Assessing the levels of intraspecific admixture and interspecific hybridization in Iberian wild goats (Capra pyrenaica)

Abstract Iberian wild goats (Capra pyrenaica, also known as Iberian ibex, Spanish ibex, and Spanish wild goat) underwent strong genetic bottlenecks during the 19th and 20th centuries due to overhunting and habitat destruction. From the 1970s to 1990s, augmentation translocations were frequently carried out to restock Iberian wild goat populations (very often with hunting purposes), but they were not systematically planned or recorded. On the other hand, recent data suggest the occurrence of hybridization events between Iberian wild goats and domestic goats (Capra hircus). Augmentation translocations and interspecific hybridization might have contributed to increase the diversity of Iberian wild goats. With the aim of investigating this issue, we have genotyped 118 Iberian wild goats from Tortosa‐Beceite, Sierra Nevada, Muela de Cortes, Gredos, Batuecas, and Ordesa and Monte Perdido by using the Goat SNP50 BeadChip (Illumina). The analysis of genotypic data indicated that Iberian wild goat populations are strongly differentiated and display low diversity. Only three Iberian wild goats out from 118 show genomic signatures of mixed ancestry, a result consistent with a scenario in which past augmentation translocations have had a limited impact on the diversity of Iberian wild goats. Besides, we have detected eight Iberian wild goats from Tortosa‐Beceite with signs of domestic goat introgression. Although rare, hybridization with domestic goats could become a potential threat to the genetic integrity of Iberian wild goats; hence, measures should be taken to avoid the presence of uncontrolled herds of domestic or feral goats in mountainous areas inhabited by this iconic wild ungulate.


