Genetic Evaluation of a Wildtype Population of Cornus mas Accessions in Austria

: Cornus mas grows in the “Cornelian Cherry Valley” in Lower Austria. The characterization of the biodiversity of this underutilized crop was undertaken to increase its visibility and to develop breeding parameters. To determine the genetic relationships among Austrian Cornelian cherries, a collection of 447 mainly wild accessions originating from four regions in Lower Austria were genotyped. Analyses of seven microsatellite loci detected a total of 74 alleles, with 10.57 being the average number per locus. Shannon’s index was calculated to be 0.95 in locus CM008 and 1.80 in locus CM020. The average values of observed and expected heterozygosity were 0.54 ± 0.11 and 0.70 ± 0.12. The analyses revealed a high level of diversity. All accessions could be distinguished from each other, and wild and cultivated accessions could be separated. The pattern of grouping of the accessions based on the STRUCTURE (version 2.3.4) analyses suggested a greater number of subpopulations (K = 9). Karyotyping consistently yielded diploid chromosome numbers 2n = 2x = 18 for a millennial accession. The evaluations yielded a high degree of diversity, which provides excellent starting conditions for considerations of breeding strategies for the genetic enhancement of C. mas in Austria.


Introduction
Cornus is a genus of about 70 species of woody plants in the Cornaceae family, commonly known as dogwoods, most of which are used as ornamental plants, distinguished by their blossoms, berries, and distinctive bark [1].
Cornelian cherry (Cornus mas L.), together with C. officinalis, is one of the few species grown for the organoleptic characteristics of its fruits and its medicinal properties [2][3][4][5][6] The fruits are commonly olive-shaped, but they can also be found as being cylindrical, spherical, and pear-shaped [7], carrying a stony kernel.They are usually shiny red, varying in intensity from dark red and light red to yellow [8].
Cornelian cherry is a shrub or small tree distributed over Southern and Central Europe, the Black Sea basin, and the Caucasus [9,10] which occupies forest margins, woodland clearings, and light-exposed sites.Today, the Cornelian cherry is a minor wild fruit tree in Western Europe, but it represents a massively exploited natural resource in the Black Sea region.It is a very ancient crop plant in Ukraine and was mainly distributed by monasteries [10].
In the Late Iron Age, wild fruits like Cornelian cherry, together with Crataegus sp. and Prunus spinosa, contributed to the daily diet of settlements in Southern and Central Europe and can now be considered part of the so-called synanthropic vegetation [11].The presence and importance of C. mas in the Danube region as a follower of human settlements (and thus as a highly valued forest species) has been described, even in the archeological literature.
There are indications that the red color of the fruits played an important role in mortuary practice [12].In the Mesolithic cremation graves of Vlasac in the Danube Gorges, Cornelian cherry stones were recovered in high numbers, suggesting that Cornelian cherries could have had a role in complex funerary rituals performed by Mesolithic communities (9500 to 6500 BC).The dual function of Cornelian cherries in daily life at Vlasac, as a fresh fruit and a source of oil, was extended to the ritual sphere [12].In Tuscany, the use of Cornelian cherries was also reported as being widespread in the Early Bronze Age and in the Etruscan period (3500-1000 BC) [13].C. mas was described as a species that thrives well in floodplain habitats [14], which is in contradiction to the assumption that the plant is drought-tolerant and can grow and produce a valid yield under drought conditions.In fact, in the very dry years of 2017 and 2018, it was not possible to collect fruits from about half of the accessions.According to Werneck [14], C. mas has existed as a wild fruit but also as a cultivated fruit from the Neolithic period until 1850, i.e., for more than 4000 years.
In 1960, Werneck [14] stated that C. mas is conspicuously neglected in botany as well as in horticulture, which has led to its inclusion in the list of neglected and underutilized species (NUS) [15][16][17].
Given its natural area of distribution, its native status in many areas of Central Europe and the Eastern Alps has been doubted.Hegi [18] described C. mas as an element of the Southern European Pontic flora, which seems to have spread into Central Europe after the last Ice Age from Southeast to Central Europe through the Danube Gorges, a view that has been corroborated by archeological findings [12,19,20].The Cornelian cherry has native status in Lower Austria [14].
Presently, the largest area of diversity of C. mas is in Lower Austria at the northern edge of the Flysch zone from Klosterneuburg to Amstetten and at the eastern edge of the Vienna basin; here, accessions with black-purple, red, yellow, and white fruits can be found [8,14].In the Lower Austrian valleys of Pielach, Gölsen, and Triesting, Cornelian cherries, popularly named "Dirndl", are spread in hills, open pastures, and meadows [8,16].In the region, known as the "Cornelian cherry valley" (Dirndltal in German), active intervention to preserve and characterize the biodiversity of C. mas is being implemented [16,21] as changing climatic conditions have a considerable impact on wild-growing plants.
