The Salt River, known to the Akimel O’Odham people as On’k Akimel, has been dammed, diverted, mined, clear-cut, restored, and ignored over more than a century of intensive manipulation. Although the Hohokam and their descendants irrigated fields with its waters for centuries, diversion intensified dramatically in the early 1900s when the Salt River Project, backed by federal funding, constructed dams and levees that isolated the river from its floodplain and left much of the channel dry. What water remains flows through a handful of reaches where human infrastructure unintentionally or deliberately sustains year-round inundation. In the Sonoran Desert, where riparian and wetland areas cover only about 0.4 percent of Arizona’s land yet support up to 80 percent of its wildlife species, these wet oases are disproportionately important for regional biodiversity.

The research team selected three perennially wet sites representing fundamentally different management philosophies. The first, called Tonto, sits upstream of the urban sprawl within Tonto National Forest, where Stewart Mountain Dam regulates flow and recreationists, kayakers, anglers, and growing herds of feral horses exert ongoing pressure on the ecosystem. The second, Price, is what ecologists call an accidental wetland: it formed beneath and around the intersection of two major highways in the early 2000s and is sustained largely by storm-drain runoff from the roads, agricultural fields on the Salt River Pima-Maricopa Indian Community reservation, and occasional reclaimed wastewater discharge. Beyond access restrictions and occasional harvests of culturally significant plants by Indigenous Elders, Price receives essentially no management. The third site, Rio Salado, is a former dumping ground in downtown Phoenix that the City of Phoenix, the US Army Corps of Engineers, and Audubon Southwest transformed in a restoration project exceeding 110 million dollars, involving litter removal, invasive plant eradication, groundwater wells, and native plantings.

The monitoring itself was methodical and demanding. Bird surveys were conducted seasonally at fixed points along transects crossing the river channel, with a trained observer recording all birds seen or heard within a 50-meter radius over fifteen minutes during calm, clear mornings. Herpetofauna were documented through visual encounter surveys in nine plots per site, conducted by two observers once morning temperatures reached about 21 degrees Celsius, the peak of reptile activity. Vegetation was sampled intensively at the study’s outset in 2012 and again in 2022, with researchers assessing plant cover by species across thirty plots per site using standard Daubenmire cover classes. The team analyzed species richness, abundance, and evenness using the Shannon Diversity Index and analysis of variance frameworks, treating the sites as case studies rather than statistical replicates, a candid acknowledgment that landscape-scale ecology often demands messy but necessary long-term observation.

The results paint a portrait of divergence between taxonomic groups. Bird communities along the river proved remarkably stable, and at the restored Rio Salado site, average annual bird abundance actually rose by 56.7 percent, from 68.5 individual observations in 2014 to 107.3 in 2022, with gains across every habitat preference guild. Aggregated across all three sites, bird species richness showed a statistically significant increase over the study period. Herpetofauna told a different and more troubling story: reptile and amphibian abundance and diversity both declined significantly, with the trend more pronounced at the urban sites. The researchers caution that COVID-19 pandemic disruptions reduced sampling effort and statistical power, but parallel analyses of the same data using different methods have detected similar declines, and the known hazards of urbanization for reptiles, including roads, free-roaming cats, contaminated soils and water, and intensifying heat, lend biological plausibility to the pattern.

The plant communities underwent the most dramatic transformation of all. At both urban sites, vegetation became substantially more abundant, increasing by roughly 50 percent in total cover, and more species-rich by 2022, yet also less evenly distributed, with a few prolific cosmopolitan species such as giant reed, California bulrush, and broadleaf cattail dominating the landscape. The non-urban Tonto site, by contrast, grew more evenly diverse even as drought and horse grazing reduced its obligate wetland plants by about half and its tall trees by roughly two-thirds. This contrast between homogenizing urban sites and diversifying non-urban ones aligns with the well-documented global phenomenon of biotic homogenization, in which cities worldwide increasingly share the same suite of urban-adapted species.

Perhaps the study’s most provocative finding concerns the unmanaged accidental wetland at Price. Despite being wedged beneath a highway interchange and receiving no ecological stewardship whatsoever, Price harbored the two rarest plant species encountered in the entire survey: Ammannia coccinea, a wetland plant more typical of higher elevations, and Ludwigia erecta, a tropical species found nowhere else within 1000 kilometers. Native plants increased from roughly 23 percent cover to 31 percent, and plants culturally significant to the Akimel O’Odham and Piipaash peoples climbed from about 1 percent to 4 percent cover, even as invasive giant reed and tamarisk proliferated. These observations suggest that accidental wetlands, often dismissed as degraded wastelands, may quietly complement formal restoration efforts and deserve consideration as legitimate components of urban ecological infrastructure.

The findings also complicate the conventional wisdom surrounding invasive species management. At Rio Salado, managers invested heavily in eradicating species such as chaste tree, giant reed, globe chamomile, and floating water primrose, yet these plants recolonized from the surrounding urban matrix despite the costly efforts. Meanwhile, the authors found no evidence that introduced plants were suppressing native or culturally important species, which increased in abundance even at sites dominated by invaders. The team argues that management rationales should be critically evaluated before expensive eradication campaigns are launched, and that holistic goals such as increasing habitat complexity, supporting culturally significant plant populations, and protecting accidental wetlands may deliver greater conservation value than reactive removal programs.

The project’s most visible innovation may be its art. Working with City of Phoenix land managers, the researchers harvested the three most abundant marsh plants targeted for removal, which would otherwise have been sent to landfills, and pulped their biomass using field retting, caustic cooking, and a Hollander beater to create handmade paper. Each study site was represented by two sheets, one made from the dominant marsh species of 2012 and one from 2022, and the team hand-carved linoleum blocks to print community composition graphics directly onto the paper, with each icon representing one percent of annual plant cover. The resulting artworks were exhibited at the Nina Mason Pulliam Rio Salado Audubon Center, a university, and an art center between 2023 and 2025, reaching more than 2000 people and sparking public conversations about the river’s ecological past and future. For the researchers, this embodied mode of science communication demonstrates that ecological data, ecosystem stewardship, and community engagement can be woven together, and that the plants managers routinely discard may hold unexpected value as both scientific evidence and cultural material. As the United Nations Decade on Ecosystem Restoration unfolds, the Phoenix wetlands offer a compelling lesson: in social-ecological systems, surprises thrive where humans least expect them, and listening to a decade of data, sometimes printed on cattail paper, is the first step toward stewarding them wisely.

The study’s framing draws on a social-ecological perspective in which large-scale pressures, such as damming and urban expansion, interact with rapid pulses like drought, recreation, and management interventions to shape what lives in a given reach of river. The authors constructed a Press-Pulse diagram to situate their case studies within this broader context, treating community composition as a visible clue to underlying ecosystem processes that are otherwise difficult to observe directly.

The research also addresses a recognized gap in the urban wetland literature. A review cited by the team identified the mitigation of urban wetlands acting as ecological traps, where animals are attracted to habitats that ultimately reduce their survival, as one of the field’s largest unresolved questions. By monitoring birds, herpetofauna, and vegetation simultaneously rather than focusing on a single taxonomic group, the study captures some of the trophic complexity that single-group surveys miss, since vegetation structure creates habitat for animals while animals, including humans, disperse seeds that reshape the landscape. Invertebrates and fungi, however, were excluded from the analysis.

