Infra-specific morphological , anatomical and genetic variations in Lallemantia peltata ( L . ) Fisch . & C . A . Mey . ( Lamiaceae )

Lallemantia is a small genus with 5 species, which one of them is Lallemantia peltata. This aromatic plant grows in a limited area of the Iran and is known as medicinal herb in folk medicine. We have no information on the infraspecific variations on morphological, anatomical and molecular features of this species. Therefore, in the current study; we investigated these variations on six geographical populations of it. These plants were used for morphological, anatomical and molecular investigations. GenAlex, GenoDive and PAST softwares were used for statistical analyses. In morphological study, we used fifteen characteristics. PCA bi-plot revealed some characteristics had diagnostic value in separation of the populations. The studied populations were clustered into two groups. We examined fifteen qualitative and quantitative anatomical variables. PCA loading revealed that some anatomical variables had the highest positive correlation and were the most variable anatomical characters among the studied populations. Based on the anatomical traits, theses populations divided into two groups. Ten ISSR primers were used for examination of genetic variations among the populations. The STRUCTURE analyses of ISSR data revealed two groups in theses populations that they were divided into two groups in MDS plot. The results of morphological, anatomical and also ISSR studies were similar. Clustering and ordination of the studied populations separated some of these populations from the others and suggest the existence of two groups within this species.


Introduction
Lallemantia Fisch.& C.A. Mey. is a genus of family Lamiaceae.It is a small genus with five species that are herbaceous annual or biennial plants, characterized by simple leaves; a thyrsoid, spike-like or oblong, often interrupted inflorescence; ovate to rotund or sometimes linear, aristate-toothed bracteoles; and oblong, trigonous, smooth and mucilaginous nutlets (Jamzad, 2012).Some of these aromatic species are of medicinal value (Mahmood et al., 2013), while others have been used as the source of food and industrial plant (Rivera, Obón, 1992).
Lallemantia species are widely distributed in various parts of the world such as Afghanistan, China, India, Kazakhstan, Kyrgyzstan, Pakistan, Iran, Russia, Tajikistan, Turkmenistan, Uzbekistan, SW Asia and Europe, with the Caucasian region as the center of its origin.All five Lallemantia species occur in Iran (Rechinger, 1982;Jamzad, 2012).In general, limited biosystematics and molecular studies have been performed on this genus (see for example, Koohdar et al., 2015).
Infra-specific taxonomic entities can be identified after a detailed population-based investigation within plant species.In general, extensive morphological and genetic divergence among populations is the advent of speciation events.Intra-specific variation occurs within both wild as well as cultivated crop plants (Maxted, Hunter, 2011).Within the crop plants, variation may be in morphological, anatomical, karyological, ecological, physiological, biochemical and molecular characteristics.Explanatory surveys of genetic diversity can be useful preliminaries to germplasm collecting (Maxted, Hunter, 2011).
Molecular methods for characterizing genetic diversity developed enormously and techniques such as ISSRs, microsatellites or AFLPs, have been extensively applied particularly to geographically differentiated populations (Sheidai et al., 2012;Sheidai et al., 2014;Koohdar et al., 2015,).Nevertheless, at this level, collecting strategies in wild species is still a matter of determining the size of the sample per site (population/ subpopulation) and the number of sites that need to be sampled to ensure appropriate allelic retention.
Lallemantia peltata (L.) Fisch.& C. A. Mey.grows in limited area of the country and is known as medicinal plant that contains volatile and essential oil (Sefidkon et al., 2006).During our extensive Lallemantia plant collection from 2015 to 2018, we encountered morphological variation in different geographical populations of L. peltata.Therefore, we performed detailed morphometric, anatomical and molecular investigation of these populations with the aim to identify potential infra-specific taxonomic rank within L. peltata.To our knowledge this is the first time report on intra-specific variability in this species.

Plant material
In total 47 plant specimens of L. peltata were collected randomly from six geographical populations and were identified based on the descriptions provided in Flora of Iran (Jamzad, 2012).These plants were used for morphological, anatomical and molecular investigations.Details of populations are provided in Table 1.
Table1.Populations studied their locality and their localities adress.

Morphometry
For morphological analyses, we used fifteen vegetative and reproductive morphological characteristics.Data were obtained from 5 to 10 randomly collected plants in each population.The studied morphological traits were: plant height, petiole length, basal leaf length and width, basal leaf length/width ratio, length and width of bract, bract length/width ratio, bract arista length, calyx and corolla length, calyx / corolla length ratio, nutlet length and width and nutlet length/width ratio.
Morphological variables were standardized (mean = 1, variance = 0), and used for multivariate statistical analyses.We used different clustering as well as ordination methods to study morphological divergence of the populations.These methods are UPGMA (Unweighted paired group using mean average) and PCoA (Principal coordinate analysis).PCA biplot (Principal components analysis) was used to identify the most variable morphological characters among the studied populations (Podani, 2000).

