domestic dog ancestry in the dingo population, particularly in south-eastern Australia. Our survey of 307 wild dingoes from across Australia revealed that the occurrence and extent of domestic dog ancestry within the dingo population was limited and no feral domestic dogs were observed (Figures 2 and 3). In Victoria, where previous studies using microsatellites suggest that the dingo population is small and highly admixed Stephens et al., 2015), we observed that 87.1% (n = 54) of animals tested were pure dingoes, and 6.5% (n = 4) were historical dingo backcrosses, with >93% dingo ancestry (Table 2). Similarly, in New South Wales and Queensland (n = 171), where dingo × domestic dog hybridisation had been considered pervasive Stephens et al., 2015), we only observed two wild canids with <70% dingo ancestry. We found that most of the sampled New South Wales and Queensland animals were pure dingoes (n = 101) or historical dingo backcrosses (n = 31). In the Northern Territory, South Australia and Western Australia, we found little evidence of domestic dog admixture in the dingo population.

F I G U R E 7
Principal coordinates analysis (PcoA) of Nei distances carried out in SambaR of 320 dingo and historical dingo backcross samples with less than 7% dog ancestry based on 195,474 genomic SNPs. Samples are coloured according to primary dingo population cluster identity as assigned by FastStructure. Our genome-wide SNP data demonstrate that dingoes and domestic dogs are genetically distinct (Figures 3 and 4).  (Figures 4-6). Dingoes from southern Australia exhibit greater phenotypic variation than those elsewhere, particularly in terms of pelage (Cairns, Newman, et al., 2021;Jones, 1990Jones, , 2009Newsome & Corbett, 1985). While variable pelage was previously hypothesised to be the result of domestic dog admixture (Newsome & Corbett, 1985), our finding of rare domestic dog admixture makes this unlikely, and suggests that diverse coat colours may represent standing ancestral variation or perhaps local adaption (Cairns, Newman, et al., 2021).
It was hypothesised that dingo × domestic dog admixture patterns in Australia fit a unimodal hybrid swarm with extensive genetic swamping (Allen et al., 2017;Claridge et al., 2014;Daniels & Corbett, 2003;Stephens et al., 2015). However, our data suggest that the pattern of genetic admixture observed in dingoes fits a bimodal hybrid zone, with a majority of the population being pure dingoes or dingo backcrosses with low levels of intermediate dingo × dog hybrids (Figures 3 and 4, Table 2). The low occurrence of both firstgeneration dingo × dog hybrids and low dingo ancestry backcrosses further supports this hypothesis as it suggests the presence of behavioural barriers or assortative mating preferences (Allendorf et al., 2001;Fitzpatrick et al., 2015;McFarlane & Pemberton, 2019).
It has also been suggested that lethal management strategies such as aerial 1080 baiting, particularly during the dingo breeding season, increase the risk of dingo × dog hybridisation by fracturing social structures and reducing the availability of dingo mates Wallach et al., 2009).
This might explain why dog ancestry is more commonly observed in dingo populations where intensive or aerial baiting strategies are widespread, such as NSW and QLD (Table 2). Further research is needed to identify whether behavioural barriers to hybridisation or assortative mating preferences occur in dingo populations and to understand the impact of lethal control on patterns of dog admixture in dingo populations. miscategorised pure dingoes as dingo backcrosses compared to analysis based on 195,474 SNPs ( Figure 9). Interestingly, dingoes from the Big Desert, South, East or Captive population clusters were more often miscategorised by microsatellite-based methods than those from the West (Table 5). Our findings are consistent with reports from a wide range of other species demonstrating that genome-wide SNP analysis outperforms microsatellites in terms of accurate identification of ancestry and admixture (Dziech, 2021;Mattucci et al., 2019;McFarlane et al., 2020;Steyer et al., 2018;Stroupe et al., 2022;Szatmári et al., 2021;Vaha & Primmer, 2006;Väli et al., 2010;Zimmerman et al., 2020).

| Dog ancestry detection methodology
While microsatellite-based markers remain useful for short evolutionary timeframes, or in taxa that are slow evolving or clonal (Putman & Carbone, 2014) frequencies between a small core set of mixed breed domestic dogs and dingoes from captivity or western and central Australia Elledge et al., 2008;Stephens et al., 2015;Wilton, 2001). We hypothesise that the under representation of dingoes from eastern and southern Australia in reference populations during both marker selection and analysis has confounded microsatellite-based dingo ancestry analyses. In light of these data, we urge wildlife managers to interpret microsatellite-based DNA testing data cautiously, particularly of dingoes from southern and eastern Australia.

| The identity and management of dingoes
While dingo × dog hybridisation is not a pervasive threat to most dingo populations, we acknowledge that in parts of Australia, where there is a long history of suppressing dingo populations, levels of hybridisation are, by comparison, high. Thus, dingo × dog hybridisation may be a threat to specific regional dingo populations. There is ongoing debate concerning when dog admixture in dingo populations poses a threat to conservation and at what genetic threshold a dingo backcross becomes a dingo for purposes of conservation policy (Allen et al., 2017;Claridge et al., 2014;Crowther et al., 2020).
We encourage policymakers to adopt a flexible framework such as that outlined by vonHoldt et al. (2018) which assesses local conservation goals, ecological function and a genomic synthesis of species concepts. At a minimum, it would be appropriate to expand the definition of dingoes in conservation policy and legislation to include historical dingo backcrosses using a genetic threshold of 93% dingo ancestry, as suggested by Allen et al. (2017).
In Australia, the term 'wild dog' is widely used in policy and legislation based on the logic that the wild canine population is dominated by mixed ancestry dingoes or feral domestic dogs (Bird & Bowman, 2016;DEPI, 2013;Fleming et al., 2001;NWDAP, 2020).
Given our demonstrated findings of limited domestic dog admixture in dingoes, future use of the terms 'dingo' and 'feral domestic dog' will more accurately reflect the identity of wild canids in Australia. Furthermore, such a shift in terminology aligns with calls from Australian Aboriginal people for government policy to better acknowledge and respect the value of dingoes as a native animal and culturally significant species (Costello et al., 2021).

ACK N O WLE D G E M ENTS
We would like to thank the numerous private citizens, pest control officers, National Park rangers, government agencies and conservation groups who contributed samples to this research project. We

O PE N R E S E A RCH BA D G E S
This article has earned Open Data and Open Materials badges. Data and materials are available at https://doi.org/10.5061/dryad.vq83b k3ww.

B EN EFITS S H A R I N G
We consulted with many citizens, government employees, conservation volunteers and landholders who participated in the collection of biological samples from wild or privately kept dingoes in Australia. Individual research reports have been provided to relevant stakeholders concerning their contributed samples to inform conservation, management or community priorities. This research addresses a priority concern regarding the occurrence of domestic dog admixture in dingoes across Australia as well as the presence of biogeographical population structure. As described above, all data have been publicly shared via appropriate biological databases.