Assessment of Genetic Diversity among Wild Ruta chalepensis L. from the North of Jordan

Abstract

Ruta chalepensis, known as Fringed Rue, is a small shrub of the Rutaceae family. To date, there is no record of its natural distribution across Jordan, a country located in the eastern part of the Mediterranean basin, and there are no previous studies on its genetic diversity in the region. Therefore, this study was conducted to assess the genetic diversity of R. chalepensis in the northern parts of Jordan using morphological trait and amplified fragment length polymorphism (AFLP) analyses. For the morphological traits, the analysis of variance indicated that there were significant differences between the identified populations. The Shannon diversity indices showed relatively high values, indicating the existence of a high variability among the identified populations. The principal component analysis explained 82% of the variation between the collected plants, and a clear separation of the collected individuals from the Jarash-A, Jarash-B and Ajloun-B populations from the rest of the populations was observed. The heatmap clustering was in general agreement with the results of the principal component analysis, with the plant height, rachis length and plant width considered as the discriminative traits. The AFLP analysis using eight different primer combinations generated 59 polymorphic bands, with an average polymorphism information content value of 0.32. The phylogenetic analysis identified three main clusters, with the first cluster including 65% of the individuals collected from the Jarash and Ajloun provinces, with a clear separation of the Jarash-B population. The AMOVA revealed that the genetic variation between the populations contributed 30% of the total genetic variation, while the variation within the populations explained 70%. In conclusion, morphological traits and molecular markers were used successfully to assess the genetic diversity among wild R. chalepensis from the north of Jordan, and such data can be used for future conservation plans and utilization purposes.

1. Introduction

Fringed Rue (Ruta chalepensis L.) is a small evergreen shrub that belongs to the Rutaceae family. It is grown in the wild throughout the Mediterranean region [1]. The plant is characterized by its distinct and powerful odor [2], and it is widely used as a traditional medicine by the indigenous population in the region [3] for the treatment of various diseases and disorders [4]. The plant is known for its abortifacient, anthelmintic and emmenagogue effects and its depressant effect on the central nervous system [5]. Furthermore, R. chalepensis is used as an anthelmintic, antipyretic, anti-inflammatory, analgesic, antispasmodic and abortifacient [6]. It is also used to treat snake and scorpion bites, as a hair tonic and as an insect repellent [7]. In addition to its medicinal properties, R. chalepensis is also used as a spice and flavoring agent in food and beverages [8].

The pharmacological properties of R. chalepensis are attributed to its high content of flavonoids, furocoumarins, phenols, amino acids, saponins and alkaloids that are present in its young stems and leaves [9]. Rutin was identified as the main flavonoid in R. chalepensis [10], and it is mainly produced in the leaves, followed by the stems and roots [11]. The plant is also considered a rich source of several quinoline alkaloids and acridones, as well as coumarins, such as marmasin, chalepin and chalapensin [12]. The flavonoids, glycosides and tannins found in R. chalepensis were identified as potent inhibitors of pro-inflammatory signaling molecules [5]. Furthermore, the extracts of R. chalepensis have distinguishable oxidative characteristics [13], which can be used to treat colon cancer [9]. It was reported that treatment with the ethanolic extract of R. chalepensis can reduce oxidative stress, as well as inflammation, in the hypercholesteremic organism and improve liver function [14]. Furthermore, the biological activity of R. chalepensis was proven to be a potential nematicide [15], insecticide [16] and herbicide in plant protection [17].

Biodiversity issues are acquiring global interest, and this interest is growing due to the growing need for novel natural resources that can be utilized for human welfare purposes [18]. In Jordan, a country located in the eastern part of the Mediterranean region, the biodiversity of its flora is not fully explored, and our knowledge about this flora is still fragmentary and requires more detailed studies to uncover all its components. The authors of [19] stressed the need for scientific research on the economic and medicinal values of the Jordanian flora and studies aiming to investigate the potential of wild species for medicinal applications. However, the major difficulties in assessing the utilization and developing a strategy for conservation are our insufficient knowledge about the species used and the level of genetic variation and geographical distributions [20]. Therefore, understanding the genetic variation within and between populations is essential for the establishment of effective and efficient conservation practices for rare plants [21].

