📄 agriculture

Beyond the Breast Height: Leveraging Stump Diameter as a Robust Predictor of Biomass and Carbon Stocks in Tropical Acacia Plantations

This study demonstrates that stump diameter measured at 30 cm above ground serves as a robust and practical predictor for estimating biomass and carbon stocks in *Acacia* plantations in Bangladesh, offering a viable alternative to conventional breast-height measurements for effective post-harvest monitoring and climate change mitigation.

Original authors: Niamjit Das, Md. Qumruzzaman Chowdhury

Published 2026-07-29

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Niamjit Das, Md. Qumruzzaman Chowdhury

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Forest's Hidden Ledger

Imagine the world's forests as giant, living bank accounts. Every tree is a deposit, storing carbon (a key ingredient in the air that warms our planet) in its wood, leaves, and roots. Scientists and governments need to know exactly how much "money" is in these accounts to fight climate change and protect nature. To do this, they usually play a game of "measure and guess." They measure a tree's width at a specific height—about as high as a person's chest, known as "breast height"—and use that number to calculate how much wood and carbon the whole tree holds. It's like guessing the size of a watermelon just by measuring its middle.

But what happens when the tree is gone? If a forest is dense and tangled, or if the trees have already been cut down for timber, you can't measure the chest anymore. The tree is either too crowded to reach or just a pile of logs on the ground. This leaves a big gap in our bank ledger: we can't count the carbon if we can't measure the tree. That's where a new idea comes in: looking at the stump. The stump is the base of the tree left in the ground after a harvest. The big question for scientists is simple: Can we measure the width of this stump and use it to figure out how big the whole tree was, even if we never saw the top?

The Stump Detective: Solving the Forest's Missing Puzzle

In a recent study, researchers Niamjit Das and Md. Qumruzzaman Chowdhury decided to play detective in the plantations of Bangladesh. They were investigating three types of fast-growing trees: Acacia auriculiformis, Acacia mangium, and a special "hybrid" mix of the two. These trees are like the sprinters of the forest world, growing quickly to provide wood and soak up carbon. The team wanted to see if they could swap the traditional "chest measurement" for a "stump measurement" without losing accuracy.

To solve the mystery, they didn't just guess; they went full forensic. They selected 1,050 trees across ten different sites. For 75 of these trees, they took the drastic step of cutting them down, chopping them into sections, and drying them in an oven at 105°C until they were bone dry. They weighed every single piece to get the exact amount of biomass (wood weight). This gave them the "truth" to compare against their measurements. For the rest of the trees, they measured the stump diameter (the width of the stump 30 cm above the ground) and the traditional chest height (DBH).

The researchers then built a digital toolbox filled with 14 different mathematical models. Some were simple straight lines, some were curved, and others were fancy computer programs called "machine learning" that try to find patterns in data. They asked the models: "If you only know the stump size, can you tell us the tree's total weight?"

The results were a happy surprise. The study found that the stump diameter is a super-reliable stand-in for the chest measurement. When they used a specific type of math (called a log-log regression), the models were incredibly accurate, with a success rate (R²) of over 0.98. This means the stump size predicted the tree's size almost perfectly. In fact, the stump-based models worked just as well as the traditional chest-based ones.

However, the study also tested some high-tech "smart" models. One computer program, called Random Forest, agreed with the researchers that the stump was the most important clue. But another program, called Gradient Boosting, got a little too excited about the data and started to "overfit"—basically, it memorized the specific trees it studied so well that it couldn't guess correctly about new ones. This taught the researchers that while fancy computers are cool, sometimes a simple, sturdy formula is better.

The team also discovered that the hybrid Acacia trees were the heavyweights of the group, storing the most carbon per hectare, followed by A. mangium and then A. auriculiformis.

Why does this matter? Because this method is a game-changer for forests that have already been harvested. If a forest is cut down, you can't measure the chest anymore, but the stumps are still there. By measuring the stumps, foresters can now quickly and accurately calculate how much carbon was removed or how much biomass was harvested. This helps countries keep better track of their "carbon bank accounts," which is crucial for global climate goals. The study suggests that this simple trick—measuring the stump instead of the chest—could be a powerful, practical tool for managing tropical forests and fighting climate change, turning a pile of leftover wood into a clear picture of the forest's past and its future.

