Morphological Characterization of the ECPGR Wild Brassica Species Collection F. Branca, G.L. Chiarenza and L. Ragusa Department of Agriculture and Food Science University of Catania Catania Italy S. Argento CNR - Institute for Mediterranean Agriculture and Forest Systems, UOS Catania Italy Keywords: Brassicaceae, characterization, populations, bio-morphological Abstract A project planned by the Brassica WG in the frame of Phase VIII (2009-2013) of the European Cooperative Programme for Plant Genetic Resources (ECPGR) allowed to characterize several populations of wild Brassica in the same environmental condition, with the aim to characterize a collection of accessions conserved in several European genebanks. We characterized 26 accessions of wild Brassica species including Brassica barrelieri, Brassica balearica, Brassica bourgeaui, Brassica cretica, Brassica desnottesii, Brassica drepanensis, Brassica fruticulosa, Brassica hilarionis, Brassica insularis, Brassica incana, Brassica macrocarpa, Brassica montana, Brassica oleracea wild type, Brassica rapa, Brassica rupestris and Brassica villosa. The biomorphological characterization was performed by the main IBPGR descriptors. The data obtained were processed using SPSS program by applying the hierarchical cluster analysis with the method of the complete linkage (furthest neighbour). The cluster analysis classified the studied accessions in seven main groups with different characteristics. INTRODUCTION The centres of origin of many wild Brassica species are Asia, the Mediterranean, the Balkans and the Middle East (Vavilov, 1926). Over the years the centres of origin indicated by Vavilov were modified, but according to all theories, the Mediterranean remains an important area for the origin of many of the current species of agricultural interest (Zhukovskij, 1968). Brassica wild species have been largely utilized for their resistance/tolerance to biotic and abiotic stress and for obtaining new cultivars with high levels of glucosinolates (Mithen et al., 2003; Branca et al., 2012). In some less developed countries the wild species represent an important food source and they contribute to meet the nutritional needs, even though advanced agriculture systems have contributed to their demise (Branca and Candido, 2007; Branca, 2008). In situ conservation and protection of wild species represent the most interesting perspective in view to conserve for the future generations the existing gene pools contained in the crop wild relatives (Baker, 1972; Branca and Cartea, 2008). It is convenient to use the wild species, because many of them contain minerals, proteins (Franke, 1985) and vitamins A and C (Souci et al., 1986; Schneider and Reinartz, 1987; Fritz, 1989). On the other hand, the presence of nitrates is a downside which may cause some problems to human health (Schneider and Reinartz, 1987). Brassica wild species (n=9) represent an interesting objective of study owing to their agronomic and qualitative traits, as well as their resistance to biotic and abiotic stress and their potential use in breeding programs of Brassicaceae crops. They are also widely used for human diet and for oil industry (Branca et al., 2005, 2010). As part of the European Cooperative Programme for Plant Genetic Resources (ECPGR) Phase VIII (2009-2013), the Brassica Working Group (BWG) carried out a project to characterize various European populations of wild Brassica and of B. rapa for their bio-morphologic, metabolic and genetic traits. The Department of Agriculture and Proc. VIth IS on Brassicas and XVIIIth Crucifer Genetics Workshop Eds.: F. Branca and A. Tribulato Acta Hort. 1005, ISHS 2013 157 Food Science (DISPA) of the Catania University was involved in bio-morphological and genetic characterization of the wild Brassica populations, whereas glucosinolate (GLS) compounds were analysed by the Institute of Sustainable Agriculture (CSIC) of Cordoba, Spain. MATERIALS AND METHODS We characterized 26 accessions provided by four