The analysis considered RCP 4.5 and RCP 8.5, identified in the study as moderate- and high-emissions pathways, and evaluated both present and future climatic suitability. Model performance was reported as robust, with an area under the receiver operating characteristic curve of at least 0.78 and a true skill score of at least 0.60. These metrics support the reported predictive performance of the suitability framework, although they do not establish the feasibility of individual projects.

Under RCP 4.5, the models projected a significant expansion of climatically suitable territory, particularly across central India. The newly suitable areas were expected to form a contiguous corridor, indicating that moderate warming could extend the climatic envelope for some agro-voltaic systems over a broader connected region. Its maps identify locations where climatic conditions are estimated to be suitable, not where all non-climatic requirements for deployment have been met.

The assessment under RCP 8.5 produced a more constrained picture. Suitable areas were projected to undergo spatial redistribution, while the geographic centers of suitability and the range of climatic conditions tolerated by agro-voltaic systems would change. The high-emissions results therefore do not simply extend the moderate-warming pattern.

They indicate that expansion of the total suitable area can coexist with reduced tolerance ranges and changing environmental limitations. The findings are also not uniform across the three models, which emphasized distinct environmental drivers, including the concentration of precipitation, thermal variability and interactions between moisture and temperature. Another important result was a change in the variables that most strongly controlled suitability.

Diurnal temperature variability was the leading influence in the earlier assessment, but temperature seasonality and precipitation extremes became more important in the projected future. This transition implies greater sensitivity to seasonal climatic fluctuations and monsoon variability in the modeled suitability patterns. Together, the results show why a national increase in suitable area should not be interpreted as a uniform reduction in environmental risk.

Overall, the study provides a spatially explicit, national-scale assessment of potential climate-driven suitability shifts in India’s agro-voltaic systems. Within those limits, the results support land-use planning that accounts for changing spatial patterns rather than assuming that future climatic suitability will remain stable.

Subject of Research: Climate-driven suitability and spatial shifts of agro-voltaic systems across India

Article Title: Environmental assessment of climate-driven suitability and spatial shifts of agro-voltaic systems in india using ensemble modeling

Article References: Mathur, M., & Poonia, S. (2026). Environmental assessment of climate-driven suitability and spatial shifts of agro-voltaic systems in india using ensemble modeling. Theoretical and Applied Climatology, 157(11), Article 733. https://doi.org/10.1007/s00704-026-06656-3

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06656-3

Keywords: agro-voltaic systems, India, climate change, ensemble modeling, RCP 4.5, RCP 8.5, bioclimatic suitability, Earth System models, spatial redistribution, land-use planning, Environmental, assessment

Researchers Niantao Jiao, Yuxin Zhang, and Hao Li have now built a spatial decision framework designed to fix exactly that problem. Published in Frontiers in Ecology and Evolution, the study applied the framework to the mainland coastline of Shandong Province, China, using more than four decades of Landsat satellite imagery from 1984 to 2025 at 30-meter resolution. The work identifies Priority Areas of Coastline Restoration, or PACR, by integrating three complementary dimensions: Coastline Fragmentation Degree (CFD), which diagnoses historical degradation; Stability of Coastline (SC), which assesses whether restoration can physically persist; and Coastal Ecological Carrying Capacity (CECC), which defines the ecological boundary for long-term sustainability. Together, these dimensions form a hierarchical decision pathway that transforms coastline restoration from a length-based compensation exercise into an ecosystem-function-oriented optimization strategy.

The technical foundation of the framework rests on quantifying how broken up a coastline has become. The researchers measured fragmentation using two indicators: the Index of Coastline Type Diversity, which captures how many different artificial and natural shoreline types crowd into a given stretch, and Patch Density, which quantifies how densely fragmented those shoreline units are. When only a single coastline type dominates a segment, fragmentation drops to zero; as ports, reclamation, aquaculture embankments, salt fields, transportation corridors, and urban development proliferate, fragmentation climbs toward one. Critically, the team discovered that fragmentation measurements follow a robust power-law scaling relationship with observational scale, with R-squared values exceeding 0.98 across all years from 1984 to 2020. This means that how fragmented a coastline appears depends heavily on the spatial resolution at which you measure it—a finding with profound implications for anyone attempting to compare fragmentation assessments across studies or regions.

Testing seven observation scales ranging from 1 to 30 kilometers, the researchers found that the 5-kilometer scale captured 66 percent of homogeneous coastline segments, including 52 percent of segments within the 1-to-5-kilometer range, enabling effective characterization of dominant fragmentation patterns while avoiding the diminishing returns seen at coarser scales. They therefore adopted 5 kilometers as the optimal spatial unit, dividing Shandong’s roughly 2,727 kilometers of mainland coastline into 546 segments for detailed analysis. Over the four-decade study period, fragmentation followed a distinct rise-then-fall trajectory: values climbed from 0.2256 in 1984 to a peak of 0.3195 around 2000, driven by extensive aquaculture expansion, reclamation, and port construction in Binzhou, Weifang, and Qingdao, then gradually declined to 0.2753 by 2025 as coastal spatial management policies, including the occupation-restoration balance mechanism, began to rein in uncontrolled artificial coastline expansion.

The second dimension of the framework, coastline stability, quantifies whether a given stretch of shore can maintain its structural characteristics under the combined pressures of natural processes and human activity. The researchers measured this using two indicators: the Index of Coastline Utilization Degree, which assigns human activity intensity scores from 1 to 4 to each coastline type (natural coastlines score lowest at 1, while ports, reclamation, transportation, and urban coastlines all score 4), and Area Changes in the Coastal zone, which tracks seaward expansion or erosion by comparing coastline positions between 1984 and 2025 using profile lines drawn perpendicular to the shore. Stability scores ranged from 1 to 5, with higher values indicating more stable coastlines. The results revealed pronounced spatial heterogeneity: high-stability areas accounted for 33.74 percent of the total coastline and extremely high-stability areas for another 25.92 percent, concentrated mainly in the Yellow River Delta of eastern Dongying, southern Yantai, and the northern and southern coasts of Weihai. Low-stability zones clustered in northeastern Binzhou, northeastern Dongying, and northern Weifang, where intensive aquaculture embankments and reclamation structures have driven substantial seaward expansion—cumulatively around 2,050 square kilometers along the most affected stretches.

