MARSHAL

MARSHAL simulates maize root system hydraulic architectures to link detailed root architectural traits with water flow dynamics for virtual phenotyping and breeding.

Key Features:

• Model integration: Combines CRootBox (Schnepf et al., 2018) for simulating root architecture with algorithms for solving water flow in root system hydraulic architectures (RSHA) as described by Meunier et al. (2017).
• Macroscopic parameter estimation: Implements methods from Couvreur et al. (2012) to compute macroscopic parameters that link local root traits to plant-scale interpretations.
• Dynamic hydraulic-architectural representation: Represents dynamic hydraulic and architectural properties parameterized specifically for maize to capture architectural and functional root system traits.
• Inverse optimization and sensitivity analysis: Supports inverse optimizations and sensitivity analyses to identify key parameters affecting root system conductance, uptake depth, volume, and convex hull.
• Ensemble generation and validation: Generates ensembles of diverse hydraulic architectures for comparison with observed architectural (root length density) and hydraulic (root system conductance) measurements.

Scientific Applications:

• Virtual phenotyping and breeding: Enables identification and optimization of maize root architectural and hydraulic traits for virtual breeding studies.
• Scaling from local to plant scale: Translates local root hydraulic and architectural traits into macroscopic parameters for plant-scale analysis.
• Sensitivity and parameter analysis: Quantifies multivariate sensitivities of mature root system conductance, mean depth of uptake, root system volume, and convex hull to input parameters.
• Model validation and comparison: Facilitates comparison of simulated architectures with observed root length density and root system conductance for validation of root and soil-root system models.

Methodology:

MARSHAL uses a dynamic hydraulic-architectural model parameterized for maize, combines CRootBox for architecture with RSHA water-flow solvers (Meunier et al., 2017), applies methods from Couvreur et al. (2012) to compute macroscopic parameters, and employs ensemble runs, sensitivity analyses, and inverse optimizations as described in the source material.