New insights into the genetic composition and phylogenetic relationship of wolves and dogs in the Iberian Peninsula

Abstract This study investigates the gene pool of Portuguese autochthonous dog breeds and their wild counterpart, the Iberian wolf subspecies (Canis lupus signatus), using standard molecular markers. A combination of paternal and maternal molecular markers was used to investigate the genetic composition, genetic differentiation and genetic relationship of native Portuguese dogs and the Iberian wolf. A total of 196 unrelated dogs, including breed and village dogs from Portugal, and other dogs from Spain and North Africa, and 56 Iberian wolves (wild and captive) were analyzed for nuclear markers, namely Y chromosome SNPs, Y chromosome STR loci, autosomal STR loci, and a mitochondrial fragment of the control region I. Our data reveal new variants for the molecular markers and confirm significant genetic differentiation between Iberian wolf and native domestic dogs from Portugal. Based on our sampling, no signs of recent introgression between the two subspecies were detected. Y chromosome data do not reveal genetic differentiation among the analyzed dog breeds, suggesting they share the same patrilineal origin. Moreover, the genetic distinctiveness of the Iberian wolf from other wolf populations is further confirmed with the description of new mtDNA variants for this endemism. Our research also discloses new molecular markers for wolf and dog subspecies assignment, which might become particularly relevant in the case of forensic or noninvasive genetic studies. The Iberian wolf represents a relic of the once widespread wolf population in Europe and our study reveals that it is a reservoir of unique genetic diversity of the grey wolf, Canis lupus. These results stress the need for conservation plans that will guarantee the sustainability of this threatened top predator in Iberia.