The third dimension, coastal ecological carrying capacity, was assessed through a hierarchical framework structured across three criterion layers following the Pressure-State-Response model, four factor layers encompassing environmental stress, demographic pressure, coastal resources, and economic development with social support, and twenty quantified indicators. Weights were determined using the entropy method combined with the analytic hierarchy process, and a state-space model calculated the composite carrying capacity. Because indicators such as population, GDP, environmental protection investment, and per capita water availability are only reported at the administrative scale, carrying capacity was evaluated at the prefecture-level city resolution and integrated as a continuous soft constraint with a relatively low weight rather than as a hard threshold. The temporal trajectory proved striking: from 2010 to 2025, provincial carrying capacity followed a U-shaped path, declining sharply before 2016—with an annual change rate of minus 15.34 percent during 2010 to 2013—then recovering with annual growth rates of 10.75 percent and 18.10 percent during 2016 to 2019 and 2020 to 2025 respectively. The year 2016 thus marks a critical transition point, suggesting a shift in the balance between coastal development pressure and ecological management effectiveness across the region.

Integrating all three dimensions through entropy-weighted aggregation produced the Coastline Restoration Priority Index, with fragmentation receiving the highest weight of 0.6, reflecting its central role in diagnosing the artificialization that the occupation-compensation policy primarily targets, while stability and carrying capacity each received 0.2. Higher index values unambiguously denote lower restoration priority. The spatial results were dramatic: extremely high-priority segments represented only 7.52 percent of the total coastline, concentrated in Jiaozhou Bay, southeastern Qingdao, and eastern Rizhao, where intensive port activities, industrial development, and urban expansion have driven high coastline artificialization and fragmentation. Extremely low-priority segments dominated at 28.12 percent, mainly in the Yellow River Delta, northwestern Yantai, and southern Weihai, where protected ecosystems have maintained extensive natural coastline and ecological continuity. At the municipal scale, northern Weifang, southern Qingdao, and eastern Rizhao emerged as the highest-priority restoration areas, whereas Dongying exhibited the highest average index value of 4.1, attributed to the Yellow River Delta National Nature Reserve where relatively weak anthropogenic disturbance has preserved high ecosystem integrity.

Spatial statistical analysis using the Getis-Ord Gi* statistic revealed that both high-priority and low-priority segments cluster significantly rather than distributing randomly. High-priority coldspots at the 99 percent confidence level concentrated along the southern and southeastern coasts of Qingdao, while low-priority hotspots clustered in the Yellow River Delta and the northeastern and southeastern coastal areas of Weihai. This spatial dependence means restoration planning should move beyond isolated site-based interventions and adopt spatially coordinated strategies targeting continuous priority belts, which can be delineated by buffering core clusters and reconciling them with administrative boundaries and marine functional zoning. The framework can also be directly translated into the operational mechanisms of China’s coastline occupation-compensation balance policy: high-priority segments can be designated as candidate offset supply zones where restoration is most likely to yield ecologically recovering shorelines, while extremely low-priority segments within intact areas should be safeguarded as no-occupation zones under natural-coastline retention targets.

The findings arrive at a moment when China’s coastal transformation offers a sobering national context. From 1990 to 2020, the country’s artificial coastline length increased from 10,022 kilometers to 17,661 kilometers, while natural coastline decreased from 20,019 kilometers to 15,317 kilometers, reducing the national natural coastline retention rate from 66.7 percent to 42.3 percent. Shandong itself mirrors this pattern: its natural coastline retention rate dropped from 64 percent to 38 percent between 2000 and 2020, even as the province’s marine GDP exceeded 1.72 trillion yuan in 2023, accounting for 16.8 percent of the national total. The region also sustains critical ecological infrastructure as the core of the Yellow-Bohai migration corridor, providing essential staging grounds for 60 percent of species along the East Asian-Australasian flyway. The tension between economic engine and ecological lifeline could hardly be sharper.

The researchers acknowledge limitations, including uncertainties from statistical yearbook data at finer resolutions, coastline extraction accuracy with a mean error of 10.14 meters that may propagate into fragmentation and stability calculations, and the absence of climate-driven variables such as sea-level rise projections and storm surge frequency. Future applications at smaller scales should recalibrate indicator systems and weighting schemes according to local geomorphological characteristics and data availability. Nevertheless, the framework represents a meaningful step toward adaptive, climate-resilient coastal governance. By shifting the focus from how much coastline should be compensated to where restoration can achieve ecological equivalence, physical persistence, and long-term sustainability, the study offers a transferable spatial optimization approach for any coastal region grappling with the challenge of restoring degraded shorelines while accommodating sustainable development—a challenge that grows more urgent as sea levels rise and coastal populations expand worldwide.

Subject of Research: coastal restoration prioritization

Article Title: A spatial decision framework for coastline restoration prioritization: integrating ecosystem dynamics, stability, and ecological capacity

Article References: Jiao, N., Zhang, Y., & Li, H. (2026). A spatial decision framework for coastline restoration prioritization: integrating ecosystem dynamics, stability, and ecological capacity. Frontiers in Ecology and Evolution, 14, Article 1975765. https://doi.org/10.3389/fevo.2026.1975765

Image Credits: AI Generated

DOI: 10.3389/fevo.2026.1975765

Keywords: coastline restoration, spatial decision framework, coastal ecological carrying capacity, coastline fragmentation, Shandong Province, occupation-compensation balance, remote sensing, coastal zone management, spatial, decision, framework, coastline

The research, published in the journal One Ecosystem, tackles a blind spot that has long plagued land-use policy across Europe. Governments typically track land-take — the conversion of natural, forested or agricultural land into urban functions — as a simple accounting problem: hectares in, hectares out. The European Union’s headline target of No Net Land-Take by 2050 is built on this logic. Yet a hectare of shallow soil on a steep slope and a hectare of deep, fertile loam are not interchangeable, and treating them as such can mask the ecological damage done when the wrong hectares disappear.