The authors emphasize that urban wetlands deliver a wide array of ecosystem services in arid settings where such features are spatially rare, including flood mitigation, water quality improvement, phytoremediation, carbon storage, heat reduction, denitrification, and habitat corridors, and even shelter for people experiencing houselessness. Infrastructure itself can generate habitat, as bridge overpasses create shaded niches for plants and animals seeking respite from the sun. Because long-term monitoring and citizen engagement are considered essential for managing these systems rather than working against them, the project’s blend of decade-scale surveys, collaboration with land managers, and public art exhibitions offers a model for how ecological research can remain connected to the communities it studies.

Subject of Research: Long-term changes in bird, herpetofauna, and plant community composition across managed, restored, and accidental urban wetlands of the Salt River in Phoenix, Arizona

Article Title: A portrait of changing community composition in three urban wetlands of the Salt River, Phoenix AZ USA

Article References: Ramsey-Wiegmann, L., Bateman, H. L., Childers, D. L., Green, H., & Makings, E. (2026). A portrait of changing community composition in three urban wetlands of the Salt River, Phoenix AZ USA. Discover Ecology, 2(1), Article 18. https://doi.org/10.1007/s44396-026-00030-3

Image Credits: AI Generated

DOI: 10.1007/s44396-026-00030-3

Keywords: urban ecology, wetlands, Salt River, Phoenix, biodiversity, herpetofauna, invasive species, restoration ecology, accidental wetlands, plant communities, long-term monitoring, science art

The research, led by Andrew Dumit and colleagues at Watershed Technology Inc. in San Francisco, addresses a problem that distinguishes LCI mapping from most standard machine learning benchmarks. In supervised classification, a misclassified item typically produces some anomalous signal that can be detected downstream. In LCI database mapping, by contrast, an incorrect mapping produces no such anomaly. A wrongly matched dataset will dutifully return an emissions number, and that number will look perfectly plausible in a report. The only reliable way to catch such errors has traditionally been item-level expert review, which is prohibitively expensive at the scale of modern corporate procurement data, where organizations may need to map hundreds of thousands of line items. The result has been an industry-wide reliance on category-average emission factors, which sacrifice the resolution that activity-based accounting is supposed to provide.

The team’s solution is a deliberately structured two-agent system that separates the act of proxy selection from the act of quality assessment through what the authors call an information barrier. The first agent, the mapper, proposes LCI database matches using iterative, tool-augmented retrieval, searching the database and refining its candidates much as a human practitioner would. The second agent, the judge, is architecturally prevented from seeing anything the mapper did. It observes only the original input item, the proposed activity, and the activity’s metadata, and then scores the proposed mapping along two independent dimensions: emissions similarity and material similarity. This enforced separation is not an implementation convenience but a methodological choice, designed to prevent the judge from inheriting the mapper’s biases or rationalizing its choices after the fact.

The performance gains reported in the study are striking. On an evaluation set of 1,039 items spanning seven product categories, the mapper achieved 90.7 percent defensible accuracy — defined as the share of items mapped to an option that a domain expert would approve — with zero abstentions. By comparison, retrieval-based baselines reached only 19 to 43 percent, and prior automated systems, while sometimes avoiding outright errors, abstained on 70 to 73 percent of items, effectively punting the hard decisions back to humans. Because the mapper always commits to an answer, every procurement line item receives a concrete lifecycle-based emission factor rather than a coarse category average, which is precisely what item-level carbon accounting requires.

Even more consequential is what the judge component accomplishes. Because proxy errors are silent, the practical value of any automated mapping system depends on how well it can tell its own confident successes from its quiet failures. At a simulated expert review budget of 20 percent, the judge captured 67 percent of all mapping errors, compared with only 37 to 40 percent for heuristic baselines. When the analysis focused on severe errors — cases where the chosen proxy’s emissions deviated by more than 100 percent from the correct value — the judge caught 74 percent of them. This means that organizations deploying the system can concentrate their scarce expert review capacity on exactly the items where a wrong proxy would most distort the reported footprint, rather than sampling randomly or reviewing everything.

The information barrier also yields a property that is rare in applied carbon accounting tools: calibration. Because the judge never sees the mapper’s reasoning, its quality scores function as an independent audit of each mapping, and these scores turn out to be well calibrated against actual correctness. In practical terms, the system can auto-accept 30 percent of its mappings while incurring an error rate of just 0.3 percent on that auto-accepted subset. This selectivity framework connects the work to a broader literature on selective classification and learning to defer to experts, in which models must know not only how to predict but when their predictions deserve trust. The authors note that large language models are often overconfident and biased self-evaluators when asked to grade their own outputs, which strengthens the case for a structurally independent assessor rather than self-reflection.

The technical architecture draws on several strands of recent research. The mapper’s iterative retrieval approach reflects agentic patterns such as ReAct, in which a language model interleaves reasoning steps with tool calls to ground its decisions in external data — in this case, the LCI database itself. The judge’s evaluation role builds on work on LLM-as-a-judge methods, while its scoring dimensions echo established lifecycle inventory practice, notably the data quality indicators introduced by Weidema and Wesnæs in the 1990s and subsequent proxy selection methodologies for choosing the most appropriate LCI dataset. Earlier machine learning applications in life cycle assessment largely focused on classification or regression tasks; the present work differs by treating the mapping problem as one of defensible proxy selection with explicit, auditable quality control, consistent with the requirements of ISO 14044.

The evaluation combined proprietary internal data from Watershed’s technical assessments with a public benchmark. A reproducible subset of 275 items from the Amazon Parakeet dataset of emission factor recommendations has been released, allowing outside researchers to compare their own systems on identical ground truth. The remainder of the evaluation data derives from proprietary procurement records and cannot be published, and the system code itself is proprietary to Watershed, whose carbon accounting products the company commercializes. All authors are Watershed employees, and the paper states that no external funding was received. These disclosures situate the work within a growing wave of industry-led research into AI-assisted sustainability measurement, alongside recent efforts to apply generative AI to product carbon footprint estimation and LCA data quality assessment.

The broader implications reach well beyond one company’s product pipeline. Accurate Scope 3 accounting has become a central battleground in corporate climate accountability, as regulators, investors, and voluntary disclosure frameworks increasingly demand granular, activity-based figures rather than spend-based approximations. By demonstrating that automated mapping can approach expert quality while providing calibrated signals for where human judgment is still needed, the study sketches a plausible division of labor between machines and domain experts at a scale that neither could achieve alone. If such systems mature, the long-standing trade-off between the resolution of a carbon footprint and the cost of producing it may finally begin to loosen, moving organizations from category averages toward genuinely item-level transparency across their supply chains.

The study arrives at a moment when the infrastructure for lifecycle inventory data itself is expanding. Databases such as ecoinvent, which releases documented change reports with each version update, and newer entrants like China’s HiQLCD, continue to grow in geographic and sectoral coverage, yet the fundamental mismatch between what companies purchase and what databases contain persists. This is why proxy selection has long been recognized as a methodological challenge in its own right, with earlier work proposing structured selection methodologies and expert elicitation to patch inventory gaps for specific product categories such as laundry detergents.

The reporting context also matters. Under the Greenhouse Gas Protocol’s Corporate Value Chain standard and its technical guidance, companies are expected to prioritize the Scope 3 categories most relevant to their sector, and sector-specific technical notes from disclosure platforms such as CDP reinforce that purchased goods and services rank highest in relevance for most industries. Spend-based estimation, which multiplies procurement spending by sector-average emission factors, satisfies reporting requirements cheaply but obscures the physical processes actually driving emissions, limiting the value of the resulting figures for procurement decisions and supplier engagement.

