Morphological and molecular characterization of an Elaeis oleifera (H.B.K) Cortes germplasm collection located in Ucayali, Peru

The African oil palm (Elaeis guineensis Jacq) is a crop that is widely distributed in tropical regions around the world; however, this crop is subject to limitations such as rapid trunk growth and susceptibility to bud rot and red ring diseases particularly in South America. To overcome these limitations, national breeding and conservation programs have been established, and there is a need to identify parental palms from natural populations of the American oil palm (E. oleifera H.B.K. Cortes) with desirable yield and morphological traits (i.e., yield production and bunch number) and with high genetic diversity. However, in Peru the morphological and genetic data related to this important crop is limited. In this study, we characterized the morphological and yield and estimated the genetic diversity using 12 neutral microsatellite markers (simple sequence repeats, SSRs) across 72 oil palm individuals belonging to the E. oleifera germplasm collection located in the tropical region of Ucayali, Peru. Our results showed that morphological and yield traits explained approximately 40.39% of the variability within the Peruvian germplasm. Furthermore, Yield Production was highly correlated with two yield traits: Bunch Number (0.67) and Average weight per bunch (0.78). Based on the yield and morphological traits, a clustering analysis was performed and three phenotypic groups were identified (1, 2 and 3) in which groups 1 and 3 showed high scores associated primarily with yield traits. Microsatellite markers revealed 143 alleles, 11.92 ± 4.72 alleles per locus (A) and an expected heterozygosity (He) of 0.69 ± 0.045. A structural analysis identified three populations (k = 3), that were not related to the phenotypic groups. Interestingly, a multiple allele background was identified within the groups using multilocus and phylogenetic relationship analyses. This is the first Peruvian report regarding E. oleifera that shows preliminary data of the morphological and yield traits and genetic data, and highlight the importance of this information to set up future steps to national breeding strategies and improve the conservation of genetic material of E. oleifera. Overall, these novel findings could contribute to the development of the local oil palm industry in Peru.


Response and change:
We agree with the comment. We have revised diligently our manuscript and adapted it to PLOS ONE's style requirements.
2. Comment: Thank you for stating the following in the Competing Interests section: 'The authors have declared that no competing interests exist.' We note that one or more of the authors are employed by a commercial company: Independent research.
2.1.Please provide an amended Funding Statement declaring this commercial affiliation, as well as a statement regarding the Role of Funders in your study. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. You can update author roles in the Author Contributions section of the online submission form. Please also include the following statement within your amended Funding Statement. "The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the 'author contributions' section." If your commercial affiliation did play a role in your study, please state and explain this role within your updated Funding Statement.

Response and change:
We thank the Editor's comment. We have updated our Funding Statement in the Financial Disclosure on the submission system.
Original text: This work was financially supported by the Programa Nacional de Innovación Agraria  Response to Reviewer 1: 1. Comment 1: The study focuses on morphological, yield traits and genetic diversity of E.oleifera. However, the study on fruit related traits, which is an important commercial trait in oil palm is missing. Furthermore, according to descriptive statistics (Table 2), morphological traits has reduced variability than yield traits whereas yield traits has reduced variability than morphological traits in PCA analysis. There was a conflict between these two analyses. Is there any explanation for the observed results?