| 4405
PIRES Et al. in the early 20th century (Álvares, 2011). Although management and conservation unit status are recognized for this wolf subspecies (Chapron et al., 2014;Pimenta et al., 2005), different policies are in place in Portugal and Spain. In Portugal, where the last census estimated a population size of only 300 individuals (Pimenta et al., 2005) and the current trend is unknown (Torres & Fonseca, 2016), the conservation status of the Iberian wolf is "Endangered," and it is fully protected by law (Portuguese Law 90/88). In Spain, the Iberian wolf conservation status is "Vulnerable" and its protection varies regionally.
In some areas, it is fully protected and in others, such as northwestern Spain, hunting is legal. Most wolves occur in the northwestern part of the Iberian Peninsula and south of the Douro river, with an uncertain small group in Sierra Morena (south of Spain) (Álvares, 2011;López-Bao et al., 2015).
In the Iberian Peninsula, both dog and wolf subspecies have been the focus of several studies where molecular markers were used to describe genetic composition, population structure, and phylogeny.
Microsatellite and AFLP analyses of modern dog suggest that the genetic variability of dog breeds is structured according to breeds and geography (Pires et al., 2009) and that Portuguese dogs have multiple mtDNA haplotypes in agreement with the mtDNA diversity found in other European dogs (Pires et al., 2006).
The occurrence of gene flow between domestic dogs and Iberian wolves has been documented and estimated (van Asch et al., 2010;Fan et al., 2016;Godinho et al., 2011Godinho et al., , 2015Ramirez et al., 2006). However, wolf-dog hybrids in Iberia have only been reported for marginal and recently expanded wolf nuclei (Godinho et al., 2011) with the possible involvement of hunting or livestock guard dogs, rather than feral dogs. Another study in the Caucasus region reports the existence of gene flow between wolves and livestock guard dogs (Kopaliani et al., 2014).
Information from other genetic markers, such as mitochondrial DNA (mtDNA) or genomewide SNPs, is also available for Iberian wolves (van Asch et al., 2005;Godinho & Ferrand, 2007;Godinho et al., 2011;Pilot et al., 2010Pilot et al., , 2014Vilà et al., 1997Vilà et al., , 1999. Briefly, these latter studies concluded that for Canis found in Iberia, particularly in Portugal: (1) Iberian wolves exhibit both unique mtDNA haplotypes and haplotypes that are shared with other European and Asian wolves; (2) some wolf haplotypes are widespread, while others are rare with a restricted distribution, but there is no obvious mtDNA haplotype geographic structure; (3) dog and wolf mtDNA haplotypes are mostly distinct, with few haplotypes shared; and (4) Iberian wolf genome analysis shows signatures of long-term isolation from other European wolf populations, evidence of diversifying selection due to local adaptation and admixture with dogs.
For Canis, the most recent studies using Y chromosome markers are from Brown et al. (2011), Godinho et al. (2011), Ding et al. (2012), Gomerčić et al. (2013) and Sacks et al. (2013). Brown et al. (2011) studied Y-STRs and Y-SNPs in a large population of 633 worldwide dogs (including village and breed dogs) and concluded that modern European dog breeds possess roots in southeastern Asian dogs and that African dogs have different patrilineal origins. Godinho et al. (2011) used six Y-STR loci in their study of Iberian wolf-dog hybridization and found six different Y-linked haplotypes in 99 male wolves, and 14 different Y haplotypes in 78 male dogs (including breed and feral dogs). To further investigate the phylogeography of the domestic dog, Ding et al. (2012) studied a 14,437 base pairs (bp) sequence fragment of Y chromosomal DNA in 151 dogs distributed worldwide, which revealed 28 haplotypes distributed across five haplogroups. Additionally, they observed that a region in Asia, south of the Yangtze River (ASY), has nearly the full range of genetic diversity supporting previous evidence from mtDNA (Pang et al., 2009) that ASY was the main region where wolves were domesticated. Gomerčić et al. (2013) also used Y chromosome sequences and concluded that dogs and gray wolves from Croatia could not be differentiated as they shared a single haplotype. Sacks et al. (2013) used dog patrilines, by combining rapidly mutating Y-STRs and slower mutating Y-SNPs, to date the origin of modern western dogs.
As referred by Lescureux and Linnell (2014), more studies are needed to clarify the relationship between wolves and dogs. In line with this need, our research aims to analyze the overall genetic variability and the phylogenetic relationship between wolves and native dogs from the Iberian Peninsula. Our main objectives were as follows: (1) to capture unreported genetic variants for the Portuguese Canis populations; (2) to understand the strength of genetic divergence between paternal lineages of Portuguese native dogs and the Iberian wolf; (3) to investigate whether Iberian dogs and wolves represent genetically distinct reservoirs from their worldwide counterparts; and (4) to provide data to assist the conservation of the locally adapted Iberian wolf subspecies. were collected from specimens whose phenotypes could not be assigned to any breed, including common worldwide breeds, nor resembled the phenotype of wolf-dog hybrids (Table 1).

| Iberian wolves
Wolf sampling included 56 individuals: 44 randomly collected wild dead animals from Portugal (n = 42) and Spain (n = 2); and 12 live captive specimens kept at the Iberian Wolf Recovery Centre (three specimens originally from Spain and nine from Portugal). Wild specimens' samples were either tissue or blood (from necropsied animals), obtained from the wolf tissue bank (SMLM) managed by the Portuguese Institute for the Conservation of Nature and Forest (ICNF); blood from live captive animals was collected by a veterinarian. Samples included 11 females, 44 males, and one specimen of undetermined sex and were collected between 1987 and 2008 ( Figure 1).
All forward primers were fluorescently labeled (6-FAM, Hex, or NED from Applied Biosystems, Foster City, CA, USA). PCR and electrophoresis conditions are described in Supplementary Materials.

| Primer design
Sequences of newly designed primers for PCR amplification of Y chromosome fragments (Natanaelsson et al., 2006) and for SNaPshot extension reactions are presented in Tables S1 and S2, respectively.  Natanaelsson et al. (2006) (Table S2).

| PCR amplification
PCR amplifications were carried out individually for each Y chromosome fragment using the primer pairs described in Table S1. All reaction mixtures contained 25-50 ng of genomic DNA, 1× Biomix PCR master mix (Bioline), and 0.2 μmol/L of each primer in a total volume of 15 μL.

