Morphological and agronomical characterization of common wheat landraces ( Triticum aestivum L . ) from the National Wheat Collection of Bulgaria

The knowledge about the extent of variability, the distribution and the relationship between descriptors within local germplasm collection are a high value for the improvement and the efficient genetic diversity maintenance and utilization of plant species. The objective of this study was to assess the morphological and agronomic characteristics of original germplasm of common wheat (Triticum aestivum L.), maintained in ex situ collection in IPGR-Sadovo. Fifty-five accessions of Triticum aestivum L. stored for more than 10 years in the National gene bank of IPGR-Sadovo were planted under field condition and their agro-morphological characters such as plant shape (at tillering), leaf-flag attitude (at the beginning of heading), spike attitude (at full ripeness), spike awnedness, spike color and spike shape, length of vegetative growth phase, plant height, length of spike, spikelets per spike and 1000 grain mass were recorded. The variation analysis showed that the most relative variable character during the period of study is the length of spike (C.V. %=15.09%), following to 1000 grain mass (C.V. %=8.04%) and spikelets per spike (C.V. %=7.66%). PC-analysis was applied to group accessions according to similarity on the basis of five traits (length of vegetative growth phase, plant height, spike length, spikelets per spike and 1000 seed weight) in two components in the factor plane. The analysis shows that the first component explains 30.349% of the total variation and the second -26.001%. Two factors explain total 56.350% of the variation in the experience. A database with assessment information of regenerated accessions was created. The results of this study will support efforts of conservation and utilization of landraces in winter bread wheat breeding programs.


Introduction
In the recent years increasingly recognize the importance of genetic resources on the prosperity of breading selection, agriculture and ecology (Anonymous, 1999;Stoyanova et al., 1998).Narrowing the range of genetic variation observed in the common and durum wheat as a result of using of conventional breeding selection practices reducing the ability to improve productivity of crops (Hadjiivanova et al., 2010).By approaching the limits of biological productivity of wheat in the recent years has greatly increased the need of new initial material (Hailegiorgis, 2011;Graybosch and Peterson, 2010;Lanning et al., 2010).In this regard the formation of the current gene pool of wheat, its planned and targeted research has been and is a major priority in the researching activity.
Collections of wheat as a genetic resource are maintained in the National Gene bank in the Institute of Plant Genetic Resources-Sadovo (12,539 accessions), (http:// eurisco.ecpgr.org).The collections contain a large genetic diversity: local populations, breeding materials (mutants, hybrids, introduced cultivars and lines) and wild species originating from the country as well of all over the world (Odzhakova et al., 2007;Kolev and Stoyanova, 2005;Popova, 2003).The National collection of common wheat includes 9,591 accessions, from which 1,051 accessions originating from Bulgaria.The study of ex situ collections is done according the international descriptors lists (Anonymous, 1984).Comprehensive evaluation of collections is a major source of information to create a database with assessment information of plant genetic diversity (Kolev, 2001;Angelova and Popova, 1998).This assessment is carried out on a number of morphological and agronomical characteristics of the samples and ranges from 10 to 68 indicators depending on the type of culture and biology.
The objective of this study was to assess the morphological and agronomic characteristics of original germplasm of common wheat (Triticum aestivum L.), maintained in ex situ collection in IPGR-Sadovo.

Materials and Methods
During the period 2010-2012 in the experimental field of Institute of Plant Genetic Resources-Sadovo are reproduced 55 accessions of common wheat (Table 1).All accessions originating from Bulgaria and characterized as "original germplasm" created in this region.The seeds been stored for more 10 year for long-term conservation under -18°C.
Sowings were made in the optimal time for this area: 10-15 October.Six rows were planted from each accession with row spacing 20/5 cm and 1 m length.The standard variety -Sadovo1 was sown as a systematic check.Regular field management operations were performed during the cropping season.
Observations and evaluations of morphological, biological and agronomic traits were carried out according to international descriptors lists (Anonymous, 1984).The agronomic characters were taken after harvesting the plants.From each accession, 20 plants were collected for biometrical measurements.
Statistical analyses were performed using the statistical program SPSS 13.0.PC-analysis was applied to group accessions according to similarity on the basis of five traits (length of vegetative growth phase, plant height, spike length, spikelets per spike and 1000 seed weight) in two components in the factor plane.

Morphological characters
During the study, accessions were characterized by following morphological characters: plant shape (at tillering), leaf-flag attitude (at the beginning of heading), spike attitude (at full ripeness), spike awnedness, spike color and spike shape.
According to the plant shape investigated accessions were divided into 3 groups: semi-erect, drooping and strongly declined (Table 2).The largest numbers of accessions were with drooping plant shape (37 accessions) and strongly declined (17 accessions).Only one accession (2001-TRT-AE-296) was with semi erect plant shape (Table 2).The leaf-flag attitude of 45 accessions was semiupright (15-45°) and upright (< 15º) at 7 accessions.Dominated the accessions with white color and cylindrical shape of spike and 39 accessions were without awns.In thirteen of them the lengths of the awns were less 21 mm.Only two accessions were with length of awns within the borders of 61-100 mm (2000-TRT-AE-235, 1997-TRT-AE-1).
According to the spike attitude accessions were grouped into 4 groups: erect, semi-erect, horizontal and nodding.The accessions with semi-erect attitude of spike were thirty and these with nodding attitude of spike only four (Table 2).

