Comparative Productivity and Adaptive ability of Forage Pea ( Pisum sativum L . ) and Vetch ( Vicia sativa L . ) Cultivars

Adaptive potential of forage pea and vetch accessions was estimated based on seed yield and yield components. А varietal-specific reaction of the studied accessions to environmental conditions was established. The conducted assessment of selective value of genotypes gave accurate data for breeding of highly productive forms. Among studied pea varieties the following were characterized to possess adaptive ability and selective value―Glyans for numbers of seeds (1000 seed weight and seed yield), Kamerton for plant height, and Svit for seed yield. In terms of general adaptive ability and stability in plant height with the greatest value as a starting material was vetch variety Vilena. Moldovskaya was of interest in terms of 1000 seed weight. Liya is preferred vetch genotype for improving the number of seeds per plant and for selection of new forms for improving grain yield.


Introduction
Pea and vetch are ones of the main annual legumes used in fodder production (Sarviro & Martynyak-Pshibyshevska, 2001;Vishnyakova, 2008).These species have wide application in modern agriculture due to their high productivity and relatively less rigorous growing conditions (Goncharov, 2011).To the present moment considerable attention is devoted for assessment of the separate components of yield (Debeliy et al., 2011).
An important task in breeding is development of cultivars that can realize a high productive potential in wide range of soil-climatic conditions (Vorobyov, 2009).The stabilization of productive potential of the crops requires a complex application of breeding and agrotechnical approaches among targeted at development of new cultivars that realize high productive potential (Medvedev, 2011;Vishnyakova, 2012).
Developing ecologically resistant varieties is priority in breeding of agricultural crops.The transition toward adaptability is possible on condition that the cultivated species and varieties are capable to effectively use native and artificial resources to the greatest extent (Medvedev, 2011;Vishnyakova, 2012).The increasing requirements for new varieties to possess tolerance to stress factors determine the necessity of development of adaptive and ecological direction in breeding process.The presence of perspective initial material having high-adaptive potential in some traits has an important role in future hybridizations (Anohina et al., 2007;Batalova, 2013).
Since the yield is determined by the interaction of genotype and environment, there is a necessity of comprehensive approach in designing breeding programs aimed at increasing the potential of ontogenetic adaptability of genotype.There are many methods for statistical analysis of the genotype-environment interaction which allow assessment of adaptability of genotypes towards environment.Kilchevskiy andHotyleva (1985, 1989) have proposed a method to analyze the genotype-environment interaction which allows simultaneous analysis of general and specific adaptability (Kilchevskiy, 2005).This method allows to determine the breeding value of genotype and to conduct natural selection depending on breeding task.The authors underline that when considering stability is expediently to be used more than one mutually complementary parameters (Fadeev, 2014).
Practice has shown that the assessment of the parameters of adaptive ability, stability and breeding value of varieties and lines allows with precision to determine the best genotype for given area for cultivation (Vorobyov & Vorobyov, 2011).
The aim of the study was to conduct an assessment of adaptive potential of vetch and pea accessions with regards to traits determining the productivity and yield.

Establishment of a Field Experiment
The field trial was carried out from 2012 to 2014 at the Institute of Forage Crops, Pleven (43.41ºN,24.61ºE),situated in the central part of the Danube hilly plain.The trial was set up as a complete block design method with three replications and with a plot size of 4 m 2 .The sowing was done by hand, at a depth of 5 cm and with a rate of 120 and 220 seeds m -2 for pea and vetch, respectively.The plants were cultivated in organic farming conditions without use of any fertilizers and pesticides (Table 1).Weeds were controlled mechanically during the growing period.Ten randomly selected plants from each unit were marked and used to estimate yield components-plant height (cm); 1000 seed weight (g); number of pods per plant and number of seeds per plant.Seed yield (kg ha -1 ) was found based on seed yield per plot at the end of growing season at a standard seed moisture content of 14%.For all traits an arithmetical average was calculated.

Plant Material
Four spring cultivars of pea (Pisum sativum L.) were included in the trial namely Glyans, Svit, Kamerton and Modus and four spring vetch (Vicia sativa L.) cultivars namely Liya, Lorina, Vilena and Moldovskaya.The origin of pea cultivars is Ukraine, and of vetch cultivars, Moldova.

Weather Conditions
Single trial years were used as the factor of environmental conditions.Some monthly meteorological data during the period of crop growth are given in Figure 1.The meteorological conditions during experimental period were different considerably.The average daily air temperature in 2012 (March-April) was 1.2 and 1.5 o С higher as compared to 2013 and 2014, respectively.The high temperature in 2012 coupled with lower rainfall (as compared to 2013 and 2014) determined this year as drier and unfavorable for plant development.Considerably greater amount of rainfalls in 2014, their even distribution, as well as lower temperature (not exceeding 24 o С) created favorable conditions for plant growth and development.The year 2013 was intermediate.

Statistical Analyses and Methods
The obtained data were processed by two-factor analysis of variance for each trait for determine of effects of genotypes (G), (E) environments and genotype environment interaction (G × E).The estimation of the ecological stability of the tested cultivars was done through the application of regression analysis according to Finlay and Wilkinson (1963).
The method of Kilchevsky and Hotyleva (1985a;1985b) was used for the characterization of: general adaptive ability (GAA: characterized average meaning of trait in different environmental conditions) and specific adaptive ability (SAA: characterized deviation from the GAA in certain environment), relative stability of the genotypes (Sgi: indicator which for comparing different crops and varieties); variance of specific adaptability (σ 2 CACi); organism's reaction to the cultivation place (Igi), criterion for estimation of the genotype ability to enter into interaction with environment (σ(GxE)gi) and selective value of genotype (SVG: used for determining the variety which combines high-productiveness and stability).Stress resistance (Y) (Rossielle & Hamblin, 1981), homeostatic (Hom) (Hangildin, 1984) and stability index of the variety (SV) (Nettevich et al., 1985;Nettevich, 2001) were determined.The interdependence between yield and parameters of stability and adaptability were determined by correlation analysis (Dimova & Marinkov, 1999).All experimental data were statistically analyzed by MS Excel ( 2003) for Windows XP and computer software PBSTAT 1.2.

