The core problem the dataset addresses is deceptively simple. Modern very high-resolution satellite imagery, combined with machine-learning image analysis, has become remarkably good at finding and outlining individual tree crowns from space. An analyst, or increasingly an algorithm, can delineate the shadowy green footprint of a single mango tree in a Malawian maize field or a lone acacia on Kenyan rangeland, and measure the area that crown projects onto the ground. But satellites cannot see through the canopy to the trunk. And the trunk, specifically the diameter at breast height, or DBH, is the single most important measurement in conventional forestry. Decades of research have produced well-established allometric equations that convert trunk diameter into estimates of aboveground biomass and, from there, into the carbon stored in wood. Without a trunk measurement, a remotely detected crown is just a shape; with one, it becomes a quantifiable store of carbon.

This measurement gap matters most precisely where Africa’s restoration future is unfolding. In landscapes dominated by trees outside forests, trees are spatially dispersed, which makes the conventional plot-based inventory approach expensive and difficult to scale. Walking transects to find and measure every scattered tree across thousands of hectares of farmland is simply not practical, and traditional national forest inventories were designed for closed-canopy forests, not mosaic agricultural landscapes. At the same time, ground datasets containing concurrent measurements of both crown dimensions and stem dimensions have remained scarce, particularly in Africa. That scarcity has left a methodological chasm: remote sensing scientists could map crowns, and foresters could measure trunks, but there was little empirical ground truth connecting the two for the kinds of trees that dominate African restoration landscapes.

The new dataset was built to be that bridge. Field teams working across restoration landscapes in all four countries measured each tree following a single standardized protocol, recording the diameter at breast height, two perpendicular crown diameters, total tree height, species identity, geographic location, and identifiers linking each tree to its plot and broader landscape. The sampled landscapes deliberately span the range of systems where trees outside forests occur, including agroforestry, reforestation plantings, assisted natural regeneration, and other restoration and land-management contexts. After quality control, the final dataset contains 3,493 individual trees, each carrying the full suite of concurrent crown and stem measurements. Crucially, the field measurements were designed from the outset for spatial linkage with individual tree crowns delineated from very high-resolution satellite imagery, so that each ground-measured tree can serve as a calibration point for translating remotely sensed crown projected area into an estimate of stem diameter.

To demonstrate the dataset’s power, the researchers fitted power-law transfer functions relating DBH to crown area and evaluated how well these models predicted trunk diameter across sites and countries. The results were statistically robust: every site-specific relationship was significant at the p < 0.001 level, with coefficients of determination, or R² values, ranging from 0.44 to 0.85 and generally small prediction bias. In practical terms, this means that a measurement of a tree’s crown area, obtainable from a satellite image, can be converted into a usable estimate of its trunk diameter, which can then be fed into established biomass equations. The chain of inference runs from pixels to crowns, from crowns to trunks, and from trunks to carbon, with each link anchored in field data rather than assumption.

Yet the study also delivers an important caution. The variation in model coefficients and predictive error among landscapes was substantial, revealing deep ecological and structural heterogeneity in how crown size relates to trunk size across Africa. A tree growing in the humid highlands of Rwanda may allocate its growth differently than a savanna species in Ghana or a farm-grown shade tree in Malawi, and those differences show up in the crown-to-stem relationship. The authors argue that this heterogeneity is precisely why geographically distributed calibration data are so valuable, and why relying on a single universal allometric relationship would be a mistake. A crown-to-stem equation fitted in one country cannot simply be exported to another without risking systematic error in carbon estimates, particularly when those estimates feed into carbon accounting and payment systems where accuracy carries financial consequences.

Beyond the site-specific models, the team demonstrated a hierarchical synthesis of 19 site-specific transfer functions, showing how the data can be used to derive a generalized crown-to-stem relationship while still retaining information about variation among landscapes. This hierarchical approach offers a template for how the field can move forward: rather than choosing between one global equation and dozens of disconnected local ones, researchers can build models that share strength across sites while honestly representing uncertainty. For monitoring programs, that means carbon estimates for scattered trees can improve as more calibration landscapes are added, without discarding what has already been learned elsewhere.