European genebanks. These were sown in trays in April 2009 and transplanted in June in cold greenhouse at the experimental farm of the Catania University (Table 1). Sowing was carried out in alveolar containers but the seed samples and their quality did not permitted to obtain the same amount of plants for each accession which varied from seven to twenty-one. The plantlets were transplanted at 3-4th leaf stage, in rows 100 cm apart and 50 cm apart along the row (2 plants per m2). The experimental design was randomized blocks. Characterization was made after one year after transplanting. Several IBPGR descriptors were utilised for morphological characterization (Table 2). The mean values of the parameters analyzed for each accession were used for the preparation of a numerical matrix and for the identification of related parameters. This matrix was analysed to define the contribution of each character in determining the total variability. The major components of variability have been identified, enabling to distribute the studied accessions on a system of Cartesian axes, taking into account the first three principal components. The data matrix obtained was processed using SPSS 8.0 program by applying the hierarchical cluster analysis (Chatfield and Collins, 1980) with the complete linkage (furthest neighbour) method (Massie et al., 1996). RESULTS After one year only 20 accessions reached the reproductive stage. The characters showing higher variability were plant height, plant diameter, leaf blade shape and length, petiole length and width and the number of leaves per main stem. The cluster analysis classified the studied accessions into seven main groups with different characteristics. Accessions belonging to the same species were placed in different groups; this is caused by the fact that plants from the same species showed marked differences in some characters. The two accessions of Brassica cretica (CRE1 and CRE2) were different for vegetative stem width, ratio length/diameter, leaf blade length and width, petiole length, width, enlargement and colour. Brassica incana (HRIGRU 6691 and BRA 2918) were different for vegetative stem width, plant height and branching, and leaf anthocyanin coloration, blade length and shape. Brassica macrocarpa (MAC1 and MAC2) are different for plant branching, leaf lobes, and color, petiole length, number of leaves for plant main stem and flowering time. Brassica montana (MON1 and MON2) differed mainly for the ratio plant length/diameter, plant branching, plant height, leaf anthocyanin colour, petiole length and colour and number of leaves for plant main stem. Brassica rupestris (RUP1, RUP2, RUP3) were different for the ratio length/diameter, plant shape, leaf shape, plant branching and number of leaves for plant main stem. Brassica villosa VIL3 is different from VIL1 and VIL2 for ratio plant length/diameter, plant height, plant shape, leaf blade length, plant branching and number of leaves for plant main stem. Group A is mainly different from the other groups for the light green colour of the leaves and the white colour of the petiole; group B is different mainly for a more green coloration of the petiole; group C is different for larger width of the petiole and number of scars; Group D is different for the larger vegetative stem width, higher relation length/diameter, plant height, leaf blade length and width, higher angle of petiole, with leaf lamina more straight and less branching of the plant; Group E is mainly different for lower height and relation length/diameter of the plant, lower plant diameter, anthocyanin coloration and blade width of the leaf, lower petiole length and width and lower number of leaves and scars. Group F mainly differed for longer diameter, more elliptic leaf blade 158 shape and longer petiole length. Group G was mainly different for lower vegetative stem width, leaf blade length, leaf blade blistering and for higher branching of the plant and average number of leaves per plant (Fig. 1). The