To manage germplasm resources effectively, in terms of the utilization and conservation of biodiversity, knowledge of the genetic diversity existing in natural populations is required. Genetic diversity has been studied based on morphological characterization, which is valuable for identifying genotypes. However, the results are not completely reliable, because morphological traits can be influenced by environmental factors [22]. The ability to estimate genetic variation could be greatly enhanced by using appropriate molecular markers. These techniques are considered neutral and directly measure variation at the DNA level and are not affected by the environment. Molecular markers are commonly used to assess the genetic variation in natural plant populations and to estimate genetic diversity [23]. From this perspective, amplified fragment length polymorphism (AFLP) markers have been used effectively to study the genetic diversity of different medicinal, spice and aromatic plant species [24]. Despite the many studies that have investigated the pharmacological and chemical composition of R. chalepensis, there is no study related to the assessment of the genetic variation in natural populations of this species using molecular tools. Therefore, this study aimed to detect and assess the genetic variation within and between natural populations of R. chalepensis collected from different locations in the north of Jordan using morphological trait and AFLP analyses. For this purpose, 90 individuals were collected from the north of Jordan, representing nine natural populations. Twenty-two morphological traits were recorded, and the obtained data were analyzed using different tools. Molecular markers were generated and used to construct a phylogenetic tree and to dissect the genetic variation between and within the R. chalepensis populations using AMOVA. The information presented in this study is expected to provide basic knowledge of the genetic variation between natural populations of R. chalepensis and to develop suitable strategies for its conservation and future utilization.

2. Materials and Methods

2.1. Plant Material and Collection Site Descriptions

A collection mission was conducted to survey different locations in the northern parts of Jordan (Figure S1) in the spring of 2015 in order to collect plant tissues and characterize different morphological traits of the naturally growing R. chalepensis (Figure 1). The geographical data and site descriptions for the identified populations are shown in

Table 1. Site name, province, habitat description, rainfall (in mm), altitude (m), latitude (N) and longitude (E) of the Ruta chalepensis populations collected from northern Jordan.

Site	Province	Habitat Description	Rainfall	Altitude	Latitude	Longitude
Beit Idis	Irbid	Mountainous evergreen oak.	450	650	32°26′14″.	35°42′29″
Massom	Irbid	Rocky cliff in cultivated land, near Jordan Valley.	380	82	32°40′12″	35°39′22″
Aqraba	Irbid	Low mountainous, rocky, evergreen oak.	450	215	32°44′56″	35°47′55″
Kofer Assad	Irbid	Valley bottom.	470	172	32°38′07″	35°43′38″
Taibeh	Irbid	Cliff near Jordan Valley, perennial bulbs.	470	353	32°34′32″	35°44′06″
Ajloun-A	Ajloun	Mountainous, evergreen oak.	530	500	32°23′42″	35°40′15″
Ajloun-B	Ajloun	Mountainous, evergreen oak.	600	940	32°22′04″	35°44′19″
Jarash-A	Jarash	Near stone fruit orchards.	400	600	32°16′23″	35°52′25″
Jarash-B	Jarash	High mountainous, mixed pine and oak.	550	750	32°09′59″	35°49′06″

. From each location, 10 plants were selected randomly, and the final number of collected samples was 90, representing 10 R. chalepensis natural populations. From each identified plant, fresh leaf tissue was collected and kept in an ice box and then transferred to, and preserved in, the laboratory of Jordan University of Science and Technology (JUST) for molecular analysis.

2.2. Morphological Data

The morphological traits were recorded using the 90 collected samples, and their description and measurement methods are presented in

Table 2. Morphological traits of R. chalepensis populations identified in northern parts of Jordan and the measurement methods used in this study.

