Genetic diversity analysis in the whole population
A total of novel fifteen microsatellite markers were used to analyze the genetic diversity of the captive Rhesus monkeys population, and the characteristics of microsatellite loci were summarized in Table 1. The number of observed alleles (Na) is one of the most important indicators of gene differentiation, which is directly related to population, type, and geographic location. The analysis results of 15 microsatellite loci showed 155 alleles were identified in the 104 individuals, with the number of alleles each locus ranging from 7 (C4 001 and C17 010) to 17 (C13 016), giving an average number of 10.3 alleles per locus. The number of effective alleles in each locus ranged from 3.401 to 10.989, the mean effective number of alleles was 5.602. Heterozygosity present in populations reflects the genetic variation of the population at different loci, which can be used as an optimal parameter to evaluate the genetic diversity [28]. The observed ranged from 0.409 to 0.891 and expected heterozygosity ranged from 0.706 to 0.909, respectively. Mean Ho and He were 0.7297 and 0.8016, respectively. When the heterozygosity is between 0.5 and 0.8, it can be considered that the genetic diversity of the population is high [29]. The PIC ranged from 0.663 to 0.897 with an average value of 0.771. Moreover, nine out of fifteen loci presented null alleles frequencies F(Null), ranging from - 0.0289 to 0.3801 (Table 2).
Table 2
Total genetic diversity parameters of 15 microsatellite markers
Chr ID | Primer | Size (bp) | N | Ne | Ho | He | PIC | P value | F(Null) |
1 | C1 033 | 223-255 | 8 | 3.401 | 0.591 | 0.706 | 0.663 | 0.0464 | 0.0903 |
2 | C2 038 | 220-248 | 8 | 4.201 | 0.755 | 0.762 | 0.725 | 0.1627 | -0.0005 |
3 | C3 001 | 254-282 | 8 | 4.807 | 0.809 | 0.792 | 0.759 | 0.6071 | -0.0169 |
4 | C4 001 | 129-153 | 7 | 4.347 | 0.755 | 0.770 | 0.727 | 0.3922 | 0.0057 |
5 | C5 009 | 191-231 | 10 | 5.181 | 0.691 | 0.807 | 0.776 | 0.0198 | 0.0808 |
6 | C6 001 | 190-246 | 13 | 4.672 | 0.655 | 0.786 | 0.758 | 0.0000 | 0.0963 |
8 | C8 013 | 196-228 | 9 | 4.405 | 0.809 | 0.773 | 0.734 | 0.5527 | -0.0278 |
9 | C9 002 | 196-256 | 15 | 8.000 | 0.891 | 0.875 | 0.857 | 0.7632 | -0.0121 |
10 | C10 016 | 188-220 | 9 | 3.875 | 0.764 | 0.742 | 0.714 | 0.7230 | -0.0127 |
11 | C11 019 | 225-285 | 15 | 8.403 | 0.745 | 0.881 | 0.865 | 0.0051 | 0.0811 |
13 | C13 016 | 187-261 | 17 | 10.989 | 0.409 | 0.909 | 0.897 | 0.0000 | 0.3801 |
14 | C14 011 | 233-269 | 10 | 7.194 | 0.827 | 0.861 | 0.841 | 0.1998 | 0.0170 |
15 | C15 019 | 200-236 | 10 | 5.917 | 0.745 | 0.831 | 0.806 | 0.1439 | 0.0515 |
17 | C17 010 | 232-256 | 7 | 4.854 | 0.836 | 0.794 | 0.759 | 0.0000 | -0.0289 |
18 | C18 021 | 157-189 | 9 | 3.787 | 0.664 | 0.736 | 0.694 | 0.1200 | 0.0525 |
Mean | - | - | 10.3 | 5.602 | 0.7297 | 0.8016 | 0.7716 | 0.745 | 0.718 |
Comparison of genetic diversity between HS and XJ population
The number of alleles each locus (Na), the number of effective alleles (Ne), Shannon index (I), observed heterozygosity (Ho), expected heterozygosity (He), P-value for Hardy-Weinberg equilibrium for HS (N = 65), and XJ (N = 39) were summarized in Table 3. The number of alleles on each locus in HS was greater than or equal to that in XJ; in particular, locus C3 001, C4 001, and C18 021 shared the same alleles. The Ne ranged from3.409 to 11.45 in HS, and ranged from 2.931 to 8.199 in XJ, respectively. The mean Ho of the HS group (Ho = 0.741)was higher than that of the XJ group (Ho = 0.718). Besides, the mean He ( ~ 0.745༉and PIC ( ~ 0.761) were also lower in the XJ group than in the HS group (He = 0.801, PIC = 0.767). To sum up, we speculate the two captive Rhesus monkeyss possess a relatively high level of genetic diversity. HWE analysis showed that 4 loci (C5 009, C6 001, C13 016 and C17 010) in HS population and 5 loci (C4 001, C6 001, C11 019, C13 016 and C17 010) in XJ population, and the remaining loci were in accordance with HWE.