distribution of studied samples in the space described by the three main components (Tables 3 and 4) allowed to identify four main groups. Group A included more wild Brassica species. Group B included B. barrelieri and B. desnottesii. Group C included one B fruticolosa (ITA), one B. cretica (GRC), one B. montana (ITA) and B. insularis. Group D included all remaining accessions (Fig. 2). CONCLUSION This work allowed us to observe a considerable variance among wild Brassica species as well as diversity among the same species widespread in different distribution areas. It could be noted that the used morphological descriptors could not discriminate the species according to the existing taxonomy and the additional biochemical and genetic studies foreseen in the frame of the BWG project will classify. Among the characterized species, B. macrocarpa and B. balearica were markedly different from other Brassica species. Literature Cited Baker, H.G. 1972. Human Influences on Plant Evolution. Econ. Bot. 26:32-43. Branca F. 2008. Cauliflower and broccoli. p.147-182. In: J. Prohens and F. Nuez (eds.), Vegetables I. Springer, Heidelberg, Dordrecht, London, New York. Branca, F. and Candido, V. 2007. Innovazione di processo e di prodotto nella filiera del cavolfiore e del cavolo broccolo. p.654-657. In: Atti II Convegno Nazionale Piante Mediterranee. Agrigento, 7-8 October 2004. Branca, F. and Cartea, E. 2011. Brassica. p.17-36. In: C. Kole (ed.), Wild crop relatives: genomic and breeding resources, Vol. Oilseeds. Springer, Heidelberg, Dordrecht, London, New York. DOI 10.1007/978-3-642-14871-2_2. Branca, F., Bahcevandziev, K., Perticone, V. and Monteiro, A. 2005. Screening of Sicilian local cultivars of cauliflower and broccoli to Peronospora parasitica. Biodiversity and Conservation 14:841-848. Branca, F., Ragusa, L., Argento, S. and Tribulato, A. 2010. Attività per la conservazione in situ di specie spontanee del genere brassica (n=9) diffuse in Sicilia. p.175-180. In: G. Sarli, A. Alvino and C. Cervelli (eds.), Le potenzialità del territorio e dell’ ambiente. Nova Siri Marina, 7-10 October 2010. ISNB 978-1-4466-8981-3. Branca, F., Argento, S. and Tribulato, A. 2012. Assessing genetic reserves in Sicily (Italy): the Brassica wild relatives case study. p.52-58. In: N. Maxted, M.E. Dulloo, B.V. Ford-Lloyd, L. Frese, J.M. Iriondo and M.A.A. Pinheiro de Carvalho (eds.), Agrobiodiversity conservation: securing the diversity of crop wild relatives and landraces. CABI, Wallingford, UK. doi: 10.1079/9781845938512.0052. Chatfield, C. and Collins, A.J. 1980. The multivariate analysis of variance, Introduction to Multivariate Statistic. Chapman and Hall, London. p.140-161. Franke, W. 1985. Vergleichende Qualitätsbewertung von Wild- und Kulturgemüse. Vortragreihe d. 38. Hochschultagung d. Landw. Fakultät der Universität Bonn, 5-6 March. Fritz, D. 1989. Starting points for crop research to promote diversification. Acta Horticulturae 242:193-201. Massie, I.H., Astley, D. and King, G.J. 1996. Patterns of genetic diversity and relationship between regional groups and populations of Italian landrace of cauliflower and broccoli (Brassica oleracea L. var. botrytis and var. italica). Acta Horticulturae 407:45-53. Mithen, R., Faulkner, K., Magrath, R., Rose, P., Willianson, G. and Marquez, L. 2003. Development of isothiociante-enriched broccoli and its enhanced ability to induce phase 2 detoxification in mammalian cells. Theoretical and Applied Genetics 106:727734. 159 Schneider, V. and Reinartz, M.T. 1987. Nitratgehalte von heimischen Wildgemüse und Wildsalatarten. Ernährungs-Umschau 34:157-160. Souci, S.W., Fachmann, W. and Kraut, H. 1986. Die Zusammensetzung der Lebensmittel. Nährwert-Tabellen 1986-87. Wissenschaftl. Verlagsges. mbH., Stuttgart. Vavilov, N.I. 1926. Studies on the origin of cultivated plants. Bull. Appl. Bot. Genet. Pl. Breed. 