Genetic diversity of Elaeis oleifera (HBK) Cortes populations using cross species SSRs: implication’s for germplasm utilization and conservation
BMC Geneticsvolume 18, Article number: 37 (2017)
The Elaeis oleifera genetic materials were assembled from its center of diversity in South and Central America. These materials are currently being preserved in Malaysia as ex situ living collections. Maintaining such collections is expensive and requires sizable land. Information on the genetic diversity of these collections can help achieve efficient conservation via maintenance of core collection. For this purpose, we have applied fourteen unlinked microsatellite markers to evaluate 532 E. oleifera palms representing 19 populations distributed across Honduras, Costa Rica, Panama and Colombia.
In general, the genetic diversity decreased from Costa Rica towards the north (Honduras) and south-east (Colombia). Principle coordinate analysis (PCoA) showed a single cluster indicating low divergence among palms. The phylogenetic tree and STRUCTURE analysis revealed clusters based on country of origin, indicating considerable gene flow among populations within countries. Based on the values of the genetic diversity parameters, some genetically diverse populations could be identified. Further, a total of 34 individual palms that collectively captured maximum allelic diversity with reduced redundancy were also identified. High pairwise genetic differentiation (Fst > 0.250) among populations was evident, particularly between the Colombian populations and those from Honduras, Panama and Costa Rica. Crossing selected palms from highly differentiated populations could generate off-springs that retain more genetic diversity.
The results attained are useful for selecting palms and populations for core collection. The selected materials can also be included into crossing scheme to generate offsprings that capture greater genetic diversity for selection gain in the future.
Under Aracaceae, genus Elaeis consists of two important species, namely E. guineensis Jacq. and E. oleifera (HBK) Cortes. Both species yield oil used for food and non-food applications. In terms of their habitats, E. guineensis is endemic to the tropical lowlands of West and Central Africa, spreading from 16°N in Senegal to 15°S in Angola , whereas E. oleifera originates from Central and South America with natural distribution from 15°N in Honduras to less than 10°S in the north of Porto Velho in Brazil . The natural American palms grow well at the river margins, semi-swamps and generally in open habitats . The African origin, E. guineensis hybridizes with the American origin to produce F1 OxG hybrids .
Malaysian Palm Oil Board (MPOB), formally known as Palm Oil Research Institute of Malaysia (PORIM) has collected E. oleifera germplasm from Colombia, Panama, Costa Rica, Nicaragua and Honduras between 1981 and 1982. The collections were carried out on a collaborative basis, as explained in the Ethics section. Unlike the E. guineensis, the natural distribution of E. oleifera was very sparse . Within a collection site, fewer palms were found thus, only few mature bunches were sampled. Between one and six bunches were harvested per site. In total, 167 bunches were collected from 59 sites distributed across the five countries. Half of seeds from each bunch were presented to the host country while the other half was brought to Malaysia. The seeds then were sowed and field planted in a Completely Randomized Design (CRD) and now serves as the ex situ living collection of E. oleifera.
Field evaluation of the E. oleifera genetic materials collected by MPOB revealed lower height increment as compared to the E. guineensis . As low as 4.6 cm yr-1 height increment were recorded among selected E. oleifera palms particularly from Colombia, which is about one-tenth of that reported in E. guineensis . Oil from E. oleifera palms contained higher level of unsaturated fatty acids, carotenes, vitamin E and sterol contents . Selected palms collected from Costa Rica and Panama contained more than 3,000 ppm of carotene content , whereas those from E. guineensis possessed only between 400 and 1,000 ppm of carotene . Palms from Colombia, Panama and Costa Rica exhibited low palmitic acid (15.8–18.6%) and high oleic acid (56.5–61.5%) contents , resulting in higher level of unsaturation in their oil, a value-added feature that can enhance palm oil marketing in temperate countries. In terms of bunch components, palms from Panama recorded highest mean fruit weight. For mesocarp content, fruits harvested from palms collected from Costa Rica showed the highest value, followed by Panama and Honduras . Despite these, the American oil palm does not attract much interest among planters due to its extremely low oil yield (0.5 t ha-1yr-1) compared to the African oil palm that produce an average of 3–4 t ha-1yr-1 . Thus, direct commercialization of the former is not possible. Nevertheless, selected oil palm agencies worldwide have incorporated E. oleifera into their breeding programs [11, 12] via backcross breeding scheme. In South American countries, where oil palm disease caused by bud rot commonly occurs, interspecific hybrids of both species have been widely adopted. It is believed that the interspecific hybrids are relatively more tolerant to this disease than the African oil palms.
Maintaining ex situ living collection of perennial species like E. oleifera requires huge financial support and land. One round of oil palm conventional breeding cycle takes 10 years. Typically, development of new and improved oil palm varieties needs 30–40 years. Therefore, the selected palms carrying genes for interesting traits identified from the germplasm collection must be appropriately preserved to ensure continuous access, in accordance with the long-term oil palm improvement program. In addition, the present ex situ living germplasm is exposed to diseases and climate change. Valuable palms may be lost at any time. Thus, an effective conservation program should be in place to preserve palms that carry genes of economically important traits as well as those that represent the diversity of the species for exploitation and selection gain in the future. The principle of core collection  where, optimum diversity is preserved within minimum redundancy and population size offers an efficient method for conserving the E. oleifera germplasm.
