Genetic Diversity and Population Structure of Nutria (Myocastor coypus) in South Korea

Simple Summary Nutrias (Myocastor coypus) are large semiaquatic rodents native to subtropical and temperate South America. Nutrias have been introduced on all continents, except Oceania and Antarctica, and have become invasive in many countries. This study carried out a survey of nutrias in South Korea under the Nutria Eradication Project from January 2013 to August 2015. Out of 1509 habitat traces, 1497 tracks were observed in the Nakdong River basin. Based on the ecological field data, we identified concentrated areas of nutria populations. Tissue samples were collected from captured nutria for genetic analysis. According to the microsatellite marker analysis, the estimated genetic diversity of the nutria populations was low, which suggests that nutrias in South Korea originate from a single population. Abstract The nutria (Myocastor coypus) is an invasive alien species that have had major adverse effects on biodiversity and the agricultural economy in wetland habitats. Since 2014, the Ministry of Environment in South Korea has been carrying out the Nutria Eradication Project, and we investigated nutria distribution and genetic diversity of nutria populations in South Korea. We estimated that 99.2% of nutria habitats are in the mid-lower Nakdong River regions. To further analyze the genetic diversity in eight major nutria populations, we performed a genetic analysis using microsatellite markers. Genetic diversity levels of the eight nutria populations in South Korea were relatively lower than those in other countries. The probability of migration direction among nutria populations was predicted from genetic distance analysis. Genetic structure analysis showed little difference among the nutria populations in South Korea. These results suggest that nutrias in South Korea originated from a single population. Our results provide important data for establishing management strategies for the successful eradication of nutria populations in South Korea, as well as in other countries with alien invasive species.


Introduction
Environmental concerns regarding invasive alien species have been rapidly increasing with the expansion of global trade. Invasive mammalian species, especially rodents, have adverse effects on the biodiversity of wetlands, such as reducing biodiversity, disrupting ecosystems, human activities, and human welfare [1,2]. The nutria (Myocastor coypus) is a large, semi-aquatic rodent that originates Nutria tissue samples (e.g., ears or tails) were collected from eight regions containing large nutria populations in the Nakdong River basin (  Nutria habitats were also analyzed using the kernel density tool, which utilizes kernel functions to estimate the degree of concentration per unit area, in ArcGIS 10.3.1. The kernel density tool was used based on the method described by Silverman [35]. The kernel density estimation undergoes a smoothing phase of location coordinate data. Defining the bandwidth of a one-sided kernel function affects the probability density function estimation [36,37]. Methods to adjust the bandwidth include LSCVh (least squares cross-validation), CVh (cross-validation), and h_ref (reference method). As adequate bandwidth values may differ depending on the distribution of points [38], the default setting in the search radius in the kernel density tool was used (5485. Nutria tissue samples (e.g., ears or tails) were collected from eight regions containing large nutria populations in the Nakdong River basin ( (Table 1).   Figure 2D-F) [26,39]. Amplified fragments were analyzed by an ABI3130XL Genetic Analyzer (Applied Biosystems, USA) and the data were scored using GeneMarker v2.6.4 (SoftGenetics LLC, State College, PA, USA) ( Figure 2G).

Nutria Distribution and Tissue Sampling
Representative habitat traces of nutria distribution were observed in two water systems in the Han River and Nakdong River basins and one in Jeju Island. Among the 1509 habitat traces, 1497 were observed in the Nakdong River basin, five in the Han River basin, and seven in Jeju Island ( Figure 3A).
The kernel density was used to estimate a smoothing phase of location coordinate data. Nutria samples in the 0.1% high density distributed area were collected using the CVh value in the kernel density calculation ( Figure 3E). The ecological and scenery conservation areas were disregarded as viable sample collection sites, as wildlife capture in the areas is restricted by the National Environment Conservation Act of South Korea. Sample collection sites in Daegu and Jinju at the northern and western ends were included based on habitat track evidence. The majority of observed habitat traces in this study were observed in the middle and lower regions of the Nakdong River basin (99.2%). These results show that the Nakdong River basin is the most populated area for nutria individuals in South Korea. In line with previous nutria distribution studies in South Korea, the kernel density results support the middle and lower reaches of the Nakdong River having slow water flow leading to tributaries and wetlands, with a warm winter and abundant food sources, providing a good habitat for nutria populations to establish and proliferate [19,50]. A previous investigation reported that nutria management efforts are focused on the Nakdong River basin in South Korea because nutria individuals are mainly clustered in this region [22]. Although the previous management was effective to decrease nutria habitat tracks, we tried to apply genetic analysis for the developed nutria eradication project in South Korea. According to nutria distribution analysis (Figures 3 and 4), we selected eight regions (BUS, CHA, DAE, JIN, GIM, MIO, MIY, and YAN) of nutria habitat containing large nutria populations in the middle and lower reaches of the Nakdong River basin for genetic analysis (Table 1, Figure 4). A total of 93 nutria individuals were captured and genomic DNA used for genetic density analysis.