One caution raised in the machine learning literature concerns correlated errors: when multiple automated mappings fail in similar ways, headline accuracy figures can mask systematic biases within particular product categories or database regions. The authors’ emphasis on item-level expert annotation of defensible mappings, supported by a dedicated sustainability data advisory team, reflects an awareness that evaluation quality ultimately bounds the trustworthiness of any automated accounting pipeline built upon it.

Subject of Research: Quality-aware automated mapping of procurement items to lifecycle inventory databases for Scope 3 carbon accounting using a two-agent AI system

Article Title: Quality-aware automation for LCI database mapping

Article References: Dumit, A., Rao, K., Ulissi, S., Watson, S., Feintzeig, J., Joyce, P. J., & Bao, S. (2026). Quality-aware automation for LCI database mapping. Journal of Industrial Ecology. https://doi.org/10.1007/s44498-026-00157-2

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00157-2

Keywords: life cycle assessment, Scope 3 emissions, LCI database mapping, carbon accounting, large language models, two-agent AI, proxy selection, quality assessment, selective classification, emission factors, supply chain, machine learning

The research, led by Anoja N of the National Water Supply and Drainage Board in Sri Lanka, together with R. C. L. De Silva of the University of Kelaniya and J. Prabagar of the University of Jaffna, was driven by a pressing regional and global problem. Coastal waters in northern Sri Lanka carry total dissolved solids (TDS) of approximately 39,400 milligrams per liter, while hypersaline brines in the same region exceed 56,000 milligrams per liter — far above the roughly 35,000 milligrams per liter of natural seawater. With desalination facilities expanding worldwide to combat freshwater scarcity, the concentrated brines they discharge, laden with sodium, chloride, magnesium, and trace metals such as cadmium, lead, and zinc, represent a growing environmental concern. Monitoring these waters for toxic metals is essential, but conventional analytical methods buckle under the extreme salinity.

The analytical difficulty is rooted in what chemists call matrix interference. In a graphite furnace atomic absorption spectrometer, a small volume of sample is dried, pyrolyzed, and finally atomized at high temperature inside a small graphite tube, and the absorption of light by free cadmium atoms is measured at a characteristic wavelength of 228.8 nanometers. The problem is that when the sample contains enormous quantities of dissolved salts, those salts produce their own nonspecific absorption and physical effects that scatter or suppress the analyte signal. In conventional measurements without correction, blank artificial seawater produced an apparent cadmium concentration of 41.88 micrograms per liter — a massive phantom signal — while blank artificial brine yielded 54.18 micrograms per liter. Without intervention, such readings would be catastrophically misleading.

The Sri Lankan team attacked the problem on two fronts. First, they exploited the three-field Zeeman background correction system of their Analytik Jena ZEEnit 700P instrument, which uses a magnetic field to separate the true atomic absorption signal from nonspecific background absorption. The effect was striking. With Zeeman correction applied, the apparent cadmium in blank artificial seawater plummeted from 41.88 to 0.92 micrograms per liter, and in blank artificial brine from 54.18 to 0.24 micrograms per liter. For samples spiked with 10 micrograms per liter of cadmium, measured values dropped from 39.45 to 8.48 micrograms per liter in seawater and from 52.78 to 6.78 micrograms per liter in brine. The second front was the optimization of the graphite furnace temperature program itself — the sequence of drying, pyrolysis, and atomization steps that determines how cleanly the salty matrix is driven off before cadmium atoms are measured.

Using a one-variable-at-a-time approach, the researchers systematically explored drying temperatures of 110, 150, and 200 degrees Celsius; pyrolysis temperatures from 700 to 1000 degrees Celsius; and atomization temperatures from 1200 to 1600 degrees Celsius. For artificial seawater, the sweet spot proved to be drying at 150 degrees Celsius, pyrolysis at 800 degrees Celsius, and atomization at 1400 degrees Celsius. The pyrolysis step is the critical one: it must be hot enough to vaporize the salt matrix but not so hot that volatile cadmium is lost. Above 800 degrees Celsius, cadmium signals declined, evidence of analyte volatilization. For the far more concentrated artificial brine, everything remained the same except the atomization temperature, which needed to rise to 1500 degrees Celsius to achieve maximum absorbance and excellent precision, with a relative standard deviation of just one percent. Three distinct furnace programs emerged: a manufacturer default for clean standards and certified reference materials, and matrix-specific programs tuned for seawater and hypersaline brine respectively.

The payoff of this matrix-specific tuning was dramatic in the recovery studies. When the default furnace program was used both to calibrate and to measure cadmium in seawater, salt-induced signal suppression crushed recoveries to a mere 5 to 26 percent — essentially useless. But when aqueous calibration was paired with measurement under the optimized seawater program, recoveries jumped to between 85 and 107 percent in both undiluted and diluted samples. Hypersaline brine, with its stronger suppression, required an additional step: simple dilution. Undiluted brine analyzed with the optimized brine program yielded recoveries of 83.8 to 92.9 percent, and after two-fold dilution the recoveries reached 94.1 to 95.1 percent. Notably, the study found that cadmium recovery correlates strongly and inversely with electrical conductivity, a practical indicator that higher ionic strength intensifies matrix interference.

Calibration strategy mattered as much as furnace chemistry. External calibration with aqueous standards delivered excellent linearity, with coefficients of determination exceeding 0.999, and precision improved as cadmium concentrations rose, with the relative standard deviation falling from 7 percent at 2.5 micrograms per liter to 2 percent at 10 micrograms per liter. Matrix-matched calibration, though superficially linear, systematically overestimated cadmium — a signal enhancement bias that ruled it out for routine work. The standard addition method, in which samples are spiked incrementally, performed best of all, with a coefficient of determination of 0.9995 and a recovery of 92.7 percent for a fortified seawater sample measured at 0.93 micrograms per liter, confirming negligible residual matrix effects even at ultra-trace levels.

The method’s sensitivity and reliability were rigorously quantified. The limit of detection was 0.50 micrograms per liter in ultrapure water, 0.76 micrograms per liter in artificial seawater, and 0.59 micrograms per liter in artificial brine — comfortably below the concentrations of environmental concern. Measurement uncertainty, evaluated under the EURACHEM/CITAC framework by accounting for standard preparation, calibration, repeatability, and recovery bias, produced a combined standard uncertainty of 0.31 micrograms per liter and an expanded uncertainty of plus or minus 0.62 micrograms per liter at a 95 percent confidence level. The dominant contributor was day-to-day analytical repeatability rather than any systematic bias, and the fortified seawater sample measured 10.16 plus or minus 0.62 micrograms per liter — comfortably within the generally accepted plus-or-minus 10 percent performance criteria for trace metal analysis in complex matrices.

What makes the study especially timely is its explicit embrace of Green Analytical Chemistry. Traditional approaches to taming salt matrices — solid-phase extraction, dispersive liquid-liquid microextraction, cloud point extraction — consume hazardous organic solvents, generate chemical waste, and demand laborious sample preparation. Even greener alternatives such as deep eutectic solvents struggle under the extreme salinity of hypersaline brine. The new method sidesteps all of this: it requires no chemical modifier at all, uses just 20 microliters of sample per analysis, relies on ultrapure water and dilute nitric acid, and completes each determination in roughly two to three minutes. A formal assessment using the Green Analytical Procedure Index (GAPI) classified 13 of 15 evaluation criteria as green, reflecting low reagent consumption, minimal waste, and reduced handling risk. Even the graphite tubes proved durable, surviving approximately 543 analytical cycles and minimizing solid waste.