| INTRODUC TI ON
The genetic diversity of wild animal species has been modified by multiple factors related with human activity. Habitat destruction and fragmentation combined with overhunting, climate change, and the introduction of invasive animals and plants have caused severe reductions in the genetic diversity and fitness of wild species, leading, in some cases, to their extinction (Fahrig, 2003;Pimm et al., 2014).
Human activities, either intentionally or not, might have also contributed to increase the genetic diversity of wild species. Translocation, which implies the deliberate release of animals from one location to another with the goal of reinforcing, introducing, or reintroducing a species within its indigenous range (Griffith et al., 1989), can be effective in enhancing genetic diversity (Chipman et al., 2008).
Obviously, it can also have adverse effects on resident animals at release sites, including the spread of diseases that could cause drastic population bottlenecks (Chipman et al., 2008). Moreover, increased stress and mortality of released animals may limit the potential benefits of translocations (Chipman et al., 2008).
Hybridization between wild animals and livestock herds, which is largely unintentional, can also increase the genetic diversity of wild species, by introducing completely new alleles and genotypes, at the expense of decreasing adaptive potential due to outbreeding depression and behavioral changes, for example, reduced predator and human avoidance (Barbato et al., 2017;Goedbloed et al., 2013).
Even in the cases in which interspecific hybridization is a rare event, it can lead to long-lasting changes in the genomic architecture of the affected wild species (Schwenk et al., 2008). While reduction of genetic variation mediated by humans has been documented in wild animals and its consequences have been thoroughly assessed (Abascal et al., 2016;Grossen et al., 2020), few reports have addressed the potential impact of translocation and interspecific hybridization on the genetic diversity of wild species (Shackleton, 1997).
The Iberian wild goat (Capra pyrenaica, also known as Iberian ibex, Spanish ibex, and Spanish wild goat) is a wild goat ungulate native to the Iberian Peninsula which inhabits mountainous and rocky areas and feeds on shrubs, bushes, and grasses (Acevedo & Cassinello, 2009a;Granados et al., 2001Granados et al., , 2007. According to Cabrera (1911Cabrera ( , 1914, in the early 20 th century, there were four Iberian wild goat subspecies, namely C. p. hispanica (CPH, south and east of the Iberian Peninsula), C. p. victoriae (CPV, center and northwest of the Iberian Peninsula), C. p. lusitanica (CPL, Galicia and north of Portugal, extinct in the 19 th century), and C. p. pyrenaica (CPP, Pyrenees), which became extinct two decades ago . The Iberian wild goat was abundant during the Middle Ages but it experienced a sustained and strong demographic decline during the 19-20 th centuries as a consequence of the growing hunting pressure (particularly during the 1940s-1970s) and habitat loss and fragmentation (García-González, 2011;Pérez et al., 2002).
The strong reduction of genetic diversity produced by this process of demographic contraction has been previously reported (Amills et al., 2004;Angelone et al., 2018). Strong signs of genetic differentiation among Iberian wild goat populations due to reproductive isolation and substantial genetic drift associated with severe genetic bottlenecks have also been described (Amills et al., 2004;Angelone et al., 2018). In the last decades, the creation of a network of national parks and protected areas, the absence of predators, reforestation policies, and the progressive abandonment of rural activities have contributed to the recovery and subsequent expansion of Iberian wild goats (Acevedo & Cassinello, 2009b).
Iberian wild goats constitute a valuable model to explore the impact of translocation and hybridization on genetic diversity.
Restocking/repopulation translocations have favored gene flow between distant populations (Acevedo & Cassinello, 2009a;Crampe, 1991). The most comprehensive report to date analyzing the variability of 333 Iberian wild goats with a panel of 30 microsatellites did not show any evidence of genetic signatures typically associated with translocations and population admixture (Angelone-Alasaad et al., 2017). However, this outcome might be caused by the limited resolution of the microsatellite panel employed in such study.
Excellence Severo Ochoa 2020-2023, Grant/Award Number: CEX2019-000902-S, CGL2012-40043-C02-01, CGL2012-40043-C02-02 and CGL2016-80543-P; Spanish Ministry of Education, Grant/Award Number: BES-C-2017-0024 and FPU15/01733; CAPES Foundation-Coordination of Improvement of Higher Education, Ministry of Education of the Federal Government of Brazil BeadChip (Illumina). The analysis of genotypic data indicated that Iberian wild goat populations are strongly differentiated and display low diversity. Only three Iberian wild goats out from 118 show genomic signatures of mixed ancestry, a result consistent with a scenario in which past augmentation translocations have had a limited impact on the diversity of Iberian wild goats. Besides, we have detected eight Iberian wild goats from Tortosa-Beceite with signs of domestic goat introgression. Although rare, hybridization with domestic goats could become a potential threat to the genetic integrity of Iberian wild goats; hence, measures should be taken to avoid the presence of uncontrolled herds of domestic or feral goats in mountainous areas inhabited by this iconic wild ungulate.

K E Y W O R D S
Capra pyrenaica, genetic diversity, high-density SNP arrays, Iberian ibex, introgression In addition, the impact of domestic goat introgression on the genetic diversity of Iberian wild goats is not well known yet. Alasaad et al. (2012) reported the mating of one captive Iberian wild goat male with domestic goats and the obtaining of viable offspring.
Hybrids between Alpine ibexes (Capra ibex) and domestic goats have been also described (Giacometti et al., 2004). Moreover, Angelone et al. (2018) reported the segregation, in Iberian wild goats from four Southern Spain locations (Sierras de Cazorla, Segura and las Villas Natural Park, El Hosquillo in Serranía de Cuenca Natural Park, Sierra del Mencal, and Cabañeros National Park), of one major histocompatibility complex class II DRB1 allele, MHC DRB1*7, identical to another one reported in domestic goats. They hypothesized that this result could be due to either the maintenance of ancient polymorphisms by balancing selection or, alternatively, introgressions from domestic goats through interspecific hybridization, and they concluded that this matter should be clarified in future . By using a high-throughput single nucleotide polymorphism (SNP) genotyping approach, we expect to answer this question and find out whether domestic goat introgression has had a significant impact on the genetic diversity of Iberian wild goats.
In summary, the main goal of the current work is to investigate the impact of intraspecific (translocations) and interspecific (hybridization between wild and domestic goats) gene flow on the diversity of Iberian wild goats by genotyping 118 individuals with a SNP assay.

