Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: November 17, 2006
Publication Date: November 17, 2006
Citation: Jaradat, A.A., Surek, D., Archer, D.W. 2006. Phenotypic plasticity and fractal dimension are strong determinants of grain yield in soybean [abstract]. PMA06: The Second International Symposium on Plant Growth Modeling, Simulation, Visualization and Applications. p. 40. Technical Abstract: The normalized geometric structure and vegetative growth characteristics of two genotypically different soybean [Glycine max (L.) Merr.] varieties were quantified using digital imagery during two cropping seasons under eight combinations of two management systems. Midday differential canopy temperature (dT) was the single most important environmental factor in predicting the fractal dimension (FD) of both varieties (R2 range 0.40 - 0.76) and was a reliable indicator of plant stress under different management systems. A multilayer perception neural network with back propagation identified plant dry weight, plant volume, plant circularity, and leaf area per plant in decreasing order as reliable (R2=0.76) predictors of FD. However, the simplest neural network model accounted for 61.0% of the variation in FD and was limited to plant dry weight and number of pods per plant, the latter is an estimate of the number of fruiting nodes per plant. The fractal dimension was the most important predictor in a generalized regression neural network followed in decreasing order by plant dry weight, plant volume and plant circularity, in predicting grain yield m-2 (R2=0.64). The management implications of manipulating phenotypic plasticity and FD to optimize grain yield are presented.