Response and change:
We thank the reviewer for the comment.
We have focused the main idea of this article in the principal morphological and yield traits and some yield traits were related with fruit such as: i) the weight of fruits at harvest, ii) the number of fruits per palm during harvest, iii) weight in kg of 10 fruits, and iv) fruit diameter (Material and Method section; Phenotyping). These traits were described in American oil palm for genetic studies using molecular markers such as SSRs, in order to set up future studies in Peru that figure out the relationship between morphological and yield traits in this important crop new manuscript is being prepared to present information related the fruit traits and quality in oil belong from this peruvian E.oleifera germplasm.
Regarding with the next sentence: "Furthermore, according to descriptive statistics (Table 2), morphological traits has reduced variability than yield traits whereas yield traits has reduced variability than morphological traits in PCA analysis. There was a conflict between these two analyses. Is there any explanation for the observed results?". We have rephrased the whole paragraph in order to avoid the misinterpretation the results of variability. We emphasized in the results of descriptive statistics (Table 2) and PCA analysis, also we added a PCA results as a supplementary material to support our findings (S1 Table in  To compare the variability among traits, a principal component analysis (PCA) was performed to identify the main traits that contributed to the dispersion of the values (Fig 1). Of the total variation in the Peruvian E.oleifera germplasm, was explained by two components: PC1 (26.01%), which was associated with four morphological traits (FA, LEL, LXL and LD) and two yield traits (YP and ABW) and PC2, which was primarily associated with two morphological traits (TH and CC) and explained the 40.39% of the variance in the Peruvian germplasm. The PCA analysis demonstrated that the morphological traits explained more of the variability in the oil palm individuals in this study than the yield traits.
Revised sentence: Results section; page 9-10; lines 215-231 Descriptive statistics for the morphological and yield traits of the Peruvian Elaeis oleifera germplasm (n=72) are shown in Table 2. The morphological trait that exhibited the highest "Decenio de la igualdad de oportunidades para mujeres y hombres" "Año de la universalización de la salud" A principal component analysis (PCA) was performed to identify the main traits that contributed to the dispersion of the values (Fig 1). Of the total variation in the Peruvian E.oleifera germplasm, was explained by two components: PC1 (26.01%), which was associated with five morphological traits (FA, LEL, LL, LXL, and LDW) and two yield traits (YP and ABW) and PC2, which was primarily associated with three morphological traits (TH, TD, and CC) and one yield trait (FW) explained the 40.39% of the variance in the Peruvian germplasm. The PCA analysis demonstrated that both the morphological and yield traits explained the variability in the oil palm individuals in this study (S1 Table in S2 File).
2. Comment 2. The author described the importance of relationship between the morphological traits and genetic diversity. According to their results, groups 1 and 3 were more closely related in morphological and yield traits whereas groups 1 and 2 were more closely related in genetic diversity analysis (number of alleles). Would appreciate more explanation for these results.

Response and change:
We appreciate the revisor´s comment. Regarding the genetic differences between groups 1, 2 and 3 (number of alleles), we had tested that difference to He and A between the groups using statistic parameter and we had not found any evidence to the differentiation between groups and we reported it in the first version of this manuscript Discussion section; page 16; lines 390-394. However, we have considered to include some lines to explain the lack of differences due to the low number of individuals for each group. And, we would like to emphasize in the use of neutral microsatellites to identified a correct set of diversity genetic parameters. Discussion section; page 17; lines 403-405.
3. Comment 3. The abstract can be made more reflective of the total paper. The authors need to highlight what are altogether the novel results of their study.

Response and change:
We thank the reviewer for the comment. We proceed to improve the abstract in order to highlight our novel results and its impact in the Peruvian industry. Abstract section; pages 2-3; lines 46-71.
"Decenio de la igualdad de oportunidades para mujeres y hombres" "Año de la universalización de la salud" 4. Comment 4. Although previously published methodologies were referred, the methods and the number of replicates for measurement of phenotypic data in materials and methods section should be briefly described.