To move beyond this surface-level view, the researchers combined cadastral land-use records with detailed soil and topographic maps of Wallonia. They first identified every parcel that shifted from non-artificialised in 2007 to artificialised in 2018, drawing on four official regional and national datasets, including the Walloon land-use map and a one-metre-resolution land-cover layer. The comparison was painstaking: the team cross-checked the 2007 baseline against a national map, correcting roughly 126 square kilometres of misclassified land, and then visually inspected 286 large parcels flagged by their automated classification. Nearly a third of those — 91 parcels — turned out to be falsely labelled as artificialised and were corrected, substantially improving the reliability of the final dataset.

With the converted areas mapped, the researchers characterised each one using abiotic properties that govern how ecosystems function: soil wetness, texture, depth, hydromorphy (the signs of prolonged waterlogging), slope and aspect. From these they built a map of ecologically constrained areas — peat soils and permanently waterlogged ground, temporarily wet soils, steep slopes exceeding 20 degrees, and shallow substrates. These are places where biophysical limitations restrict intensive agriculture and construction, but which, precisely because of those constraints, often retain semi-natural vegetation and a fine-grained mosaic of habitats that supports a disproportionately high capacity for regulating services.

The numbers confirm what policy has long feared about farmland. Agricultural areas — sown pasture, annual croplands and intensive orchards — accounted for 76.10 percent of all newly-artificialised surfaces over the study period, with forests contributing a further 16.42 percent. Most conversions produced built-up, impervious surfaces such as buildings and transport infrastructure (43.30 percent), while 37.85 percent became residential gardens and parks. The pattern reflects decades of Walloon zoning policy, which designates buildable land near existing urban and village centres — historically located on fertile agricultural plateaus and in valley bottoms. Nearly 90 percent of land currently available for residential development in the region is still agricultural.

But the study’s quieter revelation concerns the soils that planners rarely fight for. Overall, 12.31 percent of land-take occurred on ecologically constrained areas — slightly less than their 15.93 percent share of the region, but far from negligible. Among these, hydromorphic alluvial and colluvial soils and non-alluvial hydromorphic soils were the most affected subtypes. These waterlogged soils, often dismissed as having limited agronomic value, are exactly the ones that excel at regulating water flow, mitigating floods and maintaining water quality. Their conversion, the authors note, echoes a long-documented pattern in which urban development has repeatedly degraded wetlands and moisture-dependent environments.

To quantify what was lost, the team built an ecosystem service capacity matrix — a table crossing land-use and soil-condition categories (the service-providing units) against sixteen ecosystem services selected by a fifteen-member expert steering committee from the regional administration. Forty-three experts from universities, government and civil society then scored each cell from 0 (not relevant) to 5 (very high capacity) during workshops in October 2021, following individual pre-scoring rounds and consensus discussions. The approach refines the widely used matrix method of Burkhard and colleagues, which traditionally relies on land-cover classes alone, by embedding soil and topographic information directly into the rows of the matrix.

The results are stark. Provisioning services were essentially annihilated: all five evaluated — from food crops to fodder and livestock production — suffered capacity reductions of 98 to 100 percent, statistically highly significant. Regulating services fared scarcely better where water was concerned. Flood risk mitigation fell by 70 percent, soil erosion control by 64 percent and surface water quality regulation by 63 percent. Pollination dropped 63 percent and global climate regulation through carbon sequestration fell 61 percent. Cultural services declined between 42 percent (recreation) and 61 percent (cultural and spiritual inspiration). Only visual screening — the buffering of views — showed a minor, statistically insignificant change.

The analysis also exposed how much expert judgement itself shapes such assessments. Variation partitioning showed that land-use classes explained about 19.1 percent of the variance in capacity scores, while soil and topographic classes accounted for only 0.43 percent — a statistically non-significant contribution. In other words, the experts weighted differences between land uses far more heavily than variation within them, even though soil condition demonstrably matters: expert scores for timber production in coniferous forests, for example, climbed from 1 on permanently waterlogged peat soils to 4.8 on unconstrained ground. The team also found that artificialised land is anything but homogeneous — sealed surfaces scored zero for erosion control, while parks and gardens scored 3.0, cereal cropland on good soils 1.8 and permanent grassland 4.8 — yet the area-weighted balance of conversions still left regulating capacity in net decline, because gains from urban green space could not offset the losses from sealing and the disappearance of high-performing grasslands.

The study’s implications reach well beyond Wallonia. Its land-take estimate of 21.43 square kilometres per year is considerably higher than the roughly 5.7 square kilometres per year reported for Belgium by the European Environment Agency, a discrepancy the authors attribute to methodology: European figures rely on satellite-derived land-cover data with minimum mapping units that miss small-scale changes, whereas regional statistics use parcel-level cadastral records. As the EU’s new Soil Monitoring Law tightens definitions of land-take and soil artificialisation, the message of this research is that counting hectares is no longer enough. Whether a converted field sits on deep fertile loam or on a waterlogged valley-bottom soil determines whether a region loses little more than a line on a map — or the invisible machinery that keeps its water clean, its floods contained and its landscapes alive.