PCR and electrophoresis conditions are described in Supplementary
Materials. Six female samples from both subspecies were also PCRtested for all loci to check the Y-specificity of the primer pairs.

| SNaPshot reaction for probe primer extension
This step consisted of a multiplex reaction for single-base primer extension using the ABI PRISM ® SNaPshot ™ Multiplex Kit (Applied Biosystems) according to the manufacturer's protocol. Two separate multiplex reactions were prepared: one included pooled PCR products and extension primers for loci Ydog_21, Ydog_28, Ydog_G (part 1) and Ydog_B (part 2) (SNPs position here named Ydog_B), covering seven dog-specific SNPs; and the other multiplex included the PCR products and extension primers for the remaining loci mentioned in Table 2 (Ydog_20, Ydog_29, Ydog_30, Ydog_N), covering four SNPs.

| Y chromosome DNA sequence analysis
In the search for new polymorphisms, sequencing reactions were performed for each locus in dogs and wolves. PCR products were sequenced on both directions by BigDye Terminator v1.1 Sequencing Kit and analyzed on an automated fluorescence-based ABI PRISM 3130 Genetic Analyser (Applied Biosystems).

| Y chromosome microsatellites
Four Y chromosome microsatellite (short tandem repeats, STRs), loci-990-35, MS41A, MS41B, and MS34TTR (Bannasch et al., 2005), were analyzed. All loci were amplified in a single multiplex reaction using the primers presented in Table S3. PCR and electrophoresis conditions are described in Supplementary Materials.

| Mitochondrial DNA PCR amplification and sequencing
The

Iberian and North African dogs' SNPs Iberian wolves' SNPs
Ydog_20

| Autosomal STRs data analysis
Standard measures of genetic diversity and the F ST value were calculated in GenAlEx 6.5b (Peakall & Smouse, 2006) using a set of 122 dogs and 52 wolf genotypes. Genetic differentiation was further investigated using a Bayesian clustering procedure included in the STRUCTURE software version 2.3 (Falush, Stephens, & Pritchard, 2003;Pritchard, Stephens, & Donnelly, 2000). Five independent runs [K = 2; 100,000 Markov Chain Monte Carlo (MCMC) iterations; burnin period of 10,000 interactions] were set to run using the 'admixture' model and assuming that allele frequencies are correlated among populations for a K = 2. The STRUCTURE PLOT software (Ramasamy et al., 2014) was used to construct the STRUCTURE bar plot.

| Y chromosome markers
Regarding Y-STRs, we estimated haplotype diversities and analyzed molecular variance (AMOVA) using GenAlEx v6.501 (http://biologyassets.anu.edu.au/GenAlEx/Welcome.html), and the level of genetic divergence between the dog and wolf subspecies was determined through the fixation index PhiPT for Y-STRs alone and in combination with Y-SNPs. p-values were estimated after 9,999 permutations.
A combined analysis of Y chromosome SNP and STR markers was performed with the software NETWORK 5 (http://www.fluxusengineering.com) using only the polymorphic loci. The Y-STR loci were weighted inversely to their variance, and considering the much lower rate of substitutions relative to autosomal STR mutations; SNP loci were weighted as follows: transitions w = 10 and transversions w = 30.
We first constructed a reduced-median network and then applied a median-joining analysis to create the final networks.
The haplotype data generated by the detection of 11 SNP loci in this study were also analyzed at a global scale with other worldwide Canis SNPs haplotypes described in Ding et al. (2012) and Brown et al. (2011).
For the network analysis, we used the same method as described above.
Analyses at the geographic scale of the Iberian Peninsula were performed considering 420 bp of the mtDNA fragment sequenced.
Genetic diversity and differentiation analyses were performed with DnaSP v5.10.01 (http://www.ub.edu/dnasp/) (Librado & Rozas, 2009) for Iberian dogs and wolves for which mtDNA haplotype frequency information was available (135 dog sequences from a previous study (Pires et al., 2006) and Iberian wolves from this study). Dogs and wolves were considered as two populations and haplotype diversity (HD) and nucleotide diversity (π) were calculated for each population separately. The extent of divergence among these two populations and gene flow was also computed and statistically assessed. Because there are some indels in the sequence alignment, gaps were treated as a fifth character.
To test if differences in haplotype richness between wolves and dogs found in Iberia were due to unbalanced sampling, the total number of haplotypes for each population was estimated using a rarefaction resampling method as implemented in the software Analytic

