Genetic Singularity Coefficients of Common Vetch ...

ISSN 1022-7954, Russian Journal of Genetics, 2008, Vol. 44, No. 11, pp. 1309–1316. © Pleiades Publishing, Inc., 2008. Original Russian Text © E.K. Potokina, T.G. Aleksandrova, 2008, published in Genetika, 2008, Vol. 44, No. 11, pp. 1508–1516.

PLANT GENETICS

Genetic Singularity Coefficients of Common Vetch Vicia sativa L. Accessions Determined with Molecular Markers E. K. Potokina and T. G. Aleksandrova Vavilov Institute of Plant Industry, St. Petersburg, 190000 Russia; e-mail: [email protected] Received December 20, 2007

Abstract—Organization and practical application of ex situ collections require estimation of genetic differences between numerous accessions of local cultivars and field weed forms collected from the same ecological and geographical region and similar in their morphophysiological characteristics. A mathematical algorithm for estimating the degree of genetic singularity of a specimen in the system of local gene pool determined with the help of molecular markers is described. The utility of this algorithm is demonstrated by the example of classification of 677 common vetch accessions from the collection of the Vavilov Institute of Plant Industry from 11 ecological–geographic regions of Russia analyzed using AFLP. The proposed classification of accessions is the result of processing the AFLP data by weighting the marker traits based on their frequency in particular regions. This allowed each accession to be characterized according to the ratio of rare and frequent alleles as a genetic singularity coefficient. The proposed method is appropriate for any types of molecular markers. A practical result of its application is the classification of accessions using a five-point score scale, which can be added to descriptors of certificate databases and used for optimization of the work with collections. DOI: 10.1134/S1022795408110094

INTRODUCTION Common vetch (Vicia sativa L. sensu stricto) is among the most valuable fodder legume crops in Russia. The traditional local cultivars and field weed populations of common vetch, collected during over 80 years by expeditions of the Vavilov Institute of Plant Industry (VIR), form the basic collection for the breeding programs in Russia. The efficiency of these programs significantly depends on the genetic knowledge about the preserved collection accessions. Maintenance of ex situ collections is an expensive issue; therefore, a goal of gene banks is to find the maximum genetic diversity in the minimum number of accessions for creation of active working collections. Recently, molecular markers are successfully used to optimize collections of genetic resources [1]. The data on genetic diversity obtained using molecular markers ensure development of an optimal strategy for supplementing collections and detection of duplicates [2]. The information about genetic similarity and differences between accessions is also necessary for creating core collections [3]. The genetic diversity of the common vetch grown on the territory of Russian Federation and neighboring countries and preserved in the VIR Gene Bank was analyzed by restriction and amplified fragment length polymorphism (AFLP) of genomic DNA [2]. This method utilizes selective amplification of individual genomic DNA fragments obtained after digesting DNA with two restriction endonucleases. Nucleotide inser-

tions, inversions, and deletions in the genotypes accumulate in individual populations, thereby changing the distances between the sites for restriction endonucleases in individual genotypes. This makes it possible to assess the genetic polymorphism in theaccessions. The genetic diversity of the Russian common vetch gene pool from the VIR World Collection has been earlier compared with that in the worldwide gene pool of V. sativa from the collection of the Institute of Plant Genetics and Crop Plant Research (IPK, Gatersleben, Germany) [5]. It was found that various geographical regions of the vetch range can be better characterized with a certain frequency ratio of AFLP fragments rather than according to the presence of any allele specific of a certain region. The differences in the frequencies of AFLP fragments over the range were considerable (up to 90%). In each geographical region, 10–15% of vetch accessions displayed similarity to the accessions from other regions of the area, frequently rather distant. Thus, despite a similar geographical origin, the accessions could show different genetic characteristics. The results of this work can be applied to screening of collections and detection of accessions carrying the alleles rarely met in a particular region. Here, we describe a simple mathematical algorithm for estimating the degree of genetic singularity of accessions from the ex situ collection using molecular markers. This algorithm was used to classify 677 accessions of common vetch local cultivars and field weed populations from the VIR collection according to the degree of their genetic singularity in the system of local gene pool.

