Forests play a foundational role in maintaining ecological balance, regulating the Earth’s climate, and supporting economic and social well-being. As major carbon sinks, they store vast amounts of atmospheric CO₂, harbor diverse species, and provide critical goods and services for both human societies and natural ecosystems. Gaining a deep understanding of forest growth is essential for evaluating ecosystem integrity, guiding disturbance management, and anticipating how forests will respond to ongoing climate change. Yet, accurately predicting forest dynamics remains a formidable challenge. This complexity arises from intricate ecological processes, feedback mechanisms, and uncertainties linked to shifting climatic conditions. Moreover, legacy effects, such as the long-term impacts of previous climate extremes, land use, or forest management practices, can obscure cause-and-effect relationships by producing delayed, nonlinear impacts on forest growth and resilience.
This complexity has led to the development of numerous forest growth models, each with distinct assumptions and representations of key processes. Many models struggle to replicate recent trends in growth decline and mortality linked to climate extremes, emphasizing the need to reassess modeling approaches and underlying theories. A recently published article in Current Forestry Reports (Boukhris et al., 2025; http://doi.org/10.1007/s40725-025-00249-5) provides a comprehensive review of both established and emerging theoretical frameworks that have shaped forest growth modeling over the past two decades. In addition to synthesizing advances made in understanding and simulating forest growth processes, the paper also highlights key challenges and outlines promising future directions for research in this field.
Evolution of Forest Growth Modeling
Historically, forest models have moved from empirical methods to mechanistic, process-based approaches grounded in conservation laws and causal relationships. While these models are praised for their robustness under non-stationary climates, many still fail to capture observed patterns of growth under stress and disturbance, revealing gaps in current understanding.
The Boukhris et al. (2025) review explores 18 representative forest growth models, integrating theoretical foundations, empirical data, and emerging tools like machine learning. It assesses their ability to simulate forest dynamics under changing environmental conditions and suggests new directions for future research.
Forest Growth Theories
Tree growth, defined as biomass accumulation, involves cell division, expansion, and differentiation. Net growth at the stand or landscape level incorporates both biomass gain and losses from mortality and recruitment. Several theories explain how plants regulate growth:
- Functional Balance Theory: Growth depends on balancing resource supply with plant tissue demands.
- Local Determination Theory: Growth patterns aim to optimize resource capture.
- Optimality Principles: Plants allocate resources to maximize long-term fitness and short-term physiological efficiency.
- Coordination Theory: Internal regulation coordinates multiple processes in response to environmental conditions.
Growth models generally follow two main paradigms: photosynthetic-centric (“source-driven”) and non-photosynthetic-centric (“sink-driven”).
Photosynthetic-Centric Models
These models link growth directly to carbon assimilation through photosynthesis, minus respiration costs. Growth results from net photosynthesis, adjusted for maintenance and construction of tissues. Environmental variables like temperature, light, and nutrient availability are critical drivers.
Examples include 3-PG, 3D-CMCC-FEM, LPJ-GUESS, SEIB-DGVM, and FATES. They use methods like Light Use Efficiency (LUE) or the Farquhar biochemical model to simulate photosynthesis. These models assume that limited photosynthesis constrains growth, reflecting the functional balance and local determination theories.
Non-Photosynthetic-Centric Models
These models emphasize the role of sinks, organs demanding resources, and how plants allocate resources based on physiological needs and stress responses. Growth is shaped not just by resource availability, but by the ability of tissues to use resources effectively. This approach supports optimality and coordination theories.
Models like ForClim and LandClim fall into this category. They simulate growth through empirical relationships rather than direct photosynthesis, capturing long-term dynamics in resource-limited environments.
Source-Sink Integration
Some models integrate both paradigms, adjusting dynamically between source-driven and sink-limited behaviors depending on environmental stress. For instance, under optimal conditions, photosynthesis may dominate, while under stress, growth may depend more on non-structural carbon reserves. This hybrid approach acknowledges that both supply and demand control growth and improves model adaptability.
Modeling Growth Across Scales
Models vary in complexity and scale, from local stand-level simulations to global terrestrial ecosystem representations. They differ in temporal resolution, spatial detail, and process representation. The 18 models reviewed cluster into three major groups:
1. Stand-Scale Models
These models simulate fine-scale processes within forest stands, often using high-resolution time steps (daily or sub-daily). They track individual trees or cohorts and may include species competition and succession. Examples: ForClim, 3PGmix, 3D-CMCC-FEM. Their spatial simplicity limits application to broader-scale forest heterogeneity.
2. Landscape-Scale Models
Operating at regional scales, these models integrate disturbances (e.g., fire, pests) and demographic processes (e.g., dispersal, establishment) across forest patches. They offer a balance between complexity and scalability but often simplify growth mechanisms. Examples: iLand, LandClim, LANDIS-II. Useful for regional forest planning, they are less accurate at simulating individual tree dynamics.
3. Terrestrial Ecosystem-Scale Models
These large-scale models (e.g., LPJ-GUESS, FATES, CLM) simulate biome shifts, land-use change, and climate feedback. They use mechanistic representations of photosynthesis and carbon allocation but aggregate vegetation types, limiting their ability to simulate detailed forest processes. They are critical for assessing global carbon budgets but are less reliable for local management.
Trade-Offs and Model Selection
Each modeling scale offers specific advantages and limitations. Stand-scale models provide high-resolution insights ideal for managed forests but lack landscape interactions. Landscape-scale models handle spatial heterogeneity and disturbances but simplify individual tree dynamics. Terrestrial models inform climate policy and global trends but cannot represent species-specific or stand-level dynamics. Selecting a model depends on the research goal, scale, and required detail.
Conclusion and Future Directions
Forest growth modeling must navigate immense complexity, from physiological theory to ecosystem-scale feedbacks. Climate change amplifies these challenges, with models struggling to predict emerging dynamics, legacy effects, and nonlinear responses.
However, integrating machine learning with process-based models offers a promising path forward. These tools can analyze vast datasets, uncover hidden patterns, and improve predictive power. Combining empirical learning with theoretical understanding may enhance model flexibility and ecological realism.
Future efforts should prioritize interdisciplinary collaboration, improved data integration, and the development of models that dynamically balance source-sink interactions. As climate change reshapes forest ecosystems, adaptive and scalable modeling approaches will be essential for informed decision-making and effective conservation.