| Study areas and historical description of populations
In this work, we have investigated three CPH populations (Tortosa-Beceite, Muela de Cortes, and Sierra Nevada), two CPV populations (Gredos and Batuecas) and one CPP individual (Ordesa and Monte Perdido). All these Iberian wild goat populations underwent strong bottlenecks during the 19 th and 20 th centuries but, as shown in Table 1, during the last six decades they have experienced an accelerated demographic expansion due to the lack of predators, human depopulation in rural areas, and protected status (Acevedo & Cassinello, 2009b). Hybridization with domestic goats has not been reported in any of the five populations mentioned before. While the cohabitation of domestic goats and Iberian wild goats has been described as a risk factor for the transmission of certain diseases (Astorga Márquez et al., 2014), to the best of our knowledge the spatial proximity between wild and domestic goat populations has not been thoroughly investigated in Spain. Part of the translocations among Iberian wild goat populations have been documented, and such information can be found in Figure 1.

| Isolation of genomic DNA from Iberian wild goat samples
We used two different batches of CPV and CPH samples. The first batch was reported by Jiménez et al. (1999), as well as by Amills et al. (2004), and consisted of (1) CPV: liver samples from seven and 14 Iberian wild goats from Batuecas and Gredos, respectively; (2) CPH: blood samples from 27 and five Iberian wild goats from Tortosa-Beceite and Sierra Nevada, respectively, and five liver samples from Iberian wild goats inhabiting Muela de Cortes. A second batch included blood or solid tissue (muscle, spleen, or ear cartilage) samples from 59 CPH individuals from Tortosa-Beceite (N = 43, 2010-2019), Muela de Cortes (N = 7, 2017-2019), and Sierra Nevada (N = 9, 2006. Finally, one muscle sample from one of the last CPP representatives was collected in the location of Ordesa and Monte Perdido in 1996, before the extinction of this subspecies. Genomic DNA was isolated from blood samples as previously reported (Amills et al., 1996), while a standard phenol-chloroform protocol was used to purify genomic DNA from solid tissues (Sambrook & Russell, 2006). The second batch of samples and the CPP sample had not been analyzed in previous genetic studies.    Huesca, 1960s). A few years later, only 2 males and 4 females were left. While one of the males was transferred to Gredos and the other one died, the whereabouts of the females are unknown although it is highly unlikely that they reached Ordesa (García-González, 1989). Iberian wild goats from Gredos were also translocated to Las Batuecas (Salamanca, 1970s), Regional Hunting Reserve of Riaño (León), and La Pedriza (current Sierra de Guadarrama National Park, Madrid-Segovia) during the 90s. Before 1995, the introduction of specimens from Gredos to private farms in the Montes de Toledo is also documented (Acevedo et al., 2011). There were also translocations of individuals from Batuecas to Sierra de Guadarrama (Madrid) and to Riaño (León). The Natural Park of Invernadero also received Iberian wild goats from Batuecas (Prada & Herrero, 2013) and possibly from Gredos (Crampe, 2020), although this latter translocation event is not completely confirmed. An unsuccessful reintroduction attempt was made, between 1957 and 1962, in the Covadonga National Park (Asturias) with 14 individuals from Gredos and Cazorla (Jaen), as reported by Arenzana et al. (1964). At the end of the 90s, translocations from Riaño to Los Ancares (León) and, in 2005-2007, to Mampodre (León) are also known. In 2018, CPV individuals from Guadarrama were transferred to the Pyrenees National Park and the Ariège Pyrenees Regional Park, both in France, and to the Valle de Arán (Lleida). Concerning Capra pyrenaica hispanica (CPH), Iberian wild goats from Cazorla were taken to private farms in the Montes de Toledo, Sierra Morena (Jaén-Ciudad Real), Serranía de Cuenca (Hosquillo), Sierra de Guara (Huesca), Sierra de Baza (Granada), and Muela de Cortes (Valencia) during the 1960s and 1970s. More recently animals from Cazorla were transferred to an enclosure in the Serra del Montgrí (Girona), but they escaped and formed a population of more than one hundred specimens. At the end of the nineties, Iberian wild goats were taken from Tortosa to Montserrat (Barcelona). Finally, CPH from Sierra Nevada were introduced in the Serranía de Ronda (Málaga) and in the Sierra de Baza (Granada) during the 1970s and 1980s, and to Sierra de Mágina (Jaén). CPH from Sierra Nevada were also brought to enclosures in Almuñecar (Granada), Garcipollera (Huesca), and Cumbres Mayores (Huelva) at the end of the 90s. Finally, during the first decade of this century, Iberian wild goats from the Sierra Nevada have been brought to the Moratalla and Caravaca mountains in the Murcia region and Sierra de Orce in the north of the province of Granada iques/ micro array s/array -data-analy sis-exper iment al-desig n/genom estud io.html).

