Investigations of genetic diversity and domestication in South American camelids (SAC) have relied on autosomal microsatellite and maternally-inherited mitochondrial data. We present the first integrated analysis of domestic and wild SAC combining male and female sex-specific markers (male specific Y-chromosome and female-specific mtDNA sequence variation) to assess: (i) hypotheses about the origin of domestic camelids, (ii) directionality of introgression among domestic and/or wild taxa as evidence of hybridization and (iii) currently recognized subspecies patterns. Three male-specific Y-chromosome markers and control region sequences of mitochondrial DNA are studied here. Although no sequence variation was found in SRY and ZFY, there were seven variable sites in DBY generating five haplotypes on the Y-chromosome. The haplotype network showed clear separation between haplogroups of guanaco-llama and vicuña-alpaca, indicating two genetically distinct patrilineages with near absence of shared haplotypes between guanacos and vicuñas. Although we document some examples of directional hybridization, the patterns strongly support the hypothesis that llama (Lama glama) is derived from guanaco (Lama guanicoe) and the alpaca (Vicugna pacos) from vicuña (Vicugna vicugna). Within male guanacos we identified a haplogroup formed by three haplotypes with different geographical distributions, the northernmost of which (Peru and northern Chile) was also observed in llamas, supporting the commonly held hypothesis that llamas were domesticated from the northernmost populations of guanacos (L. g. cacilensis). Southern guanacos shared the other two haplotypes. A second haplogroup, consisting of two haplotypes, was mostly present in vicuñas and alpacas. However, Y-chromosome variation did not distinguish the two subspecies of vicuñas.

The domestication of wild vicuña and guanaco by early pre-Inca cultures is an iconic example of wildlife management and domestication in the Americas. Although domestic llamas and alpacas were clearly selected for key, yet distinct, phenotypic traits, the relative patterns and direction of selection and domestication have not been confirmed using genetic approaches. However, the detailed archaeological records from the region suggest that domestication was a process carried out under significant control and planning, which would have facilitated coordinated and thus extremely effective selective pressure to achieve and maintain desired phenotypic traits. Here we link patterns of sequence variation in two well-characterised genes coding for colour variation in vertebrates and interpret the results in the context of domestication in guanacos and vicuñas. We hypothesise that colour variation in wild populations of guanacos and vicunas were strongly selected against. In contrast, variation in coat colour variation in alpaca was strongly selected for and became rapidly fixed in alpacas. In contrast, coat colour variants in llamas were of less economic value, and thus were under less selective pressure. We report for the first time the full sequence of MC1R and 3 exons of ASIP in 171 wild specimens from throughout their distribution and which represented a range of commonly observed colour patterns. We found a significant difference in the number of non-synonymous substitutions, but not synonymous substitutions among wild and domestics species. The genetic variation in MC1R and ASIP did not differentiate alpaca from llama due to the high degree of reciprocal introgression, but the combination of 11 substitutions are sufficient to distinguish domestic from wild animals. Although there is gene flow among domestic and wild species, most of the non-synonymous variation in MC1R and ASIP was not observed in wild species, presumably because these substitutions and the associated colour phenotypes are not effectively transmitted back into wild populations. Therefore, this set of substitutions unequivocally differentiates wild from domestic animals, which will have important practical application in forensic cases involving the poaching of wild vicuñas and guanacos. These markers will also assist in identifying and studying archaeological remains pre- and post-domestication.

The vicuña ( Vicugna vicugna ) is the most representative wild ungulate of the high Andes of South America with two recognized morphological subspecies, V. v. mensalis in the north and V. v. vicugna in the south of its distribution. Current vicuña population size (460,000–520,000 animals) is the result of population recovery programs established in response to 500 years of overexploitation. Despite the vicuña’s ecosystemic, economic and social importance, studies about their genetic variation and history are limited and geographically restricted. Here, we present a comprehensive assessment of the genetic diversity of vicuña based on samples collected throughout its distribution range corresponding to eleven localities in Peru and five in Chile representing V. v. mensalis , plus four localities each in Argentina and Chile representing V. v. vicugna. Analysis of mitochondrial DNA and microsatellite markers show contrasting results regarding differentiation between the two vicuña types with mitochondrial haplotypes supporting subspecies differentiation, albeit with only a few mutational steps separating the two subspecies. In contrast, microsatellite markers show that vicuña genetic variation is best explained as an isolation by distance pattern where populations on opposite ends of the distribution present different allelic compositions, but the intermediate populations present a variety of alleles shared by both extreme forms. Demographic characterization of the species evidenced a simultaneous and strong reduction in the effective population size in all localities supporting the existence of a unique, large ancestral population (effective size ∼50,000 individuals) as recently as the mid-Holocene. Furthermore, the genetic variation observed across all localities is better explained by a model of gene flow interconnecting them rather than only by genetic drift. Consequently, we propose space “continuous” Management Units for vicuña as populations exhibit differentiation by distance and spatial autocorrelation linked to sex biased dispersal instead of population fragmentation or geographical barriers across the distribution.