The Unrecognized Determinants of Stalk Strength.

The many determinants of stalk / stem strength are difficult to precisely control in experimental studies. Consequently, the relative contribution of numerous physiological features to stalk strength remains ambiguous.  These limitations can be overcome, in part, with advanced in silico models.  Not only do in silico models operate on a much reduced time scale as compared to experimental studies but they also allow precise, independent control of all modeled plant features, thus removing confounding effects.   

To investigate the influence of physiological features on stalk strength, an in silico structural engineering model of maize stalk was developed. Model morphology was acquired from high resolution micro CT scanning. A sensitivity analysis was performed to investigate the relationships between model inputs (morphology, tissue properties) and model outputs (mechanical stress). This analysis resulted in 68 unique model simulations.

Changes in stalk geometry were 18 times more influential on stalk mechanical stresses than changes in tissue properties.  The analysis also identified several previously unrecognized morphological features of corn stalk act as mechanical stress concentrators that can cause the plant to break prematurely.  These features include but are not limited to divots associated with undeveloped brace roots, sharp changes in diameter occurring at the node line, voids in the pith tissue, and altered surface roughness of rind tissue. Further analysis confirmed that small modifications to local stalk morphology near the node (i.e. smoothing rapid geometric changes) can significantly increase overall plant strength in the absence of modifications to stalk diameter, rind thickness, or tissue properties. These findings are corroborated by additional engineering analyses and recent experimental studies of maize stalk lodging.

Detailed results from the preceding study will be presented and the pros and cons of in silico modeling techniques as they pertain to plant physiology and hypothesis generation and testing will be discussed.