Hi-sAFe: A 3D Agroforestry Model for Integrating Dynamic Tree–Crop Interactions

Christian Dupraz, Kevin J. Wolz, Isabelle Lecomte, Grégoire Talbot, Grégoire Vincent, Rachmat Mulia, François Bussière, Harry Ozier‐Lafontaine, Sitraka Andrianarisoa, N. D. Jackson, Gerry Lawson, Nicolas Donès, Hervé Sinoquet, Betha Lusiana, Degi Harja, Susy Domenicano, Francesco Reyes, Marie Gosme, Meine van NoordwijkSustainability 2019 · DOI

TL;DR

This paper introduces Hi-sAFe, a novel 3D biophysical model that simulates the complex interactions between trees and crops in agroforestry systems. By integrating a detailed tree growth model with a crop model, Hi-sAFe allows researchers to explore various agroforestry designs and management strategies, ultimately aiming to promote sustainable land-use practices.

What Problem Does This Paper Address?

Agroforestry, the practice of intentionally integrating trees with crops or livestock, offers numerous ecological and economic benefits. However, understanding the intricate dynamics of tree-crop interactions has historically relied on time-consuming and expensive field experiments. These experiments are often limited by slow tree growth, funding constraints, and the difficulty of capturing the full complexity of these systems over long periods. Existing models have struggled to adequately represent the spatial and temporal heterogeneity of agroforestry, particularly the dynamic competition for resources like light, water, and nitrogen. This gap hinders the ability to design and optimize agroforestry systems for diverse environments and management goals.

Methods

Hi-sAFe is a mechanistic, process-based model built on a 3D, spatially explicit framework. It couples the established STICS crop model with a new tree model called sAFe-Tree. The integration occurs through three interaction modules that govern light, water, and nitrogen dynamics. Key methodological innovations include:

  • 3D Spatial Representation: The model divides the agroforestry scene into a grid of "cells," each containing homogeneous crop and potentially one tree. These cells are further discretized into "voxels" to capture belowground soil structure and resource distribution. This 3D approach is crucial for accurately simulating competition and facilitation processes.
  • Dynamic Tree Plasticity: The sAFe-Tree model incorporates plasticity mechanisms that allow trees to respond dynamically to resource availability. This includes adjustments in aboveground architecture (e.g., crown shape) and belowground root growth in response to competition for light, water, and nitrogen.
  • Integrated Resource Competition: The model simulates competition for light using a ray tracing algorithm and for water and nitrogen using algorithms adapted from existing models (WaNuLCAS), but extended to a 3D context.
  • Daily Time-Step Simulation: Hi-sAFe operates on a daily time-step, allowing for the simulation of dynamic processes throughout the entire lifetime of an agroforestry system, from planting to harvest.
  • Management Interventions: The model can simulate various management practices, including fertilization, irrigation, tillage (via STICS), and specific tree interventions like thinning, branch pruning, and root pruning.

Datasets

Not discussed in the paper. The paper focuses on the model's development and capabilities, not on specific datasets used for training or validation.

Key Insights and Ideas

  • The Importance of 3D Spatial Dynamics: The paper strongly emphasizes that accurately representing agroforestry requires a 3D, spatially explicit approach. Simple 1D or 2D models can miss crucial interactions, especially those occurring belowground or in complex canopy structures.
  • Tree Plasticity is Key: The model's ability to simulate tree plasticity in response to resource availability is highlighted as a critical factor in understanding overyielding and niche complementarity in agroforestry. Trees aren't static entities; their growth and resource acquisition adapt.
  • Bridging the Gap Between Experimentation and Prediction: Hi-sAFe offers a powerful way to extrapolate findings from limited field experiments to a wider range of scenarios, saving time and resources in exploring new designs and management strategies.
  • Facilitating Sustainable Land Use: By providing a tool to better understand and optimize agroforestry systems, the model aims to accelerate their adoption as a sustainable land-use practice.

Key Results

The paper primarily describes the model's architecture, components, and capabilities rather than presenting specific simulation results. However, it outlines the types of analyses Hi-sAFe can perform:

  • Simulation of Long-Term Dynamics: The model can simulate the entire lifetime of agroforestry systems, from tree planting to harvest, on a daily time-step.
  • Calculation of Land Equivalent Ratio (LER): By simulating both agroforestry systems and monocultures, Hi-sAFe can calculate LER to quantify land-use efficiency.
  • Exploration of Design and Management: The model can be used to explore the impact of various factors, including tree spacing, crop type, tree row orientation, thinning, pruning, fertilization, and irrigation.
  • Environmental Response Analysis: Hi-sAFe can assess system responses to environmental variations like climate change, soil depth, and fluctuating water tables.
  • Example Outputs: Figure 7 provides illustrative examples of model outputs, showing daily and annual fluxes of carbon, water, and nitrogen, as well as tree carbon storage over time. Figure 8 shows example illustrations of relative yield, light capture, water uptake, and nitrogen uptake contributions.

