Path coefficient analysis of quality of two-row spring barley

• Published in: Genetika (Beograd) Vol. 45; no. 1; pp. 21 - 30
• Main Authors: Przulj, Novo, Momcilovic, Vojislava, Crnobarac, Jovan
• Format: Journal Article
• Language: English
• Published: 2013
• Subjects: Hordeum vulgare
• ISSN: 0534-0012; 1820-6069
• DOI: 10.2298/GENSR1301021P

Malting quality is composed of numerous interacting traits with a high complexity concerning their biochemical and genetic basis. Malt extract is key indicator of barley malting quality and it is a mega-trait since it is influenced by a number of independent component traits. Understanding genetic and non-genetic factors that effects grain quality and grain yield is crucial in developing new cultivars, seed and mercantile production. Path analysis is one of the reliable statistical techniques which allow separation of the direct effect of each component trait on malt quality from the indirect effects caused by the interdependence component trait. The aim of this study was to investigate spring two-row barley quality as mega-trait depending on the component traits in the conditions of the Pannonian environments. Regression analysis with extract (EXT) as dependant and other traits (yield-YIL, test weight-TW, grain weight-GW, grading-GRA, grain protein concentration-GPC, viscosity-VIS, Kolbach index-KOL, Hartong number-HAR) as independent traits was performed out. Simple coefficient of correlations were calculated between independent traits and EXT in all pair combination and then used as inputs for path coefficient analysis. The quadratic curve fitted the best relationship between EXT and the independent traits. EXT was in positive (P<0.01) relationship with GW, GRA, KOL, and HAR with simple correlation coefficient of 0.47, 0.42, 0.39 and 0.50, respectively and in negative (P<0.01) relationship with GPC and VIS with simple correlation coefficient of -0.72 and -0.51, respectively. Path analysis explained more than 70% of the variation in EXT of which 34.3% was determined by direct negative path coefficient (P<0.01) of GPC without significant any indirect path effect. VIS negatively directly, (P<0.01) and negatively indirectly via GPC effected EXT. KOL did not have significant direct effect on EXT, but had rather prominent indirect effect via GPC, VIS and HAR. HAR positively directly (P<0.01) and positively indirectly via GPC effected EXT. The direct effect of VIS and HAR determined 13.0% and 14.1% of the variation, respectively. Kvalitet pivskog jecma odredjuje veliki broj medjusobno povezanih osobina, kompleksne biohemijske i geneticka osnove. Ekstrakt slada je osnovni pokazatelj kvaliteta pivskog jecma i moze se smatrati mega osobinom, posto zavisi od delovanja velikog broja drugih osobina. Poznavanje genetickih i negenetickih faktora pod cijom kontrolom se nalazi kvalitet i prinos pivskog jecma sustinsko je za stvaranje novih sorti i uspesnu semensku i merkantilnu proizvodnju. Path analiza je pouzdana statisticka metoda koja omogucava razdvajanje direktnih i indirektnih efekata nezavisnih osobina na mega osobinu ili mega osobine. Indirektni efekti su posledica medjuzavisnosti i zajednickog delovanja nezavisnih osobina. Cilj ovoga istrazivanja je proucavanje ekstrakta slada, kao mega osobine, u zavisnosti od osnovnih osobina zrna i slada, koje su u odredjenoj vezi sa sadrzajem ekstrakta u sladu. Istrazivanje je obavljeno u periodu 1998-2004 na osam sorti dvoredog jarog jecma u agroekoloskim uslovima Panonske nizije. Odnos izmedju sadrzaja ekstrakta u sladu (EXT) kao zavisne osobine i prinosa (YIL), zapreminske mase (TW), mase hiljadu zrna (GW), sortiranja (GRA), sadrzaja proteina u zrnu (GPC), viskoziteta (VIS), Kolbah indeksa (KOL), Hartong broja (HAR) kao nezavisnih osobina odredjeni su regresionom analizom, prostom korelacijom i path analizom. Kvadratna regresija je najbolje predstavljala funkcionalnu determinisanost EXT navedenim nezavisnim osobinama, mada nisu utvrdjene statisicki znacajne vrednosti koeficijenta determinacije. Vrednosti prostih koeficijenta korelacije izmedju EXT s jedne strane i GW, GRA, KOL, i HAR sa druge strane bile su pozitivne (P<0.01) i iznosile su 0.47, 0.42, 0.39 i 0.50, respective. EXT je bio u negativnoj (P<0.01) korelacionoj vezi sa GPC, -0.72 i VIS, -0.51. Path analiza je objasnila 70% varijabilnosti kod EXT, od cega je 34.3% bilo pod uticajem negativnog direktnog path koeficijenta GPC, koji nije imao znacajnih indirektnih efekata. VIS je imao znacajan (P<0.01) direktan efekat na EXT i indirektan efekat preko GPC. KOL nije imao znacajan direktan efekat na EXT ali su bile visoke vrednosti indirektnih efekata KOL preko GPC, VIS i HAR. HAR je imao znacajan pozitivan (P<0.01) direktan efekat na EXT i indirektan efekat preko GPC. Pod uticajem VIS bilo je 13.0%, a pod uticajem HAR 14.1% od ukupne varijabilnosti EXT.