Unlinked microsatellite markers developed from E. guineensis are available publicly at http://tropgenedb.cirad.fr/tropgene/JSP/interface.jsp?module=OILPALM . These SSRs have successfully been used for assessment of genetic diversity of E. guineensis , fingerprinting and construction of genetic linkage map for E. guineensis  and their interspecific hybrids [17, 18]. We initiated an effort to evaluate the genetic diversity of selected E. oleifera populations using E. guineensis SSR markers . Our analysis included genetic materials assembled from the species’ natural distribution in selected countries in Central and South America. This work provides an overview of the genetic variation, structure and relatedness among individual and populations of the species that may help breeders and germplasm managers prioritize individual palms and populations for establishment of core collection.
Transferability of E. guineensis microsatellite markers to E. oleifera
Of the 18 SSRs tested, 14 (77.78%) revealed scorable amplification products in the screening panel and were then genotyped across the entire samples set in the study. The amplified products observed in the oleifera samples were within the expected size as reported in  (Table 1). The four SSRs that failed to generate scorable profiles were mEgCIR1753, mEgCIR3785, mEgCIR3300 and mEgCIR3574.
Genetic diversity parameters
Fourteen microsatellite markers generated 109 scorable alleles across 532 E. oleifera palms. The number of alleles observed ranged from 3 (mEgCIR0832) to 13 (mEgCIR2387) (Table 1). The G”st value for each marker ranged from 0.005 to 0.812. Six markers namely, mEgCIR0802, mEgCIR3282, mEgCIR0173, mEgCIR2387, mEgCIR1730 and mEgCIR0353 revealed high G”st values (>0.250) indicating reasonable power for population discrimination.
The fundamental genetic diversity parameters for each population are summarized in Table 2. The mean number of different alleles (Na) was 3.0. The palms originating from K14 recorded highest Na (4.4). The lowest Na (1.8) was exhibited by population C1, located at the other end of the population distribution. H2 recorded the least Shannon’s Information Index, I, (0.220) whereas population K14 revealed the highest (0.602). C6 showed highest observed heterozygosity (Ho = 0.218), while H2 exhibited the lowest (0.053) with an average of 0.141. H2 also recorded lowest unbiased He (He = 0.112) while K14 showed the highest (0.293) with overall mean of 0.221. The observed heterozygosity values were generally lower than the expected which lead to the positive fixation index (F) for most populations, except for population C6. F values ranged between -0.030 (population C6) and 0.587 (population K2). At the country level, Costa Rica showed higher values, whereas Honduras had the least for most of the genetic diversity parameters.
The plot of allelic richness against the positions of the 19 populations revealed a decreasing trend of Na(rar) towards the east of the populations’ distribution (Fig. 1). The linear regression analysis further confirmed this result (p < 0.05). Approximately, 29.9% of the Na(rar) variation can be explained by the distances in the linear regression.
Genetic structure and population differentiation
The PCoA diagram revealed a single big cluster that consisted of palms from all countries (Fig. 2). Palms from Honduras are dispersed within the distribution area of Costa Rica palms. Some of the palms from Colombia overlapped with those from Panama. The first and second principle component respectively explained 39.3% and 14.1% of the molecular variance. Results from AMOVA showed high and significant overall variation among population (0.290) (Table 3). Pairwise Fst values among populations were in the range of 0.008–0.338 (Table 4 – above diagonal). A heatmap was prepared using these values to provide a more comprehensive overview of the genetic differentiation pattern (Fig. 3 – below diagonal). Population C1 was highly differentiated from four populations namely, H2, H3, P13 and K8 with Fst values above 0.250. Similarly, population C9 also revealed strong genetic differentiation against H2. Twenty-three pairwise Fst values (between 0.150 and 0.250) indicated moderate differentiation, between populations from Colombia and those from Panama, Honduras and Costa Rica. The remaining pairwise Fst values were <0.150 signifying low differentiation between the populations.
The genetic distance among the population was highest (0.315) between C1 and K2 while the lowest was 0.003, between H2 and H3 (Table 4 – below diagonal). The phylogenetic tree constructed according to these estimates revealed two main groups: populations from Costa Rica, Panama and Honduras formed one group while, populations from Colombia established another cluster, together with P8 and K21 (Fig. 4).
The LnP(D) result from Structure v2.2 showed gradual increment from K = 1 to K = 10 without clear peak (Additional file 1). Thus, we further examined the results attained for K = 2, 3, 4 and 5 (Fig. 5a, b, c and d). For K = 2, populations from Honduras, Costa Rica (except K21) and three Panama populations (P3, P5 and P13) formed one subpopulation whereas the remaining populations (mainly from Colombia, together with P8, P10, P12 and K21) created the second subpopulation. At K = 3, three subgroups were attained due to the separation of P8 and P10 from the second subpopulation. Three subgroups were also retained for K = 4. At K = 5, the populations in subpopulation one broke up into two groups; populations C1, C9 and K21 differentiated from C5, C6, C8 and P12. Populations P8 and P10 remain distinct at K = 5.