The implications extend well beyond the models demonstrated in the paper. The dataset provides a reusable observational resource for developing and validating alternative crown-to-stem and crown-to-biomass models, for analyzing structural and species-level variation in African trees, for ecological scaling studies, and for calibrating individual-tree remote-sensing products. It also maintains compatibility with established forest mensuration and biomass estimation methods, meaning that carbon estimates derived through this crown-based pathway remain legible to the national and international reporting frameworks that govern climate finance. For countries participating in AFR100, many of which have pledged to restore tens of millions of hectares, the ability to monitor progress and verify carbon storage in landscapes where trees are dispersed rather than clustered could transform what is reportable and, ultimately, what is fundable.

There is a broader scientific shift embedded in this work as well. The framework extends individual-tree allometric scaling from the field plot to the landscape-scale tree census, a transition made possible only recently by the combination of sub-meter satellite imagery and machine-learning crown delineation. Where earlier generations of carbon mapping relied on coarse pixels and forest-area multipliers, the emerging paradigm treats every detectable tree as an individual entry in a landscape-wide inventory. Datasets like this one supply the empirical calibration that makes such censuses scientifically defensible. As the authors note, the approach extends monitoring, reporting, and verification capacity across African landscapes while keeping one foot firmly planted in the proven methods of traditional forestry.

For a continent racing to meet restoration commitments under climate pressure, the timing could hardly be better. The dataset, which the authors have also deposited in a public repository for reuse, turns a fundamental limitation of satellite remote sensing into a solvable calibration problem. Satellites will keep seeing crowns; with this new empirical bridge, scientists can now estimate what those crowns conceal, tree by tree, across the farms, rangelands, and regenerating landscapes where Africa’s restoration story is actually being written.

Subject of Research: Crown–stem allometry of African trees outside forests for satellite-based biomass and restoration monitoring

Article Title: An African Tree Crown–Stem Allometry Dataset for Monitoring Trees Outside Forests and Landscape Restoration

Article References: Alimo, T., Skole, D., Samek, J., Ndalowa, D., Oeba, V., Kamoto, J., Chioza, A., Papa, C., Spore, J., Levin, D., Brandt, J., & Tumeo, T. (2026). An African Tree Crown–Stem Allometry Dataset for Monitoring Trees Outside Forests and Landscape Restoration. https://doi.org/10.5194/essd-2026-756

Image Credits: AI Generated

DOI: 10.5194/essd-2026-756

Keywords: trees outside forests, allometry, remote sensing, carbon accounting, AFR100, forest landscape restoration, crown diameter, DBH, satellite imagery, Africa, biomass estimation, dataset

The research, published in Discover Forests by Niamjit Das of Shahjalal University of Science and Technology and Md. Qumruzzaman Chowdhury, examined three of the most widely planted timber taxa in South and Southeast Asia: Acacia auriculiformis, Acacia mangium, and the interspecific hybrid between them. Working across ten plantation sites in Sylhet Division, a humid monsoon landscape receiving roughly 5,000 millimeters of rain each year, the team measured 1,050 trees spanning ages of 8 to 15 years and stand densities of 1,100 to 1,400 trees per hectare. Crucially, they did not stop at tape measurements. They felled 75 trees, 25 of each taxon, cut the stems into sections, and dried them in ovens at 105 degrees Celsius until their weight stopped changing, producing the kind of ground-truth biomass data that allometric modeling depends on.

The central finding is that stump diameter, measured just 30 centimeters above the ground, predicts breast-height diameter almost perfectly. A simple log-log regression achieved adjusted R-squared values exceeding 0.98 across all three taxa, with root mean square errors as low as 0.048. In practical terms, this means a forester walking through a recently harvested compartment can measure the remaining stumps and reconstruct the diameter distribution of the trees that once stood there with negligible error. The relationship makes intuitive biological sense: the base of the stem and the breast-height section are connected by the tree’s taper, and within a given species and growth environment that taper follows a consistent geometric pattern.