The authors are careful to note that further validation with natural seawater samples from different geographical locations would strengthen the method’s broader applicability, and the instrument’s instantaneous energy demands are higher than those of low-energy extraction techniques — though the short atomization cycle keeps total energy use modest. Still, the study demonstrates something analytically satisfying: that a conventional, widely available instrument can be coaxed into reliable performance in one of the harshest sample matrices in environmental chemistry simply by understanding and exploiting the volatility differences between the salt matrix and the analyte. As desalination expands across water-stressed coastlines worldwide, this modifier-free, Zeeman-corrected approach offers monitoring laboratories a practical, sustainable route to safeguarding marine waters against cadmium contamination — no exotic reagents required.

The findings carry practical weight for environmental regulators. Cadmium is listed by the World Health Organization and national agencies as a priority pollutant, and drinking water guidelines typically sit in the low microgram-per-liter range, so a detection limit of 0.76 micrograms per liter in seawater places the method squarely within the sensitivity window needed for compliance monitoring. Because the procedure relies on equipment already installed in many water quality laboratories, adoption barriers are low compared with techniques requiring dedicated inductively coupled plasma instrumentation.

The study also illustrates a broader trend in analytical chemistry: rather than adding complexity to overcome interferences, careful exploitation of fundamental chemistry — here, the volatility gap between a chloride-rich salt matrix and a relatively volatile metal — can achieve cleaner results with fewer reagents. The one-variable-at-a-time optimization strategy, though simpler than modern multivariate designs, proved sufficient to isolate the critical temperature thresholds, and the confirmation experiments demonstrated that the resulting furnace programs were robust across repeated cycles.

Alignment with the United Nations Sustainable Development Goals on responsible consumption and life below water underscores the motivation. As brine discharges intensify in arid coastal regions, routine, low-waste monitoring of trace metals becomes a necessity rather than an aspiration. Extending the validated approach to other trace metals such as lead and zinc, and to natural field samples across diverse oceanographic settings, represents the logical next step for this line of research.

Subject of Research: A modifier-free, Zeeman-corrected GFAAS method for trace cadmium determination in seawater and hypersaline brine

Article Title: Environmentally sustainable modifier-free, Zeeman-corrected graphite furnace atomic absorption spectrometry for trace cadmium determination in seawater and hypersaline brine

Article References: N, A., De Silva, R. C. L., & Prabagar, J. (2026). Environmentally sustainable modifier-free, Zeeman-corrected graphite furnace atomic absorption spectrometry for trace cadmium determination in seawater and hypersaline brine. BMC Environmental Science, 3(1), Article 20. https://doi.org/10.1186/s44329-026-00062-w

Image Credits: AI Generated

DOI: 10.1186/s44329-026-00062-w

Keywords: cadmium, graphite furnace atomic absorption spectrometry, seawater, hypersaline brine, matrix interference, Zeeman background correction, green analytical chemistry, desalination brine, GAPI, trace metal analysis, measurement uncertainty, environmental monitoring

The research, led by Atia Elahi of the University of Karachi together with Kamran Khan and Abdul Jameel Khan of Iqra University and Saqib Ur-Rehman of the University of Karachi, covers 161 districts of Pakistan over a twenty-four-year period from 1998 to 2021. Its central question is deceptively simple: how do the relationships between fine particulate matter, or PM2.5, and six key meteorological variables — precipitation, temperature, relative humidity, wind speed, surface pressure, and aerosol optical depth — vary across space, across seasons, and across the country’s dramatic topographic gradient, which stretches from sea level to some of the highest peaks on Earth?

The answer, the authors report, is that these relationships vary enormously, and that capturing that variation requires a statistical instrument purpose-built for the task. The method they developed, called 4D-GTWR, extends an established technique known as geographically and temporally weighted regression, or GTWR, by explicitly adding elevation to the spatiotemporal weighting function. In conventional GTWR, regression coefficients are recalculated for every location and time point, weighted by proximity in space and time, so that relationships can differ from one district to the next and from one month to another. The Pakistani team’s innovation is to make the weight between two observations depend not just on how far apart they are horizontally and chronologically, but also on how different their elevations are. The logic is physical as much as mathematical: the processes that control whether a molecule of pollution accumulates or disperses behave very differently at 200 meters above sea level than at 2,000 meters, so observations from the valley floor should not carry the same influence over a mountain district as they do over a neighboring plain.

The stakes of this modeling challenge are not abstract. PM2.5 — particles smaller than 2.5 micrometers in diameter — is among the most consequential environmental health hazards in the world. Because of its tiny size, it penetrates deep into the lungs, crosses into the bloodstream, and is linked to cardiovascular disease, respiratory illness, cancer, and premature death. Pakistan consistently ranks among the countries with the worst air quality on the planet, with Lahore and other Punjabi cities regularly topping global rankings for pollution during the winter smog season. Meteorology plays a central role in those episodes: temperature inversions trap cold air and its pollutant load in valley basins, low wind speeds allow particles to accumulate rather than disperse, and humidity drives hygroscopic growth of particles, increasing both their mass and their optical signature. Conversely, monsoon rainfall washes particles from the atmosphere, and the summer monsoon season often brings a respite — a pattern the new analysis confirms quantitatively.

The data underpinning the study combine satellite-derived PM2.5 and aerosol optical depth, the standard proxy for aerosol loading measured by instruments observing how much sunlight particles scatter, with meteorological fields from modern reanalysis datasets that blend observations with numerical weather models. The districts of Pakistan form the spatial units, and the seasons are defined in the conventional climatological shorthand: DJF for December, January, and February, the winter smog months; MAM for March, April, and May, the spring; JJA for June, July, and August, the summer monsoon; and SON for September, October, and November, the post-monsoon autumn.

When the 4D-GTWR model was fitted to those data, its performance against traditional alternatives was striking. The model explained roughly 70.28 percent of the variance in PM2.5 during winter, 80.70 percent during spring, 80.87 percent during the summer monsoon, and 72.41 percent during autumn. Those seasonal coefficients of determination were consistently higher than those achieved by ordinary least squares regression, the classical global model that assumes one set of relationships holds everywhere, and by standard GTWR, which accounts for space and time but not altitude. Equally important, the residuals — the portion of variability the model fails to explain — showed less spatial autocorrelation, measured by Moran’s I, a statistic that detects whether errors cluster geographically in ways that signal missing structure. A model that leaves a coherent spatial pattern in its errors is, in effect, telling the researcher what it has overlooked. The lower Moran’s I of the 4D-GTWR residuals indicates that including elevation absorbed a substantial part of that missing structure.

What does the model reveal about the physical system? The authors describe significant seasonal and regional variation in how meteorology drives PM2.5, with distinctly different responses in the plains, the mountains, and the monsoon-influenced zones of the country. In the Indo-Gangetic plains of central Punjab and Sindh, where much of the population and industry is concentrated, winter relationships dominate: cool, stagnant conditions, weak winds, and high surface pressure combine with intense emissions from heating, traffic, and seasonal crop residue burning across the border to build the notorious smog episodes. In these lowland districts, the local coefficients — the district-specific slopes the model estimates for each meteorological variable — paint a picture of an atmosphere primed for accumulation, with humidity and pressure conditions amplifying particle persistence.

In the northern mountainous regions, the picture changes. There, elevation modifies everything: temperatures fall steeply with height, precipitation arrives as snow as often as rain, valley winds follow terrain channels rather than broad synoptic patterns, and the boundary layer — the lowest slice of the atmosphere where pollution mixes — is compressed and sometimes sealed by cold-air pools. Studies of mountain valleys elsewhere have shown that persistent, multi-day inversions can lock PM2.5 in place for days or weeks, and the Pakistani analysis confirms that the coefficients linking meteorology to pollution in such terrain differ systematically from those in the plains. This is precisely the effect that the fourth dimension of the new model is designed to capture: a temperature-PM2.5 relationship calibrated in Lahore is not transferable to Gilgit, and the model now knows that.

