Conservation implications for dingoes from the maternal and paternal genome: Multiple populations, dog introgression, and demography

Abstract It is increasingly common for apex predators to face a multitude of complex conservation issues. In Australia, dingoes are the mainland apex predator and play an important role in ecological functioning. Currently, however, they are threatened by hybridization with modern domestic dogs in the wild. As a consequence, we explore how increasing our understanding of the evolutionary history of dingoes can inform management and conservation decisions. Previous research on whole mitochondrial genome and nuclear data from five geographical populations showed evidence of two distinct lineages of dingo. Here, we present data from a broader survey of dingoes around Australia using both mitochondrial and Y chromosome markers and investigate the timing of demographic expansions. Biogeographic data corroborate the presence of at least two geographically subdivided genetic populations, southeastern and northwestern. Demographic modeling suggests that dingoes have undergone population expansion in the last 5,000 years. It is not clear whether this stems from expansion into vacant niches after the extinction of thylacines on the mainland or indicates the arrival date of dingoes. Male dispersal is much more common than female, evidenced by more diffuse Y haplogroup distributions. There is also evidence of likely historical male biased introgression from domestic dogs into dingoes, predominately within southeastern Australia. These findings have critical practical implications for the management and conservation of dingoes in Australia; particularly a focus must be placed upon the threatened southeastern dingo population.


| INTRODUCTION
The effect of removing large socially complex apex consumers such as whales, big cats, bears, wolves, and dingoes from ecosystems is poorly documented (Estes et al., 2011). Apex predators are in decline, globally, which has lead to and threatens continuing impacts to entire ecosystems (Estes et al., 2011;Morris & Letnic, 2017;Ripple et al., 2014Ripple et al., , 2016Ripple et al., , 2017. Estes et al. (2011) suggest that worldwide large apex consumer declines can cause extensive trophic cascading, exacerbated by agricultural land management, widespread habitat degradation, pollution, and ultimately climate change. On the Australian continent, indigenous apex predators went extinct thousands of years ago, leaving the dingo as the sole remaining apex predator on the mainland. As such, the dingo plays a central ecological role. Today, dingoes are threatened by extensive lethal control programs, habitat fragmentation, and genetic dilution from hybridization with domestic dogs (Stephens, Wilton, Fleming, & Berry, 2015).
In this study, we explore the evolutionary history of Australian dingoes with a goal of informing management and conservation decisions ( Figure 1). Since 1788, dingoes have been subject to hybridization pressure from modern domestic dogs brought by Europeans, particularly in regions where human populations are high (Stephens et al., 2015). The observation of hybridization in species and populations is an increasingly common conservation concern; well documented examples include bison, coyotes, wolves, wild cats, and even Galapagos tortoises (Garrick et al., 2012;Halbert & Derr, 2007;Hertwig et al., 2009;vonHoldt et al., 2016;Reich, Wayne, & Goldstein, 1999).
Historically, the dingo is thought to have arrived on mainland Australia approximately 5,000 years before present (BP) (Gollan, 1984;Macintosh, 1975;Savolainen, Leitner, Wilton, Matisoo-Smith, & Lundeberg, 2004). The minimum arrival time of dingoes is approximately 3,500 years BP based upon the oldest fossil observed in southern Western Australia (Macintosh, 1964). The dingo is frequently blamed for the thylacine extinction on mainland Australia 2,000-3,000 years BP Johnson & Wroe, 2003;, although new modeling suggests climate change and human population growth may have played a more significant role (Prowse, Johnson, Bradshaw, & Brook, 2014). Dingoes and New Guinea Singing Dog (NGSD) likely came to Oceania with humans. However, there is uncertainty concerning which human colonization event they accompanied and whether dingoes were brought directly to Australia or immigrated via the prehistoric land bridge between Papua New Guinea and Australia prior to 8,000 years BP. It is commonly presumed that dingoes were brought to Australia 5,000 years ago as part of the Neolithic human expansion (Sacks et al., 2013;Savolainen et al., 2004). However, Fillios and Taçon (2016) argue that the Neolithic people are unlikely to be responsible for the arrival of the dingo because Australia lacks key Neolithic cultural indicators.
An alternative hypothesis is that dingoes and NGSD are part of an older dog radiation that immigrated into Australia via the land bridge between Australia and Papua New Guinea, which flooded 6,000-8,000 years BP (Cairns & Wilton, 2016). Some ethnographic evidence supports this hypothesis, for example, the lack of Neolithic cultural items, such as chickens and pigs, in Australia prior to European colonization (Larson et al., 2010;Oskarsson et al., 2011), lack of human genetic signatures indicating contact between South East Asia and Indigenous Australians (Brown, 2013;Haak et al., 2010;van Holst Pellekaan, 2001Karafet et al., 2005;McEvoy et al., 2010;Pugach, Delfin, Gunnarsdóttir, Kayser, & Stoneking, 2013), and the finding that dingoes only carry the two ancestral Amylase gene copies, consistent with their having diverged from modern domestic dogs before the agricultural era (Arendt, Cairns, Ballard, Savolainen, & Axelsson, 2016;Freedman et al., 2014). More recent molecular dating efforts based on mitochondrial divergence time suggest that dingoes could have arrived in Australia approximately 8,000-10,000 years BP (Cairns & Wilton, 2016;Oskarsson et al., 2011). Bayesian skyline modeling may help inform this debate by testing when dingoes underwent population expansion and/or contraction.
Knowledge concerning the population biology of dingoes can provide insight into the arrival patterns and origin of this enigmatic canine, which may have important conservation implications. In 2016, Cairns and Wilton identified the presence of at least two dingo lineages, southeastern (SE) and northwestern (NW), in Australia with a pattern of geographical subdivision. Early studies using mitochondrial DNA were unable to elucidate continent-wide biogeographic patterns, as they restricted genetic sampling to the mitochondrial control region Savolainen et al., 2004). Y chromosome studies either had samples from mostly Western Australia or mostly eastern Australia, and differences in genetic marker sampling made it difficult to compare datasets (Ardalan et al., 2012;Sacks et al., 2013).
Ecological theory predicts that shifts in the distribution and abundance of apex predators and herbivores may result in sizable changes in ecosystem dynamics (Fretwell, 1987;Hairston, Smith, & Slobodkin, 1960