Table 3
Genetic diversity indices of 15 microsatellite loci in two captive populations
Pop | Locus | Na | Ne | I | Ho | He | uHe | F | P-value |
HS | C1033 | 8 | 3.409 | 1.473 | 0.646 | 0.707 | 0.712 | 0.086 | 0.1076 |
(N=65) | C2038 | 7 | 3.815 | 1.534 | 0.815 | 0.738 | 0.744 | -0.105 | 0.6449 |
| C3001 | 8 | 3.967 | 1.609 | 0.754 | 0.748 | 0.754 | -0.008 | 0.6857 |
| C4001 | 7 | 4.191 | 1.526 | 0.769 | 0.761 | 0.767 | -0.010 | 0.9942 |
| C5009 | 10 | 4.904 | 1.794 | 0.692 | 0.796 | 0.802 | 0.130 | 0.0102 |
| C6001 | 12 | 4.763 | 1.842 | 0.708 | 0.790 | 0.796 | 0.104 | 0.0090 |
| C8013 | 9 | 4.499 | 1.670 | 0.800 | 0.778 | 0.784 | -0.029 | 0.2860 |
| C9002 | 15 | 8.086 | 2.259 | 0.908 | 0.876 | 0.883 | -0.036 | 0.2664 |
| C10016 | 9 | 3.641 | 1.672 | 0.708 | 0.725 | 0.731 | 0.024 | 0.1547 |
| C11019 | 15 | 8.527 | 2.304 | 0.800 | 0.883 | 0.890 | 0.094 | 0.2179 |
| C13016 | 17 | 11.45 | 2.601 | 0.369 | 0.913 | 0.920 | 0.595 | 0.0000 |
| C14011 | 10 | 7.119 | 2.063 | 0.831 | 0.860 | 0.866 | 0.033 | 0.2499 |
| C15019 | 10 | 6.006 | 1.960 | 0.785 | 0.833 | 0.840 | 0.059 | 0.3239 |
| C17010 | 7 | 4.782 | 1.656 | 0.846 | 0.791 | 0.797 | -0.070 | 0.0013 |
| C18021 | 7 | 3.539 | 1.474 | 0.677 | 0.717 | 0.723 | 0.056 | 0.1916 |
| Mean±SE | 10.067±0.842 | 5.513±0.598 | 1.829±0.088 | 0.741±0.032 | 0.794±0.017 | 0.801±0.017 | 0.062±0.042 | - |
XJ | C1033 | 7.000 | 2.931 | 1.372 | 0.513 | 0.659 | 0.667 | 0.222 | 0.0849 |
(N=39) | C2038 | 7.000 | 4.237 | 1.600 | 0.641 | 0.764 | 0.774 | 0.161 | 0.2125 |
| C3001 | 8.000 | 5.633 | 1.876 | 0.872 | 0.822 | 0.833 | -0.060 | 0.8360 |
| C4001 | 6.000 | 4.156 | 1.550 | 0.769 | 0.759 | 0.769 | -0.013 | 0.0188 |
| C5009 | 7.000 | 5.263 | 1.774 | 0.667 | 0.810 | 0.821 | 0.177 | 0.0537 |
| C6001 | 12.000 | 4.768 | 1.913 | 0.641 | 0.790 | 0.801 | 0.189 | 0.0038 |
| C8013 | 6.000 | 4.111 | 1.521 | 0.846 | 0.757 | 0.767 | -0.118 | 0.8867 |
| C9002 | 9.000 | 5.805 | 1.900 | 0.846 | 0.828 | 0.838 | -0.022 | 0.5575 |
| C10016 | 7.000 | 3.583 | 1.546 | 0.821 | 0.721 | 0.730 | -0.138 | 0.8702 |
| C11019 | 10.000 | 7.493 | 2.101 | 0.641 | 0.867 | 0.878 | 0.260 | 0.0017 |
| C13016 | 12.000 | 8.199 | 2.239 | 0.513 | 0.878 | 0.889 | 0.416 | 0.0000 |
| C14011 | 9.000 | 6.170 | 1.970 | 0.795 | 0.838 | 0.849 | 0.051 | 0.3967 |
| C15019 | 9.000 | 5.130 | 1.863 | 0.718 | 0.805 | 0.816 | 0.108 | 0.4845 |
| C17010 | 6.000 | 4.651 | 1.640 | 0.821 | 0.785 | 0.795 | -0.045 | 0.0134 |
| C18021 | 7.000 | 3.935 | 1.547 | 0.667 | 0.746 | 0.756 | 0.106 | 0.0771 |
| Mean±SE | 8.133 ± 0.515 | 5.071 ± 0.368 | 1.761 ± 0.063 | 0.718 ± 0.030 | 0.789 ± 0.014 | 0.799 ± 0.014 | 0.086 ± 0.040 | - |
Analysis of Molecular Variance(AMOVA)is a method to detect population differentiation utilizing molecular markers [30]. This procedure was initially implemented for DNA haplotypes, but applies to any marker system. The implementation of AMOVA requires two very basic components: (1) A distance matrix derived from the data and (2) a separate table used to partition the data into different stratifications. AMOVA showed that the genetic variance was 91% within individuals, while it was 9% and 0% among populations and individuals, respectively (Table 4).