Biomass prediction proved nearly as strong. Stump-based models performed comparably to the conventional DBH-based equations, with the parsimonious log-log stump model reaching adjusted R-squared values between 0.945 and 0.968 for stem biomass. The authors were careful to emphasize validation statistics rather than goodness-of-fit alone, a discipline that forest scientists have repeatedly called for after decades of unreliable generic equations being applied across unrelated ecosystems. Observed and predicted values clustered tightly around the one-to-one line, paired t-tests found no significant bias, and residuals were balanced across the size spectrum. For the hybrid Acacia, mixed-effects models that accounted for random variation among species and sites reduced prediction bias even further.

The study also put machine learning to the test, and the results offer a cautionary tale for the current enthusiasm surrounding artificial intelligence in environmental science. Random Forest, an ensemble method that averages hundreds of decision trees, consistently identified stump diameter as the strongest predictor of stem biomass and produced stable, generalizable predictions under five-fold cross-validation. Gradient Boosting, by contrast, assigned the highest relative importance to DBH and showed a telltale 12 percent gap between training and validation accuracy, the classic signature of overfitting. The authors are refreshingly candid about what this means: stump diameter’s dominance in the Random Forest reflects its strong correlation with overall tree size, not some newly discovered biological law, and variable importance rankings can shift with model settings and data structure.

Converting biomass into carbon introduced a species-specific dimension that generic accounting often misses. Rather than applying the default IPCC carbon fraction, the researchers used conversion factors drawn from published Acacia studies: 0.475 for A. auriculiformis, 0.480 for A. mangium, and 0.485 for the hybrid, reflecting real differences in wood chemistry among the taxa. The resulting carbon stock estimates ranked the hybrid Acacia highest, followed by A. mangium and then A. auriculiformis, a pattern consistent with the hybrid’s reputation for superior growth and wood quality. The authors stress that these figures cover stem carbon only, excluding branches, foliage, and roots, and therefore should not be read as total ecosystem carbon stocks.

The methodological rigor underlying these numbers deserves attention. Measurements were restricted to the dry season to minimize moisture-related variability, instruments were calibrated daily, and any readings differing by more than 0.5 centimeters in diameter or 0.2 meters in height triggered re-measurement. Oven-drying continued for 48 to 72 hours with weights checked at 12-hour intervals, and data entry was cross-checked by two independent researchers. The validation design was two-tiered: 975 trees held back from model development served as an independent test set, and five-fold cross-validation stratified by species and site confirmed that the models were not merely memorizing the quirks of particular plantations. This is precisely the kind of quality assurance that earlier reviews of forest biomass estimation have found lacking in much of the published literature.

Why does this matter beyond Bangladesh? Tropical forests hold nearly a third of terrestrial carbon, and programs such as REDD+ depend on credible, verifiable measurements of how much carbon is stored and, crucially, how much is removed when trees are harvested. Under current practice, post-harvest carbon accounting is notoriously difficult because the trees that would provide the measurements are gone. Stump-based equations change that calculus entirely: the stumps remain in the ground, they are easy to locate and measure, and they now carry enough information to reconstruct the stand’s biomass and carbon content. The authors note that this capability could inform measurement, reporting, and verification frameworks under REDD+ and support IPCC Tier 2 and Tier 3 approaches, though they are careful to say that broader policy applications require independent validation.

The limitations are stated with unusual honesty. The study covered healthy trees in even-aged Acacia plantations on fertile alluvial loams in a single region; damaged or suppressed trees were excluded, which may limit applicability in operational forestry where such trees are common. Buttressed stems and mixed-species stands remain challenging, and the models were calibrated on trees aged 8 to 15 years, so extrapolation to older rotations or different environments is untested. The improvements offered by mixed-effects and machine learning approaches, while statistically significant, were described as modest in operational terms, and the authors insist that DBH remains the conventional standard with stump diameter serving as a complementary surrogate rather than a replacement.