The results from PowerCore are presented in Table 5. A total of 34 palms were shortlisted. These include 7 palms from Colombia, 10 from Panama, 13 from Costa Rica and another 4 from Honduras.
In this study, 77.7% of the E. guineensis microsatellites recorded successful cross-amplification in E. oleifera. Successful cross-species amplification of SSRs developed from genomic information of E. oleifera  and E. guineensis  have been reported previously. Cross-species amplification of SSRs has also been reported in many plant species for instance napier grass , Eucalyptus , Jathropa , Rhododendron , Lavandula species , sugarcane  and Dendrobium . The average transferability rate of SSRs among species within the same genus was reported at approximately 73% , a figure that is comparable to the value attained in this study. Mating system and life span are among the factors influencing cross amplification of SSRs. Oil palm is an out crossing species with long generation time. Such species is expected to record higher SSR transferability than selfed- and short-lived species . Therefore, the high cross amplification rate attained in the current work is expected.
The high transferability rate of SSRs among the Elaeis species and the comparable size of the amplified products attained in the oleiferas samples could also be the result from the highly conserved sequences flanking the microsatellite regions. Moreover, the success of SSR locus amplification across-species is higher when genetic distance between the species is small. This was demonstrated in a phylogenetic analysis carried out based on annotated subset of proteins from E. guineensis and E. oleifera, where the two species revealed close genetic relationship . Furthermore, there is also evidence on the successful application of E. guineensis SSRs for constructing the genetic linkage map involving oil palm interspecific hybrids (oleifera x guineensis)  as well as their backcrosses .
However, a number of SSRs (34%) failed to generate amplified product in E. oleifera. Mutation has been reported as one of the possible causes of failure in SSR amplification . Mutation that occurs at the SSR primer binding sites may prevent annealing and subsequently result in no amplification of PCR products . These can include indels or nucleotide substitutions in the primer binding sites .
Genetic diversity of the E. oleifera
Previous studies on evaluating E. oleifera genetic materials by means of molecular markers are quite limited.  analyzed natural populations collected from the Amazon forest in Brazil using ninety-six Random Amplified Polymorphic DNA (RAPD) markers. The group reported moderate level of diversity compared to E. guineensis accessions. The group also found that the palms were clustered according to their distribution along the Amazon River rather than geographic distances. The river network provided means for seeds dispersal. Therefore, palms along the river were grouped together.
A primary attempt to develop simple sequence repeats (SSRs) from DNA extracted from selected E. oleifera was reported in 2010 . Successful amplification of the oleifera SSRs was attained across DNA of E. guineensis, Cocos nucifera and Jessinia bataua . Further, marker data analysis revealed considerable level of genetic diversity among the populations and clear grouping of samples according to the species.
Recently,  described the genetic diversity of Elaeis oleifera natural populations assembled from four countries, namely Peru, Brazil, Colombia and Ecuador as well as two hybrid populations created from E. oleifera and E. guineensis. Using 13 SSRs developed by , four genetic groups could be distinguished. These groups corresponded with the country where the accessions were assembled. The two hybrid populations were clustered respectively into Colombia and Brazil groups, in agreement with the origin of the female parents used to create them. Significant differences between countries were reported for several phenotypic data such as mesocarp-to-fruit and oil-to-bunch ratios. Further analysis of both, the molecular and phenotypic data revealed the number of entries per country that are needed in the core collection for long term conservation.
In this study, we have determined the genetic diversity of 19 E. oleifera populations using 14 E. guineensis SSRs. Previously,  and  had applied 16 and 13 SSRs respectively across populations of E. guineensis and E. oleifera with satisfactory results. The average number of alleles per locus and expected heterozygosity in E. oleifera populations (Na = 3.00; He = 0.221) analysed in this study were lower than that detected in E. guineensis (Ao = 5.0; He = 0.644) , indicating a rather low genetic diversity of the American oil palm populations. Similar observation was also reported by [19, 35] in a study carried out on small number of E. guineensis and E. oleifera samples. The oleifera populations were reportedly less diverse too, in terms of the phenotypic traits especially palm height, mesocarp ratio and nut weight . Nevertheless, E. oleifera offers genes that are not available in the guineensis. Genes for low height increment, unsaturated fatty acids and high carotene content [6, 8] can be exploited for improvement of the E. guineensis.
