Determination of the best blackberry cultivar using various statistical techniques

Abstract

This study was conducted to determine the best blackberry cultivar using jointly various statistical techniques such as Chi-square, G, and Correspondence statistics. For this aim, data of pomological traits such as fruit weight, cane number, cane diameter, cane height, cane yield per plant of Ness, Cherokee, Arapaho, Chester Thornless, Navaho, Black Satin, Dirksen Thornless and Jumbo cultivars were collected during 2002-2006. With respect to chi-square and G statistics, associations between cultivar and each pomological trait (fruit weight, cane number, cane diameter, cane height, and cane yield per plant) were found more significant (P < 0.0001). The relationship between year and cane number was significant (P < 0.0001 and year and cane diameter was significant (P < 0.05). The highest cane number was produced in 2004, followed by 2005. The relationship between fruit weight and cane diameter or cane yield per plant were more significant (P < 0.0001). Although Blacksatin and Jumbo blackberry cultivars had the highest cane number, Chester Thornless from blackberry cultivars had the highest cane diameter, the highest cane yield per plant and the highest fruit weight. It was concluded that Chester Thornless cultivar was the most appropriate cultivar for Central Anatolia region. Key words: Adaptation, blackberry, cane, fruit, Chi-square statistic, G statistic, correspondence analysis.