Morphological Trait	Measurement Method
Plant height (cm)	Measured from ground level to the tip of the plant.
Plant width (cm)	The mean of two measurements taken perpendicularly.
Internode length (mm)	Measured as the distance between the third and the fourth nodes of a young shoot.
Rachis length (mm)	Measured as the length of the main axis of the compound leaf.
Number of leaves	Determined by counting the leaves.
Compound leaf area (cm2)	Determined by using a leaf area meter.
Number of leaflets per leaf	Determined by counting the leaflets in the compound leaf.
Ultimate segment length (mm)	Measured from the base to tip of the terminal leaflet of the compound leaf.
Ultimate segment area (cm2)	Measured from the leaflet tip to the base of the terminal leaflet of the compound leaf.
Ultimate segment width (mm)	Measured as the widest part of the terminal leaflet of the compound leaf.
Young shoot diameter (mm)	Measured in the middle part.
Leaflet thickness (mm)	Measured as the thickened part of the leaflet.
Petal length (mm)	Measured from the tip to the base of the petal.
Petal width (mm)	Measured as the widest part of the petal.
Number of flowers per inflorescence	Counting the flowers in the inflorescence.
Number of fruits per inflorescence	Counting the fruit in the cluster (corymb).
Pentamerous fruit length (mm)	Measured from base to tip.
Pentamerous fruit diameter (mm)	Measured at the middle distance between the end and the tip.
Tetramerous fruit length (mm)	Measured from base to tip.
Tetramerous fruit diameter (mm)	Measured at the middle distance between the end and the tip.
Number of seeds in pentamerous fruit	Counting the seeds.
Number of seeds in tetramerous fruit	Counting the seeds.

. The collected data were recorded for each plant in each population using 10 replicates per individual, except for the plant height and width, which were based on one replicate. The leaf, fruit and flower measurements were based on samples taken from young shoots from the middle portion of the shrub. Analysis of variance (ANOVA) and multivariate analysis were performed for each trait to estimate the variance between populations using STATISTICA software V 10. [25]. A principal component analysis (PCA) was performed and the relationships between populations were analyzed using the ‘FactoMineR’ version 2.4 and ‘factoextra’ version 1.0.7 packages for R. For the heatmap construction, the pheatmap package v1.0.8 in R (https://cran.r-project.org/web/packages/pheatmap/index; accessed on 1 April 2022) was used.

Using the morphological data, the diversity was estimated using the morphological data mean (µ) and standard deviation (SD) for the quantitative traits. The µ and SD were used to classify the individuals into five groups, and the Shannon diversity index (H′) [26] was used to describe the phenotypic diversity of each trait and population. The following formula was used for the H′ calculation:

$${\mathrm{H}}^{\prime }=-{\displaystyle \sum _{\mathrm{i}=1}^{\mathrm{n}}\mathrm{Pi}\ln \mathrm{Pi}}$$

(1)

where n is the number of classes for a character and Pi is the proportion of the total number of entries in the ith class.

2.3. Molecular Analysis

Total genomic DNA was extracted from young leaves (0.1 g) collected from each R. chalepensis sample using the Wizard^® Genomic Purification Kit (Promega, Madison, WI, USA) following the manufacturer’s instructions. The quantity and quality of the extracted gDNA were assessed using a 1% agarose gel stained with Red Safe (Intron, Bio-tek, Seoul, Korea) and using a spectrophotometer (BIO-RAD, SmartSpec^TM Plus, Hercules, CA, USA). A stock solution of gDNA (30 ng/µL) was prepared for each collected sample with sterile distilled water and was stored at −20 °C until further use.