Genetic Diversity Analysis of Nutria Populations
Previous studies have reported successfully settled nutria populations in the middle to lower reaches of the Nakdong River basin, as well as in other sites, such as the Han River and Jeju Island [19,50]. Estimations of the patterns of migration among nutria populations may help the eradication programs prevent the re-colonization of nutria [26]. However, information on the genetic diversity among populations is unknown for nutrias in South Korea. Invasive species, established in a relatively small population, can undergo bottlenecks that may decrease its genetic diversity [51]. Genetic diversity monitoring of invasive species can provide information about invasion history and possible management strategies [52]. Allelic richness (AR) and observed heterozygosity (HO) are generally used to estimate the genetic diversity among populations [23,26,27,34,39,53]. To evaluate the genetic diversity among nutria populations, we calculated allelic richness (AR), observed heterozygosity (HO), and expected heterozygosity (HE). The average values of AR, HO, and HE were 1.806, 0.298, and 0.305, respectively ( Table 2). The values of genetic diversity were affected by the size of the nutria population; however, the genetic diversity of South Korean nutrias did not correlate with the nutria population size (R 2 = 0.0137). Nutria populations in South Korea had relatively lower levels of HO (0.298) than those from other locations, such as Louisiana, USA (0.624, 9 markers) [27]; Czech Republic (0.504, 11 markers) [39]; Maryland, USA (0.46, 27 markers) [26]; and the Argentinean Pampas (0.658, 16 markers) [53]. Furthermore, nutrias in South Korea had lower AR values (1.806) than those in other locations, such as Louisiana, USA (3.8) [27] and the Czech Republic (3.6) [39]. These low levels of HO and AR indicate that nutria in the Nakdong River basin have a low level of genetic diversity. The low level of genetic diversity and lack of heterozygosity in the McoD228 marker may suggest that nutria in South Korea spread from one population source. This means that, even though nutrias were imported twice, only the secondary import from Bulgaria successfully settled in the Nakdong River, and the population from France [6] did not.