The selection and acceptance of the Cornelian cherry as an important healthy functional food [15,[21][22][23] have been approached by establishing an in vitro collection of valuable germplasms of centenary Cornelian cherries [16], in parallel to a thematic garden as ex situ collection (https://info.bml.gv.at, accessed on 28 August 2023).In addition, the phenotypic characterization of a wildtype population of C. mas was carried out on fruits to evaluate their nutritional values [21,24].
Population genetics analyses can provide data on a variety of important evolutionary parameters, including standing levels of genetic variation, the partitioning of this variability within/between populations, overall levels of inbreeding, selfing versus outcrossing rates, effective population sizes, and the dynamics of recent population bottlenecks [25].Beyond providing basic evolutionary insights, such analyses are also an important tool for developing effective management strategies for endangered and/or invasive species [26,27].Owing to their codominant and highly polymorphic nature, simple-sequence repeats (SSRs) have been widely utilized in the study of genetic variation [28].
Phenotypic characterization by microsatellites has been used for inferring genetic diversity and population structure of species [25,29].They have been used to determine DNA diversity, gene flow, and genetic relationships in Cornus species [30][31][32][33].Perceiving the potential of Cornelian cherries in Austria, available genomic markers (SSRs) [32] were used to assess the genetic diversity of 447 accessions from Lower Austria.
Since the different fruit sizes might be related to increasing levels of ploidy, karyotyping of a millennial C. mas accession [16] was performed.This will provide useful information to develop novel commercial cultivars from a locally adapted population.The characteristics considered to be desirable attributes are mainly based on different fruit shapes, smaller stones, and valuable bioactive chemical composition [21,32].Plants in these regions represent a landscape-defining element in a predominantly cattle farming and forestry-shaped environment.Centennial individuals are typically found in open pastures, on borders between properties or toward forest areas, on very dry, stony hilltops, and even in closed forest stands.In most cases, these stands are rather steep hillsides with an exposition towards the south, with rather difficult accessibility.In only a few cases, vigorous plants were found in sites with an exposition towards the north or with high humidity (Table 1).A few sites could be considered as plantations, with or without irrigation.They yielded important information during the sampling years exhibiting particular conditions of drought.
The control population consisted of grafted cultivars of C. mas, which are very popular in Austria, e.g., 'Bulgarico', 'Flava', 'Kanzanlak', 'Schönbrunner Gourmet' and 'Schumener', but also novel breeding lines and plants from the Botanical Garden Vienna and other sites in the Vienna region, e.g., Lobau, an alluvial forest habitat.One large-fruited sample from Iran was included in the genetic analyses as a potential outgroup.
For any breeding program, it is a prerequisite to select the most appropriate crossing partners.For this purpose, it is necessary to determine the genetic distance/vicinity of the wild-growing population.Since the main aim was to identify the genetic variability of the huge population of C. mas present in the region, the selection of the accessions attempted to encompass both positive as well as negative traits that could be observed.Plants were included in the sampling process based on distinguished properties, e.g., attractivity and farmers' preferences or particular traits described as unfavorable, like not fruit-bearing, the formation of black fruits, or exceptional anthocyanin levels.Individuals sampled per site were initially planned to be 10, but finally varied from 2 to 30, reflecting different locations, expositions, and the presence of interesting specimens.1).
Plants in these regions represent a landscape-defining element in a predominantly cattle farming and forestry-shaped environment.Centennial individuals are typically found in open pastures, on borders between properties or toward forest areas, on very dry, stony hilltops, and even in closed forest stands.In most cases, these stands are rather steep hillsides with an exposition towards the south, with rather difficult accessibility.In only a few cases, vigorous plants were found in sites with an exposition towards the north or with high humidity (Table 1).A few sites could be considered as plantations, with or without irrigation.They yielded important information during the sampling years exhibiting particular conditions of drought.1).
Plants in these regions represent a landscape-defining element in a predominantly cattle farming and forestry-shaped environment.Centennial individuals are typically found in open pastures, on borders between properties or toward forest areas, on very dry, stony hilltops, and even in closed forest stands.In most cases, these stands are rather steep hillsides with an exposition towards the south, with rather difficult accessibility.In only a few cases, vigorous plants were found in sites with an exposition towards the north or with high humidity (Table 1).A few sites could be considered as plantations, with or without irrigation.They yielded important information during the sampling years exhibiting particular conditions of drought.1).