1309

1310

POTOKINA, ALEKSANDROVA

MATERIALS AND METHODS Collection of common vetch accessions. AFLP was used to analyze the genetic polymorphism of 677 common vetch accessions from the VIR collection. The accessions comprised four groups of genetic material, designated according to the descriptor of the VIR Certificate Database: (1) breeding cultivars (AC), (2) local cultivars (LR), (3) breeding material obtained from regional breeding institutions (BR), and (4) field weed and wild-growing populations (WE). The gene pool of common vetch represented 64 administrative regions of Russia and neighboring countries. They were grouped into 11 regions based on the similarity in ecological and geographical conditions taking into account the regions of the State Register of Agricultural Achievements of the Russian Federation approved for use [6]. The common vetch accessions were distributed according to their origin and attribution to ecological–geographical groups and the dominant ecotypes according to Leokone [7–10]. The grouping of the studied accessions into 11 ecological–geographical regions according to their geographical origin and ecological–geographical and ecotype attribution is briefed below. (1) Baltic region (Latvia, Lithuania, and Estonia): 90 accessions, Western European ecological–geographical group, Baltic ecotype. (2) Belarus: 22 accessions, Middle Russian ecological–geographical group, Belarusian ecotype. (3) Central Ukraine (Kharkov, Zhitomir, Khmel’nitsk, Poltava, Sumy, Chernigov, Kiev, and Dnepropetrovsk oblasts): 54 accessions, Middle Russian ecological–geographical group, Eastern Ukrainian ecotype. (4) Western Ukraine (Ternopol, Lvov, and IvanoFrankovsk Oblasts and Transcarpathia) and Moldavia: 54 accessions, Western European ecological–geographical group, Hungarian ecotype. (5) Caucasus (Armenia, Azerbaijan, Georgia, Nagornyi Karabakh, Northern Osetia, Stavropol and Krasnodar krais, Dagestan, and Crimea): 90 accessions, Caucasian ecological–geographical group, Svanetian, Dzhavakhet, and Armenian ecotypes. (6) Russia, northwestern region (Leningrad, Pskov, Novgorod, Vologda, Yaroslavl, and Tver oblasts): 55 accessions, Middle Russian ecological–geographical group, northwestern ecotype. (7) Russia, central region (Vladimir, Ryazan, Smolensk, Kostroma, Ivanovo, Yaroslavl, Kaluga, and Moscow oblasts): 63 accessions, Middle Russian ecological–geographical group, Kursk and northwestern ecotypes. (8) Russia, central chernozem region (Kursk, Belgorod, Orel, Lipetsk, Tambov, and Voronezh oblasts): 105 accessions, Middle Russian ecological–geographical group, Kursk ecotype.