Response and change:
We thank the reviewer for the comment. we would like to rephrase a paragraph related to "Phenotyping" and add more information about the methods and the number of replicates that we employed in the study.
Original text: All experimental data obtained were related to the morphological and yield characteristics of an Each measure was made using previously published methodologies (2, 20, 21).
Revised sentence: Material and Methods section; page 6-7; lines 134-163 All experimental data obtained were related to the morphological and yield characteristics of an American palm (Elaeis oleifera H.B.K. Cortes) plantation that was laid out using a traditional pattern, a quincunx, with plant and row spacing of 9 x 9 m. Nine morphological characters were quantified: Trunk Diameter (TD was taken from the trunk circumference at the midsection with a tape from 10 cm below the crown height), Trunk Height (HT was estimated from the distance between the lowest green leaves and the fruit), Cup Coverage ((CC) The leaves were counted one by one following the phyllotaxis of the trunk), Leaf Length ((LL) the leaf located in the T: (51) 61-571831 E-mail: Pucallpa@inia.gob.pe www.inia.gob.pe www.minagri.gob.pe middle third was taken, similar to the position of leaf number 17 in E. guineensis, once it was cut, the total length of petiole and length of the rachis were added to estimate the Leaf Length)), Petiole Length (PL was measured from petiole separated of the rachis in leaf number 17), Leaf Dry Weight (The dry weight were determined by chopping the leaf number 17 finely, and drying to constant temperature at 100-105ºC, them, the dry weight was multiplied by the number of leaves produced to estimate LDW), Foliar Area (FA represents the mean area per leaf multiplied by the number of leaves per palm this measure was obtained by measuring two leaves per palm), Leaflet length (LL was determinates in the largest leaflet in the leaf number 17 by measuring from apex to the insertion base of the rachis leaf ), Leaflet diameter (LD was estimated by bending the largest leaflet in the middle and then width measure was taken) and Leaflet per Leaf (LXL, the number of leaflet per leaf was estimated by counting the leaflets in the leaf number 17 including rudimentary leaflets). The five yield characters were: Average Bunch Weight (ABW was determined by weighing the bunch at harvest time) , Bunch Number (BN was represented by the number of bunch per palm collected during the first harvest), the 10-weight fruits (FW was estimated by weighing 10 fruits at random), fruit diameter (FD) and Yield per Palm (YP, this indicator was calculated form the total production in kg of fruit per palm per year). Each measure was made using previously published methodologies (2, 20, 21). And, the number of replications for each morphological trait were done every two weeks during the year of evaluation with the exception of the LL, PL, LDW, and LD traits that were done once due to the implication of the use of destructive steps. The complete collected phenotyping data is presented in S1 file.

Comment 5.
In the results section (Page 8, Line #186-187), it is not clear whether yield traits are compared to morphological traits or not. Please re-write that part.

Response and change:
We thank the reviewer for her/his observation. We agree with the reviewer; however, we decided eliminate this sentence in order to make a response to Reviewer 1, comment 1. 6. Comment 6. Page 11, Line # 261, the word "wich" should be "which"..).

Response and change:
We thank the reviewer for the correction. We agree and we proceed to corrected it. This change is shown in yellow color. Discussion section; page 13; line 307 "Decenio de la igualdad de oportunidades para mujeres y hombres" "Año de la universalización de la salud"  Table 4, the word "morphologycal" should be "morphological".

Response and change:
We thank the reviewer for the correction. We agree and we proceed to corrected it. This change is shown in yellow color. Table 4 8. Comment 8. Page 262, Line 262-263, please clearly describe the sentences.

Response and change:
We thank the reviewer for the observation. Also, we considered to add a sentence to explain the differences between Colombian and Peruvian germplasm.
Original text: The primary phenotypic differences within this Peruvian germplasm were based on two morphological traits, trunk height (TH) and cup coverage (CC), wich explained 40.4% of the total variation in contrast, the variation in an Elaeis oleifera Colombian germplasm was explained by for four traits (trunk height, leaf area, foliar area and leaf dry weight) (2) Revised sentence: Discussion section; page 13; lines 305-310 The primary phenotypic differences within this Peruvian germplasm were based on three morphological traits, trunk height (TH), trunk diameter (TD) and cup coverage (CC), and one yield trait (FW, Fruits 10-weight)which explained 40.4% of the total variation in contrast, the variation in an E. oleifera Colombian germplasm was explained by four morphological traits (trunk height, leaf area, foliar area, and leaf dry weight) (2) however, the age of both plantations could explain this differences.
9. Comment 9. English is sometimes odd, please improve it.

Response and change:
We thank the reviewer for his/her comment. We considered improving the English employing a professional language service. We attached the certificate of this service.
1. Comment 1. The manuscript describes genetic diversity of Elaeis oleifera assembled from an area in Peru by means of 12 SSRs and selected morphological traits. The manuscript was written in readable English language but with minor spelling mistakes and some grammatical errors.
These errors are however correctable. I have specific concerns as below. The phenotypic data was collected at very young age, when the palms are 4 years old, as described by authors in method section. It should be cautioned that analysis and interpretation of such data can be erroneous and misleading. It is very important to note that For oil palm specifically, standard phenotypic data collection is carried out for at least 8 years, starting from palms at 2 or 3 years old. Measuring vegetative traits such as height, frond length is carried out once on palms at 8th year old. I am doubtful on the results presented especially if it is going to be used for the purposes of selection and breeding.