Subject of Research: Quantifying ecosystem service capacity loss from land-take in Wallonia using soil and topographic ecosystem condition data

Article Title: Beyond surface change: assessing ecosytem services capacity loss from land-take in Wallonia

Article References: Pairon, M., & Dufrêne, M. (2026). Beyond surface change: assessing ecosytem services capacity loss from land-take in Wallonia. One Ecosystem, 11, Article e172396. https://doi.org/10.3897/oneeco.11.e172396

Image Credits: AI Generated

DOI: 10.3897/oneeco.11.e172396

Keywords: land-take, ecosystem services, soil sealing, Wallonia, No Net Land-Take, ecosystem condition, hydromorphic soils, flood regulation, capacity matrix, land-use planning, soil health, Belgium

The study’s lead finding is stark: roughly 19 percent of Nepal’s lowland areas fall into medium to very high flood susceptibility zones, threatening approximately 900,000 people and more than 3.4 million buildings. In the hillier terrain, the picture is equally sobering, with about 40 percent of the country susceptible to slope failure, endangering around 200,000 people and some 600,000 buildings. These are not abstract statistics in a nation where, over the past half century, disasters have killed an average of two people every day and affected more than 300 families annually. Landslides alone claim an average of 113 lives each year, while floods account for more than nine percent of disaster deaths and touch over sixty percent of affected families.

What sets this research apart from previous hazard assessments is its methodological machinery. Traditional susceptibility mapping in Nepal has relied heavily on multi-criteria analysis, a GIS technique in which experts assign subjective weights to variables such as slope, elevation and land cover. That approach struggles to capture the tangled, non-linear relationships that govern where water flows and where slopes fail. The ICIMOD team instead deployed a Random Forest model, an ensemble machine learning algorithm that builds hundreds of decision trees, each trained on random subsets of the data, and averages their predictions into a continuous probability score between zero and one. Hyperparameter tuning across 400 to 600 trees showed that 600 delivered the best performance, and the resulting models achieved strong discrimination, with an area under the receiver operating characteristic curve of 0.84 for floods and 0.85 for landslides.

The variable lists themselves tell a geomorphological story. For floods, accessibility and flow accumulation ranked as the most important predictors, followed by landform, land cover and slope, confirming that monsoon runoff concentrates in river corridors where human development presses closest to the water. For landslides, slope steepness and aspect dominated, with vegetation greenness, captured through the Normalized Difference Vegetation Index, coming in third. Validation against independent datasets added credibility: the flood model agreed 74 percent with an established hydrological inundation product, while the landslide model matched a manually digitized inventory of landslides triggered by the devastating 2015 Gorkha earthquake with 81 percent agreement.

The truly novel step came after the single-hazard maps were complete. The researchers fed the flood and landslide probability surfaces into an unsupervised k-means clustering algorithm, which partitioned every spatial unit of the country into three natural groupings without any expert-defined thresholds: low-hazard zones, flood zones and landslide zones. A subsequent spatial adjacency analysis, a GIS operation that identifies polygons sharing boundaries, then carved out a fourth category: interaction zones, where landslide-susceptible slopes physically touch flood-susceptible river corridors. Nationally, this data-driven classification assigned 81 percent of the territory to the low-hazard category, nine percent to flood zones, five percent to landslide-only zones and five percent to interaction zones.

Those interaction zones matter because of what they represent physically. In mountainous terrain, the boundary between a steep, fragile hillslope and an active river channel is not an arbitrary line on a map; it is the slope-toe interface where debris from slope failures enters the drainage network and can amplify downstream inundation. By applying the adjacency operator a second time, linking interaction zones to downstream flood-prone polygons, the team delineated what they call critical multi-hazard flow zones: roughly 7,588 square kilometres of connected corridors that funnel hazard material from unstable headwaters toward populated valleys. These corridors overlap 88 square kilometres of built-up land, about twelve percent of Nepal’s total urban footprint, and 1,722 square kilometres of cropland.

The authors are careful to note that these flow zones represent susceptibility-based spatial connectivity, a screening-level proxy, rather than physically simulated cascading processes. Full hydrodynamic modelling of landslide-dam formation, breach dynamics and downstream flood propagation was beyond the scope of the study. Yet the screening framework has already proven its relevance against real events. The catastrophic 2021 Melamchi disaster, in which intense upstream rainfall, landslides and debris flows cascaded into a flood that devastated downstream settlements and the Melamchi drinking water supply project serving Kathmandu, followed precisely the source-to-sink pathway the mapping is designed to flag. The July 2025 transboundary outburst flood from Nepal’s border with China, which swept away a critical trade bridge at Rasuwaghadi, underscores the stakes for infrastructure planning in similar corridors.

The study also exposes an uncomfortable institutional blind spot. In Nepal, flood hazards are managed by the Department of Hydrology and Meteorology while landslides fall under the Department of Mines and Geology, a siloed arrangement that treats physically coupled processes as administratively separate phenomena. Meanwhile, current national building regulations rely on uniform river-buffer setbacks, thirty metres from rivers, fifty metres from lakes, ten metres from canals, that bear no relationship to actual topographic susceptibility. Satellite imagery documenting the transformation of the Sardu Khola floodplain in Sunsari District, from sparse settlement and intact farmland in 2004 to dense housing and road networks by 2025, illustrates how rapidly exposure is accumulating in exactly the wrong places.

The researchers argue their maps can shift Nepal from reactive post-disaster relocation toward proactive risk screening. Low-hazard zones, often found on stable interfluves and elevated terraces even within high-risk basins like Melamchi, should be prioritized for settlement expansion and critical infrastructure. Interaction zones, by contrast, warrant strict regulatory controls and could even be designated no-development areas. The framework extends beyond government planning: banks and infrastructure investors, including those financing hydropower projects and strategic road corridors across the Himalaya, could use the flow zones as preliminary geospatial due diligence, embedding geological risk assessment into lending decisions before capital is committed to hazardous ground.

Limitations remain, and the authors flag them candidly. The models were trained on historical inventories that treat hazard patterns as stationary, even as climate change intensifies rainfall extremes and vegetation recovery alters slope stability, a tension visible in the divergent bias patterns between the flood and landslide validations. Detailed lithology, groundwater dynamics and real-time hydrometeorological data were unavailable, and spatial block cross-validation was not implemented. Exposure figures should therefore be read as indicative screening estimates rather than precise measurements. Still, as a replicable, computationally efficient template for data-scarce mountain nations, the framework aligns squarely with the Sendai Framework for Disaster Risk Reduction and the Sustainable Development Goals, offering a pathway for risk-informed land-use planning that could save lives before the next monsoon arrives.

