Evaluation of molecular diversity analysis and relation with environmental factors in accessions of Cumin (Cuminum cyminum L.) in Iran, revealed by inter-simple sequence repeat (ISSR) markers and start codon targeted (SCoT) markers

Cuminum cyminum (as an aromatic herb) is the sources of cumin, which has been commonly used in traditional medicine for the treatment of various diseases especially in tropical Asia. Thus, the current experiment was performed to accesses molecular genetic diversity among 22 collected accessions of cumin from different area of Iran, based on inter simple sequence repeat (ISSR) markers and start codon targeted (SCoT) markers. Analysis of molecular variance revealed that a signicant genetic difference between the studied accessions reach to 43%. Mantel test between genetic diversity and geographical distance indicated that an increase in geographical distance of collected accessions did not inuence on genetic differentiation. Results about all agro-morphological traits showed a signicant difference between the ecotypes except the number of branches. Canonical correspondence analysis ballot of genetic features and environmental factors included 5 geographic and climatic factors for seed origin’s habitat, which have been show that the mentioned parameters had no strong effect on the plant genotypes. However, difference in elevation and latitude of accession origins had effect on genetic clustering in two major groups. The consensus tree of both molecular and morphological data relieved that the accessions Chah-Dashi, Damghan, Ferdows and Nehbandan formed discreet clusters based on genetic and agro-morphological features. In conclusion, the results conrmed that both ISSR and SCoT markers were reliable and useful tools for analyzing the genetic diversity of cumin in Iran.

Introduction environment, economical bene ciary and more informative than morphological traits (SAMANTARAY et al., 2010;RAHIMI et al., 2018). Inter simple sequence repeat (ISSR) marker is preferred by many researchers due to many advantages such as simplicity, quickly, less costly and high reproducibility (PRADEEP REDDY et al., 2002;XU et al., 2018). ISSR markers rapidly reveal high polymorphic ngerprints and have been used frequently to determine the genetic diversity among date palm cultivars (AYESH, 2017). Also, SCoT (start codon targeted), as a novel marker, can be evaluating as the strong polymorphic markers which indicate differences between individuals extremely well (COLLARD and MACKILL, 2009).
As yet, genetic relationships between cumin ecotypes were investigated in terms of agro-morphological traits (BAHRAMINEJAD et al., 2011) and molecular polymorphism by markers like RAPD, SCoT, CCMP and SSR (BAHRAMINEJAD et al. 2012;BAHRAMINEJAD and MOHAMMADINEJAD, 2013;ROSTAMI-AHMADVANDI et al., 2013;PARSSHAR et al., 2014). Considering the worth of medicinal plants in industry and commerce, the present study was survey to undertaken the agro-morphological traits and genetic differentiation (by ISSR and SCoT markers) of 22 cumin accessions collected from different regions of Iran. This study would be helpful in selecting parents in the cumin breeding programs and conservation of this medicinal plant. Furthermore, relationships between ecological factors and genetic differentiation will be discussed due to poor information about environmental in uence on different accessions of cumin.

