Molecular diversity among Turkish oaks ( QUERCUS ) using random amplified polymorphic DNA ( RAPD ) analysis

The genus Quercus (Fagaceae) includes the most important woody plants with decidious and evergreen species in Northern hemisiphere. They have a problematic taxonomy because of widespread hybridization between the infrageneric taxa. Turkey is one of the most important region of the world according to oak species number and variation. In this study, species belonging to evergreen oaks in Turkey were investigated to solve taxonomic problems and to design the limit of taxa by using random amplified polymorphic DNA (RAPD) data. Here, three species of evergreen oaks known as Quercus coccifera, Quercus ilex and Quercus aucheri were studied in all area located and made the comparison within and among species studied using ten RAPD markers. As a result; it can be stated that the presence of the three species in Ilex section is clear. Furthermore, existence of two infraspecific taxa or two seperate taxa in species level within Q. coccifera may be quite possibly considered.


INTRODUCTION
The genus Quercus is one of the most diversified groups of the trees of temperate zone in north Hemisiphere with more than 500 species (Govaerts and Frodin, 1998;Tovar-Sanchez and Oyama, 2004;Olfat and Pourtahmasi, 2010;Maryam Ardi et al., 2012).Govaerts and Frodin (1998) state that the genus Quercus is represented by 531 species in the world and 250 of these species in America, 125 of these in Asia and remaining species in Europe, North Africa and Macaronesia.The area including South East Asia and Pacific islands is the center of morphological variation of Fagaceae, altough this area does not contain the most species of Quercus (Kaul, 1985).
Oaks are the woody, widespread, long-lived, outcrossing and wind-pollination species.For this reason, oaks can spread too wide geographic regions and as a result of this, they show high variations comparison to other woody plant species (Kremer and Petit, 1993;Hokanson et al., 1993;Bacilieri et al., 1996;Neophytou et al., 2010).It is well known that extensive hybridization behaviors may occur among species (Bacilieri et al., 1996;Manos et al., 1999;Samuel, 1999;Jensen et al., 2009;Neophytou et al., 2010) in the same group or section in the genus Quercus, because of weak reproductive barriers between oak species.Consequently, hybrid species spring up.Therefore the genus Quercus is taxonomically one of the most problematic groups (Bacilieri et al., 1996).The most of species in Turkey and all distributed countries have taxonomic problems.Taxonomic problems can be solved by molecular studies in addition to morphological and cytological studies and so genetic diversity and the limits of taxa can be deter-mined more clearly (Borazan and Babaç, 2003;Yılmaz et al., 2008;Simeone et al., 2009;Alam et al., 2009;Papini et al., 2011;Yılmaz et al., 2011).
Turkey is one of the most important region for oaks according to the species number and geographical distribution.Oaks in Turkey have a natural distribution of about 6.5 million ha area represented by 18 species in three different section (Davis, 1982;Yaltırık, 1984;Kasapligil, 1992) as white oaks (Quercus L.), red oaks (Cerris Loudon.) and evergreen oaks (Ilex Loudon.).Here, the species analysed were Quercus coccifera, Quercus aucheri and Quercus ilex known as evergreen oaks.These are very problematic species in Turkey in the comparison to other members of the genus.The distribution area for Q. aucheri is only south weast region of Turkey and in the Greek island like Rhodos in the world.
However; Q. aucheri is confused with the another member of Ilex section, (Q.coccifera).As a result, it can be stated that it is not very well known species for biosystematic features and species limit.Moreover, it is controversial subject that Q. coccifera and Quercus calliprinos Webb.are seperate species or Q. coccifera has two subspecies known as Q. coccifera subsp.coccifera and Q. coccifera subsp.calliprinos (Toumi and Lumaret, 2001;Salvatore and Paola, 1976).Distribution area for Q. calliprinos is east mediterranean region and seperated from Q. coccifera with different living area.These taxonomical problems indicate that the real phenetics and fylogenetics relations within Ilex section have not still been fully explained.Hybridization and vegetative variations cause problems and make difficult to determine the borders of taxa.
Random amplıfıed polymorphıc DNA (RAPD) is a polymerase chain reaction (PCR) based technique used to show polymorphism among species.Especially this method is very helpful for systematics purposes and phylogenetic relation.For this aim, RAPD was used in this study as molecular technique (Kumar and Gurusubramanian, 2011).In order to solve this problem, variations within and between populations of taxa were pointed out by using some statistical analyses such as Statistica version 8.0 for principal component analysis (PCA) and cluster analysis (CA) using an unweighted pair group method (UPGMA) analysis and Popgen 32.According to the results of statistical analyses, it was attempted to draw the most possible borders of taxa based on the DNA bands obtained from RAPD analyses (Sesli and Yegenoglu, 2009;Açık et al., 2009;Kavalcıoglu et al., 2010) and a better phenetic classification by using molecular characters showing high correlations with each other.