Agronomic characters
The biometrical measurements in this study included: length of vegetative growth phase, plant height, length of spike, spikelets per spike and 1000 grain mass (Table 3).The results show that the length of spike in the study varied from 5 to 11 cm.The highest values of this character (11 cm) were registered in 3 accessions (2000-TRT-AE-117, 2000-TRT-AE-113 and 2000-TRT-AE-84).The numbers of spikelets per spike were between 15 and 20.The mean value of plant height was 93.9 cm.Accessions with tall plants (90 to100 cm) and heavy kernels (1000 grain mass from 36 to 40 g) dominated this set of accessions (Table 3).The length of vegetative growth phase (from growing to heading) varied from 183 to 187 days.The basic statistics of the main descriptive characteristics (mean, minimum and maximum, std.error of means, std.deviation, variance and coefficient of variation) of five characters are shown in Table 4.The values of coefficient of variation (C.V. %) are from 0.6 to 15.09 %.The analysis show that the most relative variable character during the period of study is the length of spike (15.09 %), following to 1000 grain mass (8.04 %) and spikelets per spike (7.66 %).The values of these coefficients confirm that these traits are more susceptible to change under the influence of different factors.Relatively the least variable for the period of study indicated the length of vegetative growth phase.
PC-analysis was applied to group accessions according to similarity on the basis of five traits (length of vegetative growth phase, plant height, spike length, spikelets per spike and 1000 seed weight) in two components in the factor plane.The values of the two components to each of the study parameters were calculated empirically (Table 5).The analysis shows that the first component explains 30.349 % of the total variation and the second -26.001 %.Two factors explain total 56.350 % of the variation in the experience.This relatively small percentage illustrated the existence of complex relationships between the studied characters.For example the characters: length of spike, number of spikelets per spike and 1000 grain mass are associated with the first component.The second component is in correlation with length of vegetative growth phase and plant height (Figure 1).
The evaluation of phenotypic variability by multivariate analysis gives the possibility to include a large number of accessions and to identify the most suitable resources for special trait (Stoilova, 2007).Distribution of evaluated accessions in the coordinate system of PC1 and PC2, presents the grouping of accessions according to similarity of traits: length of vegetative growth phase, plant height, spike length, spikelets per spike and 1000 seed weight (Figure 2).Some of the accessions are separated as "detached" from other.The accession № 41 (2000-TRT-AE-113) in the upper right quadrant is characterized with least 1000 grain mass, long vegetative growth phase and largest length of spike.Sample № 46 (2000-TRT-AE-145) located to the opposite in the lower right quadrant is with largest number of spikelets of spike.The samples № 2, 3, 14, 49 (2000-TRT-AE-85, 2000-TRT-AE-144, 2000-TRT-AE-40 and 1978-TRT-AE-157) in the left lower quadrant are with the largest 1000 grain mass.The accessions №21 (2000-TRT-AE-98) and №5 (1978-TRT-AE-200) in the upper left quadrant are characterized with small length of spike and spikelets of spike and the bigger 1000 grain mass.These samples represent a certain interest to hybridization by different traits and can be recommended as donors in the breeding selection of winter bread wheat.1.

Conclusion
The present research provided information on genotype studied and its grouping.
The most relative variable character during the period of study is the length of spike (C.V.-15.09%),following to 1000 grain mass (C.V.-8.04%) and spikelets per spike (C.V.-7.66%).Relatively the least variable for the period of study indicated the length of vegetative growth phase.
PC-analysis was applied to group accessions according to similarity on the basis of five traits in two components in the factor plane.The accessions: 2000-TRT- AE-145, 2000-TRT-AE-85, 2000-TRT-AE-144, 2000-TRT-AE-40 and 1978-TRT-AE-157 can be recommended as donors by different traits in the breeding selection of winter bread wheat.

Figure 2 .
Figure 2. Distribution of evaluated accessions within the factor plane.The accessions presented with the conditional numbers 1-55 of the graph correspond to the description in Table1.

Table 1 .
Inventory of local (native) Triticum aestivum L. accessions stored for more than 10 years in the National gene bank of IPGR-Sadovo.

Table 2 .
Summary of morphological characters of 55 accessions of Triticum aestivum L.

Table 4 .
The basic statistics of the main descriptive characteristics in accessions of Triticum aestivum L.

Table 5 .
Weighted factors (PC1 and PC2) of descriptive characteristics on the rotated matrix with two factors.