Plant materials
At current study 22 accessions of cumin from 9 Iran province were investigated ( Figure 1, Supplementary 1). The accession number, origin and ecological condition of these accessions are presented in Table 1.
The seeds were cultivated using randomized complete block design at the Research Field of Agriculture (Semnan, Iran; Table 2). And, the information about environmental factors of the study area was given from State Meteorological Organization of Iran (http://reports.irimo.ir/jasperserver/login.html).
The measured agro-morphological traits were including: plant height, number of branches, number of umbel, mini-umbel and seed, 1000-seed weight, seed weight per plant and grain yield. These traits were measured through morphometric method. The photos of accession were given as Supplementary data 1.  Finally 72 o C for 120s, and 72 o C for 10 min. The 2 -σσCt method was used to analyze the data. Data analyzing The analysis of variance (ANOVA) of agro-morphological data was carried out for 3 plant per accessions by SPSS 19.0 software. The obtained SCOT and ISSR bands were coded as binary characters (presence=1, absence=0). The genetic diversity parameters like allele diversity (Weising et al., 1995), Nei's gene diversity, Shannon information index, number of effective alleles, and percentage of polymorphism (FREELAND et al., 2011) were determined for each population. Nei's genetic distance was used for clustering (WEISING et al., 1995;FREELAND et al., 2011). Neighbour Joining clustering were used for grouping (FREELAND et al., 2011).
The Mantel test was performed to check correlation between geographical and genetic distances of the studied accessions (PODANI, 2000). PAST ver. 2.17 (HAMER et al., 2001) and GeneALEx 6.4 (PEAKALL and SMOUSE, 2006) programs were used for these analyses.
The Pearson coe cient of correlation was determined between geographical features (longitude and latitude) and genetic diversity parameters. Genetic differentiation of the studied species and populations was studied by Analysis of molecular variance (AMOVA) test (with 1000 permutations) as performed in GenAlex 6.4 (PEAKALL and SMOUSE, 2006).
Canonical correspondence analysis (CCA) was done using PAST software v. 2.17 (HAMER et al., 2001), to determine the relative importance of geographical factors in the spatial organization of genetic diversity between accessions.
DARwin program (Version 5; PERRIER and JACQUEMOUD-COLLET, 2006) were used to compare the accessions that are genetically and agro-morphologically differentiated from the others, a consensus tree was conducted from agro-morphological and genetic obtained trees.

Results
Phenotypic data ANOVA results showed signi cant differences among the accessions for all the studied traits except number of branch (Table 4), that indicated high genetic diversity among these accessions. Accessions 17 (Boshruyah) indicated highest average seed weight per plant and grain yield, while accession 20 (Faizabad) had the lowest average. The average number of umbel and mini-umbel in accession 16 (Birjand 3) was highest and the lowest was observed in accessions 20 (Faizabad) and 8 (Talkhuncheh), respectively. On average, higher 1000-seed weight and number of seed per plant were seen in accessions In the WARD tree of morphological characters, two major clusters were formed ( Figure 2). PCA biplot supported the grouping made by WARD tree and also revealed agro-morphological traits including seed weight per plant, grain yield and number of seed per plant separated these two groups ( Figure 3). PCA analysis revealed that about 76% of total variance was described by the rst two components. The loading of morphological characters on the rst PCA axis showed that traits like seed weight per plant and grain yield had the highest positive correlation (>0.54) with the rst PCA component. Number of miniumbel in umbel and the number of seed per plant had the highest positive correlation (>0.42) with the second PCA axis.

Molecular markers polymorphism in Cumin
Genetic diversity parameters determined for ISSR, SCoT and combined ISSR-SCoT data markers are presented in Table 5. The highest value for number of effective alleles, Shannon Information Index and gene diversity, occurred in accession 10 (Kuhbanan) equal with 1.19, 0.16 and 0.11, respectively. While the lowest level of the effective number of effective alleles and Shannon Information Index occurred in accessions 16 (Birjand 1) equal with 1.02 and 0.02, respectively, and in ecotype 16 gene diversity (He) is zero.
When compared of genetic diversity parameters between SCoT and ISSR primers results indicated differences among them. So that in some accessions SCoT markers had more e ciency than ISSR marker and in others vice versa. For example, polymorphism percentage in SCoT marker was higher value for accessions 1 (Khaf), 4 (Ayesk), 5 (Ilkhchi), 6 (Shabistar), 7 (Khur), 9 (Rig), 11 (Kuhbanan), 12 (Dlazyan) and 13 (Damghan) than ISSR markers, while in other accession ISSR markers showed higher value of polymorphism percentage. The similar status was exhibited for other genetic diversity parameters (Table 5).
AMOVA results revealed that a signi cant genetic difference (PhiPT=0.56, P=0.01) among the studied accessions. It also AMOVA results indicated that 57% of total genetic variation was due to diversity within accession and 43% was due to genetic differentiation among populations. Pairwise AMOVA produced a signi cant difference among these populations (Figure 4).
PCOA plot of combined SCoT-ISSR data ( Figure 5) showed both intra-and inter-accessions. The grouping of the accessions obtained by these two procedures produced two major clusters. Accessions Khaf, Khoshab, Chah-Dashi, Ayesk, Ilkhchi, Shabistar, Khur, Mobarakeh, Rig, Kuhbanan, Dlazyan, Damghan, Sorkheh and Nehbandan com prised the rst major cluster. In this cluster, accessions Dlazyan, Damghan, Sorkheh and Nehbandan showed higher genetic similarity and were joined to each other. The same holds true for accessions Khur, Mobarakeh, Rig and Kuhbanan. Accession Shabistar had the most distinction in this cluster. In cluster 2, accessions Yazd, Birjand 1, Birjand 3, Boshruyah, Ferdows, Faizabad, Tehran, and Bojnord showed higher genetic similarity and were joined to each other. The photos of banding pattern in ISSR and SCoT analysis are given in Supplementary data 2.