| Canid sampling
In order to investigate biogeography, migration, male and female dispersal patterns, and immigration routes, we sampled 127 dingoes broadly across Australia and five NGSD from the North American captive population ( Figure 2, Table 1). Five of the dingoes were sampled from the captive dingo population. We also incorporated a dataset of Y chromosome and mitochondrial control region data from 173 male dogs, including 94 dingoes and 18 NGSD from Sacks et al. (2013).
Blood and/or tissue samples were collected, and all dingoes were screened for genetic purity, using a microsatellite-based assay for domestic dog introgression (Wilton, 2001;Wilton, Steward, & Zafiris, 1999). Only pure or genetically intact dingoes were allocated to this research project (Stephens, 2011;Wilton, 2001;Wilton et al., 1999).

| Mitochondrial gene analysis
Mitochondrial and nuclear phylogenetic analyses found that there are at least two dingo lineages, with eight diagnostic mitochondrial nucleotide differences between them (SE and NW, Cairns & Wilton, 2016).
Two mitochondrial DNA regions harboring diagnostic mutations and the mitochondrial control region were amplified and sequenced (Table 2). The two diagnostic regions were selected as they contained three of the eight differences between the SE and NW mitochondrial lineages (Cairns & Wilton, 2016). Nonrandom genetic sampling has the potential to overestimate a posteriori significance so care must be taken in interpreting results. The regions selected were 676 bp (positions 7,685-8,361 including a region of ATP6 and ATP8) and 1,028 bp (positions 14,098-15,126 including a region of cytochrome b) in length. The mitochondrial control region is 582 bp (incorporating nucleotide positions 15,458-16,039 as in Savolainen et al. (2004)).
Qiagen DNeasy kits (Qiagen Sciences, Germantown, USA) were used to extract DNA, and mitochondrial loci were amplified using PCR (Table 2). Briefly, PCR reactions were carried out in 25 μl containing water, 5× Crimson polymerase buffer (New England Biolabs Inc., MA, USA), 1.5 mmol/L of MgCl 2 , 6.25 pmol of each primer, 7.5 mmol/L of dNTPs, 2.5 U of Taq DNA Polymerase (New England Biolabs Inc., MA, USA), and 20-50 ng of DNA template. All PCR reactions were cycled using the following thermal profile: 98°C for 2 min, 95°C for 3 min (add Taq polymerase), then 95°C for 15 s, 52°C for 1 min, 65°C for 1 min for 10 cycles, then 95°C for 15 s, 52°C for 1 min, and 65°C for 1 min (increase time by 5 s each cycle) for 25 cycles followed by 65°C for 10 min.
DNA sequence chromatograms were analyzed and aligned using Sequencher 5.1 (Gene Codes corp., Ann Arbor, USA).