Limitations and Future Work

The authors acknowledge several limitations and suggest future research directions:

  • Tree Architecture Simplification: Tree crowns are approximated as simple ellipsoids, lacking explicit representation of individual branches. This limits the detailed simulation of light interception and crown interactions. Future work could involve adding more flexible crown shapes.
  • Nitrogen Fixation: The model currently does not include algorithms for nitrogen fixation by trees, which is a significant process in some agroforestry systems.
  • Fruit/Nut Production: While the model can simulate resource allocation, it doesn't explicitly model fruit or nut production, which can be a major sink for carbon and nitrogen in certain systems.
  • Environmental Sensitivity: Photosynthetic assimilation is not sensitive to atmospheric CO2 concentration or leaf temperature for trees (though it is for crops). Phenological responses to heat and drought are also simplified.
  • Crown Plasticity: While crowns stop growing when they touch, the model lacks anticipatory responses to shading (e.g., increased height growth). Crown interactions are currently limited to single-tree scenes.
  • Root Model Validation: The root growth model, while flexible, has primarily been validated using potted plants. Further field validation, especially in agroforestry settings, is needed.
  • Carbon Assimilation Algorithm: The current empirical model for carbon assimilation is sensitive to water and nitrogen stress multiplicatively. Alternative, simpler models (e.g., minimum of stresses) or more complex representations could be explored.
  • Respiration: Tree respiration is not explicitly accounted for, limiting some climate-impact investigations.
  • Crop-Tree Interactions: The model doesn't fully capture the impact of tree shade on crop temperature-humidity dynamics or wind reduction by trees.
  • Harvesting: Crop harvesting is managed independently for each cell, which may not reflect real-world scenarios where entire fields are harvested simultaneously.
  • Soil Carbon Sequestration: The STICS model's limitations in simulating deep soil carbon fate restrict Hi-sAFe's ability to model long-term soil carbon sequestration.
  • Model Complexity: The complexity of the model can make results difficult to interpret and communicate. Parameter sensitivity analysis is also noted as a challenge.

Future work could involve parameterizing for more tree species, updating algorithms as agroforestry science advances, enhancing analytical tools, and improving computational performance.

Summary updated

Abstract

Agroforestry, the intentional integration of trees with crops and/or livestock, can lead to multiple economic and ecological benefits compared to trees and crops/livestock grown separately. Field experimentation has been the primary approach to understanding the tree–crop interactions inherent in agroforestry. However, the number of field experiments has been limited by slow tree maturation and difficulty in obtaining consistent funding. Models have the potential to overcome these hurdles and rapidly advance understanding of agroforestry systems. Hi-sAFe is a mechanistic, biophysical model designed to explore the interactions within agroforestry systems that mix trees with crops. The model couples the pre-existing STICS crop model to a new tree model that includes several plasticity mechanisms responsive to tree–tree and tree–crop competition for light, water, and nitrogen. Monoculture crop and tree systems can also be simulated, enabling calculation of the land equivalent ratio. The model’s 3D and spatially explicit form is key for accurately representing many competition and facilitation processes. Hi-sAFe is a novel tool for exploring agroforestry designs (e.g., tree spacing, crop type, tree row orientation), management strategies (e.g., thinning, branch pruning, root pruning, fertilization, irrigation), and responses to environmental variation (e.g., latitude, climate change, soil depth, soil structure and fertility, fluctuating water table). By improving our understanding of the complex interactions within agroforestry systems, Hi-sAFe can ultimately facilitate adoption of agroforestry as a sustainable land-use practice.
BibTeX
@inproceedings{dupraz2019hi,
  author = {Christian Dupraz and Kevin J. Wolz and Isabelle Lecomte and Grégoire Talbot and Grégoire Vincent and Rachmat Mulia and François Bussière and Harry Ozier‐Lafontaine and Sitraka Andrianarisoa and N. D. Jackson and Gerry LAWSON and Nicolas Donès and Hervé Sinoquet and Betha Lusiana and Degi Harja and Susy Domenicano and Francesco Reyes and Marie Gosme and Meine van Noordwijk},
  title = {{Hi-sAFe: A 3D Agroforestry Model for Integrating Dynamic Tree–Crop Interactions}},
  booktitle = {Sustainability},
  year = {2019},
  doi = {10.3390/su11082293},
}
Figure