(9) Russia, Volga and Volga-Vyatka region (Saratov, Penza, Kuibyshev, Ul’yanovsk, Nizhny Novgorod, and Kirov oblasts; Chuvashia; and Republics of Mordovia and Tatarstan): 80 accessions, Southern European (steppe) ecological–geographical group, southeastern ecotype. (10) Russia, Ural region (Kurgan, Perm, Sverdlovsk, and Chelyabinsk oblasts): 42 accessions, Middle Russian ecological–geographical group, northeastern ecotype. (11) Russia, Siberia and Far East (Tomsk, Omsk, Irkutsk, and Kemerovo oblasts; Krasnoyarsk krai; Altai; and Amur region): 31 accessions, Northern and Middle Russian ecological–geographical groups, northern ecotype. AFLP analysis. The seeds of 677 V. sativa accessions were grown on experimental plots of the Institute of Plant Genetics and Crop Plant Research (IPK, Gatersleben, Germany). DNA was isolated from seedling leaves (8–15 plants per sample) using a NucleoSpin Plant kit (Macherey-Nagel, Duren, Germany) and the corresponding protocol. For AFLP, 4 µl of genomic DNA (150 ng) were added to 6 µl of the mixture containing 0.5 pmol of EcoRI adapter, 5 pmol of MseI adapter, 1 µl of T4 10× ligase buffer, 10 mM NaCl, and 80 ng/µl BSA. Then 2 µl of the restriction ligase mixture were added; this mixture contained 1 U of MseI, 4 U of EcoRI, 4 U of T4 ligase, 0.2 µl of incubation buffer for T4 ligase (10×), 50 mM NaCl, and 100 ng/µl BSA. The resulting mixture (12 µl) was incubated for 14 h at the room temperature, diluted with water to 50 µl, and used for amplification according to Vos et al. [4]. The sequences of used adapters and primers and the detailed AFLP protocol were described earlier [5]. AFLP analysis of all accessions was performed in two replicates; the AFLP fragments were further analyzed only in the case of complete reproducibility. Totally, 64 AFLP fragments produced using six most polymorphic primer combinations for V. sativa accessions (E37– M52, E38–M53, E39–M49, E40–M56, E40–M58, and E41-M57) were analyzed. The main region of fragment separation was 30–600 base pairs (bp); only the most distinct fragments falling into the range of 100–400 bp were used in the analysis. Statistical data analysis. The frequencies of AFLP fragments in regional samples were determined using the FREQ option of NTSYS-pc. The genetic distances between accessions were calculated according to Nei [11]. An UPGMA (Unweighted Pair Group Method with Arithmetic Average) dendrogram was constructed using the SAHN option. The samples were analyzed and the parameters of observation variance were estimated using the Statistica 5.0 (StatSoft, United States) software package. The χ2 test was computed using the MINITAB (Minitab Inc., 2003) software package.

RUSSIAN JOURNAL OF GENETICS

Vol. 44

No. 11

2008

GENETIC SINGULARITY COEFFICIENTS OF COMMON VETCH Vicia sativa L.

RESULTS

Baltic

Analysis of the Intraspecific Diversity Using Molecular Markers

Rus-Center Rus-Chernozemje

Tools of the so-called numerical taxonomy are used when processing the data on genetic diversity obtained using molecular markers [12]. The fact of the presence (1) or absence (0) of fragments with the same molecular weight on electrophoretic pattern is regarded as a trait. The analysis commences with constructing a binary matrix (1/0), which is used to calculate the pairwise similarity coefficients for the studied accessions. The calculation of pairwise similarity coefficient almost always is the calculation of the coincidence rate of the presence/absence of particular fragments in two accessions as a percent of the total number of observations [13–15]. The resulting matrix of similarity coefficients is visualized with the help of a dendrogram, graphically reflecting the degree of genetic similarity between the accessions (or phylogenetic relatedness between taxa). The described procedure is very efficient when comparing species or the taxa of a higher rank, as the species always contain a certain rate of species-specific alleles. The larger this rate in a particular species, the smaller is its coefficient of similarity to the remaining sample and the more isolated is this species from a phylogenetic standpoint. The specificity of analysis of intraspecific diversity is in that the populations differ in the frequency ratio of particular alleles rather than their presence or absence. Therefore, calculation of the genetic similarity coefficient for populations, based on comparing the frequency ratio of alleles, is more efficient for the classification of intraspecific diversity [11]. We have analyzed the frequencies of 64 AFLP alleles in 11 ecological– geographical regions of Russia and neighboring countries using 677 vetch accessions harvested there. The UPGMA dendrogram of the calculated genetic distances [11] demonstrated that the degree of similarity of the regions in the frequencies of AFLP fragments in the corresponding accessions was proportional to their geographical closeness (Fig. 1). The accessions from Volga region, central, central chernozem, and northwestern regions are the most similar from a genetic standpoint. The populations from Caucasian countries and Siberia are the most isolated. This analysis gives a general estimate of the degree of similarity of common vetch gene pools in different regions. However, it is more important from a practical standpoint (for organization and use of the collection) to characterize the genetic differences of accessions from the same ecological–geographical region. Classification of Collection Accessions according to the Degree of Their Genetic Singularity in the System of Local Gene Pool The proposed classification of the vetch collection accessions with the help of AFLP analysis is based on RUSSIAN JOURNAL OF GENETICS

1311

Vol. 44

No. 11

Rus-Povolzhie Rus-North-West Rus-Ural Ukr-West Belorus Ukr-Centr Rus-Siberia Caucasus 0