| Neutrality tests
To investigate whether the genetic variation present within the mitochondrial genome departs from the expectations of neutrality, Tajima's D, Fu and Li's F*, and Fu and Li's D* (Fu, 1997;Fu & Li, 1993;Tajima, 1989) statistics were calculated in DnaSP v 5.10.1 (Librado & Rozas, 2009

| Biogeographic analyses
Median spanning networks were calculated in Networks v4.6 (Bandelt, Forster, & Rohl, 1999;Forster et al., 2000) using the mitochondrial diagnostic region, mitochondrial control region, and Y chromosome datasets. As in Sacks et al. (2013), the median-joining (MJ) algorithm with default settings was used (r = 2, ε = 0). Mitochondrial networks were created for the concatenated diagnostic region and control region separately. Control region data were analyzed separately to allow incorporation of and comparison to the existing dingo control region dataset Sacks et al., 2013;Savolainen et al., 2004). As the control region is not phylogenetically informative in dingoes, it was not included in the mitochondrial diagnostic region analysis (Cairns & Wilton, 2016). Mitochondrial networks are unrooted. Y chromosome networks were calculated using concatenated SNP and STR data. Y chromosome SNPs and STRs were weighted as described by Sacks et al. (2013). Briefly, STRs were weighted as:  (Brown & Yang, 2011;Cairns & Wilton, 2016). All runs were optimized for MCMC chain steps to ensure that the estimated sampling size of all variables was above 200 in Tracer 1.5 (Rambaut & Drummond, 2007). We sampled every 5,000 steps with a 10% burn-in.
The resulting maximum clade credibility tree was midpoint rooted.

| Demographic analyses
To investigate historical patterns of demographic change in the dingo, Bayesian skyline plots were constructed in Tracer 1.5 (Rambaut & Drummond, 2007). Bayesian analyses were carried out in Beast v1.7.4 (Drummond et al., 2012) as detailed above. Skyline plots were constructed based upon the combined mitochondrial DNA dataset and each mitochondrial clade separately.

| Neutrality tests
Tajima's D statistics were calculated for all dingoes as grouped by mitochondrial lineage using the mitochondrial diagnostic region (Table 3).
Statistics could not be calculated for the NGSD as all individuals carried the same mitochondrial DNA sequence. The NW lineage statistics were found to be significantly negative, indicating the presence of purifying selection and/or population expansion. Statistics calculated for the SE lineage were negative but not significant.

| Biogeographic analyses
When ignoring indels, we observed 12 mitochondrial control region (CR) haplotypes with three novel CR haplotypes in 124 dingoes (three dingoes were excluded due to PCR difficulties) and five NGSD (Table 1). The novel haplotypes (din31, din32, and din33) were found in 1-4 individuals and differed by 1-2 nucleotide substitutions from A29. One dingo carried the A9 haplotype thought to have arisen in dingoes independently from dogs (Savolainen et al., 2004). A single dingo out of 124 carried A17, a CR haplotype hypothesized to be introgressed from domestic dogs (Savolainen et al., 2004). Incorporating all the CR data from previously published studies into our own yielded a star-shaped genetic network (Figure 3).
A total of 39 mitochondrial diagnostic region haplotypes were observed in 124 dingoes and 5 NGSD ( Table 1) To further investigate the relationship between the dingo and NGSD, a Bayesian analysis was conducted on the combined sample of 129 animals. This included 124 dingoes and five NGSD (Table 1).
This analysis corroborated the whole mitochondrial genome Bayesian phylogenetic analyses suggesting that the NGSD is more closely related to the SE dingo lineage than the NW lineage (Cairns & Wilton, 2016), with an increased posterior probability node support of 0.84 ( Figure 6).
We observed 30 Y chromosome haplotypes in our dataset of 79 dingoes and two NGSD (Table 1). Y chromosome network analysis identified three main haplogroups present within dingoes and NGSD, H1, H3, and H60 (Figure 7). A contingency table analysis, with two columns (mitochondrial lineage) and three rows (Y chromosome haplogroup), suggests that the distribution of Y chromosome haplogroups between the mitochondrial lineages was nonrandom in T A B L E 2 PCR amplification primers and conditions for mitochondrial PCR amplification and sequencing of the dingo and NGSD resulting in a total of 194 samples (Figure 8). Within the combined dataset representing 173 male dingoes and 20 NGSD, we observed an additional 6 Y chromosome SNP-STR haplotypes (Figure 8). The 20 NGSD sampled between the two datasets all carried H60 haplotypes.
When Y chromosome haplogroup information was plotted on a map (Figure 9), we observed that H1 was largely restricted to the southeastern region of Australia, H3 was restricted to the southeastern and Kimberley regions, and H60 was predominantly found throughout northern, Western, and central Australia. Of the four H3 haplogroup alleles observed in the Kimberley region, all were endemic except H3_12d, which was also observed in southeastern Australia.
To investigate the relationship between Y chromosome haplotypes observed in dingoes and NGSD and Southeast Asian dogs, a MJ network was calculated based upon a combined dataset incorporating a total of 272 samples. This comprised 173 dingoes, 79 from our dataset and 94 from Sacks et al. (2013); 20 NGSD, two from our dataset and 18 from Sacks et al. (2013); and 79 Southeast Asian dogs from Sacks et al. (2013) (Figure 10). We observed that the H1 and H3 haplotypes found in dingoes were largely unique (not shared with SEA dogs).

