Genetic Population Structure of Sus scrofa in Lithuania before the African Swine Fever Outbreak

Loreta Griciuvienė Vytauto Didziojo universiteto Zemes ukio akademija Žygimantas Janeliūnas Vytauto Didziojo universiteto Zemes ukio akademija Vaclovas Jurgelevičius Molekulinės biologijos ir genetiškai modi kuotų organizmų tyrimų skyrius Nacionalinis maisto ir veterinarijos rizikos vertinimo institutas Algimantas Paulauskas (  algimantas.paulauskas@vdu.lt ) Vytauto Didziojo universiteto Zemes ukio akademija https://orcid.org/0000-0002-6823-3754


Background
The wild boar (Sus scrofa) is among the most widespread large mammals, as its natural range extends from western Europe and the Mediterranean Basin to eastern Russian Federation and Japan, throughout southeast Asia [1,2]. For this species remarkable adaptability, wild boar populations expand their geographic range and can be found in a variety of habitats and climates [3,4]. Successful range expansion and increasing abundance of the wild boar populations is in uenced by several factors like a high ecological plasticity, high reproductive capacity, their adaptability to changing food diverse [2], lack of natural predators [5] and supplementary feeding [6]. In light of these factors, the main regulatory mechanism for the rapid increase in size of the wild boar populations is wildlife management [7,8].
Central European wild boar subspecies also abundantly distributed in Lithuania [9]. The abundance of wild boar population poses a threat to the agriculture or present a risk for livestock health. The emergence in 2014 and persistence of African Swine Fever (ASF) in Lithuania has been linked to a relatively rapid decline of the wild boar population [10,11]. The regulation of high wild boar populations levels through hunting partially solves the threat of ASF spreading to new regions, but it is still unknown how the intensive reduction of population numbers will affect the genetic structure of wild boars in the future. Molecular techniques can be applicable as valuable tools for improving the understanding of the genetic changes in populations, population structuring, and genetic differentiation [12]. Information about the state of wild boar population structure before the spread of ASF would allow us to determine and to compare how the spread of the virus may have in uenced its population genetic structure.
The main goal of this present study was to asses the genetic diversity and genetic structure of the wild boar population in Lithuania before ASF outbreak using a set of microsatellite markers.

Genetic diversity analysis of wild boars in Lithuania
Using the 15 microsatellite markers, 147 alleles were observed in the 96 wild boar samples from nine districts, ranging from 103 alleles in Vilnius to 52 alleles in Alytus (Table 1). The number of alleles for each locus (N A ) ranged from 2 to 13 with average over all loci and all sample sites of 5.02 (Table 1).
Private alleles, distinctive to a speci c population, were present in all subpopulations varying from a single in Alytus to a maximum of 8 in Kaunas subpopulation (Table 1). Overall observed heterozygosity values across all loci ranged from 0.567 to 0.650, the expected heterozygosity values ranged from 0.534 to 0.678 (Table 1). Signi cant deviation from HWE was observed in 5 out of 15 loci at P < 0.05 (Table 1). In ve subpopulations (Utena, Vilnius, Alytus, Marijampolė, Kaunas) observed heterozygosity differed signi cantly from expected heterozygosity under Hardy-Weinberg equilibrium toward heterozygosity de ciency (Table 1).     2  4  4  3  3  4  4  3  5  2  5  3  3  4  5 A   Genetic differentiation and population structure analysis Pairwise FST and Nei's genetic distances (D Nei ) among subpopulations are shown in Table 2. Nei's genetic distances and FST analysis indicated a low or no genetic differentiation between all pairs of subpopulations ( Table 2). All wild boar subpopulation pairs were not signi cantly differentiated from one another.

NS-non-signi cant population differentiation
Additionally, three-dimensional factorial correspondence analysis (3D-FCA) was also conducted in order to determine the degree of structuring of wild boar subpopulations (Fig. 1). FCA results indicated admixture between individuals from different districts. These results suggested that intensive hunting pressure, widespread distribution, no presence of geographical barriers can in uence genetic composition and population structure of the wild boar subpopulations in Lithuania.
The result of analysis of molecular variance (AMOVA) showed that 85% of the total genetic variation originated from individuals, while 15% came from differences among individuals within the populations, and 0% was observed among populations (Table 3). Statistical analysis of xation index (F ST =0.000) and analysis of molecular variance revealed no signi cant genetic differentiation between the wild boar subpopulations (Table 3). Other F-statistics revealed signi cant values for F IS = 0.150 (p < 0.001) and F IT = 0.150 (p < 0.001). These data indicate that higher genetic variability of S. scrofa is mainly distributed within individuals (Table 3). The population structure analysis showed the optimum number of subpopulations K which explained that wild boar subpopulations could be divided into two clusters (K = 2) using the Evanno method (Fig. 2). Through the graphic visualization of the population structure, there was no separation of genetic groups, and each of the 9 subpopulations had more than 1 cluster (Fig. 2). Increasing the number of structure groups beyond K = 2 did not in uence changes in population structure.