0.01

0.02 0.03 0.04 Net’s genetic distance

Fig. 1. The degree of similarity of 11 ecological–geographical regions in the frequency of AFLP fragments in common vetch accessions. Regions: Baltic countries (Baltic); Russia, central region (Rus-Center); Russia, central chernozem region (Rus-Chernozemje); Russia, Volga and Volga– Vyatka region (Rus-Povolzhie); Russia, northwestern region (Rus-North-West); Russia, Ural region (Rus-Ural); Western Ukraine (Ukr-West); Belarus (Belarus); Central Ukraine (Ukr-Centr); Russia, Siberia and Far East (RusSiberia); and Caucuses (Caucasus).

the data processing by weighting the traits according to their frequencies [16, 17]. Smirnov’s method [16] was designed to estimate the degree of similarity between taxa and is widely used in systematics. According to Smirnov, the traits characterizing taxa are not equiweighted, as they can occur rarely or frequently. The coincidence of rare traits in taxa should be regarded as a more informative fact than the similarity in frequent characteristics. Smirnov proposed to calculate the relationship coefficient of taxa as an averaged value of all the traits used in comparing the taxa: the more the number of coinciding rare traits, the higher is the relationship coefficient. Following the same logic, it is possible to calculate the relationship coefficient of a taxon to itself (Txx), which actually shows the degree of singularity of this taxon. The ordinary taxa, possessing the set of frequently met traits and forming the core assumed to exist in the group, must display the least Txx value. The maximal Txx value is characteristic of singular taxa, which combine the rarest traits. The Smirnov principle is applicable to the studied group of accessions (populations) originating from the same ecological–geographical region that are to be compared according to genetic characteristics detected, for example, by AFLP markers. The more genotypes with the AFLP alleles rare for a particular region in an accession, the higher is its degree of singularity (specificity) in the system of regional populations. From the standpoint of gene pool, this can mean that one acces2008

1312


50 (a) 40 30 20 10 0

0

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

30 (b) 25 20 15 10 5 0

–2.5 –2.0 –1.5 –1.0 –0.5 0

0.5 1.0 1.5 2.0 2.5

25 (c) 20 15 10 5 0 1

5 2

4 3

Fig. 2. The principle used for classifying the accessions according to their GSC by the example of 80 accessions from Volga and Volga–Vyatka region. (a) A logarithmically normal distribution of 80 GSC values is transformed into (b) a normal distribution by taking the logarithm. (c) The five-point score scale corresponds to five ranges of the obtained distribution, namely, minimum–5th percentile, 5th–25th percentiles, 25th–75th percentiles, 75th–95th percentiles, and 95th percentile–maximum. An increase along the scale corresponds to the increase in the rate of rare alleles in accessions.

sion comprises the genotypes most typical of this region, whereas the other accession succeeded in “catching” the rare genotypes not typical of this region yet retained in the population as a variation reserve. Thus, each collection accession can be characterized according to its genetic singularity in the system of local gene pool. This forms the background for an optimal approach to use of the gene pool diversity of a culture in breeding. We have analyzed 677 common vetch accessions from 11 ecological–geographical regions of Russia and neighboring countries to identify the accessions that are most typical from a genetic standpoint of particular regions or, on the contrary, display rare untypical genotypes. The comprehensive information about the classification of the studied accessions is available on request. Here we brief the principle of the analysis by the example of one sample of overall 11 samples. Principle of Calculating the Singularity Coefficient for a Collection Accession The sample contains 80 vetch accessions collected in the Volga and Volga–Vyatka region. In this sample, 59 AFLP fragments are polymorphic. The weights of the presence (1) and absence (0) in each accession were calculated for each AFLP fragment based on its frequency in the analyzed sample (Table 1). Then the initial matrix of AFLP presence/absence in accessions was substituted with the matrix of the corresponding weighted values. At the final stage, the weights of all AFLP fragments were summed for each accession, and the sum was divided by the total number of analyzed AFLP fragments. Thus, a value was ascribed to each accession, which we named the genetic singularity coefficient (GSC). The GSC is the averaged weighed value reflecting the presence of the AFLP fragments that are polymorphic in the analyzed sample (Table 1, boldfaced column). The next task was to estimate which particular GSC values could be regarded as significantly low or, on the contrary, significantly high for the analyzed sample; in other words, how the accessions could be classified according to their GSCs. The GSC values for 80 accessions fit a logarithmically normal distribution (Fig. 2a). Consequently, their logarithms are distributed approximately normally (Fig. 2b). According to the properties of normal distribution, the GSC values that group around the mean arithmetic value within the range of one mean square deviation are characteristic of the accessions that combine the rare and frequent alleles at a ratio characteristic of 68% of observations. The higher GSC values indicate the increased rate of rare alleles. Correspondingly, the minimal GSC values denote the accessions where the rare alleles are mostly lacking. The most important parameters of the sample in the context of our goals are the following characteristics of observation variance: the minimum, maximum, and 25th, 75th, 5th, and 95th percentiles (the last two are the extreme points of the