At the population level, one of the populations from Colombia (C6) possessed negative fixation index (F) value (-0.030). This population was located at Cerete, Colombia. Bulk of the natural groves of oleiferas in Cerete was removed to give way to agriculture . The remaining genetic materials were therefore very sparse and located far from each other. In addition, during the exploration for genetic materials, the collection team noted the presence of two oil palm mills in this area  indicating extensive exploitation of the oleifera fruits by the locals. Oleifera bunches may have been transported from far for oil extraction at these mills. It has been shown that oil palm seeds from the processed bunches preserve germination ability thus, unrelated palms from these bunches could have been established in areas near the mills in Cerete. Hence, the genetic materials assembled from this area possess dissimilar genotypes, introduced by the unrelated palms, which is reflected in higher Ho value. This leads to the negative F value, as seen for the population from Cerete (C6).
Allelic richness is a useful estimate for evaluating population diversity. The plot of allelic richness Na(rar) of each population against position (Fig. 1) indicated significant decreasing trend of allele richness from Costa Rica towards two countries namely, Colombia and Honduras. High genetic diversity estimates were also attained for the populations from Costa Rica. These findings suggest that the country can be denoted as the center of diversity for E. oleifera. Analysis carried out on the E. oleifera remains across an area between the southern part of the United States and the south of Uruguay  concluded that the establishment of the E. oleifera populations in Colombia was more recent. This archaeological evidence further supports the low Na(rar) and genetic diversity among the Colombia populations. Similarly, populations from Honduras also exhibited low diversity. However, the low diversity observed among the genetic materials from Honduras is probably due to founder effect. Here, we only investigated small number of populations ie. two from this country thus, limited diversity was captured.
Population structure and differentiation
The AMOVA revealed that 29% of the variation observed in this study was attributed by the genetic differentiation between the populations. The differences between the populations are considered extensive, thus sampling a limited number of individuals from many populations should be implemented for future collection and establishment of core collection. Our results are in agreement with that reported by , based on phenotypic variation. The genetic differentiation between populations in E. oleifera is higher than that reported in E. guineensis (0.206) . Extensive distribution of E. guineensis wild groves was reported along the oil palm belt stretching from the west to the central Africa. However, unlike E. guineensis, the natural distribution of E. oleifera was highly discontinuous . In such condition, mating involving individual within a population was more common. Thus, higher differentiation among population is expected in this study.
The phylogenetic tree presented in Fig. 4 indicated grouping of the populations according to country except for P12 (from Panama) and K21 (from Costa Rica). Similar results are also observed in Fig. 5b and c. In these figures, P12 and K21 revealed different genetic attributes as compared to other populations from Panama and Costa Rica, respectively. The results presented in Fig. 5 indicated that gene flow generally occurred at different rate among individuals across populations and countries thus, no populations exhibited absolute uniformity. Gene flow was clearly evidenced across countries, particularly between P12, K21 and the Colombian populations as well as between populations from Honduras and Costa Rica. At K = 5 (Fig. 5d), P12 revealed almost similar genetic attributes with C5 and C6. P12 is the nearest collection site to Colombia, located at the east of the Panama Canal. Considerable rate of genetic exchange could have occurred among these populations which, explains their grouping.
Between 1967 and 1976, a private company in Costa Rica initiated collections of E. oleifera seeds from natural populations in Panama, Colombia, Suriname, Honduras, Nicaragua and Brazil . These genetic materials were planted in Coto for phenotypic evaluation. One of the populations analysed in this study was also assembled from Coto (K21). Our results showed that K21 is genetically similar to populations C1 and C9 from Colombia (Fig. 5d). The individual palms collected from population K21 could be the result from hybridization between palms naturally found in Costa Rica and those assembled from Colombia . This possibly explained the close relationship of K21 with the Colombian populations.
Implication for breeding and future conservation programs
The current ex situ living collection of E. oleifera occupies approximately 18 hectares of land, accommodating approximately 2,500 palms. Maintaining living collection for large crop like oil palm is expensive. Nonetheless, field evaluation of the collection revealed some interesting traits useful for oil palm genetic improvement . Achieving genetic gain through conventional breeding takes a very long time for oil palm. This is because approximately ten years is required to complete one cycle of field data collection and evaluation for the species. Therefore, appropriate planning to preserve individual palms that possess traits of interest should be in place to ensure long-term accessibility. The principles of core collection where maximum genetic variation is preserved in a reduced land size can be applied. The establishment of such collection is an ideal option for oil palm as it increases efficiency of conservation and allows for a more effective access to the genetic materials.
The results presented above facilitate the identification of unique populations or rich allelic individual palms as well as populations that exhibit high genetic variation. These results are useful for selecting genetic materials for establishing the core collection. Some genetically variable populations were identified based on the genetic diversity parameters estimated in this study. Among the interesting populations are C5, C8, P5, P12, K4, K14 and K15, as they exhibited genetic diversity estimates higher than the mean. The results from AMOVA indicated greater diversity between populations which suggests that sampling limited individuals from many populations should capture maximum diversity. Previous analysis  revealed that analyzing population size of 20 and 30 individuals resulted in comparable genetic diversity measures in oil palm. Therefore, preserving 20–30 seeds per population may cover the genetic diversity present in the population. In addition, 34 rich allelic palms were identified from PowerCore analyses. Offsprings could then be created through selfing or intercrossing of these palms. Efforts to preserve these materials further should be initiated to ensure selection gains and accessibility in the future.