Anatomical study
For anatomical investigations embedded materials were prepared as follow: adult plant samples were excised and immediately fixed in formalin-acetic acid-alcohol solution (FAA) (formalin 5%: acetic acid 5% and ethanol 90%) for 48 to 72 h, and stored at 4 °C until sectioning.Then dehydrated in a graded ethanol series and embedded in 70% ethanol.Transverse hand sections of the stem samples were washed with distilled water and placed in 5% sodium hypochlorite solution for 20 min.for clearing and rinsed with distilled water.The sections were stained with methyl blue and carmine and mounted on the slides using Canada balsam.Thin cut sections were observed using a microscope and images were taken using a digital camera.
In total fifteen anatomical traits of the stem were used.The studied anatomical features were: epidermal cell length, collenchyma cell length, parenchyma cell length, sclerenchyma cell length, xylem and phloem cell diameter, parenchyma and collenchyma cell layer number, pith cell length, stem width, xylem cell layer number.
Anatomical characters were first standardized (Mean = 0, Variance = 1) and used to establish the Euclidean distance among pairs of populations (Podani 2000).For grouping of the plant specimens, PCoA was used (Podani, 2000).PCA (Principal components analysis) was used to identify the most variable anatomical characters among the studied populations.PAST version 2.17 (Hammer et al., 2001) was used for multivariate analyses.

ISSR analysis
The fresh leaves were collected randomly in each of the studied populations and dried in silica gel powder.Genomic DNA was extracted using CTAB with activated charcoal protocol (Sheidai et al., 2013).The quality of extracted DNA was examined by running on 0.8% agarose gel.
The amplifications' reactions were performed in a Techne thermocycler (Germany) with the following program: 5 min.initial denaturation step 94°C, 30 S at 94°C; 1 min.at 50°C and 1 min.at 72°C.The reaction was completed by a final extension step of 7 min.at 72°C.The amplification products were visualized by running on 2% agarose gel, followed by the ethidium bromide staining.The fragment size was estimated by using a 100 bp molecular size ladder (Fermentas, Germany).

Morphometry
Details of morphological characteristics in six studied populations are provided in Table 2.The highest and shortest stems were recorded in Saqqez and Fasham populations, respectively.We found largest and smallest basal leaves size, petiole, corolla and calyx length in Shahrestanak and Gachsar populations, respectively.However, the biggest bract size was observed in Fasham population, but Gachsar population had the smallest bract.Fasham population had biggest nutlet, while smallest nutlet was seen in Asera population.
Table2.Details of morphological characters in L. peltata populations studied (all values are in mm).UPGMA clustering (Fig. 2) of the studied populations based on the morphological variables produced two major clusters.In the first major cluster, mostly plants of the populations Saqqez, Asera, and Gachsar are placed inter-mixed together.The studied plants of the populations Shahrestanak, Fasham, and Gajereh comprised the second major cluster.This inter-mixed placement of plants indicates some degree of morphological overlaps among the studied populations.PCoA plot (Fig. 3) of the studied populations based on the morphological features revealed morphological divergence of populations Saqqez, Asera, and Gachsar, while plants of the populations Shahrestanak, Fasham, and Gajereh were placed inter-mixed.PCA bi-plot revealed characteristics like, plant height, length of bract arista and the length of corolla can differentiate two groups.Fig. 2. UPGMA tree of the studied populations based on the morphological data (Populations 1-6 are according to Table 1).

Anatomy
The stem in cross-section was quadrangular.A one -layered epidermis was seen, which had elongated cell.The length of epidermal cell varied between the studied populations.Collenchyma cells were observed as a large area in all corners.The size of collenchyma cell and its layer number differed among the populations.The stem cortex was composed of polygonal parenchyma cell that was found between collenchyma tissue and stele.The parenchyma cell layer no. and also size of parenchyma cell varied among the populations.The vascular bundle was collateral.In all of studied populations, the vascular bundles were bigger on the corner than other parts of stem.Beneath the phloem tissue, we have a 1 to 2-layered cambium.The xylem tissue existing beneath the cambium was composed of regular trachea and tracheid cells.The pith consists of polygonal cells with intercellular spaces (Table 3).
Table3.Details of anatomical features in L. peltata populations studied (all values are in µm).The PCoA plot (Fig. 4) of the studied populations based on the anatomical characters revealed closer affinity between populations Saqqez, Asera, and Gachsar.Similarly, Shahrestanak, Fasham, and Gajereh populations were placed close to each other.This is in agreement with our morphological result.to Table 1).

Population
PCA analysis revealed that the first three PCA components comprised about 60% of the total variations.PCA loading revealed that anatomical traits like length of epidermal cell, collenchyma and parenchyma cell layers number had the highest positive correlation (>0.7).These are the most variable anatomical characters among the studied populations, which mainly differentiate the populations Asera and Gachsar, from the others.