| Autosomal STRs
Autosomal STRs analysis for dogs and wolves revealed a pairwise F ST value of 0.112 (p = .000). The partitioning analysis using the Bayesian clustering procedure clearly shows a sharp separation between dogs, (village dogs and breed dogs, these latter including livestock guard dogs), and the Iberian wolf ( Figure S1, orange and yellow, respectively).
The average Q-value (proportion of membership of each predefined population in each of the two clusters) for the studied populations is high for dogs (0.994 ± 0.0096) as well as for wolves (0.993 ± 0.0134).
We did not detect any individual with an intermediate assignment value (threshold value considered was 95%).

| Paternal lineages in Iberian Canis
Sequence data, for the studied loci for a subset of dogs (n total = 59) and wolves (n total = 44), did not reveal additional polymorphic sites (GenBank accession numbers GQ366706-GQ366793 and   Table S4A for a list of Y-SNP haplotypes found worldwide YSNP H32, YSNP H33, and YSNP H34, following the nomenclature of Ding et al. (2012).
Regarding STRs alone, a total of 25 haplotypes are described for the studied dogs (  (Figure 2a, Table S4B). The haplotypic diversity (HD) (unbiased) ± SE is 0.59 ± 0.11 for African dogs, 0.52 ± 0.11 for Iberian breed dogs, 0.46 ± 0.08 for Iberian village dogs, and 0.52 ± 0.05 for the group of all studied domestic dogs.
PhiPT pairwise values among these groupings of dogs varied between 0 (for Iberian breeds and village dogs) and 0.052 (for Iberian village dogs and African dogs). These values were not significant (p-values > .05).
The Iberian wolf displays seven Y-STR haplotypes, three of these are shared with dogs and four are exclusive to the Iberian wolf (Table   S4B). The HD (unbiased) ± SE for the Iberian wolf is smaller compared to the dog population and is 0.41 ± 0.04. The PhiPT pairwise value between dogs and Iberian wolves of this study is 0.27 (p = .00).  Table S4A). In particular, haplotype YSNP H9 is shared between two Iberian wolves (samples IbWolf-22 and IbWolf-24), and dogs from Africa (Basenji dog breed, southern Africa) and East Asia (East Siberian Laika dog breed, East Asia). The YSNP H9 haplotype differs from YSNP 10B, another haplotype shared between wolves and dogs, by a single SNP (locus B) ( Table S4A).
The single Y-SNP haplotype shared between Iberian and northwest African dogs (this study) is the most common haplotype among dogs worldwide-YSNP H1 (Figure 2b; Table S4A), also found in a high proportion of East Asian dogs (47%).

