Correlation and Principal Component Analysis on Growth, Yield and Weed Parameters of Kenaf (Hibiscus cannabinus L.) in Sudan Savanna, Nigeria

Abstract

This paper aims to show the effects of correlation and principal component analysis on growth, yield and weed parameters of Kenaf (Hibiscus cannabinus L.) in the Sudan Savanna, Nigeria. Kenaf has been used as a major source of fibres, paper, and other materials for many years. This fibre crop can be grown on poor soils with fewer inputs. Nigeria’s kenaf production figure in the year 2020 is estimated to be about 1,460 tonnes out of the 15,900 tonnes of production in Africa. The global output in 2024 is estimated at approximately 195,800 tonnes. Kenaf fibre and seed yield is mostly affected by growth, yield and weed parameters. The experiment was conducted at the Department of Crop Production Teaching and Research Farm, Faculty of Agriculture, University of Maiduguri (Latitude 11°50’N: and Longitude 13°10’E at an altitude of 319 m above sea level) during the rainy seasons of 2018 to 2019 between June and November. Data were collected on growth, yield components, and weed parameters, which were subjected to correlation analysis, and PCA was analyzed using R software (R Core Team, 2023). An AI assistant (Claude, Anthropic) was used to help write the R analysis scripts. The results of the combined year study showed that Stand count at harvest (0.748) showed the strongest significant linear correlation relationship with yield, Establishment count (0.478), and Weed Control Efficiency (0.402) had a moderate positive correlation with yield. In the year 2018, PC1 (43.66%) was the main axis of variation. PC2 (15.71%) captured additional treatment differences. Together, they explained 59.37% of the total variability. Whereas, the year 2019 PC1 explained 39.69% of the variation, and PC2 explained 11.27%; together, they accounted for 50.97% of the total variation. PC1, therefore, shows crop vigour and productivity, while PC2 was mainly associated with establishment count, stand count at harvest, seed yield and fibre yield, expressing plant population and yield formation. The interpretation shows that correlation outlined relationships existed among the variables, while PCA successfully reduced the dimensionality of the variables and identified the principal factors leading to treatment differences across years.