Environmental factors and genetic diversity
Based on the Mantel test that was performed between genetic distance and geographical distance of the origin locality of studied accessions the results showed that produced no signi cant correlation (R2=0.005, Figure 6), which have been indicate an increase in geographical distance of collected seeds of accessions did not in uence in genetic differentiation. Therefore, no isolation by distance (IBD) exists between them.
The main aims of the present study was to compare the effect of different environmental factors on the genetic features of cumin ecotypes, 5 geographic and climatic factors of seed origin's habitat were examined. CCA biplot (Figure 7) of genetic features and environmental factors showed that the mentioned parameters had no strong effect on the plant genotypes (Axis 1= 32.1% of variance; Axis 2= 28.7 %). However, it seems that difference in elevation of sea and latitude of ecotype origins had impacts on genetic clustering in two major groups. The clustering of accessions Ilkhchi and Shabistar (East Azarbayjan province) was related to in uence of altitude. These two accessions had high altitude than others accessions. Also, longitude in uence on grouping accessions Mobarakeh, Rig, Kuhbanan and Damghan.

Combined molecular and morphological results
The consensus tree of both molecular and morphological data is presented in Figure 8. The specimens of accessions Chah-Dashi, Damghan, Ferdows and Nehbandan formed separate clusters. These accessions belong to south Khorasan province except ecotype Damghan that belong to Semnan province. The results differed from the other studied populations in both genetic and morphological features. This showed that agro-morphological differences in these four accessions had genetic background.