Our results also revealed high pairwise genetic differentiation (Fst) between populations from Colombia and those from other countries. This can be visualized in Fig. 3 - below diagonal. Crossing program can be initiated using palms from two populations that revealed dark blue for instance, C1 and K8, C1 and H2, C1 and H3, C1 and P13 or C9 and H2. Based on the phenotypic data available [6, 8, 10], individual palms exhibiting interesting traits can be selected from each of these populations. Palms that possess low height increment, high mesocarp content, high carotene content and high oil unsaturation can be included in the crossing scheme. Such breeding programme can achieve two goals, first, incorporating the traits of interest into the next generation for future access and secondly, retaining the overall genetic variability for selection gains in the future.
From the current work, 77% of the E. guineensis SSRs tested showed cross amplification in E. oleifera. Of these, 40% showed discriminative power among populations thus, proving to be a reasonably useful marker resources for genetic studies of E. oleifera. The genetic diversity estimates indicated high genetic variability among the populations from Costa Rica. This result suggests that the country may serve as the centre of diversity of E. oleifera. This study also provides valuable information that help oil palm breeders and germplasm manager in identifying individual palms and populations for establishment of core collection. Implementing the suggestions described above would result in comprehensive genetic collection that retains genetic diversity as well as valuable traits for selection gains and access in the future.
Plant materials and DNA extraction
A total of 532 palms representing 19 populations were sampled from the MPOB ex situ living collection of E. oleifera. These populations were selected based on their geographic distribution, microclimates, altitudes and rainfalls in Colombia, Panama, Costa Rica and Honduras (Additional file 2). For Colombia, we included genetic materials collected from a wider range of areas as compared to that reported by . Figure 6 indicates the distribution of the populations analysed in this study. For each population, 22–30 palms were sampled depending on availability of the materials. Young unopened leaves were harvested from each palm, cleaned, and kept in -80 °C.
Genomic DNA was extracted from 3 g of fresh leaf using CTAB DNA extraction  with minor modification. The DNA samples were quantified by optical density (OD) reading using spectrophotometer (UV/VIS Spectrometer Lambda Bio, Perkin Elmer, USA). The DNA purity was further tested with DNA digestion technique using two restriction enzymes namely, EcoRI and HaeIII, followed by separation on 1.0% agarose gel in TBE buffet at 100 V for 1 h. The DNA was visualized after ethidium bromide staining under ultraviolet exposure.
Testing for Transferability of E. guineensis microsatellite markers to E. oleifera
In this study, 18 microsatellite markers designed based on , were tested for their polymorphism in E. oleifera. A screening panel that contained three DNA samples of each population was established for testing the markers. The details of the 18 microsatellite markers applied are indicated in Additional file 3. These markers were selected to represent each linkage group based on the genetic map constructed by .
The total volume of each PCR reaction was 12.5 μl, comprising 1x PCR buffer (50 mM KCl, 10 mM Tris-HCl (pH 9.1 at 21 °C), 0.1% TritonTMX-100), 1.5 mM Mg2Cl, 200 μM dNTPs mix, 1.0 U Taq DNA polymerase (Vivantis, Malaysia), 0.25 μM of each primer, 0.38 μM fluorescent dye (Applied Biosystems, USA) and 50 ng of DNA template. The PCR conditions included initial denaturation at 94 °C for 5 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 52 °C for 1 min and elongation at 72 °C for 2 min, then final elongation step of 72 °C for 10 min. The M13-tailed forward primer was labeled with 4-colour fluorescent dyes. The fluorescent dyes used in the study were FAM (blue), PET (red), VIC (green) and NED (yellow). The sample set was subdivided into four panels and each of them consisted of four microsatellite primers labeled with four fluorescent dyes, respectively.
Fragment analysis and genotyping
Prior to fragment analysis, PCR products from four different SSR primers were combined at 1:1:1:1 ratio. The multiplex (2.0 μl) was pretreated by adding 7.80 μl of Hi-DiTM Formamide (Applied Biosystems, USA) and 0.20 μl of GeneScanTM-500 LIZTM Size Standard (Applied Biosystems, UK) to make up 10.0 μl of the final volume. The mixture was vortexed thoroughly and followed by denaturation at 95 °C for 5 min. The plate was then placed on the ice immediately for another 5 min before capillary electrophoresis using ABI PRISM® 3100 Genetic Analyzer (Applied Biosystems, USA). The raw reads were retrieved and exported to GeneMapper V3.1 software. The genotype profiles were finally generated and displayed in the software for scoring purpose.
Marker informativeness was evaluated based on the G”st values determined in GenAlex version 6.5 [41, 42]. These estimates were computed after correction for small population size based on  and  as described in [41, 42]. In addition, the fundamental genetic diversity parameters such as allele frequencies, average number of different allele (Na), number of effective alleles (Ne), Shannon Information Index (I), observed heterozygosity (Ho), unbiased expected heterozygosity (He) and fixation index (F) were also estimated for each population.
