Multivariation analysis in the evaluation of the hybrids and sowing dates interaction for sunflower grain yield

Abstract

Grain yield, as a main trait in sunflower, varies depending on a hybrid, location, year of investigation, sowing date, stand density, etc. Since the interaction is very often present in agriculture, it is necessary to use suitable statistical methods in order to evaluate the causes of interaction. The aim of this study was to assess the effects of hybrids and sowing dates, so as their interaction on sunflower grain yield during three vegetation seasons. Three hybrids, eight sowing dates and three years of growing were included in the experiment. AMMI (additive main effects and multiplicative interaction) analysis is one of the mainly used multiplicative model, Gauch and Zobel 1996, that evaluates main effects and also the interaction. Based on the AMMI1 biplot for three years of investigation, it can be concluded that the sowing date R4 (April 20) was the most stable for grain yield, which was the highest in two years (2004, 2005) and it was also above the grand mean in 2006. Furthermore, in 2006, the hybrid Miro was most stable (showing the lowest interaction) in comparison to other two hybrids, and its grain yield was above the grand mean. Graphical presentation of AMMI1 in the form of biplot could facilitate the choice of stable hybrids and sowing dates for a desired trait in sunflower.

Prinos zrna, kao najvažnija osobina kod suncokreta, varira u zavisnosti od hibrida, lokaliteta, godine ispitivanja, roka setve, gustine setve i drugih faktora. Pošto je interakcija veoma često prisutna u poljoprivredi, trebalo bi primeniti odgovarajuće statističke metode, koje će što efikasnije oceniti uzroke interakcije. Cilj ovog rada je bio da se ispita veličina uticaja hibrida, rokova setve, kao i njihove interakcije na variranje prinosa zrna suncokreta tokom trogodišnjeg perioda. U eksperiment je bilo je uključeno tri hibrida, osam rokova setve i tri godine ispitivanja. Metod glavnih efekata i višestruke interakcije (Additive Main Effects and Multiplicative Interaction - AMMI) je jedan od najznačajnijih i najviše korišćenih multivarijacionih modela, Gauch i Zobel, 1996, koji osim glavnih efekata (hibrid, rok setve) otkriva i interakciju. Na osnovu rezultata AMMI1 biplota za sve tri godine ispitivanja, može se zaključiti da je najstabilniji u prinosu zrna bio rok setve R4 (20. april), koji je u dve godine (2004, 2005) dao i najveći prinos, a i u 2006. godini rok R4 postigao je prinos zrna iznad opšteg proseka. U 2006. godini hibrid Miro bio je najstabilniji tj. pokazao je najmanju interakciju u odnosu na druga dva hibrida, a prinos zrna bio je iznad opšteg proseka. Grafički prikaz AMMI1 biplota može pomoći u izboru stabilnih hibrida i rokova setve za željene osobine.