The pig is an artiodactyl mammal belonging to the Suidae family. Paleontological data suggest the earliest pig species inhabited the forests and swamps of Eurasia about forty million years ago. Endemic to the Old World, pigs arrived in the New World with European colonists. They were first introduced to the islands of Hispaniola, Puerto Rico, Cuba and Jamaica from the Canary Islands on Columbus’s second voyage in 1493. Thirty-two years later, King Carlos V ordered 300 pigs to be brought along on the Rodrigo de Bastidas expedition, leaving Hispaniola to found Santa Marta on the northern coast of South America¹. These may be the first pigs to have arrived in what was to become Colombia, although pigs were also introduced into present-day Cordoba Department during the Spanish conquest from 1500 to 1550².

Domestic pigs (Sus scrofa domestica) are still present in farming communities as an important income and protein source. Over the centuries, pigs in the region have adapted to the tropical conditions, successfully reproducing despite challenges such as an incomplete diet, lack of water, unfavorable crosses, disease and inadequate growing practices³. Population genetic data on pigs in the Colombian Caribbean and especially domestic pigs (Sus scrofa domestica) in Cordoba, is scarce, requiring special attention to identify their genetic status. For this reason, studying genetic variation in domestic pigs in Momil, Cerete and Tierralta, Cordoba (Colombia), will identify their state of genetic variability, a conclusive element to determine strategies for breeding and genetic conservation programs⁴ and rating it as genetic patrimony of the Cordoba Region.

Study of DNA polymorphisms with polymerase chain reaction (PCR) analysis involves in vitro amplification of a specific DNA fragment⁵. It is currently one of the most frequently used molecular genetics techniques, and the most used in microsatellite marker and many other genetic population study techniques. Microsatellite markers consist of from two to seven continuously repeated base pairs of DNA segments. These are called di-, tri- and tetranucleotide segments, and are a valuable data source due to their abundance throughout eukaryote genomes, high polymorphism, stability, and data reproducibility. They are also relatively simple to amplify to generate DNA fragment patterns specific to individuals that include combinations from the paternal and maternal lines⁶.

Very little data are available on pig populations in the Caribbean region of Colombia, particularly the domestic pig populations of Cordoba Department. Identifying the genetic status of these populations requires studying genetic variation within them. The present study objective was to evaluate degree of genetic differentiation among domestic pig populations in the towns of Momil, Cereté and Tierralta, Cordoba Department, Colombia, based on variation in microsatellites. Identification of genetic variability is vital to developing production strategies, and genetic conservation and improvement programs⁴, as well as evaluating the genetic heritage of the region’s pig populations.

Hair samples were collected from pigs grown on family farms in the towns of Momil (9°14’16" N; 75°36’30" W), Cereté (08°53’08" N; 75°47’48" W) and Tierralta (08°10’34" N; 76°03’46" W), Cordoba Department. Samples were also collected from pigs at a commercial pig farm using crosses between Landrace, Yorkshire, Large White, Duroc and Pietrain (Table 1). Based on previous studies^7,8, a minimum sample size per population of more than 30 individuals was applied, resulting in a total of 191 hair samples. Animals at the family farms had no genealogical data.

Table 1 Sample location, size and collection date. 

Extraction of DNA from the hair samples was done using the protocol of Sambrook and Russell⁹, with some modifications. The twenty microsatellites used in the study belong to the microsatellite panel recommended by the FAO/ISAG (International Society of Animal Genetics)¹⁰ for porcine biodiversity studies (Table 2). Each marker in the DNA extracted from the 191 samples was amplified using PCR. Amplification conditions were a final volume of 25 μl containing 10 μl 100 μM dNTPs; 2.5 μl 10X buffer; 1.0 μl 25 mM MgCl2; 3.0 μl primers specific to each 10 pmol locus; 0.3 μl Taq DNA polymerase at a 1 U/μl concentration; 4.0 μl genomic DNA at a 50 ng/μl concentration; and 4.2 μl sterilized double-distilled water. The PCR reaction was done in a thermocycler (Mycycler; Bio-Rad®; Hercules, CA, USA) in the following sequence: 95 °C for 5 min; 35 cycles of 94 °C for 30 s; optimum marker temperature (56, 58, 60 or 62 °C) for 30 s; 72 °C for 50 s; and 72 °C for 5 min. Fragment separation was done with polyacrylamide gel electrophoresis using an automatic sequencer (ABI 377XL; Applied Biosystems, Foster City, CA, USA). Fragment analysis was done with the Genescan Analysis® 3.1.2 program, and allele identification with the Genotyper® 2.5.2 program. Allele count was estimated for each marker and by population. Genetic diversity was measured in the studied populations, identifying homozygosity and heterozygosity per marker. Expected and observed heterozygosity were then evaluated by marker and population using the GDA 1.0 program¹¹.

Table 2 Genetic variability in analyzed loci of hair samples from domestic pigs in Cordoba Department, Colombia. 

n= sample size; A= number of alleles; HE= expected heterozygosity; HO= observed heterozygosity; PIC= Polymorphic Information Content; and p= Hardy-Weinberg equilibrium probability.

*Markers not in Hardy-Weinberg equilibrium (P<0.05).

A Guo and Thomson test was applied to identify deviations in Hardy-Weinberg (H-W) equilibrium and their significance levels. Genetic structure in the samples was identified by calculating the F_ST index, assuming the infinite allele model (IAM), and the Nei’s genetic identity for each population. All analyses were done with the GENEPOP 3.4 statistical package¹².

Polymorphic information content (PIC) values were calculated from the allelic frequencies for n=191, using the equation:

Where k is the number of alleles; and P_i and P_j are the allelic frequencies of the i-th and jth allele, respectively. This was done to determine if the selected markers provided sufficient information to identify genetic variability in the studied pig populations. Each microsatellite’s PIC was measured with the CERVUS v. 3.0.3 program¹³.

