cpDNA haplotype database
DNA sequencing of the choloroplast (cp) markers produced sequences of the following lengths: 573 bp (atpB-rbcL); 487 bp (petG-trnP); 500 bp (trnL1-trnL2); and 593 bp (psbM-trnD). Alignment of the 352 individuals from the 44 populations yielded a total 28 variable sites: 11 in the atpB-rbcL spacer, seven in both the petG-trnP and psbM-trnD spacers, and three in the trnL1-trnL2 spacer (Supplementary Table S1). Based on these 28 variable sites (21 base substitutions and 7 deletions) across the combined intergenic regions, a total of 22 unique haplotypes were found (Fig. 1a).
(a) Chloroplast haplotype distribution in the Shorea leprosula populations. The pie chart colours indicate haplotype distributions; and sector areas are proportional to sample size (Map was generated by ArcGIS-ArcMap version 10.8). (b) STRUCTURE analysis identified two clusters (K = 2) corresponding to Region A and B.
SSR allele frequency database
The reproducibility of SSR genotyping was confirmed by achieving consistent genotypes from five independent PCR amplifications on a single individual for each of the ten SSR loci. Individual bar plots from STRUCTURE analysis are presented in Fig. 1b. At the highest Delta K likelihood scores, the best representation of the data was K = 2 suggesting that the 44 populations in Peninsular Malaysia can be divided into two main genetic clusters: Region A and Region B. The first cluster, ‘Region A’ consists of 12 populations, namely SBadak, BPerangin, BEnggang, GJerai, RTelui, GInas, GBongsu, Belum, Piah, BHijau, Korbu and Bubu. The second cluster, ‘Region B’ consists of 32 populations, namely Behrang, Ampang, HGombak, HLangat, SLalang, PPanjang, Berembun, Angsi, Kenaboi, Triang, Pasoh, BSenggeh, GLedang, Krau, TNegara, Terenggun, SBetis, USat, CTongkat, HTerengganu, Jengai, AGading, Tekam, Beserah, Jengka, Lentang, Lesong, ERompin, GArong, Labis, AHitam and Panti. Similarly, the UPGMA dendrogram analysis also divided the 44 populations into two genetic clusters (Fig. 2) corresponding to Region A and B of the STRUCTURE result.
Dendrogram showing the relationship between 44 populations of Shorea leprosula in Peninsular Malaysia based on the UPGMA cluster analysis of SSR markers.
SSR allele frequency databases were established according to Region A and B, and characterized to evaluate the relative usefulness of each SSR marker in forensic investigation. The distribution of allele frequencies for each locus is listed in Table S2 (Region A database) and Table S3 (Region B database). Forensic parameters are shown in Table 1, with a total of 143 alleles and 174 alleles detected in the Region A and B databases, respectively. The observed (Ho) and expected (He) heterozygosity ranged from 0.3570 to 0.8346 and 0.4375 to 0.8795, respectively for populations in the Region A database; and ranged from 0.3298 to 0.8356 and 0.3469 to 0.8793, respectively for populations in the Region B database. The power of discrimination (PD) for the SSR loci ranged from 0.601 to 0.972 and 0.554 to 0.975, in Region A and B databases, respectively. The most discriminating locus was Sle605 in both the Region A (PD = 0.972) and Region B (PD = 0.975) databases. Minimum allele frequency was adjusted for alleles falling below the thresholds of 0.0066 (Region A) and 0.0024 (Region B).
Deviations from HWE were detected in four of the SSR loci for Region A (SleT11, SleT15, SleT17 and Sle465) and six SSR loci in Region B (SleT01, SleT11, SleT15, SleT17, SleT29 and SleT31). We evaluated these loci in each population independently to rule out the possible presence of null alleles. There were four populations in Region A (GJerai, RTelui, GBongsu and Piah) where a single one locus deviated from HWE; whereas there were eight populations in Region B (Behrang, HGombak, SLalang, Angsi, Klau, USat, Jengka and Panti) with a single locus and a single population (GLedang) with two loci that deviated from HWE (Table S4). Observed deviation from HWE was substantially lower in each population (either absence or not more than two loci) and thus it might be due to Wahlund effect caused by population substructuring in both Region A and B. Linkage disequilibrium (LD) testing was used to evaluate the independence of frequencies for all the SSR genotypes. A total of 13.3% and 28.9% of the 45 pairwise loci were found significant evidence of LD for Region A and B, respectively. Some of the loci might be linked as a result of population substructuring and inbreeding (inbreeding coefficient = 0.0822 [Peninsular Malaysia]). These results are in line with observations in real populations, where the assumption of completely random mating and zero migration required for HWE and LD are unlikely to be met, either in humans, animals or plants 21,22,23.
Mean self-assignment, the proportion of individuals correctly assigned back to their population, was 45.9% and ranged from 14.3% (Kenaboi) to 81.3% (CTongkat) between population (Table 2). At the regional level, correct assignment rate of individuals to their region of origin was higher, 87.4% for Region A and 90.0% for Region B, (average of 88.7%).
Conservativeness of the database
The coancestry coefficient (θ) for Peninsular Malaysia (0.0579) was higher than those of Region A (0.0454) and Region B (0.0500) (Table 3). A total of 4.54% and 5.00% of the genetic variability was distributed among populations within Region A and Region B, respectively. In terms of inbreeding coefficient (f), the value for the Region A database (f = 0.0892) was highest, followed by Peninsular Malaysia (f = 0.0822) and Region B (f = 0.0666). All the θ and f values were significantly greater than zero, demonstrated by the 95% confidence intervals not overlapping with zero. Both of the θ and f values were used to calculate the conservativeness of each database by testing the cognate database (Porigin) against the regional database (Pcombined). The databases were non-conservative at the calculated θ value. In order for both the Region databases (A and B) to be conservative, the value of θ was adjusted from 0.0454 to 0.1900 for Region A and from 0.0500 to 0.1500 for Region B. For the Region A database, the most common SSR profile frequency is 2.69 × 10–7 or 1 in 3.72 million and the rarest profile frequency is 1.84 × 10–14 or 1 in 54.3 trillion. For the Region B database, the most common SSR profile frequency is 1.06 × 10–7 or 1 in 9.43 million and the rarest profile frequency is 4.03 × 10–16 or 1 in 2.48 quadrillion.
Source: Ecology - nature.com