MATERIALS AND METHODS
Study materials are composed of three species (Q.coccifera, Q. ilex and Q. aucheri) belonging to Ilex section of Turkey oaks.Totally Aykut et al. 6359 26 populations were represented to show variations within and among species (Table 1 and Figure 1).Leaves were used as material to show the differences in the molecular study.Q. coccifera was represented by 16 populations and other two species (Q.ilex and Q. aucheri) were represented by 5 populations.Especially, fresh and young leaves were preferred as material.Collected leaves were put into plastic bags filled silica gel and dried for the DNA isolation.

DNA extraction
Firstly, leaves in plastic bags filled silica gel were ground in liquid nitrogen using a mortar.DNA was extracted using a DNAeasy Plant Mini Kit (Qiagen).Extracted DNAs were kept at 4°C.Quality of each DNA sample were controlled by running on agarose gel before being used in PCR.

RAPD-PCR and gel electrophoresis
Molecular analysis was performed using RAPD method (Williams et al., 1990;Welsh and McClelland, 1991).Totally 30 primers, studied in oaks previously, were selected to find primers that exhibit polymorphism and give reproducible results.After the initial screening, 10 primers giving the best results among 30 primers were selected for further analysis (Table 2).Amplification reactions were carried out in a 25 μl mix.The reaction mixture was prepared using PCR Buffer, MgCl 2 , dNTP mixture containing dATP, dCTP, dGTP and dTTP, 10-base RAPD primer and taq DNA polymerase.After the primer selection, PCR conditions was determined.The program consisted of 40 cycles as fallows: Denaturation at 94°C for 1 min, annealing at 36°C for 1 min, and extention at 72°C for 2 min.A final extention at 72°C for 10 min was included.
The amplification products were electrophoresed in 1.4% agarose gels with TBE buffer at 100 V for 1 h and 30 min and stained with ethidium bromide.Gels with amplification fragments were visualized and photographed under ultraviolet light.RAPD bands were estimated by reference to a 100-bp ladder (Fermentas).

Data analysis
In order to score the RAPD products, amplified fragments were recorded as present (1) or absent (0) in all individuals for each fragment.Then the tables were constructed containing number and size of the DNA fragments for each populations.Polymorphic bands were determined for all populations.Molecular diversity among populations and species was evaluated by calculating the percentage of polymorphic fragments.The comparison of genetic distance and genetic similarity were calculated according to Nei (1972).RAPD data were evaluated by using two different statistical programs.Statistica version 8.0 were used for PCA and CA using an unweighted pair group method (UPGMA) analysis.Popgen 32 was used for genetic similarity and genetic distances.