DNA Extraction and PCR
Total DNA (gDNA) was extracted from the flower buds of 447 individual plants from 41 different sites (Table 1) using the DNeasy Plant Mini Kit (QIAGEN) according to the supplier's instructions.
Seven gSSRs developed for C. florida and C. mas [32] (Table 2) were used to screen the gDNA from 447 accessions.PCR was carried out in a total 10 µL reaction mixture using 5 µL of 2XQIAGEN Multiplex (Qiagen, Hilden, Germany), 1 µL containing 2 µM of each primer, and 150-200 ng DNA of total genomic DNA.Optimized conditions (Table 2) were used to amplify microsatellite fragments using an ABI PCR cycler and fragment analyses were performed on an ABI PRISM ® 3100 DNA Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).The sizing of fragments was carried out with a GeneScan 3.7 and Genotyper ® 2.0 software (Applied Biosystems Foster City, CA, USA), utilizing the internal GeneScan™ 500 LIZ™ dye Size Standard (Applied Biosystems Foster City, CA, USA).

Chromosome Preparation
The oldest accession in this study (1000 years), a national monument tree [16] served for chromosome preparation.In vitro cultures from the PBU collection, were used for root induction [16].Feulgen staining [34] modified by Greilhuber and Temsch [35] was performed on root meristems used for metaphase spread preparation [36].Brightfield microscopy imaging (Olympus BX61 Upright Brightfield Microscope, Center Valley, USA) was conducted using standard objectives and filter sets for ultraviolet blue (DAPI), green (FITC), and red (Cy3) excitation brightfield.The appropriate pseudo-colors were superimposed using the software CellˆF version 2.8 (Olympus, Center Valley, USA).Images were digitally captured at 1000X magnification.

Data Analyses
To reveal the genetic structure of the selected accessions of C. mas, a model-based cluster analysis using STRUCTURE version 2.3.4 [29] was used.A burn-in period and Markov chain Monte Carlo (MCMC) repetitions set to 5000 and 10,000, respectively, were implemented.The admixture model with correlated allele frequencies was selected.The range of 2 to 15 discrete ancestral populations was screened (K = 1-15) with three iterations for each.The inference of the best K was performed using the delta K method [37] to determine the most appropriate K value.The results were visualized using Microsoft Excel (2019).
For each of the defined loci, Simple-Sequence Repeats (SSR) allelic composition was determined in 447 accessions.Putative alleles were indicated in alphabetical order (A for the smallest fragment).The program POPGENE version 1.32 [38] was used to calculate allele frequencies at each locus, number of observed allele (Na), effective number of alleles (Ne), Shannon's information index (I), observed (Ho) and expected (He) heterozygosity, measured as described by Levene [39] and Nei [40], inbreeding coefficients (F IS ) and Wright's statistics or fixation index (F ST ) and gene flow (Nm), estimated from the formula Nm = 0.25 (1/F ST − 1) by Nei [41].Nei's genetic identity (I) and genetic distances between populations (D) [42] were also calculated using POPGENE (version 1.32).