Even with those caveats, the study delivers something tropical forestry has needed for a long time: a cheap, fast, and demonstrably accurate way to audit carbon in the places where conventional measurement fails. As plantation forestry expands across the tropics to meet timber demand and climate goals under SDG 13 and SDG 15, the ability to turn a field of cut stumps into a credible carbon ledger could reshape how harvest impacts are monitored and reported. The next step, the authors argue, is testing these equations across wider environmental gradients, diverse plantation ages, and mixed-species stands, and incorporating belowground biomass into the accounting. Until then, the message to foresters and carbon accountants is clear: do not overlook the stump, because it remembers everything the tree knew.

Subject of Research: Stump diameter as a predictor of biomass and carbon stocks in tropical Acacia plantations in Bangladesh

Article Title: Stump diameter is a robust predictor of biomass and carbon stocks in tropical Acacia plantations

Article References: Das, N., & Chowdhury, M. Q. (2026). Stump diameter is a robust predictor of biomass and carbon stocks in tropical Acacia plantations. Discover Forests, 2(1), Article 73. https://doi.org/10.1007/s44415-026-00137-1

Image Credits: AI Generated

DOI: 10.1007/s44415-026-00137-1

Keywords: stump diameter, biomass estimation, carbon stocks, Acacia plantations, allometric models, REDD+, Bangladesh, machine learning, mixed-effects models, forest mensuration, destructive sampling, carbon accounting

The research team, led by Helen Yan, conducted their experiment on two fringing reef systems in the central Great Barrier Reef: Pioneer Bay on Orpheus Island, a sheltered no-take scientific research zone, and Indigo Bay on Fantome Island, a semi-exposed area open to most forms of recreational fishing. Separated by roughly ten kilometres, the two bays host distinct benthic and fish assemblages, making them ideal natural laboratories for testing whether census methods behave consistently across different communities. Between 11 and 17 February 2023, four trained observers surveyed fish larger than ten centimetres in total length, spanning eighteen reef fish families, with individuals identified to species, counted, and assigned to five-centimetre size bins.

The experimental design was deliberately rigorous. Each sampling block, performed by a single diver, comprised five surveys: an instantaneous one-minute point count, a continuous ten-minute point count, and belt transects of twenty, thirty and fifty metres, all using a five-metre-wide survey area. Point counts and transects within a block were conducted on non-overlapping reef areas to prevent the disturbance caused by one method from contaminating the other, and the order of methods was randomised to balance systematic diver effects. In total, the team completed 46 blocks on Orpheus Island and 48 on Fantome Island, with transect tape laid as the diver counted, a technique designed to minimise the fear response that reef fishes show toward approaching divers.

The results were striking. Point counts produced the most extreme estimates of abundance, biomass and species richness of any method tested. Instantaneous point counts recorded the lowest values, with median abundance estimates of just 0.16 individuals per square metre and standing biomass of 32.3 grams per square metre, while continuous ten-minute point counts recorded the highest, at 0.54 individuals and 93.1 grams per square metre. Transects of different lengths, by contrast, produced remarkably similar estimates, with abundance values clustering between 0.26 and 0.30 individuals per square metre regardless of whether the diver swam twenty, thirty or fifty metres. The discrepancy between the two point count variants points to a bias that is rarely recorded, let alone standardised, in reef fish censuses: total survey time. The longer a diver watches a patch of reef, the more fish swim into view, inflating counts that are then divided by area and presented as density.

But the deeper problem lay in what the methods missed, not just what they counted. Using size spectrum analyses, which examine how biomass is distributed across body size classes in a manner analogous to trophic pyramids, the researchers found that both point count methods completely failed to record the largest fishes on the reef. The reason is almost certainly the diver effect. A point count requires the diver to descend into or swim through the survey area and then remain at its centre, meaning every fish counted is within a small radius of a rotating, breathing, bubble-emitting human. Far from being a passive observer, the diver is a perceived predator, and wary large fishes simply vanish. Previous research has shown that some reef fish communities require upwards of three hours to return to their original densities after a diver passes, so it is unsurprising that even a ten-minute wait failed to bring big fish back into view.