RESULTS
In the RAPD analysis, 156 individuals representing 26 populations were used.A total 217 polymorphic bands were scored using the 10 RAPD primers.The size of the amplication products was between 150 to1600 base-pair.Table 2 shows the total number of polymorphic bands provided from each primers.The minimum and maximum size of amplification products provided from different   primers were also listed in Table 2.
In order to score the RAPD products, individuals of each population were run separately for each primer (Figure 2).Additionally, RAPD products of six individuals from every population were bulked and run together, to see all populations products in the same gel for each primer (Figure 3).CA and PCA were carried out for the analysis of variations within and among studied species.According to these results, Q. ilex and Q. aucheri were observed as close two separate groups.Populations of Q. coccifera showed more differences than populations of Q. ilex and Q. aucheri.But fundamentally, three studied species showed differences from each other.When the each species were evaluated separately, generally geo-graphically close populations showed more similarity than geographically distant ones (Figures 4 and 5).Populations belonging to Q. ilex were separated into two subgroups in CA pehenogram.The first of these was I1, I2 and I3 populations.
The second sub-group of Q. ilex was composed of I4 and I5 populations.Other species, Q. aucheri was separated into two sub-groups like Q. ilex but here C7 population of Q. coccifera showed the high similarity with the populations of Q. aucheri.Finally, when the populations of Q. coccifera were examined, it drew attention that Q. coccifera was separated into three sub-groups.When these three sub-groups are observed attentively, they were separated as geographically from each other  (Figure 4).Populations belonging to Q. coccifera evaluated in the West and South West region of Turkey are C1, C2, C3, C4, C5, C6, C7, C8, C9, C10, C11, C12 and C13 (Figure 1) and results from cluster analysis with UPGMA showed that the populations from these regions were similar in the comparison to the remaining (Figures 4 and 5).Eventually, populations having the highest differences of Q. coccifera were C14, C15 and C16.These popu-lations originated from East Mediterranean region (Figure 1).The molecular analysis with CA and PCA revealed a high degree of separation between the species.
When the tables of genetic distance and similarity was investigated, the lowest genetic distance was observed between C1-C2 and A1-A5 populations (Table 3).In other words, the highest genetic similarity was observed between C1-C2 and A1-A5 populations.The highest genetic distance was between C14-A4 and C14-I3 populations, respec-tively.Therefore, the lowest genetic similarity was between C14-A4 and C14-I3 populations.

DISCUSSION
This is the first report of RAPD data analysis for assessing relationships between these three taxa.But there are some studies that used RAPD data in different sections of oaks (Bruschi et al., 2003;Gonzalez-Rodriguez et al., 2004;Franjic et al., 2006;Ardi et al., 2012).Here, taxonomies of the studied species are not well known.Especially Q. aucheri is known as "forgotten oak tree'' (Yaltırık, 1984), because it is distributed only in South West Turkey and in a few East Aegean Islands (Davis, 1982).By this study; the lack of molecular properties of Ilex section is completed in detail.The results gave the satisfactory findings for phenetic groupings of taxas.The significant differences are found on the all studied species.Firstly, Q. ilex is separated from the other species.Due to similarities among the some Q. coccifera and Q. aucheri populations, they are well separated with each other.This is the first study that shows Q. coccifera is separated into two geographical groups in Turkey.The first group has the populations sampled from West and South West regions of Turkey.The populations sampled from C14, C15 and C16 belonging to East Mediterranean region included into the second group.This geographically separation within populations of Q. coccifera suggests that there are sub-groupings or different species in this taxon.The most common group is Q. coccifera found in many regions, while the less and restricted group is found only in the East Mediterranean region.The second group is geographically closer to Syria, Israel and Palestine.In addition, these two groups are represented as a single Q. calliprinos species and its two subspecies as Q. calliprinos subsp.coccifera and Q. calliprinos subsp.calliprinos in Flora of Palestine (Zohary, 1966).When the studied populations are compared with each other according to the genetic similarity, it can be said that genetically distant populations are also located geographically in different and far regions (Figures 4 and  5; Table 3).The most high genetic similarity are found between C1-C2 and A1-A5 populations which are also geographically close populations.On the contrary, genetically the most distant populations, C14-A4 and C14-I3 are the two different species which are geographically located very distant.
As a result of this study, it might be suggested that: (1), The results showed the presence of the second group within Q. coccifera but this needs to be supported in a study including Q. calliprinos samples from Syria, Israel and Palestina; (2), the groupings based on molecular studies support the presence of the three species in Ilex section; (3), the two groups showing geographical differentiations within Q. coccifera may strengthen the existence of two infraspecific taxa such as Q. coccifera subsp.coccifera and subsp.calliprinos or two different taxa at species level.

Figure 5 .
Figure 5.The resulting projection of principal component analysis.
Figure 1.Distribution of studied populations of Q. coccifera, Q. ilex and Q. aucheri in Turkey.

Table 2 .
The list of primers used in RAPD and analysis of PCR amplification products by selected primers.