Vol. 44

No. 11

2008

AFLP, (bp) Number of 285 score 0

AFLP, (bp)

292

300

248

228

...

Number of 285 score 0

292

300

248

228

...

Σ

GSC = Σ/n

log GSC

GSC class 1

Vol. 44 No. 11 2008

k00843 k31049 k34622 k00811

0 1 0 0

1 1 1 1

1 1 1 1

0 0 0 0

0 0 0 0

... ... ... ...

k00843 k31049 k34622 k00811

0.23 4.33 0.23 0.23

0.05 0.05 0.05 0.05

0.31 0.31 0.31 0.31

0.14 0.14 0.14 0.14

0.05 0.05 0.05 0.05

... ... ... ...

56.32 40.45 72.82 16.20

0.95 0.69 1.23 0.27

–0.07 –0.55 0.30 –1.87

3 3 4 1

k00833

0

1

1

1

0

...

k00833

0.23

0.05

0.31

7.00

0.05

...

44.18

0.75

–0.42

3

1 1 1 1 1 ... 1 1 1 1 1 1 1 1 1 1 76 4 4/ 76 76/ 4

1 0 1 1 1 ... 1 1 1 1 1 1 1 1 1 0 61 19 19/ 61 61/19

0 0 1 1 0 ... 0 0 1 1 0 0 0 0 0 0 10 70 70/10 70/10

0 0 0 0 0 ... 0 0 0 0 0 0 0 0 0 0 4 76 76/4 4/76

... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...

k00839 k00841 k00842 k00846 k00847 ... k34421 k34547 k30949 k32209 k35994 k32521 k27793 k33300 k33301 k34538 Number 1 Number 0 Weight 1 Weight 0

0.23 0.23 0.23 0.23 0.23 ... 0.23 0.23 0.23 4.33 0.23 0.23 0.23 4.33 0.23 0.23 15 65 4.33 0.23

0.05 0.05 0.05 0.05 0.05 ... 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 76 4 0.05 19.0

0.31 3.21 0.31 0.31 0.31 ... 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 0.31 3.21 61 19 0.31 3.21

0.14 0.14 7.00 7.00 0.14 ... 0.14 0.14 7.00 7.00 0.14 0.14 0.14 0.14 0.14 0.14 10 70 7.00 0.14

0.05 0.05 0.05 0.05 0.05 ... 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 4 76 19.0 0.05

... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...

60.92 49.47 89.73 76.09 247.14 ... 26.38 47.37 32.31 69.01 38.25 50.69 68.76 30.23 51.78 95.52

1.03 0.84 1.52 1.29 4.19 ... 0.45 0.80 0.55 1.17 0.65 0.86 1.17 0.51 0.88 1.62

0.05 –0.25 0.61 0.37 2.07 ... –1.16 –0.32 –0.87 0.23 –0.63 –0.22 0.22 –0.97 –0.19 0.70

3 3 4 4 5 ... 1 3 2 3 3 3 3 2 3 4

k00839 0 k00841 0 k00842 0 k00846 0 k00847 0 ... ... k34421 0 k34547 0 k30949 0 k32209 1 k35994 0 k32521 0 k27793 0 k33300 1 k33301 0 k34538 0 Number 1 15 Number 0 65 Weight 1 65/15 Weight 0 15/65