| Distinctiveness of Iberian wolf paternal lineages
We detected Y chromosome SNP haplotypes exclusive to the Iberian wolf. The three distinct Y chromosome SNP haplotypes of the Iberian wolves-YSNP H32, H33, and H34-segregate in the network among haplogroups HG1 & HG3, HG9, and HG23 where several unsampled nodes were previously detected by Brown et al. (2011) andDing et al. (2012) (Figure 2b). Moreover, it is the first time that the haplogroup HG9, represented by the haplotype YSNP H9, is reported for wolves. Independent molecular markers, such as autosomal STRs and mtDNA, support that these HG9 samples were collected from wolf specimens.
Considering the larger sequence fragments available for Iberian wolves, our findings increase the total number of mtDNA control region I haplotypes reported for Iberian wolf to 24 (see Vilà et al., 1997Vilà et al., , 1999Randi et al., 2000;Valière et al., 2003;Björnerfeldt, Webster, & Vilà, 2006, unpublished GenBank accession numbers EF380226-9, and Godinho & Ferrand, 2007). Of the 24 haplotypes described for Iberia, two are found in both Portugal and Spain, nine are found in Portugal (five of them exclusive to Portugal), and 13 are found in Spain (11 of them exclusive to Spain).
Five of the seven 420 bp mtDNA haplotypes found in this study were identified for the first time in Iberian wolves (wolf haplotypes wH-1B, wH-1C, wH-2, wH-3 and wH-4 in this study) and were submitted to GenBank (accession numbers JX845621-JX845625, where JX845621 may correspond to Godinho & Ferrand, 2007 Valière et al., 2003;and haplotype MIT2), and one was only found in Portugal (this study haplotype wH4) (Figure 3b). All the wolf samples sequenced in this study, when trimmed to 230 bp, segregate within the wolf clade previously reported for Iberia (Pilot et al., 2010), except for one individual (this study haplotype wH-5) (Figure 3b). This exception corresponds to the only mtDNA haplotype that is shared between Iberian wolves and dogs, representing one of the five cases of wolf/dog haplotype sharing detected in our worldwide wolf analysis ( Figure 3b). The same occurs for the 420 bp fragment (this study wolf haplotype wH-5; and dog haplotypes H10 and H48 from Pires et al., 2006) (Figure 3a).
In both phylogenetic analyses, either considering the 420 bp or the 230 bp fragment (model of evolution GTR+I+G and TPM1uf+I+G, respectively), the Iberian wolf sequences generated in this study cluster with other wolf sequences forming clades with high node support (posterior probability ≥0.90), except for: haplotype wH-2 in the 230 bp worldwide analysis that forms a group with low support with other wolf sequences; and haplotype wH-5 (shared with dogs) that forms a group with high node support with other dog sequences (dog haplogroup A) (posterior probability 0.94) (Figures 3a,b).
For the 230 bp fragment, the wolf samples sequenced in this study diverged by 1 to 10 nucleotides compared to other wolf sequences worldwide, and the most divergent Iberian wolf haplotype sequence was also found in other European and Asian wolves (this study wolf haplotype wH-2 and haplotype W3- Pilot et al., 2010).
Concerning dogs and wolves found in Iberia for which we have haplotype frequency information (135 dogs, 56 wolves), the rarefaction analysis revealed that haplotype richness was approximately 2.5 times higher for dogs compared to the wolves (18.8 vs. 7). Haplotype diversity (HD) for the Iberian dogs was also approximately three times higher than the value computed for Iberian wolves, 0.92 ± 0.008 (mean ± SD) and 0.32 ± 0.080, respectively. The difference in nucleotide diversity (π) for dogs and wolves found in Iberia is even higher, 0.014 ± 0.0006 (mean ± SD) and 0.002 ± 0.0008, respectively.
The statistical test indicates that Iberian dogs and wolves are strongly genetically differentiated (χ 2 = 186.428, p = .0000), where the F ST value between the two subspecies is 0.486.