| Genotyping with the Goat SNP50 BeadChip
Preprocessing and filtering of data were carried out with the PLINK v.1.7 software (Purcell et al., 2007). More specifically, markers with a GenTrain score (Illumina descriptive statistic related to clustering quality) lower than 0.8, unmapped SNPs, as well as SNPs that mapped to the X chromosome on the goat reference genome assembly Capra hircus-ARS1 (Bickhart et al., 2017, https://www.ensem bl.org/Capra_hircu s/) and those with a minimum allele frequency (MAF) lower than 0.01 (-maf 0.01) were filtered out. Markers with an individual missingness rate with more than 50% missing genotypes SNPs (-mind 0.5) and SNP with a missingness across samples greater than 1% (-geno 0.01) were also removed. After applying these filtering criteria, 21,621 SNPs were retained for genetic analyses.
To compare the diversity of Iberian wild goats and domestic goats, we used a previously published caprine data set corresponding to 50 domestic goats (ten individuals per breed) from Northern Spain (Bermeya and Blanca de Rasquera) and Southern Spain

| Population genetics analyses
The in the non-hybrid individuals so they cannot be used). For the sake of clarity, the dataset used for each one of the analysis carried out in our study is specified below. We considered as putative hybrids the eight individuals that did not collapse in the MDS plot shown in Figure S1 and that, in addition, showed signatures of domestic goat introgression in the admixture analyses (see below). The remaining 110 Iberian wild goats were considered as non-hybrids, although we cannot completely rule out the possibility that a number of them may carry a domestic goat genetic component not detectable with our methods.

| Multidimensional scaling and estimation of diversity parameters in Iberian wild goats
The PLINK v1.7 software (Purcell et al., 2007) was used to carry out sample clustering based on the multidimensional scaling (MDS) of allele information from retrieved SNPs (-cluster -mds-plot 2 eigendecomp eigvals). We did four MDS analyses: (1) (Purcell et al., 2007) and the data set of 1001 SNPs. We chose F hat2 as an estimate of inbreeding because, in a previous study focused on domestic goats, this statistic showed a high correlation (r = 0.88, p-value = 1.00E-04) with F ROH (Cardoso et al., 2018). In contrast with other inbreeding coefficients, F hat2 can take negative values (when the count of observed homozygotes is lower than the expected count of homozygotes) because it is not defined as a probability but as an excess of homozygosity-based inbreeding estimate (Purcell et al., 2007). The -hardy command was used to compute H o and H e , while the -ibc command was used to estimate the F hat2 coefficient. Nucleotide diversity was computed for each population on a per-site basis (π, command: -site-pi) using the VCFtools software (Danecek et al., 2011). Confidence intervals (CI) for each parameter were calculated according to the following formula: where X is the sample mean of the parameter for each population, 1.96 is the Z-score corresponding to a 95% confidence interval, and SE is the standard error of the mean (Sim & Reid, 1999).
Genome-wide identity by descent (IBD) between pairs of samples was estimated with the PI-HAT coefficient, which describes the probability of sharing 0, 1, or 2 alleles IBD by pairs of individuals from the same homogeneous random-mating population (Purcell et al., 2007). Heatmap plots were built in R software by using the ggplot2 package (Wickham, 2016).

| Examining the ancestry of Iberian wild goats with admixture
We used the Admixture software (Alexander et al., 2009)  (1) CI = X ± 1.96*SE for admixture proportions were inferred with Equation 1. The optimal K-value was the one with the lowest cross-validation error, as determined with the method of Alexander and Lange (2011). The Pophelper package for R (Francis, 2017) was used to process the output results from the Admixture analysis.