| Demographic analyses
Bayesian skyline plots constructed on the combined dingo mitochondrial dataset indicate that the population was stable until approximately 5,000 years ago when it began increasing steadily (Figure 11).
There was some evidence of a small decline in dingo numbers in the last 200 years. Skyline plots modeled for the individual mitochondrial clades separately suggest differences in demographic histories. The SE clade plot indicates a historically stable population size, which began increasing rapidly in the last 1,000-2,000 years ( Figure 12).
The NW clade plot depicts a more gradual population increase from about 6,000 years ago stabilizing approximately 3,000-4,000 years ago ( Figure 12).

| DISCUSSION
Understanding the ecological roles of apex predators often comes after their populations have declined to endangered levels, necessitating precautionary management (Estes et al., 2011;Ripple et al., 2014). In the case of the dingo, the findings documented here suggest the potential for sufficiently long-standing population structure to support management for multiple locally adapted populations.
Understanding the population biology, demography, and biogeography of dingoes across Australia is central to the question of how best to manage and conserve them, whilst limiting hybridization.

| Multiple immigrations-different evolutionary lineages
Mitochondrial and Y chromosome data corroborate the presence of at least two discrete populations of dingo, NW (H60/H3), and SE (H3/ H1) (Ardalan et al., 2012;Cairns & Wilton, 2016;Sacks et al., 2013). A lack of intermediate haplotypes despite broad geographical sampling suggests that the observed pattern of population structure is due to historical events. Previous studies did not observe the presence of population structure in dingoes, due to restricted sampling of the mitochondrial control region and limited geographic sampling across the Australian continent Sacks et al., 2013;Savolainen et al., 2004).
Our data suggest that the two divergent Y chromosome lineages observed in dingoes have different geographical origins and are plausibly the result of multiple immigrations into Australia, as postulated by Cairns and Wilton (2016). Notably, the Y haplogroups H3 and H60, which are both observed in dingoes, are not immediately related ( Figure 10; Natanaelsson et al., 2006;Ardalan et al., 2012;Sacks et al., 2013). The H3 Y chromosome haplogroup is also observed in Southeast Asia. However, seven of the eight haplotypes observed in dingoes were endemic to Australia, indicating shared ancestry with a history of isolation between dingoes and Southeast Asian dogs ( Figure 10; Sacks et al., 2013). On the other hand, the H60 haplogroup is unique to dingoes and NGSD and most closely related to H5, a haplogroup found in Taiwan (Figure 10; Ardalan et al., 2012;Sacks et al., 2013). The distribution of Y chromosome haplotypes between the two mitochondrial lineages is nonrandom, corroborating the presence of strong geographic subdivision in dingoes (Ardalan et al., 2012;Cairns & Wilton, 2016).
These data have intriguing implications for the movements of canids, and presumably humans, in Australasia. Cairns and Wilton (2016) postulate that dingoes immigrated into Australia via the now flooded Asia and Oceania, the movements of dogs, and presumably humans, are much more complex than assumed. Indeed, genetic research finds little evidence of Neolithic or Austronesian gene flow into Australia (Bergström et al., 2016;Haak et al., 2010;van Holst Pellekaan, 2013;Karafet et al., 2005;McEvoy et al., 2010;Rasmussen et al., 2011).
Intriguingly, human mitochondrial research found a pattern of continuous strong geographic subdivision dating back to approximately F I G U R E 8 Median spanning network based upon Y chromosome SNP and STR haplotypes for 173 dingoes and 20 NGSD. Black coloration indicates dingoes from this study, red indicates NGSD individuals, and white indicates dingo samples from Sacks et al. (2013). Strokes across branches indicate the presence of Y chromosome SNP mutations differentiating between Y chromosome haplogroups. Branch lengths are relative to the number of STR mutations between Y chromosome haplotypes H1 H21 H3 H60 F I G U R E 7 Median spanning network based upon Y chromosome SNP and STR haplotypes for 79 dingoes and two NGSD. Black coloration indicates NW mitochondrial lineage individuals, orange SE mitochondrial lineage individuals, red NGSD individuals, and white unknown mitochondrial lineage. Strokes across branches indicate the presence of Y chromosome SNP mutations differentiating between Y chromosome haplogroups. Branch lengths are relative to the number of STR mutations between Y chromosome haplotypes H1 H21 H3 H60 n24 n25 n27 50,000 years BP, after Australians first spread into the continent, with little evidence of migration between populations (Tobler et al., 2017).

