Reassessing the estimation of leaf area in oil palm (Elaeis guineensis Jacq.) by linear regression equation

Abstract The area of individual leaves in oil palm has been conventionally estimated from a regression equation that is based on the size and number of leaflets. The aim of the present study is to verify the accuracy of this equation, which became standard in oil palm research. Therefore, true leaf area, measured with a video camera, was estimated from the product of number of leaflets per leaf (n) with mean length (l) times mid-width (w) of six of the longest leaflets (nlw). The database was assembled, annually for the first 4 years after planting, from 2961 leaves of dura × pisifera testcrosses descending from six distinct pisifera origins. The regression coefficients of the regression lines of nlw plotted against true area did not show a trend with age of the palms or a difference among pisifera origins. The common regression equation fitted through all data of this study accurately estimated true leaf area of the testcrosses and also the areas of 2- to 3.5-year-old dura palms of three distinct origins as well as 18-year-old tenera palms. These outcomes are at odds with the conventional regression equation that overestimates the true leaf areas by about 24%. A more recently-developed variant underestimates true area of the young tenera and dura palms by 28%, while overestimating true area of old tenera palms by 19%. Possible causes for these deviations from true area are discussed. The paper argues that parameters depending on leaf area of previous physiological studies need to be reassessed.