| Performance of an f3 test of admixture in eight putative hybrid Iberian wild goats
In the Admixture analysis, eight individuals from Tortosa-Beceite showed genomic signatures of introgression by domestic goats. We used the qp3pop program, included in the ADMIXTOOLS software package (Patterson et al., 2012), and the set of 21,621 SNPs to carry out a 3-population test in the form f3(admixed Tortosa-Beceite; Tortosa-Beceite, Malagueña), that is, we selected the non-admixed

| Population structure and diversity of Iberian wild goats
We have investigated the population structure of 118 Iberian wild goats with 1001 markers segregating in both non-hybrid and putative hybrid individuals. We did not use the remaining 20,620 mark- As expected, the genetic diversity of Iberian wild goats was lower than that of domestic goats (  (Table 2). Iberian wild goats showed lower nucleotide diversity (π = 0.181) than domestic goats (π = 0.402).
Consistent with heterozygosity measurements, Sierra Nevada and Batuecas populations had the lowest nucleotide diversity values (π ≅ 0.160) among the investigated populations.
As shown in Table 2, the inbreeding F hat2 coefficient calculated with PLINK v.1.7 (Purcell et al., 2007) Table 3 In contrast, when these pairwise comparisons were made at the within-population level (Figure 4a), the degree of genetic similarity among individuals increased substantially. In domestic goats, PI-HAT coefficients reached values close to zero for the majority of pairwise F I G U R E 2 Four representative Genoplots obtained with the GenomeStudio software through the analysis of Goat SNP50 BeadChip (Illumina) data corresponding to Iberian wild goats (yellow) and domestic goats (red, purple, and blue). Genotypes are called for each sample (dots) by taking into account their signal intensity (Norm R, y-axis) and allelic intensity ratio (Norm Theta, x-axis) relative to canonical cluster positions (dark shading). (a) Genoplot showing the genotypes obtained for SNP 58771. It can be seen that this SNP is successfully called and it segregates in both the Iberian wild goat and domestic goat populations. (b) Genoplot showing the genotypes obtained for SNP 27056. This SNP is successfully called and it segregates in domestic goats (two genotypes are called) but not in Iberian wild goats (a single genotype is called). The intensity of the signal (norm R) is similar in domestic goats and Iberian wild goats. (c) Genoplot showing the genotypes obtained for SNP 2756. This SNP is successfully called and it segregates in domestic goats (three genotypes are called) but not in Iberian wild goats (a single genotype is called and with a weaker intensity than the one corresponding to domestic goats). (d) Genoplot showing the genotypes obtained for SNP 52788. This SNP is not successfully called in Iberian wild goats (norm R is much lower than that observed in domestic goats) comparisons, and even when they comprised individuals drawn from the same breed (Figure 4b).

| Examining the ancestry of Iberian wild goats and detecting genomic signatures of admixture
The results of the Admixture analysis corresponding to 118 Iberian wild goats and 50 domestic goats characterized with 1001 SNPs are shown in Figure 5 (K = 2-7, K = 7 is the number of clusters with the lowest cross-validation error, Figure S2). The percentages of ancestry for each one of the Iberian wild goat populations are displayed in Figure S3 for K = 7. In the majority of Iberian wild goat populations, admixture was low or non-detectable ( Figure 5 and Figure S3). The only exception were Iberian wild goats from Gredos and Batuecas, which showed signs of a common ancestry even at K = 7 ( Figure 5).
The CPP sample also seemed to have different ancestries ( Figure 5 and Figure S3) but in this case, results are not reliable because allelic frequencies cannot be inferred from a single individual. According to the Admixture analysis, two individuals from Tortosa-Beceite (Tortosa-Beceite_22 and Tortosa-Beceite_23) showed evidence of having Sierra Nevada ancestry, while one individual from Sierra Nevada (Sierra_Nevada_9) displayed signs of Muela de Cortes ancestry ( Figure 5). We did not calculate f3-statistics for these three potentially admixed Iberian wild goats because they cannot be reliably estimated with just 1001 SNPs.

| Performance of an f3 test of admixture in eight putative hybrid Iberian wild goats
The eight putative hybrids from Tortosa-Beceite displayed negative f3 values, indicative of admixture between the two Tortosa-Beceite and Malagueña source populations ( Figure 6). The Z-scores were high and significant (Table S2). These results were consistent even when different source populations were selected, for example, with Gredos and Bermeya, and Batuecas and Malagueña as source populations (Table S3).