The consequences of these size-based biases ripple directly into ecological interpretation. On Fantome Island, the instantaneous point count produced a size spectrum slope statistically indistinguishable from zero, a flat profile typically read as a sign of an inverted trophic structure, the kind of pattern associated with disrupted energy flows or external nutrient subsidies. Yet every other survey method produced clearly negative slopes, the expected bottom-heavy signature of a healthy size-structured community. In other words, one of the most common census techniques, applied for just one minute, would have led ecologists to conclude that the reef’s energetic architecture was fundamentally distorted when, according to all other methods, it was not. The 50-metre transect was the only technique that consistently captured the entire range of fish body sizes detected across both islands, and it also produced the least variable community composition in multivariate analyses.

The study also quantified how much replication any method requires. By repeatedly subsampling their data and measuring the standard error of estimates, the researchers found that every technique needed at least 31 to 36 samples to reach stable precision, defined as estimates within ten percent of the asymptotic value. This threshold matters because many monitoring programs operate with far fewer replicates per site, meaning their estimates may be too imprecise to detect real ecological change. Even the best-performing method, the 50-metre transect, has limits: highly mobile giants such as reef sharks and the bumphead parrotfish Bolbometopon muricatum are so rare and wide-ranging that transects hundreds of metres long would be needed to quantify their densities reliably.

These findings land at a delicate moment for reef science. Coral reefs are transforming under the pressure of climate change, and researchers are racing to assemble macroecological datasets that combine surveys from disparate sources, different organisations and sometimes different decades. Some of the world’s largest monitoring programs, including the United States National Oceanic and Atmospheric Administration’s National Coral Reef Monitoring Program and the Australian Institute of Marine Science’s Long-Term Monitoring Program on the Great Barrier Reef, each rely on their own protocols. The new study shows that pooling such data without accounting for methodological bias risks amplifying artefacts to continental scales, mistaking the fingerprint of a survey technique for a genuine demographic or ecological process. Statistical corrections and detection modelling may help, but the authors caution that when effect sizes are as large as those documented here, even sophisticated analytical tools are unlikely to fully separate methodological bias from real biology.

Fishing pressure adds another layer of complexity. Fishes in heavily fished areas tend to be warier and exhibit greater flight-initiation distances from divers, a well-documented behavioural response. Consistent with this, the maximum fish size recorded by point counts on fished Fantome Island was smaller than on the protected Orpheus Island, suggesting that the size bias of point counts is not fixed but worsens along gradients of human disturbance. A global dataset mixing surveys from pristine reserves and heavily exploited reefs could therefore encode systematically different biases in different places, further muddying macroecological inference.

The authors’ prescription is straightforward, if demanding: adopt uniform survey methodologies, use at least 30 replicates of a minimum 50-metre transect per site, and treat survey time as a standardised variable rather than an afterthought. In an era of big data, they argue, more data is not necessarily better. A smaller dataset collected consistently, with biases that are at least uniform and therefore quantifiable, will yield far more robust conclusions than a sprawling patchwork of incompatible counts. For a field whose findings feed directly into conservation policy, the message is that how we count fish determines not just the numbers we get, but the reefs we think we are saving.

Subject of Research: Methodological biases in underwater visual census techniques for coral reef fish community surveys

Article Title: Common census methods can lead to fundamentally different ecological interpretations of coral reef fish communities

Article References: Yan, H. F., Crisp, S. K., Bellwood, D. R., Siqueira, A. C., & Tebbett, S. B. (2026). Common census methods can lead to fundamentally different ecological interpretations of coral reef fish communities. Coral Reefs. https://doi.org/10.1007/s00338-026-02959-z

Image Credits: AI Generated

DOI: 10.1007/s00338-026-02959-z

Keywords: coral reef fish, underwater visual census, point counts, belt transects, size spectrum, diver effect, Great Barrier Reef, survey methodology, macroecology, biomass estimation, species richness, ecological monitoring