Discussion
The genetic structure of wild boar population from Lithuania has not yet been thoroughly studied. The microsatellite analysis could lead to new and important inferences about the structure and differentiation of S. scrofa population before and after African swine fever.

Genetic diversity and variation
Most of the loci used in this work had been analyzed in previous studies with wild boar in Europe. We have detected similar genetic variation patterns in the investigated loci of wild boars as compared with other genetic studies. The analysis of microsatellite polymorphisms revealed that the level of genetic diversity obtained in Lithuanian wild boar population (H O =0.622) was in the same range of that reported in Bulgarian wild boar (H O =0.63) [13], Italian populations (Ho = 0.63), Hungarian wild boar populations (Ho = 0.75) [14], in wild boar populations from Portugal (Ho = 0.627) [15] and Poland (H 0 = 0.51) [16].
A similar level of heterozygosity (Ho = 0.63) was reported in wild boar populations inhabiting East Asia [17].
Analysis of molecular variance (AMOVA) revealed high intra-population genetic variation in wild boar population in Lithuania (Table 3). A similar trend involving genetic variation that mainly distributed within population has also been reported for Bulgarian populations [13]. The high intra-population variability and genetic homogeneity are in uenced by gene ow, which is impacted by distribution and connectivity of populations [18,19].
FST analyses revealed no genetic structure among subpopulations of wild boar in Lithuania suggesting high rates of gene ow or little separation in time between populations. Conversely, the obtained higher genetic differentiation (FST = 0.0816) among Bulgarian populations occurs due to such geographical barriers as mountain ridges and human impact [13]. One possible explanation for the low FST values we observed is that the wild boar is migratory species and has a relatively large home range [16].

Population structure
In this study, we analyzed the population structure of S.scrofa population in Lithuania. At the individual level, FCA and STRUCTURE cluster analysis were performed. STRUCTURE analysis provided evidence for two genetic groups. The phenomen of two genetically distinct clusters can be occurred as a result of demographic history. Nikolov et al. 2009 [13] has identi ed two subgroups in their study of Bulgarian wild boar populations and detected that Balkan Mountain Range acts as a natural migration barrier. Our results differed from those of Ferreira et al. 2009 [15], where Portuguese wild boar formed three subpopulations (North, Centre and South) for effects of a recent genetic bottleneck. A factorial correspondence analysis con rmed homogeneity and no genetic differentiation between subpopulations of S.scrofa.

Conclusions
Microsatellite loci analyses revealed that wild boar subpopulations of Lithuania before the African swine fever outbreak were still not distinguished and admixed. Future studies with extensive sampling will be helpful for understanding the detailed structure of wild boar population in Lithuania after African swine fever outbreak.

Sampling
Tissue samples of wild boar were collected during a 5-year period (2009-2013) from 9 districts in Lithuania (Fig. 3). A total of 96 S.scrofa individuals legally harvested by the licensed hunters in different parts of Lithuania were investigated. We chose to focus on a single population of Lithuania and sample size that would be su cient to characterize population-level genetic diversity when using microsatellites. To compare and reveal genetic diversity among major regions, we grouped the Lithuanian wild boar population into nine groups based on their regional origins: Utena = 18, Vilnius = 21, Alytus = 4, Marijampolė = 10, Kaunas = 19, Tauragė = 4, Klaipėda = 4, Šiauliai = 7, Panevėžys = 6. Fresh muscle, spleen and blood were sampled from unprotected wild boars and either stored in plastic tubes (5-

Statistical analysis
In order to estimate population genetic structure of wild boars in Lithuania, number of alleles per locus (N A ), the observed heterozygosity (H O ), the expected heterozygosity (H E ) under the Hardy − Weinberg assumptions were obtained in GenAlEx v6.1 [21]. Deviations from Hardy-Weinberg equilibrium (HWE) were tested with 1000 permutations across markers using Genepop v.4.0 [22]. Factorial correspondence analysis (FCA) on the microsatellite data for individual wild boars was performed using GENETIX version 4.05.2 [23]. Analysis of molecular variance (AMOVA) and F-statistics (F ST , F IS , and F IT ) were also estimated using