F I G U R E 3 Multidimensional scaling (MDS) plot of Iberian wild goats. (a) MDS plot including 118 Iberian wild goat samples. This analysis is based on 1001 SNPs from the Goat SNP50 BeadChip (Illumina), (b) MDS of Iberian wild goat individuals. This MDS plot includes 118 Iberian wild goat samples and it is based on 894 SNPs from the Goat SNP50 BeadChip (Illumina), that is, 1001 SNPs minus the SNPs showing missing
values in the CPP sample (the individual with the highest genotype missingness rate). We carried out this analysis because we wanted to test whether the centrality of the CPP sample in the MDS is caused by its high missingness rate (MDS analyses tend to "locate" samples close to the center when missingness is high). By comparing a and b, it becomes clear that this is not the case.   and exponential rates, respectively (Miller et al., 2012). Based on the data presented by Miller et al. (2012), domestic goats and Iberian wild goats, two species that diverged 1.5 Mya (Lalueza-Fox et al., 2005), should share, on average, less than 10% polymorphic sites.
In accordance with the data reported by Amills et al. (2004), we have detected reduced observed and expected heterozygosities in Iberian wild goats when compared to domestic goats (Table 2).
Nucleotide diversities were also lower in Iberian wild goats than in Diversity across the five Iberian wild goat populations studied was similar, a result that is consistent with previous microsatellite data (Amills et al., 2004;Angelone-Alasaad et al., 2017). These findings indicated that the bottlenecks suffered by this species during the 19 th and 20 th centuries had widespread effects on its overall genetic variation (Amills et al., 2004;Pérez et al., 2002). Moreover, F hat2 coefficients reached values of 0.442-0.593 in the Iberian wild goats, while in the domestic goats they were close to zero (Table 2), a result consistent with previous reports (Cardoso et al., 2018). The PI-HAT values calculated in pairwise comparisons were also substantially higher in Iberian wild goats than in domestic goats, reflecting a significant proportion of IBD between pairs of Iberian wild goats coming from the same population ( Figure 4a). goats, observed in our study is not an artifact entirely produced by ascertainment bias. Indeed, the existence of inbreeding and low variation in Iberian wild goats is consistent with the dramatic genetic bottlenecks suffered by this ungulate species (Amills et al., 2004).

| Iberian wild goat populations are strongly differentiated
The classification of Iberian wild goats in four subspecies proposed by Cabrera (1911Cabrera ( , 1914 raised controversy because it relied exclusively on a limited number of highly variable phenotypic traits recorded in a low number of individuals (Angelone-Alasaad et al., 2017). Our study demonstrated that the F ST coefficients among Iberian wild goat populations were much higher (F ST > 0.287) than those measured among domestic goat populations (F ST < 0.072), and the latter were similar to those reported by Manunza et al. (2016) for the same populations using a data set of 39,257 SNPs. Moreover,  (Table 3).