Genetic Diversity Analysis of Nutria Populations
Previous studies have reported successfully settled nutria populations in the middle to lower reaches of the Nakdong River basin, as well as in other sites, such as the Han River and Jeju Island [19,50]. Estimations of the patterns of migration among nutria populations may help the eradication programs prevent the re-colonization of nutria [26]. However, information on the genetic diversity among populations is unknown for nutrias in South Korea. As a result of the analysis of microsatellite markers, all microsatellite markers in the nutria populations were polymorphic except for McoD228, which had no heterozygosity. Some microsatellite markers were presented as monomorphic within populations, namely Invasive species, established in a relatively small population, can undergo bottlenecks that may decrease its genetic diversity [51]. Genetic diversity monitoring of invasive species can provide information about invasion history and possible management strategies [52]. Allelic richness (A R ) and observed heterozygosity (H O ) are generally used to estimate the genetic diversity among populations [23,26,27,34,39,53]. To evaluate the genetic diversity among nutria populations, we calculated allelic richness (A R ), observed heterozygosity (H O ), and expected heterozygosity (H E ). The average values of A R , H O , and H E were 1.806, 0.298, and 0.305, respectively ( Table 2). The values of genetic diversity were affected by the size of the nutria population; however, the genetic diversity of South Korean nutrias did not correlate with the nutria population size (R 2 = 0.0137). Nutria populations in South Korea had relatively lower levels of H O (0.298) than those from other locations, such as Louisiana, USA (0.624, 9 markers) [27]; Czech Republic (0.504, 11 markers) [39]; Maryland, USA (0.46, 27 markers) [26]; and the Argentinean Pampas (0.658, 16 markers) [53]. Furthermore, nutrias in South Korea had lower A R values (1.806) than those in other locations, such as Louisiana, USA (3.8) [27] and the Czech Republic (3.6) [39]. These low levels of H O and A R indicate that nutria in the Nakdong River basin have a low level of genetic diversity. The low level of genetic diversity and lack of heterozygosity in the McoD228 marker may suggest that nutria in South Korea spread from one population source. This means that, even though nutrias were imported twice, only the secondary import from Bulgaria successfully settled in the Nakdong River, and the population from France [6] did not. As the populations are quite small and isolated, inbreeding within populations can be calculated to be a factor in the low genetic diversity that we observed [54]. The inbreeding coefficients (F IS ) were calculated to estimate the population size and possibility of migration. The calculated F IS values were 0.256 and 0.232 in the MIO and YAN populations, respectively ( Table 2). The MIO and YAN populations were distributed near small reservoirs and isolated from other populations by a road and villages. Negative F IS values were observed in the DAE, GIM, and MIY populations, indicating that there is outbreeding in these populations. The net-like stream as a topographical factor in these areas can allow easy movement of the nutria. Although BUS has the best settlement conditions, the area is geologically separated from other habitats by many highway roads and a well-developed estuary, and therefore the nutrias underwent many cases of inbreeding in small habitats. These values support the theory that nutria populations originated from a small number of individuals, and that the populations are isolated ( Table 2).

Population Structure Analysis of Nutria Populations
Despite the technical advances in the invasive mammal eradication program, molecular genetics knowledge has not been fully integrated into these control programs. Although the source of Korean nutrias was estimated, F IS values suggest that the environmental conditions of each population were different. Different colonizing populations had a high genetic structure. Therefore, those populations had capacities to promote invasiveness [55]. AMOVA, STRUCTURE, and BAPS are generally used to estimate the genetic structure of a population. To elucidate the difference in each nutria population, the population structure was analyzed using AMOVA, STRUCTURE, and BAPS. AMOVA, which is a statistical analysis for molecular differences with two hierarchical levels, indicated that among-population variation only accounted for 10.7% of the total genetic variation, whereas within-population variation explained 89.3% (Table 3). These results suggest that there is no significant differentiation among the nutria populations in South Korea. To confirm these results, STRUCTURE using Bayesian clustering was conducted. Each bar represents an individual nutria and each population is separated by vertical bars. K = 2 and the optimal number of clusters was calculated using Evanno's method. Populations show similar assignment probabilities. STRUCTURE results showed a similar trend to the AMOVA pattern that each nutria Animals 2019, 9, 1164 9 of 13 population had a similar structure (K = 2) ( Figure 5A). BAPS, using a more sensitive clustering algorithm, was used to analyze the population structure among nutria populations in detail. BAPS identified four groups: (1) BUS, CHA, GIM, YAN; (2) DAE, MIO; (3) MIY; and (4) JIN. Each population separated by a vertical bar represents a different color depending on the group. These groups are not correlated with their geographic distribution ( Figure 5B). These patterns assumed that there was an artificial change in the gene pool from nutria farms near DAE and JIN in the past. According to the results of BAPS, there was a little difference among nutria populations; however, AMOVA and STRUCTURE data showed that populations of nutria in South Korea had similar structures and also indicated that nutria populations in Korea originate from one source. Animals 2020, 9, x 9 of 13 Each population separated by a vertical bar represents a different color depending on the group. These groups are not correlated with their geographic distribution ( Figure 5B). These patterns assumed that there was an artificial change in the gene pool from nutria farms near DAE and JIN in the past. According to the results of BAPS, there was a little difference among nutria populations; however, AMOVA and STRUCTURE data showed that populations of nutria in South Korea had similar structures and also indicated that nutria populations in Korea originate from one source.