The monsoon regions add a third regime. During JJA, the model achieved its highest explanatory power at nearly 81 percent, reflecting the fact that during the monsoon, rainfall acts as a powerful scavenger of particulate matter, removing aerosols through wet deposition. The dynamics of monsoon onset and withdrawal therefore impose a strong temporal structure on PM2.5 in southern and central Pakistan that the model tracks season by season.

The aerosol optical depth variable deserves particular mention. AOD serves as both a predictor and a physically grounded link to satellite observation: it measures the column-integrated extinction of light by particles and correlates with surface PM2.5, though the relationship is modulated by humidity, vertical mixing, and the height of the aerosol layer. In Pakistan, where ground-based monitoring networks remain sparse, satellite-based estimation of surface PM2.5 is the practical foundation of any national assessment, and the local coefficients of the AOD-PM2.5 relationship are exactly the kind of quantity that varies with elevation and season — hygroscopic growth in humid lowland air increases the mass associated with a given optical depth, while dry mountain air attenuates the linkage. The 4D-GTWR framework quantifies this heterogeneity directly.

The authors frame the work as a foundation for air quality management tailored to geography and season rather than imposed as a one-size-fits-all national policy. The implication is concrete: mitigation strategies for the winter smog of the plains — controls on burning, vehicle emissions, and heating fuels, alongside inversion-aware forecasting — will not look like the strategies suited to monsoon-season dynamics in coastal districts or to the terrain-governed pollution of mountain valleys. A framework that knows how the drivers of pollution differ from one district and season to the next can support targeted interventions, seasonal early-warning systems, and more credible estimates of the health burden that PM2.5 imposes on Pakistan’s more than 240 million residents.

The study’s coverage is also notable in itself. Twenty-four years of data, from 1998 to 2021, spans an era during which Pakistan’s urban population grew rapidly, vehicle fleets multiplied, and transboundary pollution flows from the neighboring agricultural regions intensified. The long record allows the model to average over year-to-year anomalies in the monsoon and the winter climate, producing coefficient estimates that are robust to single exceptional seasons.

For the international research community, the demonstration that adding a fourth dimension to spatiotemporal regression improves both fit and residual structure offers a transferable methodological lesson. Any region with significant topographic relief — from the Andes to the Himalayas to the mountain basins of the American West — faces the same challenge: meteorological drivers of pollution are not spatially stationary, and altitude is one of the strongest organizers of that nonstationarity. The Pakistani team’s 4D-GTWR, validated across nearly a quarter century of data and four distinct seasons, provides a template for capturing it.

As South Asia’s air quality crisis deepens and climate change reshapes the monsoon and winter circulation patterns that govern pollution accumulation, tools of this kind move from academic interest to practical necessity. The four-dimensional picture that this study draws of Pakistan’s atmosphere is, in the end, a picture of what it takes to see a pollution problem clearly across one of the most topographically complex nations on Earth.

The research team focused on five artisanal mining sites scattered across Tanzania’s gold-bearing regions: Mgongo, Sekenke, Nyarugusu, Rwamgasa, and Mugusu. Each of these locations has hosted decades of informal gold extraction, in which miners crush ore and mix it with elemental mercury to capture fine gold particles, then squeeze or burn off the mercury to recover the precious metal. This technique, cheap and effective for individual miners, releases mercury directly into soils, tailings, waterways, and the atmosphere. While national inventories and the Minamata Convention on Mercury have targeted reductions in mercury use, the study’s central question was different and in some ways more sobering: even if mercury use stopped tomorrow, how much of it is already embedded in the environment, and how feasible would remediation actually be?

To answer that question, the researchers measured total mercury, abbreviated THg, in soil profiles at each of the five sites, sampling from the surface down to one meter. The results paint a consistent vertical picture across locations. Mercury concentrations are highest in the top 20 centimeters of soil, the zone where spilled amalgam, burned mercury vapor condensing on particles, and contaminated tailings accumulate most heavily. Below that surface layer, concentrations decline with depth but do not disappear. At the Nyarugusu mine site, the team recorded the highest surface-layer value of the entire study: 1.48 milligrams of mercury per kilogram of soil, with a standard deviation of plus or minus 0.02 mg/kg. Even at a depth of 100 centimeters, the same profile still registered 0.12 mg/kg, a moderate but meaningful decrease that demonstrates mercury’s mobility through the soil column. At Rwamgasa, samples pulled from 70 and 80 centimeters deep still contained 0.048 and 0.082 mg/kg respectively, concentrations that, while lower than the surface hotspots, remain well above what would be expected in uncontaminated background soils.

That deep penetration matters for a technical reason: most conventional remediation approaches are designed around excavating or treating the topmost contaminated layer. Soil capping, excavation and landfill disposal, thermal desorption, and stabilizing amendments are all most practical, and most cost-effective, when contamination is shallow and spatially confined. The Tanzanian data show neither condition holds. Mercury has been transported downward through soil profiles over years and decades, likely mediated by percolating rainwater carrying dissolved and colloid-bound mercury, by bioturbation from roots and soil-dwelling organisms that churn material downward, and by physical mixing during continued mining disturbance. Once mercury reaches these depths, excavation becomes impractical, and in situ treatment options face the challenge of delivering remedial agents through heterogeneous soil matrixes where mercury may be locked into mineral lattices or bound to organic matter in forms that resist extraction.

The horizontal dimension of the contamination is equally troubling. Along the Mabubi River, which drains directly through the Mugusu mine site, the researchers observed a clear gradient of mercury in surface-layer sediments distributed along the river’s course. Mercury does not simply sit where it was spilled. Physical erosion of contaminated tailings sends particle-bound mercury into the water column; seasonal flooding resuspends riverbed sediments and redistributes them downstream; and chemical processes, including the formation of soluble mercury complexes and the partitioning of mercury onto fine suspended particles, keep the metal in transit. Biological processes add another vector, as mercury accumulates in aquatic organisms and is excreted or deposited elsewhere. The net effect is a contamination footprint that expands far beyond any disposal site boundary, following the hydrology of the landscape itself.

The downstream destination of that hydrology is what elevates this study from a local environmental assessment to a regional warning. The Mabubi River feeds into Lake Victoria, the largest tropical lake in the world and the water source and fishery for millions of people in Tanzania, Uganda, and Kenya. Prior research, cited by the authors, has already documented elevated mercury in Lake Victoria fish species and associated human exposure in lakeside communities, and a companion study by members of the same team published in 2024 found elevated total mercury in water sources influenced by artisanal gold mining across Tanzania. In lake environments, the central chemical hazard intensifies: inorganic mercury deposited in oxygen-poor sediments can be converted by anaerobic bacteria into methylmercury, the organic form that biomagnifies powerfully up aquatic food chains. A single contaminated tributary therefore represents a slow-release loading source for an entire lake system, one that will continue to deliver mercury for decades regardless of what happens at the mine sites themselves.

The study’s numbers, while specific to Tanzania, sit within a global pattern. The United Nations Environment Programme’s global mercury assessments have identified artisanal and small-scale gold mining as the single largest source of anthropogenic mercury emissions to the environment worldwide, and historical parallels are instructive. Gold mining regions of California, Colombia, Indonesia, and Burkina Faso have all exhibited similar legacies of mercury embedded in soils and sediments long after mining ceased. What the new Tanzanian study adds is a systematic vertical and horizontal mapping of that legacy across multiple active sites, demonstrating that contamination extents are wide, layered, and hydrologically connected, which is precisely the combination that frustrates remediation planning. The authors frame their results explicitly as a challenge to the feasibility of eradication: significant mercury residues detected from surface to deep soil layers and across wide areas of river sediment, driven by simultaneous physical, environmental, biological, and chemical transport processes, create what they describe as a long-term contamination legacy that will continue to degrade environmental quality in affected regions.