F I G U R E 6
Overall, these results indicated that the classification of Cabrera (1911,1914) is not well supported by genetic evidence. Consistently, Angelone-Alasaad et al. (2017) analyzed 333 Iberian wild goats with a panel of microsatellites and found that the CPH populations from Sierra Nevada and Maestrazgo had a degree of genetic differentiation similar to that observed among CPH and CPV specimens. The strong genetic differentiation between Iberian wild goat populations, irrespective of their assignment to one subspecies or another, is probably due to intense drift associated with past genetic bottlenecks combined with prolonged geographic isolation (Pérez et al., 2002). The only exception to this general trend were the CPV populations of Gredos and Batuecas, which had a weak degree of genetic differentiation (F ST = 0.035, Table 3), probably because the Batuecas population originated from restockings with individuals from Gredos (Pérez et al., 2002).
In the MDS analyses shown in Figure 3a,b, the only CPP individual did not cluster with its CPV and CPH counterparts, although it was located close to the Muela de Cortes population (CPH). This result also needs to be interpreted with caution because allele frequencies cannot be estimated from just one individual. However, CPP suffered strong population bottlenecks and an irreversible genetic erosion that enhanced genetic differentiation before extinction (García-González & Herrero, 1999). Moreover, there was no recent gene flow between CPP and the other CPV and CPH subspecies, which strengthened the progressive genetic differentiation of CPP. The analysis of additional museum CPP samples would be needed to accurately characterize the genetic relationships between this extinct subspecies and CPH and CPV.

| Genomic signatures of past restocking translocations are rarely detected in current Iberian wild goat populations
As pointed out by Acevedo and Cassinello (2009a), Iberian wild goat distribution is the result of both natural and artificial expansion processes. Most translocations were carried out after 1970, particularly during the 1980s and 1990s (Acevedo & Cassinello, 2009a).
Although there is a general consensus indicating that Iberian wild goats were the subject of numerous restocking/repopulation translocations (movement of individuals into a population of conspecifics), the majority of them are poorly documented (Angelone-Alasaad et al., 2017). In Figure 1, we provide a description of translocations that have been reported so far.
One of the main goals of the current work was to infer whether  (Dickens et al., 2010). Another factor could be competition for food resources between the released animals and the residing conspecific population, or endemic diseases that might have decimated the incoming individuals. Avian translocations have a high failure rate (Dickens et al., 2009), and documented success in carnivores is also low (Macdonald, 2009). In the case of the American marten population of Wisconsin, augmentation provided minimal genetic and demographic rescue contributions (Manlick et al., 2017).
A failed reintroduction of 14 Iberian wild goats from Gredos and Cazorla to the National Park of Covadonga in 1957-1962 has been also reported (Arenzana et al., 1964). Despite their low diversity, Iberian wild goat populations are increasing in numbers at a fast pace (Acevedo & Cassinello, 2009a;Acevedo et al., 2007). In light of this, we consider that gene flow between naturally expanding populations could be established without the need of human intervention.  According to our data, the introgression of Iberian wild goats by domestic goats does not seem to be widespread, probably because the interbreeding of these two species in the wild is a rare event, and moreover, some degree of reproductive incompatibility may exist (Herrero, Fernández-Arberas, Prada, & García-Serrano, 2013).
However, hybridization cannot be disregarded as a potential threat to the genetic conservation of Iberian wild goats, since its prevalence might increase as a result of the rapid expansion, in numbers and geographic range, of Iberian wild goat populations (Acevedo & Cassinello, 2009;Acevedo et al., 2007;Perea et al., 2015). The introgression of Iberian wild goats by domestic goats could imply a decrease of reproductive potential and fitness, the introduction of maladaptive alleles, a reduction or loss of genetic integrity, and it may also have legal implications regarding individual or population conservation status (Leonard et al., 2013). Transmission of infectious diseases by domestic goats is another factor that could have important adverse effects on the viability of Iberian wild goat populations (Brennan et al., 2014). Of particular concern are feral goats, which can adapt quite successfully to mountainous habitats (Herrero, Fernández-Arberas, Prada, García-Serrano, & García-González, 2013). For instance, in the Sierra de Guara, a population of Iberian wild goats coexists with almost one thousand feral goats (Herrero, Fernández-Arberas, Prada, García-Serrano, & García-González, 2013) descending from individuals probably abandoned by their owners. Extensive field surveys based on SNP markers should be conducted to evaluate the presence and frequency of hybrid individuals in current Iberian wild goat populations, with special emphasis on those inhabiting geographic areas in which the presence of uncontrolled herds of feral goats is well documented.

ACK N OWLED G M ENTS
This research was funded by the European Regional Development Many thanks to Laura Botigué for her help and assistance in carrying out the f3 tests of admixture.

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are openly available