Technical Summary: Beyond the Breast Height – Leveraging Stump Diameter as a Robust Predictor of Biomass and Carbon Stocks in Tropical Acacia Plantations

Problem Statement
Accurate estimation of tree biomass is fundamental to tropical forest management, climate change mitigation, and REDD+ monitoring. Conventional inventories rely heavily on Diameter at Breast Height (DBH). However, DBH measurement is often impractical or impossible in specific contexts, such as dense stands with obstructed undergrowth, trees with buttressed stems, or, most critically, in post-harvest scenarios where trees have been felled. In these situations, stump diameter (measured at 30 cm above ground) is often the only available metric. Despite its potential utility, stump diameter has been poorly studied as a systematic predictor for biomass in tropical plantation forestry, particularly for Acacia species, leading to a reliance on DBH-based models that cannot be applied in post-harvest monitoring.

Methodology
The study was conducted across ten plantation sites in the Sylhet Division of Bangladesh, focusing on three taxa: Acacia auriculiformis, Acacia mangium, and their hybrid (A. auriculiformis × A. mangium).

Sampling: A stratified random sampling design was employed, selecting 1,050 trees (350 per species) to cover a wide range of size classes.
Destructive Sampling: To calibrate biomass equations, 75 trees (25 per species) were destructively sampled. These trees were felled, sectioned, oven-dried at 105°C to constant mass, and weighed to determine precise stem biomass.
Measurements: For all sampled trees, stump diameter ( $d_{st}$ at 30 cm), DBH (at 1.3 m), total height, and stem height were recorded.
Modeling Approaches: The researchers evaluated fourteen models, including:
- Regression Models: Linear, log-log (ln–ln), and polynomial regressions relating stump diameter to DBH and stem biomass.
- Mixed-Effects Models: Used to account for random variations among species and sites.
- Machine Learning: Random Forest and Gradient Boosting algorithms were implemented to capture non-linear relationships and interactions.
Validation: A rigorous two-level validation strategy was used: (1) independent validation using a dataset of 975 trees not used in calibration, and (2) five-fold cross-validation stratified by species. Performance was assessed using adjusted $R^2$ , RMSE, AIC, Durbin-Watson statistics, and paired t-tests.
Carbon Estimation: Stem biomass was converted to carbon stocks using species-specific conversion factors (0.475 for A. auriculiformis, 0.480 for A. mangium, and 0.485 for hybrid Acacia).

Key Results

Stump Diameter as a Proxy: Stump diameter demonstrated a highly stable and strong correlation with DBH across all three taxa. The log-log (ln–ln) regression model yielded adjusted $R^2$ values greater than 0.98, with low RMSE (0.048–0.074) and unbiased residuals.
Biomass Prediction: Biomass models based on stump diameter performed comparably to, and in some cases better than, conventional DBH-based models. The parsimonious ln–ln stump-based model achieved adjusted $R^2$ values between 0.945 and 0.968.
Model Performance:
- Mixed-Effects Models: These reduced bias significantly, particularly for the hybrid Acacia, improving generalizability across sites.
- Machine Learning: Random Forest consistently identified stump diameter as the most important predictor of stem biomass. In contrast, Gradient Boosting showed signs of overfitting, despite high apparent accuracy, highlighting the need for caution when applying complex algorithms in forestry.
Validation: Independent validation and cross-validation confirmed that stump-based predictions were unbiased (p > 0.05) with narrow confidence intervals and low standard errors of estimate.
Carbon Stocks: Species-specific estimates indicated that Hybrid Acacia was the most productive in terms of carbon stock, followed by A. mangium and A. auriculiformis.

Significance and Claims
The authors claim that this study provides a practical, scalable tool for rapid biomass assessment and carbon accounting, specifically addressing the gap in post-harvest monitoring where DBH is unavailable. By validating stump diameter as a robust surrogate for DBH, the research offers a method to reconstruct stand biomass and carbon stocks after harvesting, which is critical for Monitoring, Reporting, and Verification (MRV) under REDD+ and IPCC Tier 2/3 frameworks.

The study asserts that its findings support global climate and biodiversity goals, specifically SDG 13 (Climate Action) and SDG 15 (Life on Land), by enabling effective monitoring of carbon stocks in tropical plantation forestry. The integration of destructive sampling with advanced statistical and machine learning approaches establishes a methodological framework that allows for precise, efficient, and species-specific carbon accounting in diverse field conditions, including dense stands and harvested areas. The authors emphasize that while stump-based models are robust, the use of mixed-effects frameworks and careful selection of machine learning algorithms (favoring Random Forest over Gradient Boosting in this context) is essential to minimize bias and overfitting.

The Forest's Hidden Ledger

The Stump Detective: Solving the Forest's Missing Puzzle

Technical Summary: Beyond the Breast Height – Leveraging Stump Diameter as a Robust Predictor of Biomass and Carbon Stocks in Tropical Acacia Plantations

More like this