Results
Plants were visited, samples collected, and data recorded several times a year over the past six years.DNA was successfully extracted from 447 accessions of high quality to allow further analyses.
The overall allele frequency across 7 SSR loci and all populations ranged from 0.001 in loci CF55, CM020, CM037, and CM043 to 0.65 in locus CF55.
The expected heterozygosity (H e ) is a fundamental measure of genetic diversity in a population, describing the expected proportion of heterozygous genotypes under Hardy-Weinberg equilibrium [40].The H o varied between 0.39 (locus CF55) and 0.69 (locus CM031) and the H e between 0.53 (locus CF55) and 0.84 (CM043), respectively.The average of observed and expected heterozygosity were 0.54 ± 0.11 and 0.70 ± 0.12 (Table 3).
The F IS measures the deficiency or excess of heterozygosity [43,44].Locus CM031 showed the lowest F IS = −0.02and the highest F IS = 0.38 was observed in locus CM026.The fixation index (F ST ) was between 0.05 for locus CM043 and 0.15 for locus CM037.Higher levels of gene flow were observed for locus CM043 to be 4.77 and the lowest levels 1.43 were observed for locus CM037.
Based on the genetic diversity measurements (number of effective alleles, Shannon index, expected heterozygosity) CM002 and CM031 show the highest polymorphism in these subpopulations.The expected heterozygosity is higher than the observed heterozygosity except for allele CM008, indicating a smaller effective population size.The F ST is above 0.1 for CM008, CM002, and CM037, indicating more fixation for these alleles rather than the others in these subpopulations.
The results of population structure analyses, assuming the admixture model in STRUC-TURE (version 2.3.4),showed an optimum of 9 ancestral populations (K = 9).The individuals in population 4 showed a clearer background, implying less migration and admixture (cross-pollination) compared to the other clusters.Based on genetic distance results from POPGENE (version 1.32) population 4 is identified always as a separate group (outgroup).
Based on the admixture model generated in the software STRUCTURE (version 2.3.4) with an elevated delta (K = 9) (Figure 2), K = 3, K = 4, and K = 11 were also plotted.The K = 9 was used to further describe the individuals, grouped by valley and site (Figure 3) revealing the exact distribution of genotypes.
The results of population structure analyses, assuming the admixture model in STRUCTURE (version 2.3.4),showed an optimum of 9 ancestral populations (K = 9).The individuals in population 4 showed a clearer background, implying less migration and admixture (cross-pollination) compared to the other clusters.Based on genetic distance results from POPGENE (version 1.32) population 4 is identified always as a separate group (outgroup).
Based on the admixture model generated in the software STRUCTURE (version 2.3.4) with an elevated delta (K = 9) (Figure 2), K = 3, K = 4, and K = 11 were also plotted.The K = 9 was used to further describe the individuals, grouped by valley and site (Figure 3) revealing the exact distribution of genotypes.In total three yellow-fruited accessions were found in the study region.They can be clearly distinguished from each other and the yellow-fruited cultivar Flava' (Figure 4).Fruits appear either yellow or with a light orange flush on the skin (Figure 4).This indicates the potential for the occurrence of three spontaneous mutations related to the color of the fruit skin and flesh.In total three yellow-fruited accessions were found in the study region.They can be clearly distinguished from each other and the yellow-fruited cultivar 'Flava' (Figure 4).Fruits appear either yellow or with a light orange flush on the skin (Figure 4).This indicates the potential for the occurrence of three spontaneous mutations related to the color of the fruit skin and flesh.In total three yellow-fruited accessions were found in the study region.They can be clearly distinguished from each other and the yellow-fruited cultivar Flava' (Figure 4).Fruits appear either yellow or with a light orange flush on the skin (Figure 4).This indicates the potential for the occurrence of three spontaneous mutations related to the color of the fruit skin and flesh.A representative metaphase karyotype of the millennial accession (Laimer et al., 2021) is shown in Figure 5.The basic chromosome number observed corresponded to the diploid numbers 2n = 2x = 18, as previously reported [36,[45][46].