Allelic richness (Na (rar)) for each population was estimated using HP-Rare 1.1 software [45, 46]. Generally, large population size possesses more alleles. In HP-Rare 1.1 software, allelic richness is estimated in a standardized sample size for each population which, results in more accurate valuation. The allelic richness results were plotted against the distance between the populations. The distance was estimated using the longitude information of each population against the longitude of the population located at the most west, that is K4 (population 4 of Costa Rica). Distances (in kilometers) were computed based on the longitude values using a calculator available at the USA National Hurricane Centre (http://www.nhc.noaa.gov/gccalc.shtml).
GenAlex version 6.5 was also applied in determining the genetic relatedness among the individual palm in the form of Principle Component Analysis (PCoA) plot. AMOVA was carried out to generate the overall partitioning of the genetic variation within and between the populations. Fst values and genetic distances (GD)  were estimated to provide further detailed information of relatedness among the populations. Phylogenetic tree was later constructed in ITOL website (http://itol.embl.de/upload.cgi)  using the tree scripts attained from the GD values. The pairwise GD and Fst estimates among the sampled populations were further illustrated in a heatmap using R scripts.
Structure V2.2 program  was applied to assign the palms into subpopulations. Analysis was carried out with the assumption that the number of groups or admixtures (K) was unknown. The parameters were set at burn-in of 10,000 with 100,000 repeats for K values ranging from 1–10.
We applied PowerCore software  to shortlist individual palms that collectively represent optimum diversity and reduced duplication of the E. oleifera analyzed in this study. The software was developed using a heuristic algorithm that retained all different alleles and the results were highly reproducible.
analysis of molecular variation
Completely randomized design
interactive tree of life
Malaysian palm oil board
Principle component analysis
Polymerase chain reaction
Random amplified polymorphic DNA
Simple sequence repeats
Hartley CWS. The Oil Palm (Elaeis guineensis). 3rd ed. New York: Longman; 1988.
Hartley CWS. The Oil Palm. 2nd ed. New York: Longman; 1977.
Meunier J, Boutin D. L’Elaeis melanococca et l’hybride Elaeis melanococca x Elaeis guineensis – premières données. Oléaginux. 1975;30:5–8.
Meunier J, Hardon JJ. Interspecific hybrids between Elaeis guineensis and Elaeis oleifera. In: Corley RHV, Hardon JJ, Wood BJ, editors. Development in Crop Science 1 – Oil Palm Research. Netherlands: Elsevier Scientific Publishing Company; 1976. p. 127–38.
Rajanaidu N. Collection of oil palm (Elaeis oleifera) genetic material in Central and South America. PORIM Bull. 1983;6:1–11.
Mohd Din A, Rajanaidu N, Jalani BS. Performance of Elaeis oleifera from Panama, Costa Rica, Colombia and Honduras in Malaysia. J Oil Palm Res. 2000;12(1):71–80.
Choo YM, Yusof B. Carotenes. vitamin E and sterols in oils from Elaeis guineensis, Elaeis oleifera and their hybrids. PORIM Palm Oil Tech Bull. 1999;5(1):7.
Mohd Din, A.; Rajanaidu, N.; Kushairi, A.; Mohd Rafii, Y.; Mohd Isa, Z.A.; Noh, A. PS4-high carotene E. oleifera planting materials. MPOB Information Series 2002, No. 154.
Sundram K, Sambanthamurthi R, Tan YA. Palm fruit chemistry and nutrition. Asia Pac J Clin Nutr. 2003;12(3):355–62.
Rajanaidu N. Elaeis oleifera collections in Central and South America. In: Proceedings of the International Workshop on Oil Palm Germplasm and Utilization. Selangor: Palm Oil Research Institute Malaysia, Selangor, 26-27 March 1985, Palm Oil Research Institute Malaysia; 1985. p. 84–94.
Amblard P, Noiret JM, Potier F, Kouame B, Adon B. Comparative performance of interspecific hybrids and commercial E. guineensis materials. In: Proc. of the Seminar on Worldwide Performance of DxP, Interspecific Hybrids and Clones. Selangor: Barranquilla, Colombia. 5-6 June 1995, Palm Oil Research Institute Malaysia; 1995. p. 101–7.
Hardon JJ. Interspecific hybrids in the genus Elaeis II. Vegetative growth and yield of F1 hybrids E. guineensis x E. oleifera. Euphytica. 1969;18:380–8.
Frankel OH. Genetic perspectives of germplasm conservation. In: Arber W, Llimensee K, Peacock WJ, Starlinger P, editors. Genetic Manipulation: Impact on Man and Society. Cambridge: Cambridge University Press; 1984. p. 161–70.
Billotte N, Marseillac N, Risterucci AM, Adon B, Brotteir P, Baurens FC, Singh R, Herran A, Asmady H, Billot C, et al. Microsatellite-based high density linkage map in oil palm (Elaeis guineensis Jacq.). Theor Appl Genet. 2005;110:754–65.