Table 4
AMOVA analysis of two captive populations for Rhesus monkeys
Source | df | SS | MS | Est. Var. | % | Total Fst |
Among Pops | 1 | 5.229 | 5.229 | 0.000 | 0% | - |
Among Indiv | 102 | 664.867 | 6.518 | 0.514 | 9% | - |
Within Indiv | 104 | 571.000 | 5.490 | 5.490 | 91% | - |
Total | 207 | 1241.096 | - | 6.004 | 100% | -0.002 |
Table 5
F-statistics analysis and Nm index of 15 microsatellite loci
Locus | Fis | Fit | Fst | Nm |
C1033 | 0.151 | 0.154 | 0.004 | 68.677 |
C2038 | 0.030 | 0.035 | 0.005 | 46.054 |
C3001 | -0.035 | -0.024 | 0.011 | 22.117 |
C4001 | -0.012 | -0.007 | 0.004 | 58.768 |
C5009 | 0.154 | 0.156 | 0.003 | 90.209 |
C6001 | 0.147 | 0.150 | 0.004 | 65.963 |
C8013 | -0.073 | -0.070 | 0.002 | 110.719 |
C9002 | -0.029 | -0.020 | 0.009 | 28.358 |
C10016 | -0.057 | -0.052 | 0.005 | 54.234 |
C11019 | 0.176 | 0.178 | 0.002 | 122.271 |
C13016 | 0.507 | 0.511 | 0.007 | 34.355 |
C14011 | 0.042 | 0.046 | 0.004 | 60.721 |
C15019 | 0.083 | 0.085 | 0.002 | 107.796 |
C17010 | -0.058 | -0.054 | 0.003 | 75.092 |
C18021 | 0.082 | 0.083 | 0.001 | 216.504 |
Mean±SE | 0.074±0.038 | 0.078±0.038 | 0.004±0.001 | 77.456±12.580 |
Population structure analysis
Animals in captivity are also subject to similar evolutionary forces that act on natural populations facilitating the generation of population genetic structure. Population genetic structure essentially describes the total genetic diversity and its distribution within and among a set of populations. Given that the introduction of wild Rhesus monkeys individuals, we wonder whether there existed the peculiar allele inherited and retained in some wild individuals. To address this, the Bayesian method of Structure 2.3.4 was carried out to analyze the genetic structure of the captive population. We speculated that the captive population received frequent gene flow and osmotic due to artificial selection. Given the artificial interference and the introduction of wild individuals, Structure analysis showed when K = 3, the Delta K estimator exhibited an obvious apex(Delta K = 0.689839). The assignment results show K=3, three colors represent three different genetic clusters. The three colors red, blue, and yellow, represent three ancestral blood lineages and are distributed in all the samples. Each line represents one individual, and the proportion of population assignment of each individual is relative to the given genetic cluster, which is represented by the length of each line. (Figure 2). In the ideal situation, each column represented one individual and the colors represented the probability membership coefficient of that individual for the genetic cluster, however, no obvious genetic difference was found among the captive individuals.