For the AFLP analysis, eight primer combinations were used in this study, as described previously [27]. The selective AFLP amplifications were performed using EcoRI and MseI primers, where the EcoRI primers were labeled with IRDye-700 for the ACC extension and IRDye-800 for the AGC extension. The selective step, the PCR reaction, was comprised of 10.5 µL that included 1.2 µL 10× amplification buffer containing MgCl₂, 0.06 µL Taq DNA polymerase (5 units/µL, Promega, Madison, WI, USA), 1.5 µL diluted pre-amplification DNA, 2 µL MseI primer containing dNTPs, 0.25 µL IRDye-700 labeled EcoR1 primer, 0.25 µL IRDye-800 labeled EcoR1 primer, and 5.24 µL deionized water. The PCR was performed using a touchdown program, as follows: 13 cycles of subsequent lowering of the annealing temperature from 65 °C by 0.7 °C per cycle, while maintaining the denaturation at 94 °C for 30 s and the extension at 72 °C for 60 s. This step was followed by 23 cycles of denaturation at 94 °C for 30 s, annealing at 56 °C for 30 s and extension at 72 °C for 60 s. The analysis of the AFLP products was carried out using an LI-COR Bioscience 4300 DNA Analyzer (Li-COR Bioscience, Lincoln-NE, Dearborn, MI, USA), where 1 μL of the product was loaded onto a 6% polyacrylamide gel after mixing with 0.5 μL stop solution and electrophoresis at 1500 V in the presence of 50–700 bp sizing standard (Li-COR Bioscience, Lincoln-NE, Dearborn, MI, USA).

For the genetic diversity assessment, each AFLP band was scored and then transformed to a binary matrix, where the presence of reproducible polymorphic DNA bands at a particular position on the gels was scored as 1, while 0 indicated its absence. The polymorphism indices, including the heterozygosity index (H), polymorphism information content (PIC) and discriminating power (D), were calculated using iMEC software (https://irscope.shinyapps.io/epic/; accessed on 1 May 2022) [28]. The collected marker data were used to generate a phylogenetic tree using the maximum likelihood method, including a bootstrapping analysis with 1000 replicates using RAxML (Randomized Axelerated Maximum Likelihood) software (http://www.trex.uqam.ca/index.php?action=raxml&project=trex) [29]. The obtained tree was further processed using the Interactive Tree of Life online tool (https://itol.embl.de/) [30]. The molecular variance between and within populations was examined through analysis of molecular variance analyses (AMOVA) using GenALEx program software version 6.5 [31].

3. Results and Discussion

3.1. Morphological Trait Analysis

The one-way ANOVA revealed significant differences (p < 0.05) between the natural populations of R. chalepensis for all the morphological traits (data are not shown). Among the nine populations, the morphological traits observed in R. chalepensis were highly variable (

Table 3. Mean values of 22 morphological traits and Shannon diversity index for each trait analyzed in nine R. chalepensis populations collected from northern Jordan.