1313

Note: Left: The initial matrix of the absence/presence of AFLP fragments in accessions (k00000, catalogue numbers). Center: The weighted values of the absence/presence of AFLP fragments calculated based on their frequencies in the sample. Right: Σ-Sum of the weight of all AFLP fragments for each accession. GSC = Σ/n is the genetic singularity coefficient of accession as a quotient of the obtained sum by the number of analyzed AFLP fragments (n = 59); log GSC is the GSC logarithm having 2 as a base; and GSC class, the class of GSC according to five-point score scale (see also Fig. 2).



Table 1. Scheme for calculating the genetic singularity coefficient (GSC) for an accession in the system of local gene pool by example of 80 accessions of the Volga and Volga–Vyatka region

1314


Table 2. Ratio/absolute value (%) of five GSC classes of vetch accessions among field weed, local, and bred forms GSC class

Accession status; WE

LR*

1 6(0.07) 29(0.07) 2 15(0.16) 97(0.23) 3 44(0.48) 201(0.48) 4 19(0.21) 73(0.18) 5 8(0.09) 16(0.04) Number of 92 416 accessions

BR

AC*

1(0.02) 7(0.16) 26(0.58) 7(0.16) 4(0.09) 45

4(0.03) 13(0.11) 66(0.55) 26(0.22) 11(0.09) 120

* The difference between LR and AC is statistically significant (χ2, p = 0.003).

data1. Using these parameters based on GSC logarithm distribution, the overall sample of 80 GSC values can be grouped into five classes, reflecting a successive increase in the fraction of rare alleles in accessions: (1) minimum–5th percentile, (2) 5th–25th percentiles, (3) 25th–75th percentiles, (4) 75th–95th percentiles, and (5) 95th percentile–maximum (Fig. 2c). This makes it possible to use the five-point score estimates of the GSCs, common for descriptors of estimation databases for many traits [18], rather than the absolute GSC values. The accessions falling into the two extreme ranges are of the highest interest. The accessions of the first class almost lack rare alleles. Such accessions can be regarded as the most typical accessions of a region, representing the basic gene pool. These accessions deserve a special attention when creating a representative core collection. The accessions with the highest GSC values (class 5) are also of special interest, as they contain the maximal number of rare alleles untypical of the corresponding region.

alleles at a lower rate (Table 2). On the other hand, the interquartile range (25–75%) demonstrates different distribution of the data over the GSC value space, namely GSC of 2 and 3 in LR and of 3 and 4 in bred cultivars (AC) (Fig. 3). The field weed and wild populations (WE) displayed the widest range of GSC variation. These results demonstrate that characteristic of the genetic structure of wild and field weed populations are both the predominant amount of the alleles typical of a particular region and certain rate of rare alleles as a variation reserve. The artificial selection of local forms decreases the genetic diversity and the rate of rare alleles; correspondingly, the accessions with a GSC of 5 are extremely rare among the local cultivated populations. The procedure of cultivar testing under high competition in the modern breeding demands that new AC display improved breeding traits as compared with LR. This cannot be achieved without involving new genetic material or rare alleles, in particular, those accumulated in the local wild and field weed populations, which are well adapted to the corresponding local conditions. Presumably, this can explain the fact that bred cultivars display the highest rate of rare, possibly introgressed, alleles. DISCUSSION When analyzing the genetic diversity of the gene bank collection with the help of molecular markers, we applied the classification of accessions according to GSC in the system of local gene pool. The theoretical background of this method is the postulate that the gene pool of a species is a common system where the indi6 5

Bred Cultivars Contain the Maximum Number of Rare Alleles Using the above scheme, we have analyzed all 11 regional samples and classified each of the 677 accessions according to the five-point score scale based on the GSC in the system of local gene pool. As each accession was concurrently ascribed to one of the four groups according to its status (bred cultivar, local cultivar, breeding material, and field weed and wild populations), we compared the distribution of five GSC classes within each group of genetic material. Although all five GSC classes are present in the accessions with any status and class 3 is the most frequent (median) in all four groups, the local cultivars (LR) significantly differ from the bred cultivars (AC), as LR contain rare 1 The

25th percentile of variable is the value when 25% of the observations in the sample fall below this value; the 75th percentile reflects 75% of the observations, and so on.

