PROTEIN AND β-ODAP CONTENT AND THEIR ASSOCIATION WITH YIELD CONTRIBUTING TRAITS IN SELECTED GRASS PEA LINES POLIGNANO1 G. B., POMA2 I., GRISTINA2 L., BISIGNANO1 V., ALBA3 V., LOSAVIO1 F., DELLA GATTA3 C. 1 Institute of Plant Genetics (IGV), C.N.R., Via Amendola 165/A, 70126 Bari, Italy 2 Department of Environmental and Territorial Agronomy, University of Palermo, Viale delle Scienze, 13, 90128, Palermo, Italy 3 Department of Biology and Agro-Forest and Environmental Chemistry, University of Bari,Via Amendola 165/A 70126 Bari,Italy Email: giambattista.polignano@igv.cnr.it Introduction Nowadays, in Europe, where diversification of cropping systems and more sustainable and environmentally safe agriculture are main concerns, Lathyrus species are seen as a good alternative to imported soya or the hazardous use of protein of animal origin in animal food. In Italy grass pea (L. sativus L.) is being rediscovered, as a traditional and local product mainly by organic farmers. It has been noted that the economic value of a plant can depend upon several traits. In grass pea, seed yield and low neurotoxin content (β-ODAP) determine the value of the crop. In light of this all around the world the breeding programmes are mainly focused on the possibility to develop superior genotypes with high yield and low β-ODAP content. Characterization programmes are the most important step in order to optimize the utilization of plant genetic resources collections. Germplasm evaluation efforts during the last three decades at Bari Gene Bank have led to develop interesting lines which have β-ODAP content of 0.2 – 0.3 % as compared to the range observed in the whole grass pea collection. In our previous researches morphometric, agronomic, biochemical and molecular data collected showed the available wide genetic variation of the grass pea germplasm and an independent variation of β-ODAP content on which to investigate. The present study has been conducted in order to evaluate the selected grass pea lines for protein and β-ODAP content and their association with yield contributing traits. Materials and methods Eight selected grass pea (Lathyrus sativus L.) lines were evaluated over three different growing seasons (2005-2008) at yhe experimental field at the “Sparacia Farm” (Cammarata-AG) in Sicily. A randomized complete-block design with three replicates was adopted. 35 viable seeds/m2 of each line were placed at the end of autumn in rows distant 50 cm. After one summer ploughing, 90 Kg ha-1 of P2O5 were filled during two harrowing before sowing; later two manual weeds control were performed. Harvest was executed at full maturity of pods (end of the spring). Temperature values and effective rainfall for each season were recorded. The following traits were measured on five plant randomly chosen in each plot: plant height, days to flowering, 100-seeds weight, seed length, seed width, seed thickness, seed yield, protein content and β-ODAP content. Data were subjected to univariate analysis. The plot means were used to represent the performances of each line in each season for the multivariate analysis. A pattern analysis approach, based on ordination and classification procedures, is presented to identify differences among lines in mean performances and response across seasons. Results In Table 1 are reported the mean values for the traits measured relative to the eight grass pea lines observed in each year of cultivation. Variance, mean square and significance for the main effects and their interactions are given in Table 2. The effects associated with lines, years and their interaction (Y x L) were most important in determining differential line responses for the traits: days to flowering, plant heigth, seed yield, 100-seed weight, protein content and β-ODAP content. On the contrary, seed length, seed width and seed thickness showed not significant differences. Using principal component analysis, we calculated eigenvalues, percentage of variation and load coefficients of the first three components for 9 traits, which accounted for 68% of the total diversity (Table 3). The first component displayed differences for seed yield, seed length, seed width, seed thickness and β-ODAP content; the second component for plant height and flowering time; while, 100-seeds weight and protein content showed high loadings in the third component. The ordination procedure allowed the relative proximity of lines to be visualized in a spatial model of reduced dimensions, and also indicate directions of major variation. A cluster analysis arranged the lines into groups that were differentiable in terms of means and stability. The cluster dendrogram was truncated for presentation at the eight fusion level because, at that level correspond the number of lines tested. All lines responses in each cluster were closely related. The results of clustering were combined with those of the principal component analysis as a visual aid for discerning clusters in subsequent graphical presentation (Fig. 2). Table 1. Mean values of eight grass pea lines estimated in three growing seasons. Sparacia Farm (Cammarata-AG) - Sicily Fig. 1 Rainfall and monthly temperatures during the growing seasons. Fig. 2 A graphic representation of the behaviour of the lines according the first three principal components and identification of clusters. The behaviour in each year is represented by a letter (line code) and a number (1=2005/06; 2=2006/07; 3=2007/08). Conclusions In general, all the grass pea lines tested showed good performances in terms of stability and adaptation during the three growing seasons in agreement with our previous experiences. There was considerable variation in all traits except for seed length, seed width and seed thickness. Under the conditions of this study the seed yield, the protein content and the β-ODAP content were most influenced by the genotype x environment interaction. In terms of seed yield all lines tended to produce greater yield in the less favourable season; a similar trend was observed for the β-ODAP content. In other words lines with both high yield and high β-ODAP content showed a better adaptation to low rainfall conditions. In light of this ed according to other authors (Hacque et al., 1993; Sarwar et al., 1996; Mehta and Santha, 1996) future research to reduce β-ODAP content could be directed toward understanding the mechanisms involved in the biosynthesis of β-ODAP especially in drought conditions. Principal component analysis provides interesting information about associations of traits that are useful to formulate better hypotheses for breeding work. In fact, the absence of strong association between protein content and β-ODAP content and their independent role in each separate principal component allows the achievement of useful recombinants in breeding work. The information here reported may be important to select promising grass pea lines with wide adaptation and stability in order to promote the expansion of the crop in sustainable and low-input agricultural systems. References POLIGNANO G. B., P. Uggenti, V. Alba, V. Bisignano, C. Della Gatta, 2005. Morpho-agronomic diversity in grasspea (Lathyrus sativus L.). Plant Genetic Resoucers 3(1): 29-34. POLIGNANO G. B., Bisignano V., Tomaselli V., Uggenti P. Della Gatta C., 2009. Genotype x Environment Interaction in grass pea (Lathyrus sativus L.) Lines. International Journal of Agronomy. Hindawi Publishing Corporation. Volume 2009, Article ID 898396, 7 pages.