VALIDATION OF THE PRINCIPAL AXIS MODEL (PAM) AND ITS APPLICATION TO GENOTYPE SELECTION IN-FIELD PEA (PISUM-SATIVUM L) CROPS

The principal axis model (PAM) uses a principal axis and an ellipse to characterize the variation in the relationship between the seed (SWT) and plant (PWT) weights of individual plants within a crop. The theor etical linkage between the magnitude and variability of plant harvest index (PHI), and therefore seed yield per unit area, and changes in th e Components of the PAM was examined using data from four field pea (P isum sativum L.) genotypes sown at 9, 49, 100, 225 and 400 plants m(-2 ). A strong linear relationship (R-2 greater than or equal to 93.8%) b etween SWT and PWT and a negative SWT-axis intercept were confirmed fo r all crops. Analyses indicated that decreased variability of PHI with in a crop would result from selection to: (a) increase the SWT-axis in tercept of the PAM; (b) increase the slope of the PAM; (c) optimize th e ellipse location; and (d) minimize the deviation around the principa l axis. The first three methods were used to explain yield differences (P < 0.05) among genotypes of different populations. A potential stra tegy for single plant selection based on the PAM is proposed. This may enable early generation (F-4) selection of small, high performing pla nts that may be ideal crop ideotypes. A theoretical example of the str ategy is presented, with differences among selections based on the PAM , SWT or harvest index highlighted. (C) 1997 Annals of Botany Company.