| Dating, demography, and dispersal
Demographic modeling and neutrality test results based on mitochondrial data should be treated cautiously but can give insight into modern and historical demographic patterns. Bayesian skyline plots based upon the individual mitochondrial lineages suggest that the SE population size has been stable until about 1,000 years ago, when it underwent rapid expansion (Figure 12). The NW population on the other hand has a history of gradual population expansion from approximately 4,000-6,000 years ago ( Figure 12). Possibly Bayesian demographic modeling is reflective of rapid range expansion of dingoes in southeastern Australia following the decline of thylacines on the mainland, which occurred approximately 2,000 years BP (Figure 12; Johnson & Wroe, 2003;Fillios et al., 2012;Letnic et al., 2012;Prowse et al., 2014). It is also possible that the pattern of population expansion in SE dingoes is the result of extensive culling and baiting practices in southeastern Australia within the last 200 years (Fleming, Corbett, Harden, & Thomson, 2001;Wallach, Ritchie, Read, & O'Neill, 2009 population could be the result of long-term but gradual range expansion after immigration into Australia. Demographic modeling on the entire dingo dataset depicts a population expansion approximately 3,000-8,000 years BP (Figure 11). Ethnographic and molecular dating suggests dingoes arrived in Australia prior to 5,000 years BP (Cairns & Wilton, 2016;Fillios & Taçon, 2016;Oskarsson et al., 2011). It should be noted that uncertainty in the demographic modeling makes it difficult to discern the approximate arrival time of dingoes or pinpoint when range expansions occurred.
Biogeographic patterns within Australia provide insight into the modern dispersal and migration of dingoes. We observed that the geographical distribution of the two mitochondrial lineages, SE and NW, exhibits strong geographical subdivision ( Figure 5). Only a single instance of discordance between mitochondrial lineage and geographic origin was observed, indicating that maternal migration is limited. The geographical distribution of three Y chromosome haplogroups, H1, H3, and H60, is similar to that of the mitochondrial lineages, but more diffuse, suggesting higher levels of paternal than maternal migration (Figures 5 and 9). Introgression between the NW and SE populations seems to be west to east biased, with few H3 haplogroup individuals found in northern, Western, or central Australia. Conversely, there are some individuals in southeastern Australia harboring H60 haplogroup types, either the result of male dispersal from the NW population into the SE population or historical distribution patterns. These patterns are likely a factor of male dispersal; male dingoes and dogs range more widely and are more likely to disperse to new areas (Pal, Ghosh, & Roy, 1998;Thomson, Rose, & Kok, 1992). Human-mediated dispersal may also be a factor in facilitating the movement of dingoes, by breaking apart pack structures through culling/baiting management practices (Corbett, 1988;Fleming et al., 2006;Glen, Dickman, Soulè, & Mackey, 2007;Thomson, 1992;Wallach, Johnson, Ritchie, & O'Neill, 2010;Wallach et al., 2009).
Contrary to demographic modeling, the neutrality test results indicate that the two dingo populations may be experiencing different demographic and/or selective pressures (Table 3 and Cairns & Wilton, 2016). These data are consistent with mitochondrial network analyses depicting a complex pattern in the SE population but a more star-like pattern in the NW population indicative of population expansion (Figures 3 and 4). There is also evidence of a west to east biased dispersal pattern which might be the result of NW population dingoes moving into vacant niches opened up by extensive culling and baiting practices in southeastern Australia ( Figure 5).