| Genetic diversity of Iberian Canis
Maternal and paternal markers used in this study effectively reveal Although an extensive study on several gray wolf populations is yet to be performed for Y chromosome markers, the Iberian wolf subspecies exhibits high diversity with three distinct Y chromosome SNP lineages not found in any other wolves or dogs sampled so far. Therefore, the contemporary Iberian wolf subspecies retains significant Y chromosome diversity.
Iberian dogs revealed one-fourth of the Y-SNP lineages detected in Iberian wolves. This result may be related to ancient events such as domestication, as a very low mutation rate is estimated for the Y For the wolf samples sequenced, longer haplotypes (420 bp) that collapsed to the same haplotype when trimmed to 230 bp were given the same root label followed by a letter, for example, wH-1A, wH-1B, and wH-1C are indistinct when trimmed to 230 bp.
T A B L E 3 Summarized description of mitochondrial DNA haplotypes based on 420 bp of control region I identified in the 56 Iberian wolves sequenced. For detailed information per individual sample, see Table S5 wH-1C Vilà1999-lu33
For dogs, a mutation rate of 1.65 × 10 −9 substitutions per year and base pair, similar to humans, was assumed (Natanaelsson et al., 2006). The higher level of variability of Y-STRs observed in dogs is associated with the nature of the marker (higher mutational rate than for SNPs) and with the demographic expansion observed for this subspecies compared to the Iberian wolf's bottleneck in recent times (Sastre et al., 2011).
Two Iberian wolf samples share a low frequency patrilineage with dogs from South Africa and East Asia, the haplogroup HG9. This finding supports a wolf origin for this dog clade firstly described by Natanaelsson and collaborators (Natanaelsson et al., 2006). It is probable that an extensive study of wolf specimens with Y chromosome markers will reveal other wolf populations carrying haplotypes segregating within the HG9 haplogroup. The possibility that these specimens are male dog x female wolf hybrids (the most probable direction of wolf-dog hybridization (Vilà et al., 2003;Godinho et al., 2011Godinho et al., , 2015 but see also Hindrikson et al., 2012) is not supported by this study, as our survey on 74 male dogs in Iberia, including village dogs, revealed that they all belong to haplogroup HG1 with no exception, which is five SNPs apart from haplogroup HG9. Therefore, no Iberian dog analyzed so far carried Y chromosome haplotypes belonging to HG9. Moreover, these two particular wolves dated to 2008 carry typical wolf mtDNA haplotypes (wH-1A and wH-1B in this study), and typical wolf genotypes regarding autosomal microsatellite data. A sharp distinction between Iberian wolves and dogs was found using our dataset, which is in accordance with previous studies reporting levels of pairwise F ST ranging from 0.193 to 0.341 (p < .0001) depending on the set of autosomal markers used (Godinho et al., 2011). Our autosomal molecular data also suggest a low level of recent admixture between both subspecies in Iberia, although historical gene flow cannot be rejected (Fan et al., 2016). Population partitioning using a Bayesian clustering procedure and based on microsatellites loci reveals complete segregation between Portuguese village dogs or livestock guard dogs and Iberian wolves genotypes ( Figure S1), in contrast to the findings of Kopaliani and collaborators (Kopaliani et al., 2014) for the Caucasus region. It is important to emphasize here that our sampling strategy constrained the sampling of potential hybrids as we did not sample free-ranging or feral dogs. Although wolf-dog hybrids are known to occur in the  (Pires et al., 2006). Posterior probability support values ≥ 0.90 are shown. Outgroup and dog haplotypes are in gray; wolf haplotypes are in black. When originally distinct haplotypes within a species were collapsed to the same 420 bp or 230 bp haplotype, the label includes a representative haplotype followed by "sha" (shared). Wolf haplotype labels, for sequences retrieved from GenBank, consist of author name and date followed by original haplotype code (see text for accession number). Iberian wolf sequences generated in this study are labeled wH-1 to wH-5 (bold); haplotypes that collapsed to the same haplotype when trimmed to 230 bp have the same root label followed by a letter (e.g., wH-1A, wH-1B and wH-1C are indistinct when trimmed to 230 bp). Dog haplotype nomenclature (H01-H49) as in Pires et al. 2006; dog clades (a-d) as in Savolainen et al. 2002. Geographic distribution within the Iberian Peninsula (Portugal, Spain, or Iberia) and overall geographic distribution (in parenthesis) is indicated for wolf haplotypes; for the worldwide BI tree (b) geographic distribution is also shown for wolf haplotypes that group with Iberian wolves (in italics); wolf haplogroup nomenclature (1 and 2) as in Pilot et al., 2010 is shown for Iberian wolves haplotypes and other wolf haplotypes that group with them. -haplotype shared between wolf and dog. (c) Median-joining network of a 