16:139-248. Zhukovskij, P.M. 1968. New centres of origin and new gene centres of cultivated plants including specifically endemic microcentres of species closely allied to cultivated species. Bot. Zh. 53:430-460. Tables Table 1. Accessions characterized. Instcode DEU 146 DEU 146 DEU 146 DEU 146 DEU 146 DEU 146 DEU 146 DEU 146 DEU 146 DEU146 DEU146 DEU146 DEU146 DEU146 DEU146 DEU146 DEU146 GBR006 GBR006 NLD037 NLD037 NLD037 UNICT 2973 UNICT 3824 UNICT 3844 UNICT3944 160 Accenumb BRA 2850 BRA 2990/K10127 BRA 2998 BRA 2848 K 6631 K 10120 BRA 2919 BRA 2923 BRA1810 BRA 1727 BRA 2918 K 5997 BRA 2944 BRA 1644 CR 2929 BRA 2945/K7690 BRA1896 HRIGRU 12483 HRIGRU 6691 CGN06903 CGN18472 CGN14116 BB S542 S401 BB7 Genus and species Brassica balearica Brassica barrelieri Brassica bourgeaui Brassica bourgeaui Brassica cretica Brassica cretica Brassica desnottesii Brassica drepanensis Brassica fruticulosa Brassica fruticulosa Brassica incana Brassica insularis Brassica macrocarpa Brassica montana Brassica rapa Brassica rupestris Brassica villosa Brassica hilarionis Brassica incana Brassica oleracea Brassica montana Brassica villosa Brassica macrocarpa Brassica rupestris Brassica rupestris Brassica villosa Working code BAL BAR BOU1 BOU2 CRE 1 CRE 2 DES DRE FRU 1 FRU 2 INC 2 INS MAC 2 MON 2 RAP RUP 1 VIL 1 HIL INC 1 OLE MON 1 VIL 2 MAC 1 RUP 2 RUP 3 VIL 3 Origin PRT ESP ESP GRC TUR ITA ESP ITA ITA ITA DEU ITA ITA ITA FRA ITA Mediterranean ITA ITA ITA ITA Table 2. IBPGRI and UPOV descriptors utilized. IBPGR descriptors Plant STDI ALDI PIAL PIDI PIRA MEFP Leaf 4.2.55 4.2.56 4.2.3 4,2,4 Vegetative stem width cm (measure diameter of widest point on stem) Length/diameter (cm) Plant height cm (measure extremity of plant) Plant diameter cm (measure extremity of plant) Branching plant (0=absent, 3=intermediate, 7=high) Average number of leaves for plant main stem (3=few, 5=intermediate, 7=many) 4.2.11 FGCO 4.2.24 FGFO 4.2.16 FGLU FGLA 4.2.12 4.2.13 FGAG 4.2.15 FGAT FGMA FGLO MEDC Petiole PELU PELA PEAL 4.2.23 4.2.21 Leaf color (1= yellow green, 2=light green, 3=green, 4=dark green, 5=purple green, 6=purple, 7=other) Leaf blade shape (1= orbicular, 2=elliptic, 3=obovate, 4=spathulate, 5=ovate, 6=lanceolate, 7=oblong) Leaf blade length (measure largest leaf including petiole) Leaf blade width (widest point on largest leaf) Angle of petiole with horizontal :1=erect (>87°), 2=open (~67°), 3=semiprostrate (~45°), 4=prostrate (<30°), 5=horizontal, 6=oblique(>-10°) Leaf lamina attitude (3=convex, 5=straight, 7=concave, drooping) Leaf blade blistering (0=none, 3=low, 5=intermediate, 7=high) Leaf lobes (1=absent, 9=present) Average leaf scars (n.) 4.2.10 Petiole length (cm) Petiole width (cm) Petiole enlargement (3=narrow, 5=intermediate, 7=enlarged) Petiole and/or midwein colour (1=white, 2=light green, 3=green, 4=purple, 5=red, 6=other) UPOV descriptors 4.2.27 PECO 4.2.33 Plant PIFO 4.2.2 UPOV_3 Plant shape (Plant growth habit) Leaf FGAN FGCU FGSE UPOV_5 Leaf anthocyanin coloration (1= absent, 9=present) Leaf blade density of curling (1=absent or very low, 3= low, 5= medium, 4.5.10 UPOV_14 7=high) UPOV_15 Folding leaf section (3=weak, 5=medium, 7= strong) Table 3. Percentage of variance of the principal components identified by factor analysis. Component 1 2 3 % of variance 24.6 16.7 13.5 % of variance cumulative 24.6 41.4 54.9 161 Table 4. Correlation coefficient of individual parameters with the principal components. Parameters STDI ALDI PIAL PIDI PIFO FGAN FGCO FGFO FGLU FGLA FGAG FGAT FGMA FGLO PELU PELA PEAL PECO PIRA MEFP MEDC FGSE TOME % of variance 162 PC1 0.489 0.613 0.877 0.663 0.524 -0.126 -0.397 -0.346 0.916 0.838 0.392 0.438 -0.240 -0.298 0.319 0.287 -0.507 -0.519 -0.167 0.385 -0.477 0.362 24.6 Principal components PC2 -0.190 0.588 -0.250 -0.235 0.160 -0.278 0.192 0.378 0.844 0.834 0.692 0.423 0.461 0.639 0.120 0.463 0.174 16.7 PC3 -0.194 -0.302 -0.555 -0.536 0.223 0.548 0.281 0.762 0.553 -0.169 0.256 -0.363 -0.543 0.731 13.5 Figurese Fig. 1. Cluster analysis of the analyzed materials. Fig. 2. Distribution of the studied materials in the space described by the first three principal components. 163 164