Principal coordinate analysis was used to establish a two-dimensional location map of two captive rhesus monkeys populations, which can visualize the difference or similarity of data. The results showed the first and second principal coordinates explained 6.77% (HS) and 6.48% (XJ) of the total variation, respectively. PCoA analysis did not clearly separate the populations, and the germplasm from different populations were mixed with each other (Figure 3).
F-statistics analysis
Understanding the extent of genetic differentiation among captive populations provides insights into industry practices and the domestication process. F-statistics showed the mean Fis and Fit were 0.074±0.038 and 0.078±0.038, respectively, which indicated the inbreeding is not obvious. Wright (1965) proposed that the genetic differentiation coefficient Fst < 0.05 is low differentiation, 0.05 ≤ Fst ≤ 0.15 is moderate differentiation, Fst > 0.15 is highly differentiated, and Fst > 0.25 is extremely differentiated [31]. In this study, the Fst was 0.001 to 0.011, less than 0.05, which was a low degree of differentiation. Gene flow can play a homogenizing role in the population and effectively resist genetic differentiation caused by selection and genetic drift. Theoretically, when Nm < 1, differentiation may occur among populations due to genetic drift. If Nm > 1, gene exchange between populations can prevent population differentiation caused by genetic drift. In this study, gene flow between all populations ranged from 22.117 to 216.504, indicating that frequent gene exchange existed in captive monkeys under artificial selection, which could prevent population structure due to genetic drift between populations.
Bottleneck effect analysis
Two captive macaque populations were tested based on the allele frequencies of the microsatellite loci and three different assumptions including IAM, TPM, and SMM. Under the IAM model assumption, both HS and XJ populations showed a highly significant excess of heterozygosity in the two-tailed test of both Sign test and Wilcoxon test (p < 0.01); Under the TPM model assumption, The HS population in the Sign test showed a significant heterozygosity excess (p < 0.05), not significant in the Wilcoxon test (p > 0.05), the XJ population showed nonsignificant heterozygosity excess in both Sign test and Wilcoxon-test two-tailed test; Under the SMM model assumption, The HS population showed a significant heterozygosity excess (p < 0.05) in both the Sign test and the Wilcoxon two-tailed test. Studies have shown that many microsatellite data are more consistent with TPM models and have now been recommended for testing for bottleneck effects in population numbers. If only the TPM model was used to test the bottleneck effect of the population, the XJ population did not experience the bottleneck effect in this study, and the bottleneck effect of the HS population was not strong (Table 6).
Table 6
Bottleneck effect analysis in two captive populations
pop | Sign test | Wilcoxon test |
IAM | SMM | TPM | IAM | SMM | TPM |
| He/Hd | P | He/Hd | P | He/Hd | P | P | P | P |
HS | 15/0 | 0.00046** | 4/11 | 0.01242 | 5/15 | 0.04362* | 0.00003** | 0.03015* | 0.3894 |
XJ | 15/0 | 0.00046** | 10/5 | 0.37827 | 10/5 | 0.39352 | 0.00003** | 0.20776 | 0.45428 |