Subject of Research: National-scale flood and landslide susceptibility and multi-hazard interaction mapping in Nepal using machine learning and GIS

Article Title: Flood and landslide susceptibility assessment and multi hazard interaction mapping using machine learning and GIS for sustainable settlement planning in Nepal

Article References: Thapa, N., Uddin, K., Thapa, R. B., & Udas, E. (2026). Flood and landslide susceptibility assessment and multi hazard interaction mapping using machine learning and GIS for sustainable settlement planning in Nepal. Discover Geoscience, 4(1), Article 287. https://doi.org/10.1007/s44288-026-00670-8

Image Credits: AI Generated

DOI: 10.1007/s44288-026-00670-8

Keywords: Nepal, flood susceptibility, landslide susceptibility, machine learning, Random Forest, k-means clustering, GIS, multi-hazard, spatial adjacency analysis, Google Earth Engine, land-use planning, disaster risk reduction

The team’s central tool was ensemble species distribution modeling, implemented in the widely used biomod2 platform. Rather than relying on a single algorithm, the researchers calibrated multiple machine-learning models against 403 quality-controlled presence records of cacao in the region, each one verified to reduce the errors that plague many mapping exercises. The predictor variables drew on three distinct data streams: bioclimatic layers describing temperature and precipitation regimes, topographic information such as elevation and slope, and edaphic data capturing soil properties. On top of the statistical machinery, the authors applied spatial exclusion criteria that removed biophysically unsuitable zones, legal conservation constraints that protect protected areas, and regulatory filters tied to the European Union Deforestation-Free Products Regulation, known as EUDR, which will increasingly govern whether Peruvian cacao can enter European markets.

The modeling pipeline was deliberately strict about quality. Each candidate algorithm was evaluated under cross-validation, and only models exceeding demanding thresholds were retained: an area under the receiver operating characteristic curve, or AUC, greater than 0.9, and a true skill statistic, or TSS, greater than 0.7. These metrics matter because species distribution models can look impressive while hiding serious weaknesses. AUC above 0.9 indicates near-excellent discrimination between suitable and unsuitable locations, while TSS above 0.7 signals strong agreement between predicted and observed distributions after correcting for the prevalence of the species. Among the individual algorithms, decision tree-based approaches performed best, with Random Forest and XGBoost standing out as the strongest single predictors. Random Forest builds many decision trees on random subsets of the data and averages their votes, while XGBoost builds trees sequentially, each one correcting the errors of the last. Their dominance suggests that the relationships between cacao suitability and environmental gradients are nonlinear and interactive, precisely the kind of structure tree ensembles capture well.

Yet the study’s most consequential finding came not from any single model but from the consensus. When the researchers merged the retained models into an ensemble, the resulting suitability map showed greater spatial stability and overall accuracy than any individual algorithm. This is a recurring lesson in predictive ecology: individual models disagree at the edges of their training data, but a well-constructed consensus dampens those idiosyncrasies. For policymakers, that stability is what transforms a scientific map into a planning instrument. A boundary drawn on a consensus surface is far less likely to shift dramatically when new data arrive than one drawn on a single model’s output.

So where does the chocolate tree want to grow in Amazonas? Under current climatic conditions, approximately 14.8 percent of the regional territory shows high suitability for cacao. That footprint is not spread evenly across the landscape. It concentrates in inter-Andean valleys and in low- to mid-elevation zones, where temperatures remain warm and relatively stable and moisture is adequate through the year. The single most important environmental predictor, according to the model’s variable importance analysis, was thermal seasonality, the degree of variation in temperature across the year. Cacao, a tropical understory species, tolerates little in the way of cold stress or wide temperature swings, so it makes intuitive sense that the steadiness of the thermal regime, more than total rainfall or soil chemistry, defines the crop’s climatic envelope in this part of the Andes-Amazon transition.

Here, however, the study delivers its sharpest and most policy-relevant twist. The 14.8 percent figure describes climatic suitability, but climate is not permission. When the researchers overlaid territorial and regulatory constraints, the effectively available area shrank substantially. Protected conservation areas came off the map, as did zones excluded on biophysical grounds. Most striking was the impact of the EUDR scenario: because the European regulation prohibits products grown on land deforested after a defined cutoff date, large portions of the climatically suitable territory become ineligible for export-oriented cacao under the new rules. The gap between where cacao could grow and where cacao may legally be grown is, in other words, wide, and it is widening as environmental due-diligence requirements harden in major consumer markets.

This regulatory squeeze does not leave the region empty-handed. The analysis identified high-priority degraded areas that emerge as focal opportunities for productive restoration through agroforestry systems. Agroforestry, the practice of growing cacao beneath and alongside native trees, offers a way to rebuild canopy cover, restore ecological function, and produce a marketable crop simultaneously. Because these degraded lands are not subject to the same deforestation concerns as intact forest frontiers, they represent the sweet spot where farmer livelihoods, restoration goals, and EUDR compliance converge. In effect, the study reframes cacao expansion not as a question of finding new land to clear, but of finding old, damaged land to heal.

The third layer of the analysis is the one least often seen in suitability mapping: plant health. The researchers conducted a phytosanitary validation covering 2021 through 2024, comparing the modeled suitability surface against field records of the prevalence and incidence of major cacao pathogens and pests. The result was a high spatial concordance that carries an uncomfortable implication. The areas of greatest climatic suitability for cacao are also the areas where the crop’s worst enemies thrive, particularly the fungi Moniliophthora roreri, the agent of frosty pod rot, and Moniliophthora perniciosa, which causes witches’ broom disease. Both pathogens have devastated cacao production across Latin America for decades, and both flourish under the same warm, humid conditions that cacao itself prefers. The finding confirms that an optimal climatic niche for cultivation is simultaneously an optimal niche for biotic pressure, meaning that any expansion strategy must build disease management, resistant varieties, and integrated pest control into its foundations rather than treating them as afterthoughts.