| Introgression from modern domestic dogs
The H1 Y chromosome haplogroup is considered to be a European domestic dog haplogroup (Ardalan et al., 2012;Ding et al., 2011;Sacks et al., 2013), and it is often observed in domestic dog breeds or Southeast Asian dogs thought to have breed ancestry. The observation of H1 haplotypes within dingoes is suggestive of paternal introgression from European domestic dogs into dingoes (Figures 7, 8, and 10). The presence of the H1 haplogroup in genetically tested "pure" dingoes suggests that there has likely been historical, post-European colonization, introgression from domestic dogs into dingoes. It is unlikely that the introgression is modern because the genetic test is capable of detecting hybridization events on a recent timescale (Cairns, Wilton, & Ballard, 2011;Wilton, 2001;Wilton et al., 1999). The uniparental inheritance of the Y chromosome means that a single hybridization event will be reflected in the paternal lineage of a dingo despite extensive backcrossing. The lack of non-dingo-like mitochondrial lineages suggests that the introgression from domestic dogs is predominately due to male domestic dogs mating with female dingoes.
The distribution of the H1 haplogroup in southeastern Australia further suggests it is likely the result of introgression (Figure 9). First, domestic dogs have been present in southeastern Australia for a longer F I G U R E 1 1 Bayesian skyline plot built using mtDNA diagnostic region (1,706 bp) sequences from 124 dingoes. Analyses constructed using a GTR + G + I substitution model and a skyline coalescent model in BEAST v1.7.5 (Drummond et al., 2012). A strict clock was enforced with a substitution rate of 7.7027 × 10 −8 mutations −1 site −1 year −1 with a standard deviation of 5.4848 × 10 −9 . The skyline plot was constructed in Tracer 1.5 (Rambaut & Drummond, 2007) (Corbett, 2001).
Second, southeastern Australia has the densest human population, and areas with dense human populations are generally associated with higher incidences of hybridization (Stephens, 2011;Stephens et al., 2015). Thirdly, lethal management strategies such as baiting and culling are widespread in southeastern Australia due to the sheep industry (Fleming et al., 2001). Fatal management strategies are believed to lead to increased levels of hybridization due to the breakdown of pack structure (Fleming et al., 2006;Wallach et al., 2009). This finding is a significant conservation concern in the context of the genetic identity and integrity of the SE dingo population, which is under threat of extinction through introgression and ecological exclusion through lethal management programs.
The presence of the H1 haplogroup in southeastern Australia has important implications for conservation and future management strategies; namely, it highlights the importance of inhibiting further hybridization. Neutering male dogs and/or restricting them from reproducing with wild dingoes may help achieve this. Particularly, best practice should dictate that pet, livestock guardian, or working dogs in rural localities should be neutered or chemically castrated to avoid further risk of hybridization. Widespread lethal control measures are shown to also facilitate hybridization by breaking apart pack structures (Wallach et al., 2009). Alternative livestock protection measures should be explored, such as livestock guardians and improved dog-proof fencing (van Bommel & Johnson, 2012;Fleming et al., 2001). This observation also suggests the need for a higher accuracy "next generation" DNA test for distinguishing dingoes from hybrids; the current method is likely sufficient for monitoring wild populations but not for captive breeding programs.
Knowledge concerning levels of genetic integrity in wild populations is necessary to inform management and conservation programs. The southeastern population of dingoes is under particular extinction pressure from both fatal management strategies and hybridization; steps should be taken to preserve this population before it is too late. A broad genetic survey of dingoes in National Parks and State Forests in southeastern states would be needed to pinpoint high dingo ancestry populations and thus where to focus conservation efforts. Currently state and federal legislation do not protect the dingo sufficiently and allow widespread fatal control measures (Davis, 2001;Downward & Bromell, 1990;Fleming et al., 2001Fleming et al., , 2006. Revision of legislation must be achieved to reflect the ecological, cultural, and taxonomic importance of the dingo, balancing the need to conserve this enigmatic canine with any agricultural concerns.

| CONCLUSIONS
Distinct populations of apex consumers can exhibit different behaviors and prey on disparate trophic niches (Paiva, Fagundes, Romão, Gouveia, & Ramos, 2016). These differences could be due to ecological plasticity or genetically inherited differences. This study corroborates the presence of at least two dingo populations in Australia. It is plausible, given the divergent evolutionary histories of these popula-