Trait	Beit Idis	Aqraba	Massom	Kofer Assad	Taibeh	Ajloun-A	Ajloun-B	Jarash-A	Jarash-B	H′ *
Plant height	80.46 a**	76.33 ab	73.14 abc	78.13 ab	75.12 ab	71.42 bc	42.23 d	65.13 c	46.28 d	0.97
Plant width	48.72 a	41.52 bc	40.58 bc	49.07 a	40.28 bc	39.67 c	20.85 b	46.27 ab	23.21 d	1.00
Internode length	46.38 a	43.60 ab	44.76 ab	45.74 a	48.69 a	46.24 a	37.21 bc	40.60 c	36.50 c	1.01
Rachis length	79.74 a	78.26 a	78.59 a	79.50 a	82.99 a	78.19 a	40.40 c	54.65 b	43.35 c	0.83
Number of leaves	40.03 a	39.33 a	36.40 ab	38.73 a	35.37 abc	30.25 cd	27.47 d	30.81 bcd	29.42 d	0.90
Compound leaf area	28.29 a	26.38 ab	25.79 ab	26.91 a	24.29 bc	22.55 c	17.20 d	18.56 d	17.30 d	1.03
Number of leaflets/leaf	51.96 a	48.97 ab	44.34 bc	46.02 bc	53.33 a	51.88 a	38.44 e	43.69 cd	39.12 de	0.91
Ultimate segment length	22.44 a	21.10 ab	20.03 abc	21.47 ab	18.82 bc	17.94 c	19.63 abc	17.17 c	17.04 c	1.01
Area of ultimate segment	4.07 bcd	3.81 cde	4.22 bc	5.00 a	4.57 b	3.54 def	3.36 ef	3.23 ef	2.99 f	0.94
Ultimate segment width	4.77 a	4.62 a	4.70 a	4.54 a	4.19 ab	3.76 b	3.65 b	3.76 b	3.60 b	1.03
Young shoot diameter	2.36 c	2.38 c	2.49 bc	3.21 a	2.43 bc	1.72 d	1.54 d	2.76 b	1.72 d	0.84
Leaflet thickness	1.16 b	1.15 b	1.10 b	1.16 b	1.10 b	1.10 b	2.03 a	2.27 a	2.03 a	0.41
Petal length	6.75 a	6.55 a	5.74 b	6.41 ab	6.29 ab	6.69 a	6.94 a	6.52 a	6.92 a	0.97
Petal width	3.64 a	3.63 a	3.04 b	3.66 a	3.47 ab	3.53 ab	3.03 b	3.55 a	3.68 a	0.94
Number of flowers per inflorescence	27.64 a	25.95 abc	24.95 a–d	26.92 ab	27.71 a	24.00 a–d	21.77 cd	22.87 bcd	20.39 d	0.88
Number of fruits per inflorescence	26.68 a	24.73 ab	23.08 abc	25.98 a	27.06 a	23.20 abc	19.48 c	21.22 bc	19.67 c	0.93
Pentamerous fruit length	8.23 a	8.11 ab	7.61 c	7.78 bc	7.70 c	7.96 abc	7.82 bc	7.58 c	7.82 bc	0.88
Pentamerous fruit diameter	7.80 ab	7.71 b	7.30 bc	7.73 b	7.62 bc	8.11 a	7.56 bc	7.46 bc	7.71 b	0.87
Tetramerous fruit length	6.75 a	6.72 a	6.44 c	6.61 abc	6.64 ab	6.61 abc	6.50 bc	6.57 abc	6.61 abc	0.97
Tetramerous fruit diameter	6.32 ab	6.33 ab	6.25 b	6.28 b	6.28 b	6.44 a	6.29 ab	6.37 ab	6.37 ab	0.94
Number of seeds in pentamerous fruit	35.59 a	33.77 ab	33.09 abc	30.37 abc	34.61 ab	32.43 abc	27.01 d	28.6 cd	28.09 cd	0.97
Number of seeds in tetramerous fruit	23.72 a	23.03 a	19.73 bc	22.07 ab	22.59 a	22.06 ab	17.98 c	17.61 c	17.42 d	0.99

* Shannon diversity index; ** mean values with different letters within each row are significantly different according to the LSD_(0.05) test.

). For instance, the Beit Idis population had significantly higher mean values of 12.66%, 32.33% and 25.45% for the plant height, number of leaves, and compound leaf area, respectively, when compared to the Ajloun-A population. For the leaflet thickness, the mean values of the Ajloun-B, Jarash-A and Jarash-B populations were significantly increased by 84.55%, 106.36% and 84.55%, respectively, when compared, for instance, with the Massom population (Table 3). The Jarash-B population had a significantly reduced number of seeds in its tetramerous fruit by 26.56% when compared with the Beit Idis population (Table 3). Moreover, the pentamerous fruit diameter in the Ajloun-A population was significantly higher (11.1%) than that in the Massom population. For the petal length, the Massom population produced the lowest mean value, and it was significantly different from those of all the other populations, except for the mean values of the Taibeh and Kofer Assad populations (Table 3). The petal length in the Massom population was significantly reduced by 17.29% when compared with the Ajloun-B population. Such differences in the morphological traits between R. chalepensis populations might be attributed to both genetic and environmental variation. Similar conclusions were obtained by the authors of [32,33] for different plant genera within the Rutaceae family. In another study [34], the authors found that there was variability in the morphological traits of Zanthoxylum zanthoxyloides (Rutaceae) that was correlated with the geographic distance, and this is in general agreement with the findings of this study. In medicinal plants, such morphological variation can significantly aid in the selection of the best genotypes for different purposes, and they are usually correlated with the quality and quantity of the target metabolites [35]. Ruta chalepensis was previously characterized as possessing different active ingredients in different plant organs [36] and, therefore, the phenotypic variability observed in this study in the size of plants, leaves, and fruits might reflect the diversity in medicinal compounds between different populations.