Discussion
The conservation and use of plant genetic resources are two closely linked, important aspects for future generations [47].As a measure for the conservation of important gene pools, in vivo and in vitro gene banks are commonly used [16,[48][49][50].A thorough evaluation of the genetic resources of the native genotypes is essential for selecting the most useful genotypes for future breeding programs, including traits such as tolerance to vari- A representative metaphase karyotype of the millennial accession (Laimer et al., 2021) is shown in Figure 5.The basic chromosome number observed corresponded to the diploid numbers 2n = 2x = 18, as previously reported [36,45,46].A representative metaphase karyotype of the millennial accession (Laimer et al., 2021) is shown in Figure 5.The basic chromosome number observed corresponded to the diploid numbers 2n = 2x = 18, as previously reported [36,[45][46].

Discussion
The conservation and use of plant genetic resources are two closely linked, important aspects for future generations [47].As a measure for the conservation of important gene pools, in vivo and in vitro gene banks are commonly used [16,[48][49][50].A thorough evaluation of the genetic resources of the native genotypes is essential for selecting the most

Discussion
The conservation and use of plant genetic resources are two closely linked, important aspects for future generations [47].As a measure for the conservation of important gene pools, in vivo and in vitro gene banks are commonly used [16,[48][49][50].A thorough evaluation of the genetic resources of the native genotypes is essential for selecting the most useful genotypes for future breeding programs, including traits such as tolerance to variable temperature ranges, e.g., increasing temperatures, fluctuating availability of humidity or even drought, resistance to diseases from wild genotypes to be crossed into cultivars [51,52].Gepts [47] therefore advocated the use of molecular markers to make informed decisions about the composition of a gene bank, allowing the identification of plant accessions even in the vegetative stage, i.e., without fruit.
Overall, the genetic analyses have shown that the chosen methods allow the distinguishing of all 447 plants studied from each other, i.e., that the genetic variation is very high (Figure 2C).This indicates that cross-pollination might predominate, regardless of the presence of insects or wind.This also means that there is a good basis for selecting breeding material.The distribution of genotypes among the valleys studied (Figure 3) could also reflect a different altitudinal distribution of the occurrence, but this would need to be investigated in more detail.Unfortunately, the efforts to define the growth habit of the individual accessions were rendered difficult in several cases due to pruning interventions of the owners.
The wild forms of Cornelian cherries occurring in Lower Austria differ from the known selections (in grey) of C. mas to be clearly distinguished (Figure 3).Interestingly, the analyses also allowed the identification of three accessions as cultivars, which the owners planted on a special occasion (as part of hedge day plantings), where the plants were provided by local institutions and nurseries (plant no.177 identified as variety 'Jolico', plant 259 and plant 298).
The higher number of alleles per each SSR locus (6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18) in relation to the ones reported by Wadl et al., (2014) could be explained by the more than 10-fold higher number of samples (Table 3).The mean He in the analyses was 0.70, while Wadl [32] reported a He mean of 0.58, and Hassanpour [33] described a mean He of 0.42.The higher values obtained with the Austrian C. mas accessions indicate that the seven regions addressed by the SSR analyses were useful in the study of genetic structure in Austrian accessions.
Generally, the variability among the accessions of the different sites was high.Although the numbers of analyzed accessions varied between the valleys (Figure 3A).Certain patterns are present in a comparable amount in the three valleys, while others show a prevalence mainly in the Pielach Valley and are absent or scant in the other two valleys.
When analyzing the distribution pattern by site (Figure 3B) an additional feature can be observed: while most of the sites contain highly variable patterns, individual sites show a high number of similar genotypes, e.g., site 11, site 38, and site 41.The formation of subpopulations observed in these sites therefore could be indicating human selective intervention.This anthropogenic influence might correspond to the preference of the individual previous owners, e.g., for dark or light red fruit.The selection made by the owners over the past generations, in combination with the phenotypic characterization [21], allows the recommendation of the selection of breeding partners.
Another interesting feature is the appearance of yellow-fruited accessions.The cultivar 'Flava' is one of the oldest cultivars, first mentioned in the 17th century, with medium-sized, translucent yellow, very tasty fruit with a light pineapple aroma [53,54].
Wadl [32] refers to the yellow-fruited selections (Alosza, Florianka, and Jantarnyi) descending from a single wild, yellow-fruited mutant discovered in the Crimean Forest by the horticulturalist Kefeli [55].Klymenko [53] also mentions that even as early as 1843 "varieties with yellow fruit were very rare" in the wild and that in Ukraine it is thought that wild genotypes no longer exist in their natural habitat.Therefore, it was with much interest that the three yellow-fruited accessions-with distinctly lighter green leaves than red-fruited accessions found in the Pielach and Traisen valleys-were studied (Figure 4).The observed lower allergen content of these accessions might be of particular interest [56].
Seed propagation and long-term human selection have given rise to a great diversity of trees [32].Only recently have breeding programs been aimed at the development of largefruited and high-yielding trees in several countries, mainly in Ukraine [52], Austria [8], Bulgaria, Poland, Slovakia, Czech Republic, Serbia, but also in Turkey, Russia, Georgia, and Azerbaijan [32].