4 3 2 Median 25–75% 5–95%

1 0

WE

LR

AC

Fig. 3. Comparison of the range and pattern of GSC variation in accessions from different groups of genetic material (“box” plot): the ordinate, the classes of GSC in five-point score scale; WE, field weed and wild populations; LR, local cultivars; and AC, bred cultivars.


Vol. 44

No. 11

2008


vidual parts can differ in the ratio of allelic frequency [19]. This provides the possibility to find both rare and frequent alleles for a particular region and perform the weighting procedure for the alleles detected using molecular markers based on their frequency. Thus, each accession can be characterized by the ratio of rare and frequent alleles, i.e., by the calculated GSC. The proposed method can use any type of molecular markers, and its practical result is classification of accessions according to five-point score scale, which can be recorded in the descriptors of certificate databases and used for optimized work with collections. The classification of collection accessions according to GSC can be useful when giving recommendations for breeders. The most critical point in plant hybridization, which is widely used in breeding, is the selection of parental pairs. The long-term experience of breeders demonstrates that the genetic control of traits in the majority of accessions even from different geographical regions is approximately similar [20]. No considerable improvement in the target traits are observed in the hybrid progeny of such accessions: the transgression forms are random in the breeding process, and breeders eventually obtain them via a multitude of invaluable hybrid combinations. To solve in part this problem, attempts have been made to mark the genetic differences between accessions with the help of morphophysiological traits, in particular, by characterizing the type of trait variation under different growth conditions [21]. It has been shown that the main part of vetch accessions display a common type of trait variation, whereas the other variation types are found in minor part of the accessions. Classification of the accessions with the help of AFLP markers confirms this conclusion. The majority of accessions display the genetic structure typical of a particular region, and increased rate of rare alleles is found in only small group of accessions. Presumably, this particular group of accessions can be of special interest for breeders from the standpoint of obtaining transgression forms. The discovered trend of decrease in the rate of rare alleles during selection of the local cultivars from field weed and wild forms agrees with the previously described bottleneck effect [22]. During domestication, only the commercially phenotypes are selected from the vast overall diversity of wild plants. Eventually this leads to narrowing of the genetic diversity in cultivated populations. It has been reported that the level of DNA variation in the cultivated plant species is considerably lower as compared with their wild relatives, because cultivated species pass a certain breeding bottleneck owing to selection of particular phenotypes, which considerably reduces the level of their variation [23, 24]. Nonetheless, the example of modern common vetch cultivars demonstrates that the actual breeding achievements can be directly connected with the introgression of new alleles untypical of a particular region into the genotypes of local cultivars, well adapted to local climatic conditions. RUSSIAN JOURNAL OF GENETICS