ISSR analysis
In total, 45 ISSR bands (Loci) were obtained.Some of the bands were private and occurred in Saqqez, Fasham, and Gajereh populations (Table 4).These bands can differentiate the studied populations.Genetic diversity parameters in L. peltata (Table 5) revealed that the highest degree of genetic polymorphism and gene diversity occurred in Fasham population (77% and 0.21 respectively), followed by Saqqez population (62% and 0.16 respectively).In general moderate to high degree of genetic variability occurred in L. peltata populations.This is probably related to out-breed nature of this species.Table 5. Genetic diversity parameters determined in the studied populations.
AMOVA produce significant genetic difference among the studied populations (PhiPT = 0.34, P = 0.01).It also revealed that 35% of total genetic difference occurred among populations, while 65% was due to within population genetic variability.This indicates that the studied populations have a good level of within population genetic variability.Pair-wise AMOVA revealed significant genetic difference among all studied populations.This is further supported by MDS plot of the populations based on ISSR data (Fig. 5).The plants of the studied populations form a separate group in the MDS plot, which indicates genetic divergence of these populations.Saqqez, Asera, and Gachsar populations showed a close genetic affinity based on ISSR data, while Saqqez, Fasham, and Gajereh populations were placed close to each other.This is in agreement with morphological and anatomical results.Table 1).
The Ivanno test performed on STRUCTURE analysis produced k = 2 as the optimal genetic group.The STRUCTURE plot based on k = 2 (Fig. 6) revealed close genetic affinity between Saqqez, Asera populations, followed by Gachsar population (they have similarly coloured segments, mostly green colour).Similarly, Shahrestanak and Gajereh populations showed close genetic similarity (mostly red coloured segments); while Fasham population showed the genetic content of both these groups (both green and red coloured).Fig. 5. MDS plot of L. peltata populations studied based on ISSR data (Populations 1-6 are according to Table 1).

Discussion
In the current study, we used morphological, anatomical and also ISSR molecular data for investigation of infraspecific variations among L. peltata populations in Iran.Classification and taxonomic consideration of within species diversity may be diverse by different authors, for example, Sheidai et al., (2013) studied 38 geographical populations in four Cirsium Mill.species and treated divergent geographical populations as different ecotypes.
This species has restricted distribution in the country and is only found in some mountain slopes occasionally with more than 900 km distance.Instances of infra-specific variation are known in different plant species with extensive geographical distribution (see for example, Sheidai et al., 2012Sheidai et al., , 2013)).
Results of morphological data showed that high morphological divergence exists among the populations and they are divided into two groups.
Although several taxa of Lamiaceae family are studied anatomically (Özdemir, Altan, 2005;Aktaş et al., 2009), we have no anatomical study of L. peltata in literature.Therefore we used stem anatomical traits for investigations of infraspecific variations among the populations of this species.Various studies proved that stem anatomical characteristics are very important in plant taxonomy.Hagstroem (1916) mainly stressed the importance of stem anatomical characters for recognizing new taxa.He partly went too far in his appreciation of stem anatomy traits, particularly with respect to the recognition of hybrids.Symoens et al., (1979) used stem anatomy for the distinction between two very similar African taxa.
Our studied populations had quadrangle stem with a multi-layered collenchyma at corners.The vascular tissue is collateral, and the thickness of xylem is bigger than phloem.Metcalfe and Chalk (1972) have stated that the anatomical characteristic traits of Labiatae family are a quadrangular stem with a well-developed collenchyma, supporting tissue at the corners of stem.The mentioned traits are previously reported in some other investigated taxa of the family by different authors (e.g.Bosabalitis, Kokkini 1997;Gönüz, Özörgücü, 1999;Dinç et al., 2008).
PCA analyses showed that some anatomical features were more variable than others.In similar to morphological analyses, the studied populations were divided into two groups.
Inter simple sequence repeat (ISSR) is a PCR-based molecular marker technique, which has been proven to be highly useful for investigating genetic diversity and population genetic structure of different plant taxa (Zietkiewicz et al., 1994;Godwin et al., 1997;Archak et al., 2003).Hence, in the present investigation, we used this technique to reveal the genetic diversity and population genetic structure of L. peltata.
ISSR data showed genetic variations among these populations and divided them into two different groups similar to morphological and anatomical data.The STRUCTURE analyses proved separation of populations.
These populations not only showed morphological divergence, but also differed in anatomical and genetic content.All these data support the idea that we have different subspecies or varieties in this plant species.
When morphological population divergence is accompanied with genetic divergence we may consider these populations as distinct infra-specific taxa.In the Alpine Alps (NW Tuscany), several plant species occur as populations that are morphologically, genetically and ecologically distinct.Such natural groups are treated as sub-specific rank.This type of taxonomic treatment has been reported for a number of taxa in the different genera Buphthalmum L. (Compositae), Rhinanthus L. (Scrophulariaceae), Astrantia L. (Umbelliferae), and Asperula (Rubiaceae) (Bechi, Garbari, 1994).

Conclusion
In the current study, morphological, anatomical and also molecular ISSR technique was used for investigating infraspecific variations among L. peltata populations in Iran.This species has very restricted distribution in Iran, however against restricted distribution of L. peltata high infraspecific variations were recorded among its populations.The results of morphological, anatomical and molecular analyses were similar.In all studies, the studied populations divided into two similar groups.In one group Saqqez, Asera and Gachsar populations were clustered closely and other populations made another groups.This pattern of arrangement confirmed existence of possible infraspecific taxonomic ranks within this species.

Table 4 .
Details of ISSR bands obtained in L. peltata populations studied.