Response and change:
We thank the reviewer for his/her comment. Regarding to the English language we answered on the same way that the reviewer 1, comment 9.
Regarding with the next sentence: We agree with the reviewer and we have done some changes in the message of our manuscript in order to provide a realistic application of our results.
i) We clarified that the term "four-year-old" of the germplasms is since of transplanting event. Material and methods section; page 6; lines 129-130.
ii) The objective of our study was rephrased as follow: Original Text In that context, the aim of this study was to characterize the E. oleifera (H.B.K) Cortes germplasm of a Peruvian collection using morphological traits and SSR markers to further national breeding programs for the oil palm.

Revised sentence: Introduction section; page 5; lines 118-121
In that context, the aim of this study was to characterize the E. oleifera (H.B.K) Cortes germplasm of a Peruvian collection using morphological traits and SSR markers to contributed with important data for setting up of further studies that will be related to the development of national breeding programs for the oil palm.
iii) We eliminated a sentence and added a new paragraph to highlight the principal limitation of our study related to the age of the germplasm.

Eliminated Text: Discussion section; page 17; lines 429-431
This information must guide us in the decision-making process when planning breeding programs that are focused on crosses to obtain OxG interspecific hybrids (E. oleifera x E. guineensis) with interesting yield traits.
"Decenio de la igualdad de oportunidades para mujeres y hombres" "Año de la universalización de la salud" iv) The final paragraph of our study was rephrased as follow: Original Text "Decenio de la igualdad de oportunidades para mujeres y hombres" "Año de la universalización de la salud" Null alleles introduce error in assessment of genetic diversity, parentage analysis etc. The authors can carry out relevant analysis to identify null alleles, exclude them from dataset prior to genetic diversity analysis.

Response and change:
We thank the reviewer for his/her comment. We agree with the comment and we tested the impact of null alleles in our analysis. Our additional analysis is described in the material and methods section and the supplementary data section, we found that five SSRs have null alleles and we excluded them and performed comparative tests between all complete set (12 SSRs) and 7 SSRs (without 5 SSRs), and our results showed that there are no significant differences. At the light of these results we decided to include this new information throughout the manuscript as follows: i) We added a new paragraph in Material and methods section (Materials and methods section; page 9; lines 207-211).
Additionally, to test the effect of null alleles on the power of discrimination genotypes was addressed by the construction of a genotype accumulation curve with 1,000 iterations with replacement of samples, while the effect on the expected heterozygosity was addressed by a Monte Carlo test with 1,000 loci resampling considering the exclusion of loci with more than 10% of null alleles. All these procedures were tested using the R v3.42 software.
ii) We added a new paragraph in Results section (Results section; page 11; lines 267-273).
"Decenio de la igualdad de oportunidades para mujeres y hombres" "Año de la universalización de la salud" The frequency of null alleles was estimated between 0.01 and 0.48 for nine microsatellites (S1 Table in  Null alleles were previously reported in Peruvian samples with frequencies around 0.06 and 0.21 (1), and in this study were reported between 0.01 to 0.48, however, the presence of five microsatellites with more than 10% of null alleles (or missing information) do not alter the estimation of the genetic diversity, in terms of He, of the germplasm of 72 samples. (S3 Table in S2 File). These alleles are produced by putative mutations that occur at the microsatellite binding sites that could prevent primer annealing and in consequence, the lack of amplification of the PCR products (8) therefore further primer development could be necessary for the local or regional E. oleifera germplasm. iv) We considered to add a Supporting Information file to support our additional results related to null alleles (S2 table, S3 table,  At the light of these results we decided to include this new information throughout the manuscript as follows: "Decenio de la igualdad de oportunidades para mujeres y hombres" "Año de la universalización de la salud" The frequency of null alleles was estimated between 0.01 and 0.48 for nine microsatellites (S2 Table in  Moreover, the limited economical resources in many laboratories in South America could interrupt the adoption of these molecular tools. Recently, there are studies in E.oleifera that show that useful information is achieved when is employed a limited set of microsatellites (<20 SSRs) (1,8) and according to our analysis could be necessary at least 7 microsatellites in the Peruvian germplasm to carry up the complete analysis of genetic diversity (S1 Fig and S3 Table in S2 File), however many studies are necessary to corroborate these findings.
iii) We considered to add a Supporting Information file (S1 Fig. in S2 file) to support our additional results related to null alleles in according to reviewer 2, comment 3.