Taken together, the study offers something rarer than another suitability map: an integrative framework in which climatic modeling, territorial management, and phytosanitary risk are assessed in a single pipeline. The authors position the work as a scientific input for public policy, land-use planning, and sustainable cacao production strategies across the Amazon. The practical chain of reasoning is clear. Climate models identify where the crop can grow; legal and conservation filters identify where it may grow; degraded-area analysis identifies where it should grow; and phytosanitary data identify where it will need the most protection once it does. Each layer disciplines the next, and the final product is a map of opportunity that is honest about its own limits.

For Peru, whose cacao has earned a reputation among fine-flavor chocolate makers, and for tropical agricultural science more broadly, the message is twofold. Ensemble machine-learning models, fed with multisource data and validated against ground truth, can now deliver planning-grade precision at the regional scale. But the era in which suitability could be defined by climate alone is over. Trade regulations written in Brussels, conservation commitments made in Lima, and fungi evolving in Amazonian pods all shape the same map. The future of chocolate in its birthplace will be decided not by where the tree can survive, but by where the full weight of science, law, and biology allows it to flourish.

Subject of Research: Ensemble predictive modeling of climatic, regulatory, and phytosanitary suitability for cacao cultivation in Amazonas, Peru

Article Title: Geospatial suitability of cacao in Amazonas (Peru): An ensemble predictive modeling approach (BIOMOD2) integrating multisource data, territorial management, and plant health

Article References: Zabaleta-Santisteban, J. A., Medina-Medina, A. J., Tuesta-Trauco, K. M., Rivera-Fernandez, A. S., Silva-Melendez, T. B., Grandez-Alberca, M. A., Puscan-Rojas, J., Salas López, R., Oliva-Cruz, M., Cotrina-Sanchez, A., Rojas-Briceño, N. B., Silva-López, J. O., Gómez-Fernández, D., Leiva-Espinoza, S. T., Huaman-Pilco, A. F., & Barboza, E. (2026). Geospatial suitability of cacao in Amazonas (Peru): An ensemble predictive modeling approach (BIOMOD2) integrating multisource data, territorial management, and plant health. PLOS One, 21(10), e0360376. https://doi.org/10.1371/journal.pone.0360376

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0360376

Keywords: cacao, Amazonas, Peru, BIOMOD2, ensemble modeling, species distribution models, EUDR, deforestation, agroforestry, Moniliophthora roreri, witches' broom, land-use planning

The framework begins with a diagnostic question that sounds simple but has proven surprisingly difficult to answer at scale: which fields are sick, which are unproductive, and which are both? Earlier studies typically treated cropland quality or yield as a single proxy for overall condition, or focused narrowly on problems already flagged by national pilot programs, such as heavy metal contamination or groundwater depletion. That approach, the authors argue, misses the internal contradictions of the cropland system itself. A field can be ecologically degraded yet highly productive, or healthy but hamstrung by poor climate, terrain, or economics. It is precisely these mismatches—where health and productivity pull in opposite directions—that signal the greatest urgency for managed rest.

To capture those contradictions, the team built a two-part evaluation system grounded in the idea of the cropland as an integrated natural-resource, socio-economic, and ecological complex. Cropland health was scored using six indicators spanning farming environment and production elasticity: soil heavy metal pollution across eight elements including cadmium, arsenic, and lead; irrigation water quality drawn from more than 2,000 monitoring sites; potential soil erosion; earthworm density as a biological signal; the year-to-year variability of grain production; and soil cation exchange capacity, a measure of the soil’s buffering and nutrient-retention power. Cropland productivity, by contrast, was assessed with twelve indicators covering climate production potential, slope, soil texture, bulk density, pH, organic matter, irrigation and flood protection infrastructure, cultivation distance, mechanization levels, rural disposable income, and cropping intensity. All data were harmonized to a 30-meter grid for the 2020 baseline year, with weights derived through an Analytic Hierarchy Process based on expert judgment.

The single-factor results paint a geographically coherent picture of China’s agricultural stresses. Substandard irrigation water clusters in the Hai, Yellow, Huai, and Songliao river basins of the water-scarce north. Severe heavy metal pollution concentrates in southern Yunnan-Guizhou Plateau, the southern Middle Yangtze Plain, the eastern Huang-Huai-Hai Plain, and central Southern China, with national shares of light, moderate, and severe pollution at 11.59, 13.35, and 2.40 percent respectively. Soil erosion intensifies in the hilly south and on plateaus, while earthworm densities drop in cold, arid, or excessively hot regions such as the Northern Arid and Semiarid Region. When the indicators were combined using Jenks Natural Breaks thresholds, high-health cropland accounted for 48.15 percent of the national total, medium for 29.97 percent, and low for 21.88 percent, with low-health land concentrated in mountainous and hilly eastern, southern, and southwestern regions.

Productivity told a different story. High-productivity cropland made up 27.54 percent of the total, medium 42 percent, and low 30.45 percent, following a clear southeast-high, northwest-low gradient. The eastern belts of the Huang-Huai-Hai Plain, Middle Yangtze Plain, Southern China, and North China Plain outperform the western plateaus and basins, where climate production potential, slope, and thin infrastructure hold yields back. Notably, the Sichuan Basin emerges as a concentrated island of high productivity within the otherwise mixed Sichuan Basin and Surrounding Regions. In the north and west, the binding constraints are largely natural—limited rainfall, coarse tillage textures, short growing seasons—compounded by socio-economic limits such as weak irrigation guarantees, low mechanization on complex terrain, and thinner rural incomes that constrain investment in tillage, pest management, and harvesting.