The Shannon diversity (H′) index was used as a measure of phenotypic diversity for each trait analyzed in the nine identified R. chalepensis populations (Table 3). The H′ index values were considerably different between the traits, considering the relatively small area covered in this study. The average H′ index for all the analyzed morphological characters in R. chalepensis (0.92) was considerably high. All the characters showed a high diversity index (H′ ≥ 0.83), except for the leaf thickness trait, which had the lowest value (0.41). The character that had the most polymorphism was the area of the compound leaf, with the highest H′ value (1.03), followed by the ultimate segment length and internode length (Table 3). The authors of [37] found that the H′ index of the medicinal shrub Capparis spinosa in the northern parts of Jordan ranged between 0.80 and 0.85. Such morphological variation is usually attributed to both genetic and environmental variation. Thus, the high phenotypic diversity indices observed in this study might enable R. chalepensis to survive and adapt to different environmental conditions with varied temperature and rainfall regimes and the different habitats and soils in which this species was found at different elevations ranging from 82 to 940 m.

Principal component analysis (PCA) was conducted to determine which of the morphological traits strongly contributed to the total variation in the morphological traits of the R. chalepensis populations. The first 10 principal components, with Eigen values of more than three, explained 82% of the variation between the 90 collected samples. The first principal component (PC1) explained 29.9% of the total variation and was positively related to 15 characters, while PC2 explained 13.5% of the total variation. The rest of the components (PC3-PC10) varied to a smaller extent and captured 38.6% of the total variance. The scatter plot of the PCA separated the individuals of the two populations from the Jarash province (A and B) and the Ajloun-B population at the negative PC1 values of the plot, whereas the positive PC1 values of the plot included the majority of individuals from the remaining five populations from the Irbid governorate (Figure 2). The individuals from the Ajloun-A population were found on both sites from the plot, with no clear clustering pattern. Furthermore, the scatter plot was unable to cluster the individuals from each population in the defined groups in which they were scattered, except for individuals from the Massom population, which were scattered in the range of negative PC2 values of the plot. The PCA highlights the possibility of exiting genetic variation in the collected populations, which is important for the selection of unique individuals from each population as a genetic resource for national conservation programs [38]. Additionally, the phenotypic diversity identified between R. chalepensis populations may enable the selection of the most valuable accessions for future studies, as reported in the case of other plant species [39].

The phenotypic data were further analyzed through a heatmap clustering analysis using the pheatmap package (Figure S2). The 22 morphological traits were divided into three main clusters. The first cluster included the plant height and rachis length traits, while the second and the third clusters included 10 different traits for each cluster. The second cluster included the number of flowers per inflorescence, pentamerous fruit diameter, petal length, tetramerous fruit length, tetramerous fruit diameter, leaflet thickness, young shoot diameter, petal width, ultimate segment width and ultimate segment area. The third cluster was divided into three sub-clusters: the first sub-cluster included the plant width, internode length and leaflet number per leaf; the second sub-cluster included the ultimate segment length, number of seeds per tetramerous fruit, compound leaf area, pentamerous fruit length and number of fruits per inflorescence; and the third sub-cluster included the number of leaves and number of seeds per pentamerous fruit trait (Figure S2).

On the other hand, the individuals grouped in the heatmap were in general agreement with the PCA findings, with two main clusters identified and individuals from the Jarash-A, Jarash-B and Ajloun-B populations identified in one of them (Figure S2). In this cluster, two main sub-clusters, or groups, can be also identified, with the first sub-cluster including individuals from the Jarash-B and Ajloun-B populations, which might indicate that both populations shared similar phenotypic traits, while the second sub-cluster included mostly individuals from the Jarash-A populations. On the other hand, no obvious grouping based on the geographical origin of the collected individuals was observed for the other cluster. Interestingly, the plant height, rachis length and plant width were considered as discriminative traits between the two main clusters (Figure S2).