Bakoume C, Wickneswari R, Siju S, Rajanaidu N, Kushairi A, Billotte N. Genetic diversity of the world’s largest oil palm (Elaeis guineensis Jacq.) field genebank accessions using microsatellite markers. Genet Resour Crop Evol. 2015;62:349–60. doi:10.1007/s10722-014-0156-8.
Billotte N, Jourjon MF, Marseillac N, Berger A, Flori A, Asmady H, Adon B, Singh R, Nouy B, Potier F, Cheah SC, Rohde W, Ritter E, Courtois B, Charrier A, Mangin B. QTL detection by multi-parent linkage mapping in oil palm (Elaeis guineensis Jacq.). Theor Appl Genet. 2010;120:1673–87. doi:10.1007/s00122-010-1284-y.
Singh R, Tan SG, Panandam JM, Rahimah AR, Ooi LCL, Low ETL, Sharma M, Jansen J, Cheah SC. Mapping quantitative trait loci (QTLs) for fatty acid composition in an interspecific cross of oil palm. BMC Plant Biol. 2009;9:114. doi:10.1186/1471-2299-9-114.
Ting NC, Jansen J, Mayes S, Massawe F, Sambanthamurthi R, Ooi LCL, Chin CW, Arulandoo X, Seng TY, Syed Alwee SSR, Ithnin M, Singh R. High density SNP and SSR-based genetic maps of two independent oil palm hybrids. BMC Genomics. 2014;15:309. http://www.biomedcentral.com/1471-2164/15/309.
Mohd Zaki N, Singh R, Rosli R, Ismail I. Elaeis oleifera Genomic-SSR Markers: Exploitation in Oil Palm Germplasm Diversity and Cross-Amplification in Arecaceae. Int J Mol Sci. 2012;13:4069–88. doi:10.3390/ijms13044069.
Singh R, Zaki NM, Ting NC, Rosli R, Tan SG, Low ETL, Ithnin M, Cheah SC. Exploiting an oil palm EST the development of gene-derived and their exploitation for assessment of genetic diversity. Biologia. 2008;63:1–9.
Azevedo ALS, Costa PP, Machado JC, Machado MA, Pereira AV, da Silva Lédo FJ. Cross species amplification of Pennisetum glaucum microsatellite markers in Pennisetum purpureum and genetic diversity of napier grass accessions. Crop Sci. 2012;52(4):1776–85.
Yashoda R, Ghosh M, Sumathi R, Gurumurthi K. Cross-species amplification of Eucalyptus SSR markers in Casuarinaceae. Acta Bot Croat. 2005;64:115–20.
Sudheer PDVN, Mastan SG, Rahman H, Prakash CR, Singh S, Reddy MP. Cross species amplification ability of novel microsatellites isolated from Jatropha curcas and genetic relationship with sister taxa: Cross species amplification and genetic relationship of Jatropha using novel microsatellites. Mol Biol Rep. 2011;38:1383–8. doi:10.1007/s11033-010-0241-9.
Qin WX, Yuan H, Lin LC. Cross-amplification and characterization of microsatellite loci for the genus Rhododendron. HortScience. 2010;45:1394–7.
Adal AM, Demissie ZA, Mahmoud SS. Identification, validation and cross-species transferability of novel Lavandula EST-SSRs. Planta. 2005;241:987–1004.
Singh RK, Jena SN, Khan S, Yadav S, Banarjee N, Raghuvanshi S, Bhardwaj V, Dattamajumder SK, Kapur R, Solomon S, Swapna M, Srivastava S, Tyagi AK. Development, cross-species/genera transferability of novel EST-SSR markers and their utility in revealing population structure and genetic diversity in sugarcane. Gene. 2013;524:309–29.
Xu W, Zhang F, Lu B, Cai X, Hou B, Feng Z, Ding X. Development of novel chloroplast microsatellites markers for Dendrobium officinale and cross- amplification in other Dendrobium species (Orchidaceae). Sci Hort. 2011;128:485–9.
Barbara T, Palma-Silva C, Paggi GM, Bered F, Fay MF, Lexer C. Cross-species transfer of nuclear microsatellite markers: potential and limitations. Mol Ecol. 2007;16:3759–67.
Singh R, Ong-Abdullah M, Low ETL, Abdul Manaf MA, Rosli R, Rajanaidu N, Ooi LCL, Ooi SE, Chan KL, Halim MA, Azizi N, Nagappan J, Bacher B, Lakey N, Smith SW, He D, Hogan M, Budiman MA, Lee EK, DeSalle R, Kudrna D, Goicoechea JL, Wing R, Wilson RK, Fulton RS, Ordway JM, Martienssen RA, Sambanthamurthi R. Oil palm genome sequence reveals divergence of interfertile species in old and new Worlds. Nature. 2013;500:335–9.