The study’s most methodologically distinctive contribution is its treatment of connectivity. Previous fallow-mapping efforts judged connectivity by simple spatial adjacency, but the researchers argue that two neighboring parcels are only truly connected for management purposes if they share the same cropland type—dryland, irrigated land, or paddy field—and fall in the same health and productivity classes. Using an eight-neighbor contiguity algorithm on the 30-meter grid, they computed the area of each contiguous multi-attribute patch and classified connectivity as high, medium, or low. Nationally, high-connectivity cropland covers 50 percent of the total, concentrated in the flat plains, while low connectivity at 11.08 percent tracks the fragmented hills, mountains, and basins of the south and southwest. The logic is economic as much as ecological: contiguous, similar parcels allow uniform fallow measures, lowering transaction and supervision costs, whereas fragmented or heterogeneous blocks raise the price of every hectare rested.

Coupling the three layers produced 27 possible health–productivity–connectivity combinations, which the team sorted under a principle they call ‘prominent contradiction, concentrated and connected.’ Combinations where health and productivity sit at opposite extremes—high health with low productivity, or low health with high productivity—on highly connected land form the Urgent Fallow Zone; the same conflicts on less connected land form the Priority Fallow Zone. Milder mismatches fall into Moderate and Postponed zones. The reasoning is deliberately counterintuitive: land that is poor on both counts is often the least attractive target, because restoration investments there can be prohibitive and the expected gains uncertain. Better, the authors argue, to concentrate public money where one dimension is strong and the other is failing, so that targeted rest can tip the balance efficiently.

The resulting map is dominated by the north. More than 60 percent of the identified fallow area lies in northern China, led by the Northern Arid and Semiarid Region at 32.17 percent of the national fallow total, followed by the Loess Plateau at 20.38 percent and the Huang-Huai-Hai Plain at 14.18 percent, with urgent and priority zones clustering at the interfaces where these regions meet. The dominant northern pattern is high health paired with low productivity—land that is ecologically sound but constrained by climate, texture, organic matter, irrigation reliability, flood protection, mechanization, and rural income. There, the policy prescription is productivity enhancement: better water infrastructure, mechanization, and soil improvement rather than long retirement. In the south, the pattern inverts. Fragmented clusters of urgent zones appear in southern Hunan and northern Jiangxi, around the Changsha–Zhuzhou–Xiangtan corridor and the Nanchang–Fuzhou–Yichun–Shangrao–Jiujiang belt, where low health driven by heavy metal pollution and poor cation exchange capacity sits atop highly productive land. There, remediation fallow—planting green manure or hyperaccumulating crops, or rotating to rebuild soil—takes precedence.

The framework’s outputs translate into three concrete policy levers. Counties with large shares of urgent and priority zones and high connectivity could receive priority fallow quotas and per-hectare payments, with differentiated rates reflecting the distinct management costs of contrasting conflict types. The health–productivity class itself provides a decision pathway between productivity-supporting fallow in the north and remediation or selective medium-term retirement in the south. And because the analysis runs on 30-meter grids, results can be aggregated to administrative units for eligibility and quota setting, while connectivity guides the block design needed to capture economies of scale. The authors are candid about limitations: the 2020 snapshot is static, groundwater depletion was not included in the index system, and a single national set of weights cannot fully capture regional variation. They plan to integrate climate scenarios, land-use simulations, and multi-year monitoring to evolve the framework from static zoning toward adaptive management.

What makes the study resonate beyond China is its reframing of fallow as a precision instrument rather than a blunt conservation tool. The team’s maps successfully reproduce many regions already targeted by national pilots, including heavy metal zones in Hunan and ecologically fragile karst and degraded areas in the southwest and northwest, and they align with ongoing rotation trials in the northern agro-pastoral ecotone. But by starting from the cropland system itself rather than from pre-designated problem lists, the framework surfaces vulnerable zones that pilot programs had not flagged. As extreme weather, pollution, and resource scarcity intensify worldwide, the Chinese case offers a transferable template: diagnose health and productivity separately, weigh their contradictions, respect the geometry of the fields, and let the land rest exactly where rest will pay off most.

Subject of Research: National-scale spatial optimization of cropland fallow in China based on coupled health, productivity, and connectivity assessment

Article Title: Rational spatial arrangement of cropland fallow incorporating connectivity alleviates health-productivity conflict in China

Article References: Xu, F., Yao, X., Gao, B., Lv, Y., Ou, C., Ye, S., Geng, Q., Jia, S., Ye, K., Yang, J., Li, L., Wang, J., Zhang, C., Li, S., & Huang, J. (2026). Rational spatial arrangement of cropland fallow incorporating connectivity alleviates health-productivity conflict in China. npj Sustainable Agriculture, 4(1), Article 83. https://doi.org/10.1038/s44264-026-00194-1

Image Credits: AI Generated

DOI: 10.1038/s44264-026-00194-1

Keywords: cropland fallow, cropland health, cropland productivity, spatial connectivity, soil heavy metal pollution, food security, sustainable agriculture, land use planning, China, soil erosion, spatial coupling, agricultural policy

The research team, led by María Mercedes Ojeda-Muñoz of Universidad Politécnica Salesiana in Cuenca, together with Juan Gabriel Mollocana-Lara and Cesar Ivan Alvarez, focused on Molleturo Parish, a vast rural territory of nearly 98,400 hectares within Cuenca Canton. The landscape is extraordinarily diverse: elevations range from low inter-Andean valleys to mountain summits above 4,000 meters, slopes reach nearly 69 degrees in places, and vegetation spans Andean forests, shrublands, and the iconic páramo, a high-elevation alpine wetland ecosystem that regulates regional water supplies and harbors exceptional biodiversity. Roughly a quarter of the parish is designated as protected forest and vegetation area, underscoring its ecological importance.

Why does fire matter so much here? The páramo and Andean shrublands depend heavily on moisture availability, and their soils and vegetation store water that feeds communities far downstream. National climate projections for Ecuador indicate temperature increases of 2.6 to 3.5 degrees Celsius by the end of the century under high-emission scenarios, alongside shifting precipitation patterns. Warmer, drier air accelerates the drying of fuels, raising both the likelihood of ignition and the intensity with which fires burn. Ecuador has already seen the consequences: between January and October 2022, more than 1,200 wildfire events were recorded nationally, affecting over 6,500 hectares, with Azuay Province among the hardest hit. In 2023, fire incidence exceeded previous records, concentrated in the August-to-October dry season.