For the miners and villagers who live at these sites, the contamination is not an abstraction. Communities at Mugusu and other locations farm contaminated soils, graze livestock on contaminated pastures, and draw water from contaminated rivers. Earlier work by Tungaraza and colleagues documented dietary mercury exposure among adults in the Mugusu mining village using a total diet approach, and separate studies have found mercury residues in free-grazing cattle and domestic fowl in the Geita district. Human health effects of mercury exposure are well established, ranging from tremor, memory loss, and kidney damage associated with inorganic and elemental forms to profound developmental neurotoxicity from methylmercury. The persistence documented in the new study means that even if occupational exposure pathways, such as direct handling of mercury during amalgamation, were eliminated tomorrow, environmental exposure pathways through food crops, livestock, fish, and dust would continue for a generation or more.

Tanzania’s policy architecture already recognizes the problem. The country has completed a Minamata Convention Initial Assessment, maintains an inventory of mercury releases, and published a National Action Plan for Artisanal and Small-Scale Gold Mining covering 2020 through 2025, all aimed at reducing and eventually eliminating mercury use in the sector. The new study does not argue against those efforts; rather, it recalibrates expectations about what they can achieve. Controlling future mercury releases and preventing new contamination are necessary and achievable goals, supported by alternative gold-extraction technologies such as gravity concentration and mercury-free capture methods. But remediation, in the sense of actively removing or neutralizing mercury already distributed through soil profiles to depths of one meter and along kilometers of river sediment, is a different order of problem. The authors’ analysis implies that for regions with similar contamination extents, resources may be better directed toward containment, monitoring, and protecting exposure pathways rather than pursuing full-scale cleanup that the science suggests is not realistically attainable.

The work, funded through the Partnerships for Enhanced Engagement in Research program of the U.S. National Academy of Sciences and USAID, also underscores the value of sustained, locally led environmental monitoring in the global south. Field sampling at artisanal mining sites depends on cooperation with the miners themselves, whom the authors thank for their assistance and guidance, and on laboratory capacity built through international partnerships, in this case with the Smithsonian Environmental Research Center. The data availability statement notes that the dataset can be shared upon request, opening the door for the kind of longitudinal follow-up that this contamination problem demands. Because mercury burdens in soil and sediment respond slowly to changes in mercury emissions, only repeated sampling over years will reveal whether the Minamata-era interventions are bending the curve of environmental mercury, or whether the legacy documented at Mgongo, Sekenke, Nyarugusu, Rwamgasa, and Mugusu is destined to spread still further through the rivers of the Lake Victoria basin.

Appolloni, along with a team of graduate and undergraduate researchers at the University of Toronto’s School of the Environment, surveyed instructors at five Canadian universities — Dalhousie, McGill, Toronto, Waterloo and Simon Fraser — to find out what, if anything, professors teaching environment-related courses are doing about the eco-anxiety in their classrooms. The study, published open access in the Journal of Environmental Studies and Sciences, contacted 157 instructors identified through university websites as teaching environmentally focused courses. Forty-one completed the survey, and six follow-up interviews were conducted, four with survey respondents and two with external scholars who had published relevant work on eco-anxiety. The results paint a picture of a teaching community that is deeply concerned but fragmented, under-supported and, in some cases, itself paralyzed by despair.

The scale of the underlying problem is well documented. A landmark global survey published in The Lancet Planetary Health polled 10,000 young people aged 16 to 25 across ten countries and found that 75 percent reported feeling frightened about the future, 45 percent said climate anxiety affected their daily functioning, and more than half reported feelings of sadness, anger, helplessness and guilt — along with a striking sense of betrayal by governments and institutions that have failed to act. A parallel survey of 1,000 Canadian youth produced even starker numbers: 48 percent of respondents believed humanity is doomed, and 39 percent expressed hesitancy about having children because of climate change. In the United States, a separate poll found that roughly three-quarters of Gen Zers and Millennials said climate change had influenced major life decisions.

Crucially, the researchers emphasize that eco-anxiety is not a pathology. Clinicians and scholars, including psychologist Panu Pihkala and psychoanalyst Anouchka Grose, frame it as a normal — even rational — response to a genuine environmental emergency. Psychiatrist Lise Van Susteren has described some young people as suffering a kind of “pre-traumatic stress disorder,” anxiously rehearsing a future catastrophe that has not yet arrived, but the prevailing view in the literature is that grieving a deteriorating natural world reflects something deep in human psychology. Drawing on E.O. Wilson’s biophilia hypothesis, the authors note that humans evolved in forests and on savannahs for the vast majority of the species’ existence; our attachment to nature is evolutionarily grounded, and its loss is therefore felt as a real, embodied bereavement. Suppressing that grief, research on emotion regulation suggests, carries its own cognitive costs: studies show that emotional suppression depletes finite attentional resources and impairs memory and performance.

The survey findings reveal an uneven patchwork of institutional response. Among the 41 respondents, only three said they had not observed eco-anxiety in their students. The researchers sorted the rest into three categories. Thirty-nine percent had taken little to some action, typically confining their response to class discussions or supplementary readings. Thirty-four percent addressed eco-anxiety directly or indirectly — naming the phenomenon, assigning journalling or creative projects, organizing nature walks, or pointing students toward positive news to counterbalance the doom-laden headlines — but treated it as an add-on rather than a core element of the course. Only 27 percent made addressing eco-anxiety a deliberate, integral focus of their teaching, incorporating multiple strategies, sometimes including mindfulness exercises designed to build coping skills and foster a hopeful outlook.

The obstacles instructors cited were varied and telling. Fourteen pointed to the sheer magnitude and multiplicity of student anxieties, which range from war and economic precarity to grades. Seven said packed curricula left no time for emotional work. Six struggled to strike a balance between presenting honest negative data and avoiding either paralysis or false comfort. Five cited a lack of training, with two fearing that intervening might make things worse. Five pointed to an absence of institutional support, and four admitted that their own sense of doom prevented them from helping students. Most strikingly, eight respondents identified countering negative messages from their own colleagues as a principal challenge. One instructor reported that a student had abandoned plans for an environmental engineering career after concluding from a course that the situation was hopeless; another recounted an engineering student saying the material was simply too depressing to study. “I had to ask myself,” that instructor said, “what am I doing? My model of teaching is about empowerment, and knowledge is not sufficient for empowerment.”

It is precisely this gap — between transmitting knowledge and building capacity for action — that anchors the study’s theoretical framework. The authors argue from a liberationist perspective on pedagogy, drawing on Paulo Freire’s concept of education as “the practice of freedom.” In a Freirean classroom, students co-learn with educators, develop critical consciousness, and participate as agents in shaping their own future rather than passively absorbing facts. The authors extend this logic with Maria Ojala’s concept of “critical hope”: not blind optimism, but a clear-eyed acknowledgment of how bad things are, combined with the conviction that change remains possible and that one has both personal agency and trustworthy allies in pursuing it. Freire himself called hope an “ontological need” that must be exercised through struggle, lest it dissipate into hopelessness. Importantly, critical hope is not a feel-good dodge. A study in Positive Psychology found that very happy people were more, not less, likely to take environmental action, suggesting that psychological well-being and serious engagement are complementary rather than opposed.