Genetic Distance of Nutria Populations
Pairwise genetic distances (FST) show the degree of genetic relationship between populations. FST has been used to infer migration rates and represent genetic differentiation among populations [56]. To confirm the relationship between the eight nutria populations, we calculated the FST of nutrias in South Korea. A low value of FST between two populations means similar genetic structures and little population differentiation. We estimated the values of FST using eight nutria populations in South Korea. The values varied from 0.013 between the BUS and YAN populations to 0.292 between the JIN and MIO populations (Table 4). Some populations, such as CHA/MIY and MIO/MIY, are separated by a short geographical distance but are more distantly related from a genetic point of view than other pairs of populations. Although a Mantel test was applied to elucidate the relationship between genetic distance and geographical distances, the data showed no significance (r = 0.28). The current distribution of nutria populations may have been influenced by humans.

Genetic Distance of Nutria Populations
Pairwise genetic distances (F ST ) show the degree of genetic relationship between populations. F ST has been used to infer migration rates and represent genetic differentiation among populations [56]. To confirm the relationship between the eight nutria populations, we calculated the F ST of nutrias in South Korea. A low value of F ST between two populations means similar genetic structures and little population differentiation. We estimated the values of F ST using eight nutria populations in South Korea. The values varied from 0.013 between the BUS and YAN populations to 0.292 between the JIN and MIO populations (Table 4). Some populations, such as CHA/MIY and MIO/MIY, are separated by a short geographical distance but are more distantly related from a genetic point of view than other pairs of populations. Although a Mantel test was applied to elucidate the relationship between genetic distance and geographical distances, the data showed no significance (r = 0.28). The current distribution of nutria populations may have been influenced by humans.
The analysis of F ST identified the same four groups of populations as BAPS. Nutria individuals that belong to BUS, CHA, GIM, and YAN populations may migrate and interbreed along the Nakdong River. However, the high value of F IS from YAN suggests that the populations likely inbreed with closely related individuals and the population sizes are extremely small ( Table 2). The YAN population may be descended from the BUS population, as they are both close to the Nakdong River. Low values of F ST (0.022) between DAE and MIO imply that nutrias can migrate between the two regions through the Miryang River and the Cheongdo Stream. As shown in Table 2, low values of F IS (-0.119) in the DAE population imply that the population may have established a sufficiently large population for outbreeding, whereas the MIO population may have a relatively small population according to the high value of F IS (0.256), as shown in Table 2. The JIN and MIY populations had the greatest genetic distances of all the populations. The MIY population had a high level of FST compared to other populations, including the MIO population, even though these two populations are geographically close. The genetic divide in the MIY population may have been facilitated by the isolated habitat near the reservoir. However, based on the negative value of F IS , the MIY population may have established a sufficiently large population for outbreeding ( Table 2).
From our results, we confirm nutria population groups in the Nakdong River basin and decided that capturing each nutria group using genetic data on migration direction in the area would be a more efficient approach than a broader eradication strategy that does not use prediction data. Having implemented this strategy to effectively capture individual nutria groups, the Government started operating additional maneuver capture teams, as well as the existing nutria capture teams in the assigned Nakdong River basin area. These strategic modifications from the primary plans have enabled a more efficient nutria elimination strategy in South Korea that reduced 88% of the nutria habitat from 2014 to 2018 [22].

Conclusions
We estimated the genetic origin of and relationships among nutria populations that successfully settled in South Korea as an invasive alien species. The Nakdong River hydrosphere is densely interconnected and resembles a net-like structure. This prevents the complete eradication of nutria, despite the progress in nutria capture techniques and ongoing efforts at the national level. For the successful nutria eradication project, fundamental genetic information on nutria population dispersion and habitat changes with respect to the Nakdong River hydrosphere is necessary for South Korea. In this study, genetic diversity and structural analysis indicated that the populations of nutria in the middle-low regions of the Nakdong River might originate from one population source, possibly from the second nutria import, and the populations can be classified into four groups using pairwise genetic distance data. These genetic analyses can be used to estimate the population structure and genetic diversity of invasive species in general, especially vertebrates. We investigated the establishment process and the population composition of nutria prior to the initiation of the Nutria Eradication Project. Using genetic diversity analysis, population structure and geographical migration were inferred, establishing a foundation for estimating habitat change and creating detailed long-term eradication plans.