To translate these climatic pressures into concrete fire behavior, the team employed FlamMap, a widely used semi-empirical fire modeling framework developed by Mark Finney of the U.S. Forest Service. FlamMap simulates potential fire behavior across entire landscapes under defined environmental conditions, producing spatial estimates of flame length, rate of spread, fireline intensity, and fire growth. The researchers built a Landscape file integrating elevation, slope, aspect, fuel models, and vegetation cover percentage, all standardized to a 20-by-20-meter grid. Land cover data from Ecuador’s Ministry of Agriculture, covering 78 distinct categories within the parish, were reclassified into the standard Scott–Burgan fuel models, a classification system that encodes fuel load, fuel bed depth, and particle size distribution. Agricultural areas, water bodies, urban zones, and eroded lands were treated as non-burnable.

The simulations drew on meteorological records from the EDA CHAUCHA station, part of the Cuenca Hydrometeorological Network, which provided continuous observations of temperature, precipitation, relative humidity, wind speed, and wind direction from 2014 to 2020. For future conditions, the team used projections under RCP 8.5, the high greenhouse gas emissions pathway from the IPCC’s Fifth Assessment Report, derived from three global circulation models: CSIRO, GISS, and IPSL. Projections were obtained for two time windows, 2030–2050 and 2051–2070, and processed in the R statistical environment to match the study area’s extent. Roads, rivers, rocky outcrops, and other non-burnable features extracted from OpenStreetMap were incorporated as fire barriers, adding realism by accounting for breaks in fuel continuity that restrict fire propagation.

Two complementary algorithms powered the analysis. The Fire Behavior module calculated potential surface and crown fire characteristics for every raster cell, while the Minimum Travel Time algorithm identified the fastest spread pathways across the heterogeneous terrain using 2,000 randomly distributed ignition points, yielding burn probability maps and estimates of potential fire size. The four fire behavior outputs were normalized and combined with equal weighting into a composite Fire Hazard Index, classified into low, moderate, high, and extreme categories using thresholds originally proposed by Andrews and Rothermel.

The results tell a nuanced story. Most of Molleturo remained in the low fire-behavior category throughout all periods, but the high-hazard footprint expanded dramatically in relative terms. The area classified as high flame length nearly doubled from 1.11 percent of the parish in the baseline period to 2.05 percent in 2051–2070, while the high rate-of-spread class more than doubled from 0.60 to 1.36 percent. Most strikingly, the proportion of land classified as high fire hazard rose from just 0.250 percent under baseline conditions to 3.256 percent in 2030–2050 and 3.713 percent by 2051–2070, a roughly fifteenfold relative increase. Among fuel types, slightly disturbed shrub páramo consistently exhibited the highest fire potential, marking these ecosystems as the landscape’s most flammable zones.

Notably, crown fire activity showed limited variation across scenarios, suggesting that projected warming will primarily intensify surface fires rather than drive destructive canopy fires. This pattern aligns with fire ecology research in humid mountain ecosystems, where discontinuous canopy structure and high background humidity constrain vertical fire transition. Burn probability also remained low in absolute terms, with mean values changing only slightly among periods and maximum values reaching only about 0.0145. Mean simulated fire size hovered between roughly 58 and 62 hectares, though the large standard deviations reveal substantial variability among individual simulated fires. The extensive non-burnable agricultural land, which covers nearly 38 percent of the parish, and the incorporated fire barriers likely help contain fire spread across the landscape.

Perhaps the most actionable finding concerns people. By overlaying burn probability maps with settlement locations, the researchers identified four communities, Cochapamba, Santa María, Río Blanco, and Huahualcay, that lie within or adjacent to areas of comparatively elevated burn probability. The authors caution that because the simulations used random ignitions, this spatial coincidence does not imply that the settlements caused the pattern; rather, the maps flag locations where exposure, access routes, and ignition processes deserve closer local scrutiny. The researchers emphasize that the hazard and burn probability maps should be treated as screening instruments for prioritizing prevention and preparedness, not as deterministic forecasts of where fires will burn.

The study’s limitations are candidly acknowledged. Fuel conditions were held static, so the model cannot capture future land-use change, grazing pressure, or dynamic vegetation-fire feedbacks that could either amplify or dampen risk. Several inputs derived from governmental cartography rather than field measurements, and fine-scale heterogeneity in steep terrain may not be fully resolved even at 20-meter resolution. The simulations also do not isolate the individual contributions of wind, temperature, and humidity to the observed changes. Nevertheless, the spatial patterns are consistent with expected fire responses to warming and drying, lending confidence to the framework. The authors recommend integrating wildfire screening into Ecuador’s land-use planning instruments, strengthening meteorological and satellite monitoring, establishing strategic fuel breaks, reducing fuel loads in high-risk shrub zones, and promoting community-based fire awareness. As the tropical Andes warm, such proactive adaptation may determine whether these water-towering ecosystems retain their resilience or cross into a more flammable future.

Subject of Research: Climate change effects on wildfire hazard in high-elevation tropical Andean ecosystems of southern Ecuador, assessed with high-resolution FlamMap fire behavior simulations

Article Title: Climate change amplifies surface fire hazard in the tropical Andes: high-resolution FlamMap simulations in southern Ecuador

Article References: Ojeda-Muñoz, M. M., Mollocana-Lara, J. G., & Alvarez, C. I. (2026). Climate change amplifies surface fire hazard in the tropical Andes: high-resolution FlamMap simulations in southern Ecuador. Natural Hazards, 122(20), Article 658. https://doi.org/10.1007/s11069-026-08431-z

Image Credits: AI Generated

DOI: 10.1007/s11069-026-08431-z

Keywords: wildfire, climate change, FlamMap, tropical Andes, Ecuador, páramo, fire behavior modeling, RCP 8.5, burn probability, Azuay Province, fire hazard, land-use planning