Results and Discussion
Data for adaptability and stability can be obtained when studying a new breeding material in different years, which shows particular reactions of genotype to the change in ecological conditions.If the value of any trait is changing during the years it is an indication for the presence of genotype × environment interaction whose effect could be analyzed as dispersion complex (Korzun & Bruylo, 2011).
The results of dispersion analysis (Table 2) of the traits confirm differences in terms of the years.In pea varieties, for all traits, the genotypes (varieties), environmental conditions and interaction between them were significant.Differences for interaction genotype-environment in regard of traits mass of 1000 seeds, number of pods and seeds per plant were not significant for vetch varieties.Note.*; **: Significant at 0.05; 0.01 probability levels, respectively; ns, non significant.
The adaptability of the varieties to the environmental conditions can be assessed through their plasticity, expressed through different criteria.In our research we used coefficient of regression (bi) of Finlay and Wilkinson (1963).The more it exceeds unity, so more the cultivar is responsive to the terms of cultivation.If the value of bi coefficient is close to unity, then the cultivar is the most pliable and well adapted to various environmental conditions.When bi is considerably less than unity, the variety is appropriate for growing under adverse conditions.Zero or near to that value of coefficient of regression (bi) shows that the variety does not respond to changes of the environment.

Stability Parameters, Yield Components and Yield of Pea Varieties
The influence of environmental conditions on the formation of traits is expressed by the coefficients of regression.They characterize the average reaction of the genotype, show its stability or responsiveness and give an opportunity to predict the change of the analyzed trait in the framework of specific growing conditions.In our studies of pea varieties (Table 3) with regards to plant height, high responsiveness in cultivation in different conditions was manifested in Glyans (bi = 1.21).In Svit (bi = 1.02) and Modus (bi = 0.99) was comparatively lower bi which meant the best adaptability of trait.Svit came close to the desired breeding compromise for traits number of seeds per plant and grain yield combining very good stability with satisfying productivity.For the trait mass of 1000 seeds the tested varieties reacted in a similar way to the changes in the environment and showed relatively close stability.From breeding perspective the most interesting cultivar was Glyans (bi = 1.04) with the highest mass per 1000 seeds, followed by Modus (bi = 1.05) and Svit (bi 1.10).*; **: Significant at 0.05; 0.01 probability levels, respectively.
Another cultivar of particular interest was Glyans (bi = 1.04) with the highest mass per 1000 seeds, followed by Modus (bi = 1.05) and Svit (bi 1.10).The cultivars reacted in a quite different way to the environment with regards to number of pods per plant.Glyans forms 6-7 pods per plant while Modus produced 4-5 pods per plant and were characterized as responsive.Svit and Kamerton were ecologically stable, but with an average rate of the trait expression.

Stability Parameters, Yield Components and Yield of Vetch Varieties
Differences for plant height and mass per 1000 seeds were not significant for vetch varieties.Lorina and Moldovskaya were ecologically stable with bi ≈ 1 for two traits (Table 4).
Number of pods per plant appears to be one of the important components of productivity.For this trait variety Liya statistically exceeded the other varieties, forming more than 8 pods per plant, but this variety along with Moldovskaya can be considered as ecologically unstable.For these varieties high yields can be obtained with intensive technology use.The trait number of seeds per plant takes part in the yield formation and therefore is of certain interest to breeders.The plants of variety Liya formed most seeds and showed the greatest responsiveness to improving growing conditions.
Variety Vilena is also distinguished by high productivity, but also has the lowest coefficient of regression (bi = 0.57), which defines it in the group of stable varieties.With regard to the comprehensive trait grain yield, Liya and Vilena showed the highest values and analogically react to the environment, which is an indication that they possess different genetic systems.Hybrids possessing high yields and ecological stability can be obtained by crossing them.Varieties Lorina and Moldovskaya were low-yielded and stable with respect to environmental changes.*, **: Significant at 0.05; at 0.01 probability levels, respectively. www.ccsen

The analy population
The

Figure 1 .
Figure 1.Climatic characterization of the experimental period Note. b i -FW: Regression coefficient Finlay and Wilkinson's regression model; GAA: general adaptive ability; SAA: specific adaptive ability; Sgi: relative stability of the genotypes; σ 2 CACi: variance of specific adaptability; Igi: organism's reaction to the cultivation place, σ(GxE)gi: criterion for estimation of the genotype ability to enter into interaction with environment; SVG: selective value of genotype.
Note.FW: Regression coefficient Finlay and Wilkinson's regression model; GAA: general adaptive ability; SAA: specific adaptive ability; Sgi: relative stability of the genotypes; σ 2 CACi: variance of specific adaptability; Igi: organism's reaction to the cultivation place, σ(GxE)gi: criterion for estimation of the genotype ability to enter into interaction with environment; SVG: selective value of genotype.

Table 2 .
Analysis of variance for stability for seed yield and yield components in pea and vetch cultivars ANOVA df MSPea cultivars plant height (cm) 1000 seed weight (g) pods per plant seeds per plant Seed yield (kg da -1 )

Table 3 .
Parameters of stability and adaptability in the studied traits of forage pea varieties

Table 4 .
Parameters of stability and adaptability in the studied traits of vetch varieties