Correlation and path analysis studies of upland rice (Oryza sativa L.) collected from Pala-U village, Prachuap Khiri Khan, Thailand

The decline in rice production experienced by ethnic minority (Pa-gha-ker-yor) farmers at Pala-U village, Prachuap Khiri Khan Province, Thailand, has multiple causes. The effect, however, is threatening the sustainability and well-being of the local populations. The objective of this study is to determine the relative yield and yield components of upland rice varieties collected from these farmers, both inside and outside of forested areas. Such information is then used to help determine the most promising breed varieties for future cultivation. Correlation and path coefficient analysis were conducted for yield, yield components, and related characteristics in seven genotypes of upland rice. The research was conducted during planting seasons in 2015 and 2016 in Prachuap Khiri Khan and Phetchaburi provinces, Thailand, respectively. Results show a significant positive correlation between grain yield hill-1and seed number panicle-1.Though non-significant, positive correlations were found between grain yield and four other criteria: panicle length, tiller number hill-1, plant height, and seed weight panicle-1. In 2015 and 2016, as for path analysis, grain yield hill-1was directly influenced byfactors either showedhigh positive effect such as: seed number panicle-1, panicle length, plant height and percent of grain filling, or high negative effect as100 seed weight. Seed number panicle-1, percent of grain filling, seed weight panicle-1, and plant heightwere shown to have an indirect effect on grain yield hill-1. However, grain yield hill-1was negative indirectly influenced through other characteristics by percent of grain filling. As such, seed number panicle-1, panicle length,and plant heightdemonstrated the greatest influence on yield may be considered as primary criteria, with percent of grain fillingand seed weight panicle-1qualifying as secondary criteria for high yield selection