The framework also demands what Megan Boler terms a “pedagogy of discomfort” — allowing students the space to sit with painful emotions rather than suppressing or bypassing them. Britt Wray, author of Generation Dread, argues that Western societies lack the emotional intelligence to “stay with the trouble,” and that pushing down despair diverts energy away from the imagination and resilience needed to envision workable futures. Meanwhile, Mezirow’s transformative learning model, often invoked in sustainability education, has been criticized for focusing almost exclusively on cognitive transformation; its signature “disorienting dilemmas” can trigger denial and despair rather than action unless the emotional dimension is deliberately addressed. The authors also cite Sharon Stein’s typology of university sustainability approaches — mainstream, critical, and “beyond sustainability” — arguing that the last, which prepares students to face wicked problems with intellectual, affective and relational stamina, is best suited to the task.

The paper’s conclusion is a call for a coordinated, university-wide conversation rather than a prescription for any single technique. The authors acknowledge clear limitations: five relatively large universities and a modest sample size preclude firm generalizations about Canadian higher education as a whole. But the pattern is hard to ignore. Individual counseling services exist on every campus studied, and innovative programs are emerging — Simon Fraser University has hired an eco-chaplain devoted to ecological grief, Dalhousie University launched an Eco-Hope program after its own survey found rising hopelessness among sustainability students, and a University of Toronto symposium combined dialogue, art and future-envisioning practices. Yet these initiatives rely on patchwork funding and isolated champions. What is missing, the authors argue, is systemic integration: shared responsibility among instructors, students, administrators and staff, sustained funding, and training that equips professors — not all of whom are comfortable or skilled with emotional labor — to foster critical hope.

Ultimately, the study reframes the purpose of environmental higher education for an age of crisis. A pedagogy that informs students about planetary decline without building their psychological resilience, the authors contend, is inadequate — and may even be perpetuating the hopelessness it should be combating. To produce what one survey respondent called “actors in their own destiny,” universities must educate the whole person: mind, body and heart.

The central mechanisms are surprisingly convergent. Nanoparticles can enter endocrine cells through circulation and lymphatic routes, bind to hormone receptors and either activate or inhibit downstream signaling pathways. Silver and zinc nanoparticles, for example, can bind estrogen receptors and display estrogenic activity, effectively mimicking natural hormones. Perhaps the most important pathway, however, is the generation of reactive oxygen species. These reactive molecules induce oxidative stress, damaging cellular structures and disrupting the signaling cascades that coordinate hormone synthesis and secretion. In parallel, nanoparticles activate innate immune responses, provoke the release of inflammatory cytokines, interfere with enzymes and transporters involved in endocrine regulation, and impair hormone metabolism. The end result of these disturbances can range from reproductive failure to metabolic syndrome, and the review stresses that the endocrine-disruptor effect of metallic nanoparticles depends heavily on their capacity to release metal ions that inhibit nuclear receptors and interfere with natural hormone action.

The thyroid gland emerges as one of the most intensively studied targets. Titanium dioxide nanoparticles administered to rats at 5 milligrams per kilogram for 21 days produced histopathological changes, including abnormal follicular cell morphology and epithelial damage, alongside a decrease in the sodium/iodide symporter, elevated caspase-3 apoptotic protein, and increased pro-inflammatory factors such as interleukin-1 beta and tumor necrosis factor alpha. In vitro, follicular cell viability declined. Silver nanoparticles affected the messenger RNA expression of thyroid hormone related genes in laying hens, raising serum triiodothyronine without altering thyroxine and prompting tissue-specific expression of genes involved in thyroid hormone metabolism, including deiodinases and the hormone transporter MCT10. Longer term exposure in female rats decreased thyroxine concentrations and caused degeneration of thyroid follicles. Intriguingly, the story is not uniformly negative: biogenic silver nanoparticles of 17 to 35 nanometers have been reported to raise triiodothyronine and thyroxine levels in experimentally hypothyroid rats, suggesting a potential therapeutic dimension.

Zinc oxide nanoparticles add another layer of complexity. Fed to albino rats at high doses for 30 days, they decreased thyroxine and thyroid stimulating hormone while producing histopathological damage. When combined with lead acetate over eight weeks, they lowered triiodothyronine and thyroxine, raised thyroid stimulating hormone and tumor necrosis factor alpha, and triggered significant overexpression of the nuclear factor erythroid 2 related factor, a master regulator of antioxidant response. The review also situates metallic nanoparticles within the broader landscape of thyroid-disrupting chemicals, noting that phthalates, bisphenol-A and polybrominated diphenyl ethers released by micro- and nanoplastics are established thyroid disruptors that enter through the gastrointestinal system and interfere with hormone secretion, transport and the hypothalamo-pituitary-thyroid axis. Transcriptomic studies further show disruption of thyroglobulin synthesis and interference with Pax8 and CERB gene expression, which are crucial for thyroid growth, differentiation and hormone secretion. By contrast, direct studies of the parathyroid gland remain scarce.

The thymus, a primary lymphoid organ where T cells mature, illustrates the overlap between endocrine and immune toxicity. Silver nanoparticles stimulate immune responses and promote cytokines such as interleukin-6, driving inflammation, while at higher concentrations they induce oxidative stress and DNA damage. Zinc oxide nanoparticles can activate human T cells and even preferentially kill cancerous T cells, with toxicity tracking the level of T cell activation, an effect attributed largely to reactive oxygen species. Orally administered zinc oxide caused degenerative changes in the thymus and spleen of rats, reduced thymus weight over three months of sub-chronic exposure, and shifted staining patterns consistent with decreased proliferation and increased p53-associated damage. Cadmium nanoparticles produce thymic atrophy, depletion of CD4-positive and CD8-positive T cells, and oxidative stress and apoptosis, with prenatal exposure producing persistent changes in immune cell repertoires. Cadmium oxide and lead oxide nanoparticles altered T cell populations in mice, while nickel nanoparticles stimulated pro-inflammatory cytokines that affect T cell activation, and titanium dioxide nanoparticles showed a striking dose dependence, provoking inflammation at low doses and cytotoxicity at high doses.

The pineal gland, source of the antioxidant hormone melatonin, faces a double-edged relationship with nanomaterials. Zinc oxide nanoparticles damage both brain and pineal tissue through reactive oxygen species, and in a counterintuitive finding, melatonin pretreatment actually enhanced zinc oxide toxicity in C6 glial cells rather than protecting them. Exposures to aluminum oxide and silver nanoparticles are associated with neurotoxicity and pineal dysfunction. Yet melatonin-loaded nanocarriers are now being deployed therapeutically, delivering the hormone with superior efficacy against disease than free melatonin alone. Melatonin complexed with nanometric gold offered testicular protection, a melatonin palladium complex induced apoptosis in human lung adenocarcinoma cells, and palladium nanoparticles served as platforms for detecting melatonin and related molecules electrochemically. This duality, harm and remedy coexisting in the same material class, runs throughout the review and underscores the difficulty of blanket risk assessments.