Conclusions
This is the first investigation using SSR markers to determine the genetic relationships among Cornelian cherry in Austria.The evaluations revealed a high degree of diversity, allowing the distinguishing of every single accession, which provides excellent starting conditions for considerations of breeding improvement.This might be highly useful for population genetic studies and for assessing the degree of clonal reproduction of C. mas in both wild and cultivated populations in Austria.In combination with the de novo assembly of a reference genome of C. mas [57], this will contribute to accelerating and promoting the genetic improvement of C. mas.

Figure 1 .
Figure 1.Geographic location of three regions sampled in Lower Austria (A) Pielach Valley (red), Traisen Valley (green), and Gölsen Valley (blue); (B) Typical exposition of Cornelian cherries in borders (between pastures), free-standing in open pastures and on the forest borders (Table1).

Figure 1 .
Figure 1.Geographic location of three regions sampled in Lower Austria (A) Pielach Valley (red), Traisen Valley (green), and Gölsen Valley (blue); (B) Typical exposition of Cornelian cherries in borders (between pastures), free-standing in open pastures and on the forest borders (Table1).

ABFigure 1 .
Figure 1.Geographic location of three regions sampled in Lower Austria (A) Pielach Valley (red), Traisen Valley (green), and Gölsen Valley (blue); (B) Typical exposition of Cornelian cherries in borders (between pastures), free-standing in open pastures and on the forest borders (Table1).

Figure 2 .
Figure 2. Genetic structure of 447 C. mas individuals evaluated by Structure (version 2.3.4)(a) K model with delta K values calculated from 2 to 15 (b) The number of subpopulations determined by the L(K) (c) Population structure analysis of the 447 individuals at K = 9.

Figure 2 .Figure 3 .
Figure 2. Genetic structure of 447 C. mas individuals evaluated by Structure (version 2.3.4)(a) K model with delta K values calculated from 2 to 15 (b) The number of subpopulations determined by the L(K) (c) Population structure analysis of the 447 individuals at K = 9.Diversity 2023, 15, x FOR PEER REVIEW 9 of 14

Figure 3 .
Figure 3. Genetic analyses (based on K = 9) of 447 C. mas individuals represented (a) by Q and by valley/region and (b) by site.

Figure 3 .
Figure 3. Genetic analyses (based on K = 9) of 447 C. mas individuals represented (a) by Q and by valley/region and (b) by site.

Table 1 .
A total of 447 samples were collected from 41 different sites at different altitudes in 4 regions:

Table 2 .
[32]acteristics of seven SSR loci and size description used in this study, developed for Cornus florida (CF) and C. mas (CM)[32].

Table 3 .
Microsatellite characterization for seven loci in 447 C. mas accessions.