Traditional agricultural models often treat a crop as a "black box" that converts sunlight and water into a single value: total biomass. While useful for predicting field-level yields, these models fail to explain the complex, adaptive architecture of the plants themselves. Why does a tree branch more in the sun? Why does a wheat plant prioritize roots during a drought?
To answer these questions, researchers like Philippe De Reffye and Véronique Letort have developed Functional-Structural Plant Models (FSPMs), most notably the GreenLab framework. These models move beyond simple accumulation, simulating the internal competition for resources that dictates how a plant builds itself.
The Source-Sink Engine
At the heart of this research is the concept of source-sink dynamics. In this framework, a plant is viewed as a collection of competing "sinks"—organs like leaves, stems, roots, and fruits—that draw from a common "source" pool of biomass generated by photosynthesis.
As described in A morphogenetic crop model for sugar-beet by Sébastien Lemaire and colleagues, biomass is not distributed evenly. Instead, each organ has a specific "sink strength," a mathematical weight that determines its priority in the resource queue. This mechanism allows the model to simulate how a sugar beet decides to "invest" in its sugar-storing root versus its light-harvesting leaves.
The Q/D Ratio: The Plant's Regulatory Signal
The most critical innovation in this field is the use of the biomass supply-to-demand ratio (Q/D) as a regulatory signal.
- Q (Supply): The amount of biomass produced by photosynthesis in a growth cycle.
- D (Demand): The sum of the sink strengths of all expanding organs.
In A dynamical model of plant growth with full retroaction between organogenesis and photosynthesis, the authors demonstrate that this ratio acts as a "brain" for the plant. When $Q/D$ is high, the plant has a resource surplus, triggering the development of new structural units (metamers) or lateral branches. When $Q/D$ is low, the plant enters a conservation mode, limiting new growth to ensure existing organs survive.
This feedback loop explains morphological plasticity—the ability of a plant to change its shape in response to its environment. For example, Amélie Mathieu and Yuntao Ma showed in The Influence of Photosynthesis on the Number of Metamers that by making organ initiation dependent on the $Q/D$ ratio, a model can autonomously adapt its branching pattern to varying light levels without needing pre-programmed geometric rules.
Emergent Rhythms and Oscillations
One of the most striking proofs that this feedback mechanism works is its ability to produce emergent biological behaviors. In Oscillations in Functional Structural Plant Growth Models, researchers found that the competition between vegetative growth (leaves/stems) and reproductive growth (fruits) naturally creates cyclic patterns.
If a plant produces too many fruits, the total demand ($D$) spikes, causing the $Q/D$ ratio to crash. This starves the vegetative buds, preventing new leaves from forming. Without new leaves, future supply ($Q$) drops, eventually leading to a year with no fruit—a phenomenon known in forestry as "alternate bearing." These oscillations emerge purely from the internal resource math, rather than external environmental triggers.
From Theory to Optimization
By capturing the "why" of plant growth, these models become powerful tools for optimization. Because the models link environmental inputs (like water) to specific organ development, researchers can use them to find the most efficient ways to grow crops.
- Irrigation: Lin Wu and Bao-Gang Hu used GreenLab to discover irrigation schedules that maximize fruit yield by precisely timing water delivery to match the plant's internal demand cycles (A Water Supply Optimization Problem).
- Wood Production: Rui Qi and Thierry Fourcaud applied Particle Swarm Optimization to the model to determine the ideal balance between leaf area and wood accumulation for stable, high-volume timber growth (Application of the GreenLab model to simulate and optimize wood production).
- Breeding: By linking genetic markers (QTLs) to these "hidden" model parameters like sink strength, researchers can design "virtual ideotypes"—theoretically perfect plant architectures for specific climates (Quantitative Genetics and Functional-Structural Plant Growth Models).
Ultimately, this research direction treats the plant not as a static object, but as a dynamic system that constantly negotiates its own form through the language of resource supply and demand.