Phylogenetic reconstruction was done based on genetic distance¹⁴ and the UPGMA methodology run with the MEGA 5.0 program¹⁵. Phylogenetic relationship validity was estimated with 1,000 bootstrap replications.

All the examined microsatellite markers were polymorphic. The number of alleles per locus varied from 2 (SW2019) to 15 (IFNG), with an average of 10.3 alleles (A). These values are within the range previously reported in genetic diversity studies of pigs: 13.31 to 24.8 alleles^16,17 and 2.0 to 3.9 alleles¹⁸. The total number of identified alleles was 207, with the highest variation observed in IFNG (A=11.3, H_Ɛ = 0.917) and SW957 (A= 10.6; H_Ɛ= 0.838). The loci with the lowest polymorphism were SW2019 (A= 3.3; H_Ɛ = 0.201) and SW1067 (A= 3.0; H_Ɛ = 0.170), values similar to those reported for criollo (local) pigs¹⁹, and wattled pigs in Uruguay²⁰. Heterozygosity in the present results was near that reported for the Cuban criollo, Mexican hairless and Argentine criollo pig breeds^21,22,23. This property reflects the polymorphisms detected for each marker in the studied populations, and its value depends on the number of alleles and their frequencies. Both the allele and heterozygosity data suggest that Cordoba’s pig populations are highly genetically variable. In addition, they indicate that the microsatellite panel, which has been tested in a number of studies, is appropriate for studying genetic diversity in this zoogenetic resource.

Values for PIC varied from 0.17 (SW1067) to 0.83 (IFNG), which correspond to the markers with the lowest and highest number of alleles, respectively. A PIC’s informativeness is based on the classification of Botstein et al²⁴: values >0.5 are highly informative; those between 0.25 and 0.50 are moderately informative; and those <0.25 are minimally informative. Only 15 of the tested markers can be considered highly informative (PIC>0.5), and therefore these microsatellites are very useful in identifying genetic variability in domestic pig populations in Cordoba. Three markers were moderately informative (PIC<0.50>0.25), and two were minimally so (PIC<0.25). Compared to previous reports, the average PIC value observed here is lower than in one report²⁵, but similar to other studies^17,20. Of the analyzed microsatellites, 80 % were found to be in genetic H-W equilibrium. This high percentage suggests that the allelic frequencies of the sampled animals tend to group, forming homogeneous populations (Table 2). Grouping may indicate that mating within the population occurs randomly (in terms of the studied markers), or that any new animals added to the population come from populations with the same genetic heritage as the sampled animals²⁶. Significant deviations (P<0.05) in H-W equilibrium were observed in 20 % of the studied loci. Deviations can be expected if individual populations are structured in subpopulations (the Wahlund effect). Marked differences would therefore exist between proximate pig populations in the markers in H-W disequilibrium, but not in the other markers. If the differences between these markers have not been eliminated, it indicates limited genetic flow between nearby populations, an aspect not shown by the other markers. Lack of H-W equilibrium can also be caused by endogamy events in general populations²⁷. However, endogamy affects the entire genome, meaning that if this phenomenon were the most salient, all the markers used in the present study would exhibit an excess of homozygotes; this did not occur in the present case. Other possible scenarios are that the loci in disequilibrium are linked to genes subject to natural selection that acts differentially at the micro- and macrospatial levels, or that the founder effect (few sires with many progeny) has occurred.

Interpopulational genetic relationships were quantified using Nei’s genetic identity¹⁴. Values ranged from a high of 0.627 between the populations at Momil and Cereté, indicating greater genetic similarity, to a low of 0.350 between Momil and the commercially-farmed pigs. Comparisons made to quantify interpopulational genetic differentiation (F_ST) showed Momil and Cereté to have low but significant differentiation (F_ST= 0.076, P=0.016). Values, and differentiation, were slightly higher between Cereté and Tierralta (F_ST= 0.098, P=0.016), and much higher between Momil and the commercial pigs (F_ST= 0.226, P=0.016). All F_ST estimates were statistically significant (Table 3), and suggested moderate genetic flow in the studied populations. Based on the loci, the F_ST values indicated that between 7 and 22 % of total genetic variation in the studied domestic pig populations was explained by interpopulational differences, while between 3 and 78 % corresponded to intrapopulational differences.

Table 3 Estimated F_ST values between population pairs (below diagonal) and Nei’s genetic identity values (above diagonal). 

* Significant after Bonferroni fit (P<0.0001).

The genetic differences in the dendrogram (Figure 1) show Momil and Cereté to be equally distant from Tierralta. The population of commercial pigs is the furthest from the three sampled local populations.

Figure 1 Dendrogram based on genetic distances between studied pig populations in Cordoba, Colombia, generated using UPGMA method 

Studies using θ (a F_ST analogue) have reported a value of 13 % for varieties of Iberian pig, 27 % for European breeds²⁸, and 18 % in Chinese breeds²⁹. Reported genetic differentiation values are lower among local breeds in the same region; for instance, among local breeds in dry and humid regions (F_ST= 4.4 %), and among Cuban criollos (F_ST= 0.12 %)^21,23.

The expected heterozygosity levels in the present study indicate that domestic local pig populations in Cordoba Department, Colombia, exhibit high genetic variability. Values for Nei’s genetic identity, F_ST and Nei’s genetic distance suggest genetic proximity between the Momil, Cereté and Tierralta populations. Given the high percentage of markers with high PIC values in this study, this technique can be applied in other studies within this species, such as genealogical research and assignment of individuals to populations.