420 bp mitochondrial fragment of the control region I of Iberian wolves (subset of the data used in Figure 3 A-see text for details). Dog and wolf haplotype labels as in Figures 3 a and b. Frequency of haplotypes is only shown for wolves sequenced in this study-circle size proportional to frequency (% of occurrence inside the circle); remaining haplotypes represent sequences downloaded from GenBank (no frequency information available). Small white circles represent missing haplotypes (median vectors), and link length between nodes is proportional to nucleotide differences (number of mutations is shown next to the link when >1) wild (Randi & Lucchini, 2002;Vilà & Wayne, 1999;Vilà et al., 2003), including in Iberia (Godinho et al., 2011(Godinho et al., , 2015 but see also Echegaray & Vilà, 2010), hybridization is still a limited phenomenon in Iberia (Fan et al., 2016;Godinho et al., 2011Godinho et al., , 2015. Moreover, the search for the melanistic K locus mutation, a three nucleotide deletion derived from dogs (Caniglia et al., 2013;Randi et al., 2014), in 42 wolf tissue samples from this study and 26 forensic swabs (swabbed livestock wounds from wolf attacks) collected between 1991-2015 north and south of the Douro river in Portugal, revealed no animals with a hybrid genotype (Quaresma, 2016).
The seven mtDNA haplotypes described in this study, five of them unique to Iberian wolves, add to the diversity previously reported for this small and isolated gray wolf population. The fact that the mtDNA sequences trimmed to 230 bp still revealed two new haplotypes for Iberian wolves found in Portugal expands diversity previously reported for Iberia within the wolf haplogroup 1 (Pilot et al., 2010) and adds representation of wolf haplogroup 2 (Pilot et al., 2010). This reinforces the evidence for unreported diversity in Iberian wolves. Low genetic diversity has been reported for wolf populations that have experienced population decline and fragmentation (Randi et al., 2000).
Wolf populations in Italy, Scandinavia, Switzerland, and France all have fewer haplotypes than the Iberian population (Ellegren, Savolainen, & Rosen, 1996;Pilot et al., 2010;Randi et al., 2000). This can most likely be explained by the fact that, although the Iberian wolf population, as other European wolf population, has been experiencing demographic decline since the Late Pleistocene (Pilot et al., 2014) and underwent a steep human-mediated decline in the early 20th century, its effective population size was never as low as in other regions. Nevertheless, this sample exhibits a wolf Y-SNP profile: wolf-typical nucleotides were detected at the described diagnostic loci. Being a male wolf, for this sample to belong to a first-generation hybrid, it could have only resulted from the cross between a female dog and a male wolf, which is considered rare. The first record of this kind of hybrid in Europe dates from 2008-2009 in Latvia, a single specimen, which was reported in Hindrikson et al. (2012). Given the high genetic distance from all other Iberian wolf mtDNA haplotypes and the fact that it has never been previously found in any other wolf population, an older hybridization event cannot be excluded. Alternatively, the shared haplotype may represent an ancestral variant preserved until recently, as this haplotype segregates within wolf haplogroup 2, which was extensively detected in ancient wolf samples from western Europe (Pilot et al., 2010). Once again, the study of historical and ancient samples from the Iberian Peninsula will help to clarify this issue and possibly to date the oldest evidence of this haplotype in Iberian wolves.

| Impacts for the conservation of the Iberian wolf
We have shown, using both nuclear and mitochondrial data, that the Iberian wolf displays greater diversity than previously reported, with new maternal and paternal haplotypes disclosed.
We hypothesize that mtDNA haplotypes shared between wolves and dogs may represent relic common ancestral Canis lineages that did not diverge after the domestication of dogs. Although hybridization has been recorded in Iberia, the fact that in our dataset the percentage of dog-wolf mtDNA haplotype sharing in this area is roughly half of what was found elsewhere supports the conclusion that hybridization in Iberia, more specifically in Portugal (as most wolf samples are from this region), is a rare event. If disturbance of wolf habitats in Iberia continues, due to both habitat loss and/or fragmentation, in addition to an increase in the feral dog population, which largely overlaps with the wolf range (Álvares, 2011;LIFE Co-EX 2008), the genetic identity of the Iberian wolf is threatened.
Wolves found in Iberia represent a reservoir of unique genetic and ecomorphologic variants of the gray wolf. Therefore, according to Hofreiter and Barnes' (2010) proposal, this stresses the need for maintenance and even intensification of multiple measures toward the long-term sustainability of this locally adapted, isolated, and peripheral population, the Iberian wolf.

CONFLICT OF INTEREST
None declared.