Montoya C, Lopes R, Flori A, Cros D, Cuellar T, Summo M, Espeout S, Rivallan R, Risterucci AM, Bittencourt D, et al. Quantitative trait loci (QTLs) analysis of palm oil fatty acid composition in an interspecific pseudo-backcross from Elaeis oleifera (H.B.K.) Cortés and oil palm (Elaeis guineensis Jacq.). Tree Genet Genomes. 2014;9(5):1207–25. doi:10.1007/s11295-013-0629-5.
Callen DF, Thompson AD, Shen Y, Phillips HA, Richards RI, Mulley JC. Incidence and origin of null alleles in the (AC)n microsatellite markers. Am J Hum Genet. 1993;52:922–7.
Li YC, Korol AB, Fahima T, Beiles A, Nevo E. Microsatellites: genomic distribution, putative functions and mutational mechanisms: a review. Mol Ecol. 2002;11(12):2453–65.
Varshney RK, Graner A, Sorrells ME. Genic microsatellite markers in plants: features and applications. Trends Biotechnol. 2005;23:48–55.
Moretzshon MC, Ferreira MA, Amaral ZPS, Coelho PJA, Grattapaglia D, Ferreira ME. Genetic diversity of Brazillian oil palm (Elaeis oleifera H.B.K.) germplasm collected in the Amazon forest. Euphytica. 2002;124:35–45.
Mohd Zaki N, Ismail I, Rosli R, Chin TN, Singh R. Development and characterization of Elaeis oleifera microsatellite markers. Sains Malaysiana. 2010;39(6):909–12.
Arias D, Gonzalez M, Prada F, Ayala-Diaz I, Montoya C, Daza E, Romero HM. Genetic and phenotypic diversity of natural American oil palm (Elaeis oleifera H.B.K. Cortes) accessions. Tree Genet Genomes. 2015;11:122. doi:10.1007/s11295-015-0946-y.
Morcote-Rio G, Bernal R. Remains of Palms (Palmae) at Archaeological Sites in the New World: A Review. Botanical Rev. 2001;67:309–34.
Escobar R. Preliminary results of the collection and evaluation of the American oil palm Elaeis oleifera (HBK Cortes) in Costa Rica. In: Pushparajah E, Chew PS, editors. The Oil Palm in Agriculture in the Eighties, vol. 1. Kuala Lumpur: Incorporated Society of Planters; 1982. p. 79–94.
Hayati A, Wickneswari R, Maizura I, Rajanaidu N. Genetic diversity of oil palm (Elaeis guineensis Jacq.) germplasm collections from Africa: implications for improvement and conservation of genetic resources. Theor App Genet. 2004;108:1274–84.
Doyle J, Doyle L. Isolation of plant DNA from fresh tissue. Focus. 1990;12:13–5.
Peakall R, Smouse PE. GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes. 2006;6:288–95.
Peakall R, Smouse PE. GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research – an update. Bioinformatics. 2012;28:2537–9.
Nei M, Chesser RK. Estimation of fixation indexes and gene diversities. Ann Hum Genet. 1983;47:253–9.
Meirmans PG, Hedrick PW. Assesing population structure: FST and related measures. Mol Ecol Res. 2011;11:5–18.
Kalinowski ST. Counting alleles with rarefraction: private alleles and hierarchical sampling designs. Conserv Genet. 2004;5:539–43.
Kalinowski ST. HP-Rare: a computer program for performing rarefraction on measures of allelic diversity. Mol Ecol Notes. 2005;5:187–9.
Nei M. Genetic distance between populations. Amer Nat. 1972;5:283–392.
Lutenic I, Bork P. Interactive tree of life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees. Nucl Acids Res. 2016;44(W1):W242–5. doi:10.1093/nar/gkw290.
Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155:945–59.
Kim KW, Chung HK, Cho GT, Ma KH, Chandrabalan D, Gwag JG, Kim TS, Cho EG, Pak YJ. PowerCore: a program applying the advanced M strategy with a heuristic search for establishing core sets. Bioinfor. 2007;23:2155–62. doi:10.1093/bioinformatics/btm313.
The authors would like to thank the Director-General of MPOB for permission to publish this paper.
This study was supported by Malaysian Palm Oil Board, under Graduate Student Assistantship Scheme. Employees of the funding body were involved in the design of the study, analysis and interpretation of data and in writing the manuscript
Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
MI and WR conceived and designed the experiments; TCK performed the experiments; MI and TCK analyzed the data; MI and TCK wrote the paper. WR reviewed the paper. All authors have read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
Ethics approval and consent to participate
The genetic materials analysed in the study were attained through Government to Government agreements, involving MPOB, source countries and the Bioversity International (previously known as International Board of Plant Genetic Resources). MPOB collected the genetic materials with consultation and approval of the source countries.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Log probability data (LnP(D)) as function of k (number of clusters) from the STRUCTURE. (DOCX 2712 kb)
Information of country, collection sites and population size included in the study. (DOCX 16 kb)
Details of the 18 microsatellite markers used in the study. (DOCX 16 kb)