Vol. 44

No. 11

1315

ACKNOWLEDGMENTS The authors are grateful to the International Plant Genetic Resources Institute (IPGRI) for the financial assistance under the Vavilov–Frankel Fellowship Foundation and to the Institute of Plant Genetics and Crop Plant Research (IPK, Gatersleben, Germany) for the provided possibility to perform experimental work. REFERENCES 1. FAO: The Role of Biotechnology for the Characterisation and Conservation of Crop, Forest, Animal and Fishery Genetic Resources in Developing Countries, Background Document to Conference 13 of the FAO Biotechnology Forum, 2005. 2. Virk, P.S., Ford-Lloyd, B.V., Jackson, M.T., and Newbury, H.J., The Use of RAPD for the Study of Diversity within Plant Germplasm Collections, Heredity, 1999, vol. 74, pp. 170–179. 3. Hodgkin, T., Brown, A.H.D., van Hintum, Th.J.L., and Morales, E.A.V., Core Collections of Plant Genetic Resources, Chichester: Wiley, 1995. 4. Vos, P., Hogers, R., Bleeker, M., et al., AFLP: A New Technique for DNA Fingerprinting, Nucleic Acid Res., 1995, vol. 16, pp. 81–86. 5. Potokina, E., Blattner, R., Alexandrova, T., and Bachmann, K., AFLP Diversity in the Common Vetch (Vicia sativa L.) on the World Scale, Theor. Appl. Genet., 2002, vol. 105, no. 1, pp. 58–67. 6. Gosudarstvennyi reestr sel’skokhozyaistvennykh dostizhenii, dopushchennykh k ispol’zovaniyu: Sorta rastenii. Gosudarstvennaya komissiya RF po ispytaniyu i okhrane selektsionnykh dostizhenii (State Register of Agricultural Achievements Accepted for Use: Plant Cultivars. State Committee of Russian Federation for Testing and Legal Protection of Breeding Achievements), Moscow: MSKh RF, 2004. 7. Leokene, L.V., Ecogeographical Classification of Cultivated Common Vetch (Vicia sativa L. subsp. sativa) Cultivars, Tr. Prikl. Botan/ Genet. Selekts., 1978, vol. 63, no. 1, pp. 108–122. 8. Katalog mirovoi kollektsii VIR: Posevnaya vika (All Union Institute for Plant Cultivation Catalog of World Collection: Common Vetch), Leningrad, 1970, issue 63. 9. Katalog mirovoi kollektsii VIR: Vika posevnaya (VIR Catalog of World Collection: Common Vetch), Leningrad, 1990, issue 529. 10. Katalog mirovoi kollektsii VIR: Vika posevnaya (VIR Catalog of World Collection: Common Vetch), Leningrad, 1993, issue 645. 11. Nei, M., Genetic Distance between Populations, Am. Nat., 1972, vol. 106, pp. 283–292. 12. Sneath, P.H.A. and Sokal, R.R., Numerical Taxonomy, San Francisco: Freeman, 1973. 13. Jaccard, P., Nouvelles rescherches sur la distribution florale, Bull. Soc. Vaud. Sci. Nat., 1908, vol. 44, pp. 223–270. 14. Dice, L.R., Measures of the Amount of Ecologic Association between Species, Ecology, 1945, vol. 26, pp. 297–302. 15. Nei, M. and Li, W.H., Mathematical Model for Studying Genetic Variation in Terms of Restriction Endonu2008

1316

16. 17. 18. 19. 20.


cleases, Proc. Natl. Acad. Sci. USA, 1979, vol. 76, pp. 5269–5273. Smirnov, E.S., Taksonomicheskii analiz (Taxonomic Analysis), Moscow, 1969. Smirnov, E.S., About Trait Coding for Taxonomic Analysis, Zh. Obshch. Biol., 1971, vol. 32, no. 2, pp. 224– 228. Mezhdunarodnyi klassifikator SEV vida Vicia sativa (COMECON International Classificatory for Vicia sativa), Leningrad, 1985. Timofeev-Resovskii, N.V., Yablokov, A.V., and Glotov, N.V., Ocherk ucheniya o populyatsii (Outline of a Study on Population), Moscow: Nauka, 1973. Kurlovich, B.S., Rep’ev, S.I, Shchelko, L.G. et al., Genofond i selektsiya zernovykh bobovykh kul’tur

(lyupin, vika, soya, fasol') (Gene Pool and Legume Grain Crops (Lupine, Common Vetch, Soya Beans, and Haricot) Breeding), St. Petersburg: VIR, 1995. 21. Rep’ev, S.I., Theoretical Problems of Common Vetch Breeding, Byull. VIR, 1989, issue 190, pp. 3–6. 22. Tanksley, S.D. and McCouch, S.R. Seed Banks and Molecular Maps: Unlocking Genetic Potential from the Wild, Science, 1997, vol. 22, no. 277, no. 5329, pp. 1063–1066. 23. Doebley, J., Mapping the Genes That Made Maize, Trends Genet., 1992, vol. 8, no. 9, pp. 302–307. 24. Buckler, E.S., Thornsberry, J.M., and Kresovich, S., Molecular Diversity, Structure and Domestication of Grasses, Genet. Res., 2001, vol. 77, no. 3, pp. 213–218.

SPELL: OK


Vol. 44

No. 11

2008