The two methods used in this study (PCA and heatmap analysis) have been also shown to be useful for numerous genetic resources [40,41]. Furthermore, Ref. [42] showed that accessions collected from the same region showed an important rate of similarity, whereas accessions collected from geographically distant regions showed a weak rate of similarity. Recently, PCA and genotype cluster analyses were used successfully to study the genetic diversity of wood apple (Limonia acidissima L., Rutaceae), where a clear grouping of the studied populations based on their morphological traits and geography was observed [33], which is also in general agreement with the findings of this study.

3.2. Molecular Data Analysis

The analysis of the AFLP molecular data of the eight primer combinations produced 903 bands, of which 59 were considered polymorphic bands that ranged in size from 71 bp to 540 bp (

Table 4. AFLP marker data analysis.

Primers Combinations	TN *	SR (bp)	NM	HI	PIC	EMR	AMH	MI	DP	RP
M-CTT/E-AAC	114	71–540	14	0.451	0.349	8.53	0.00037	0.00321	0.56938	4.7608
M-CTT/E-AGC	140	78–253	8	0.492	0.371	3.06	0.00076	0.00234	0.80863	2.4347
M-CTA/E-AAC	136	127–364	6	0.419	0.339	1.00	0.41898	0.41898	0.18246	0.00
M-CTA/E-AGC	93	169–374	3	0.352	0.290	0.45	0.00191	0.00087	0.94885	0.9130
M-CTG/E-AAC	117	119–527	12	0.325	0.272	8.75	0.00032	0.00281	0.36741	1.6739
M-CTG/E-AGC	89	120–410	8	0.499	0.374	3.41	0.00077	0.00264	0.76265	2.2608
M-CAG/E-AAC	124	155–272	3	0.496	0.383	1.00	0.49556	0.49556	0.50047	0.00
M-CAG/E-AGC	90	110–360	5	0.168	0.153	3.63	0.00045	0.00165	0.17647	0.7391

* TN: total number of fragments; SR: size range of polymorphic marker (bp); NM: number of polymorphic markers; and HI: heterozygosity index; PIC: polymorphic information content; EMR: effective multiplex ratio; AMH: arithmetic mean of H; MI: marker index; DP: discriminating power; RP: resolving power.

). Additionally, the polymorphic markers data did not reveal unique loci that were present or absent at the individual and population levels, with few exceptions. The primer combination M-CTT/E-AAC generated the highest number of polymorphic bands, followed by M-CTG/E-AAC, while the lowest numbers of polymorphic bands were produced by the primer combinations M-CTA/E-AGC and M-CAG/E-AAC. The primer combinations M-CAG/E-AAC and M-CAG/E-AGC produced the highest and lowest polymorphic information contents (PIC), respectively, and the PIC values ranged between 0.153 and 0.383 (Table 4).

The AFLP marker data were used to generate a phylogenetic tree based on the maximum likelihood using RAxML software (Figure 3). The phylogenetic tree had three main clusters. The first cluster included a unique sub-cluster, with all 10 individuals from the Jarash-B population grouped in addition to seven individuals each from the Jarash-A and Ajloun-A populations and two individuals from the Ajloun-B population, accounting for 65% of individuals from the Jarash-A and Ajloun populations. In the second cluster, a sub-cluster that included the remaining individuals from the Ajloun-A population grouped with four individuals from the Ajloun-B population and an individual from the Jarash-A population was observed (Figure 3). In addition, the second cluster included a sub-cluster that grouped three individuals from the Aqraba population with three individuals from the Taibeh population. For the third cluster, a higher genetic variation was between among grouped individuals, with a clear grouping of five individuals from the Aqraba population with four individuals from the Taibeh population (Figure 3).