Discussion
Generally, in the plant breeding programs and genotype conservation the population structure and genetic diversity information are valuable keys. Agro-morphological traits are prerequisite to classify and describe genotype diversity (LOU et al., 2015). In the current study, all agro-morphological traits showed a signi cant difference between the ecotypes (except the number of branches). The results of the present study a line with some previous studies that it's were done about genetic diversity of cumin in Iran (AMINPOOR and MOUSAVI, 1997;BAHRAMINEJAD et al., 2011;ROSTAMI-AHMADVANDI et al., 2013). There is an obvious differentiation between studied accessions on the basis of agro-morphological traits. These accessions divided into 2 main clusters. The grouping of these accessions is not thoroughly math to origin geographic region of seeds. It seems that planting all ecotypes in Semnan locality and difference in cultivated condition in uence on agro-morphological traits. In contrast, examining of 49 cumin ecotypes by BAHRAMINEJAD et al. (2011) indicated categorizing of them in to 3 groups in line with geographical distance provinces. Basically, morphological variation does not always indicate true genetic variation. Because of interaction between genotype and environment, and large unbeknown genetic control of polygenic morphological and agronomic traits (XU et al., 2018). However, morphological traits are advantageous for primary assessment, because of their fast and simple application and can be used as a typical way for estimating genetic diversity among morphologically differentiable populations (RAHIMI and KORDROSTAMI, 2017). In this study, the most important traits in discriminate accessions including seed weight per plant, grain yield and number of mini-umbel that indicated the best potential for increasing seed yield in cumin breeding in most part of Iran.
To improve the cultivars, some information about genetic polymorphism and relationship by using biotechnological tools like molecular markers, are needed. At the present experiment, the mean number of polymorphism (ISSR=25.74 and SCoT=29.67) was rather lower than ROSTAMI-AHMADVANI et al. (2013) ones that reported polymorph bands for ISSR markers 67% and for RAPD markers 54%, these accessions belong to Kerman, Esfahan and Khorasan provinces of Iran among cumin accessions. BAHRAMINEJAD and MOHAMMADINEJAD (2013) BAHRAMINEJAD et al. (2012) represented high variation between and within Iranian cumin accessions using phenotypic traits and RAPD markers. Also, the results of genetic variation between Iranian cumin accession using ISSR and RAPD markers indicated different molecular and morphological groups (ROSTAMI-AHMADVANDI et al., 2013).
The results of AMOVA test showed out of total genetic variation, 57% was due to within population genetic variability that similar results have been report ed in different plant species and should be related to outcrossing nature of this plant species. The existence of high within population genetic variability is useful tool for adapting to local environmental changes (SHEIDAI et al., 2013).The same may hold true for the cumin accessions. Among-population differentiation in molecular variation (43%) is expected to occur as a consequence of isolation, drift, founder effects, and local selection (JOLIVET and BERNASCONI, 2007).
In Iran, cumin have various distribution regions like arid and semi-arid areas in eastern, south-eastern, central and western geographical regions that lead to adaptation to ecological impacts and variation in genetic traits (HASHEMIAN et al., 2013 ).In this study, the results of Mantel test has shown no signi cant correlation between geographical and genetic distances. Hence, in spite of genetic differentiation among the studied cumin accessions, they are not totally isolated and yet some amount of gene ow has occurred between them. The relationships within infra-species specimens sometimes are consistent with their geographical distance. While, in species with a wide distribution region, sometimes do not consistent with one (QIU et al., 2004;DING et al., 2013;LIU et al., 2013).

Based on our results investigation of relationship between genetic diversity and environmental factors
showed impact of ecological conditions on adaptation and differentiation of cumin accessions. Also, genetic diversity of these accession were assessed in relation to ecological parameters of origin locality.
The relationship between accession genetic diversity was not well accorded with ecological conditions of seed origin including latitude, altitude, longitude, the mean year temperature and rainfall. However, altitude and latitude effect on genetic variation of these accessions on the basis of CCA data.
Climate has an important role in local adaptation for plant species that could be lead to alter the genetic diversity of plan populations (MOSCA et al., 2012;AVOLIO et al., 2013). Increasing of genetic differentiation could be consequence of the environmental changing (WANG et al., 2009). These results are in line with the relationship between genetic diversity and ecological parameters in Stipa grandis from Inner Mongolia (ZHAO et al., 2006). Also this point was considered by HUANG et al. (2016) in Caragana microphylla (Fabaceae) with regard to relationships between climatic factors and genetic diversity that showed positive in uence of some climatic factors on genetic diversity. Similar to our results were reported in Artemisia halodendron (Asteraceae) that indicated by HUANG et al. (2014).

Conclusions
The present molecular analyses with 2 genetic markers (ISSR and SCoT), indicated high genetic differentiation between cumin accessions. Also a signi cant variation was observed between the accessions in terms of genetic data and agro-morphological traits. Evaluation of effective ecological parameters on genetic differentiation shows in uence of altitude and latitude on genetic grouping. In spite of genetic differentiation among the studied cumin accessions, they are not totally isolated and yet some amount of gene ow has occurred between them. Therefore, no IBD exists between them.

Figure 1
Distribution map of cumin accessions studied in Iran. 22 accession belong to 9 province of Iran from Northwest, Northeast, center and southeast. The accessions were cultivated in plots with 4 m long. The row spacing and distance between plants were 50 and 5 cm, respectively.