The phylogenetic results are consistent with the PCA and heatmap clustering data, where individuals from Jarash-A and Ajloun populations were found to be separated from the remaining populations (Figure 2, Figure 3 and Figure S2). Furthermore, clear genetic variation between the individuals from the other populations was observed, with no clear grouping based on the collection sites, with minor exceptions, as in the case of the Aqraba and Taibeh populations, though a high degree of genetic variation within these populations was also observed. Similarly, Ref. [43] found that the phylogenetic analysis was in agreement with the PCA results. However, Ref. [44] indicated that the genetic differentiation in AFLPs is not associated with the morphological differences between populations. Several studies have compared the utility of morphological and molecular markers to investigate genetic variation, and many have indicated that the relatedness between the two approaches was low [45,46,47].

To understand the genetic variation between and within R. chalepensis populations, AMOVA was performed using the GenAlEx6.5 genetic analysis software (

Table 5. AMOVA for comparisons within and between R. chalepensis populations collected from northern Jordan.

Source	Df	Sum of Squares	Mean Square	Est. Var.	Percentage
Among populations	8	364.022	45.503	2.038	30%
Within populations	171	811.800	4.747	4.747	70%
Total	179	1175.822		6.785	100%
	Value	p-Value
Fst	0.300	0.001

). The genetic variation between populations contributed 30% of the total calculated variation, while the variation within populations explained 70% of the total genetic variation. The authors of [48] found that most of the genetic variation in Ruta species was within-population variation (88%), while only 9% was identified between populations. The level and distribution of the genetic variability between and within populations can be a result of several factors, including pollen and seed dispersal, mating systems, the colonization history and natural selection [49,50]. Ruta chalepensis is characterized as possessing hermaphrodite flowers, having both stamens and ovary in the same flower. It is reported to be highly self-compatible, and the mating system is mixed, i.e., it allows for both outbreeding and selfing [51,52]. Accordingly, a high gene flow is expected to increase the genetic diversity between and within natural populations [53]. On the other hand, the inbreeding implied by the self-compatibility of R. chalepensis can decrease the genetic variation and increase the similarity within the population [54].

The fixation index (Fst) was used to estimate the degree of differentiation between the nine R. chalepensis populations (Table S1). Interestingly, the Jarash-B population had the highest Fst values when compared to the other populations, indicating that this population was strongly differentiated (Fst ranged from 0.367 for the Jarash-B and Ajloun-B populations to 0.557 for the Jarash-B and Ajloun-A populations) (Table S1). On the other hand, the lowest Fst value (0.087) was observed between the Aqraba and Beit Idis populations followed by the Jarash-A and Ajloun-B populations (0.120), which indicates little differentiation between these populations. These results were in general agreement with the findings of [48], who found that the overall genetic differentiation between R. corsica and R. lamarmorae populations was significant, with Fst = 0.097 and Fst = 0.086, respectively.

4. Conclusions

In this study, 90 individuals of R. chalepensis representing nine natural populations were collected from the north of Jordan and then subjected to a comprehensive morphological and molecular analysis to assess the genetic diversity among them. To the best of our knowledge, this was the first study to report on R. chalepensis in Jordan. The morphological traits analysis identified a high level of phenotypic variation between the natural populations that were location-dependent. Both the PCA and heatmap clustering analyses confirmed these results, in which a clear sub-cluster from the Jarash-B and Ajloun-B populations was identified. Using molecular markers, the phylogenetic analysis identified three main clusters, in which 65% of individuals from the Jarash and Ajloun populations were found in a separate cluster. Within this cluster, individuals from the Jarash-B population formed a unique sub-cluster, highlighting a clear separation from the rest of populations that was associated with their geographical location. The AMOVA indicated the presence of considerable genetic variation between and within the R. chalepensis populations, and the Fst values varied from the strongly differentiated population of Jarash-B to the minimal differentiation between the Aqraba and Beit Idis populations. In conclusion, the morphological diversity and genetic variation among the populations observed in this study were relatively high and informative, and it is highly recommended that researchers survey wider areas across the country in future studies.