At the top of the hormonal hierarchy, the hypothalamo-pituitary axis is vulnerable in ways that ripple across the entire endocrine system. Copper nanoparticles inhibited pituitary gonadotropin secretion in male mice, suppressing luteinizing hormone and follicle stimulating hormone alongside degenerative changes in the anterior pituitary. Nanoparticles interfere with the ability of glial cells to regulate gonadotropin-releasing hormone, and trans-generational effects have been reported, with lithium carbonate nanoparticles affecting both the pituitary-gonadal and pituitary-thyroid axes in female rat progenies. The adrenal glands are similarly susceptible. Nanoparticles can activate the hypothalamic-pituitary-adrenal axis, surging adrenocorticotropic hormone and driving structural and functional changes, while also affecting key steroidogenic enzymes including aromatase, 5-alpha reductase, CYP11A1 and CYP11B1. Silver nanoparticles increased cortical functional activity and proved cytotoxic to adrenal pheochromocytoma cells through ATP depletion, cytoskeletal changes and oxidative stress, and in zebrafish they suppressed the cortisol response to acute stress by altering the hypothalamic-pituitary-interrenal axis at the transcriptomic level. Short-term oral titanium dioxide exposure impaired adrenal cortex and medulla structure in rats without apparent general toxicity, in vitro studies confirmed uptake of magnetic iron oxide nanoparticles by adrenal cortical cell lines, and nano-arsenic was reported to induce adrenal tumors through oxidative tissue damage and hormonal disturbance.

The pancreas and placenta complete the picture. Titanium dioxide nanoparticles caused histopathological and biochemical changes in the rat pancreas, with effects attributed to oxidative stress, inflammation and apoptosis of beta cells, decreased insulin secretion and disrupted carbohydrate metabolism, collectively raising diabetes risk. Yet low-dose zinc oxide nanoparticles normalized pancreatic function in diabetic rats by protecting beta cells through GLP-1 and oxidative stress pathways, and nanoselenium protected endocrine and exocrine pancreatic function in a rat model of acute pancreatitis through antioxidant and anti-inflammatory actions. Nanoparticles are also being pursued for targeted insulin delivery, with polymeric nanoparticles proposed for oral insulin administration and oligonucleotide-loaded particles explored for gene and cell therapy to stimulate insulin production. At the placenta, nanoparticles can cross the blood-placental barrier, disrupt placental function and the release of signaling factors needed for fetal development, and dysregulate the placental secretome with consequences for angiogenesis and vascularization. Ex vivo human placental perfusion studies have illuminated these risks, though the review notes that current regulatory guidelines do not sufficiently address placental toxicity, while smart nanoparticles engineered to target the placenta could eventually improve its function.

The review closes with a call for methodological renewal. Because nanoparticle toxicology involves multiple organs and interdependent molecular mechanisms, no single experimental model can simulate a realistic scenario; the author advocates microfluidic and organ-on-chip systems combined with single-cell culture, multi-omic approaches and in silico models to achieve accurate in vivo evaluation. Attention to biomarkers, mitochondrial versus cytoplasmic sources of reactive oxygen species, and the combined effects of multiple nanoparticles is warranted. A new generation of redox nanoparticles, including cerium oxide, boron cluster and silica-containing particles, offers promising antioxidant properties and the ability to cross the blood-brain barrier for treating endocrine disruption. The overarching conclusion is sobering but constructive: endocrine toxicity must remain central to risk assessment if nanotechnology is to be safely integrated into healthcare, theranostics, industry and the environment.

Subject of Research: Mechanisms of endocrine gland toxicity induced by metallic nanoparticles

Article Title: Endocrine gland specific mechanistic paradigms of toxicity induced by metallic nanoparticles

Article References: Rana, S. V. S. (2026). Endocrine gland specific mechanistic paradigms of toxicity induced by metallic nanoparticles. Discover Toxicology, 3(1), Article 20. https://doi.org/10.1007/s44339-026-00064-y

Image Credits: AI Generated

DOI: 10.1007/s44339-026-00064-y

Keywords: metallic nanoparticles, endocrine glands, oxidative stress, thyroid toxicity, hormone disruption, hypothalamo-pituitary axis, reactive oxygen species, adrenal glands, pancreas, placenta, nanotoxicology, antioxidants

The study, published in the journal Discover Conservation, examined 481 unique comments drawn from the two Facebook posts, both dated October 7, 2025. The researchers deliberately waited about two months before scraping the data using a cloud-based automation service, ensuring that reactions and comments had stabilised before analysis. The provincial government page, with roughly 182,000 followers, attracted a more localised audience, while SunStar Davao’s 1.3 million followers gave the story a broader national reach. From the full set of 806 comments, the team excluded nearly 60 percent for reasons of relevance and substance, leaving a core of 481 substantive remarks. Many commenters appeared to have direct or indirect connections to Davao Oriental, evident from local expressions, self-references, and vivid descriptions of local ecosystems, livelihoods, and mining-related impacts, although the authors caution that these users cannot be treated as statistically representative of the general population.

The analytical approach was grounded in Social Representations Theory, a framework that explains how shared beliefs, values, and meanings are socially constructed and circulated through communication, shaping how communities understand contested issues. Rather than treating each comment as an isolated opinion, the researchers interpreted them as manifestations of collectively negotiated understandings of environmental risk, governance, livelihood, and community well-being. Comments written in Cebuano and Filipino were translated into English with the assistance of artificial intelligence tools and then carefully reviewed by team members fluent in the local languages, who worked to preserve tone, sarcasm, and culturally embedded meanings. Using Braun and Clarke’s six-step reflexive thematic analysis, two independent groups of researchers coded the comments separately, then reconciled discrepancies by consensus to strengthen reliability.

The results coalesced into six themes of public sentiment. The largest, community voice and protest, captured 139 comments and revealed a widespread sense that local voices were ignored or marginalised by those in authority, breeding scepticism toward official claims about biodiversity conservation, environmental rehabilitation, and responsible mining. Governance and accountability followed with 107 comments, expressing frustration over perceived government complicity, accusations of corruption and bribery, and anger that mining near the sanctuary has already cleared approximately 200 hectares of forest land. Environmental degradation drew 94 comments, with users describing stripped-bare mountains, dried rivers, declining biodiversity, and fragmented forest cover, and insisting that no amount of money or replanting can fully restore these sites. Calls for action and hope accounted for 55 comments, social and health impacts 20, and economic and livelihood impacts 15.

That final theme, though the smallest, captures perhaps the sharpest tension. Some commenters acknowledged that families rely on mining for their livelihoods and worried about job losses should operations cease, while others argued that communities survived long before the industry arrived and cautioned against dependence on a single sector. Calls mounted for alternative livelihood programmes, alongside dissatisfaction with the distribution of mining benefits, which many felt were disproportionately enjoyed by politicians, officials, and companies while communities bore the environmental costs. The study situates this local dispute within a global debate, noting that the Philippines ranks fifth in untapped reserves of valuable metal ores, with an estimated value of one trillion US dollars, and that mining productivity has grown exponentially worldwide over the past 150 years as the backbone of industrial and digital development.

What makes Davao Oriental exceptional, and the dispute so charged, is the ecological stakes. The Mount Hamiguitan Range Wildlife Sanctuary is home to the critically endangered Philippine Eagle, the national bird, and continues to yield new species to science, including the long-horned beetle Paraskeletodes hamiguitanensis and the recently described endemic plant Begonia corazoniae. The sanctuary’s ecological integrity is linked to downstream ecosystems that connect to Pujada Bay, a marine protected area recognised as one of the Most Beautiful Bays of the World and tentatively listed as a UNESCO World Heritage extension. The bay is an ancient nesting site for green, hawksbill, and olive ridley turtles, shelters the dugong, and lies within the Coral Triangle, the global epicentre of marine biodiversity. Parts of the mining tenements, which span roughly 5,000 hectares of coastal and mountainous concessions, overlap with the ancestral domain of the Mandaya indigenous people, whose cultural identity and ecological knowledge are deeply tied to the land.

From these findings, the team developed the RESPECT framework, a community-based action model that translates public sentiment into practical conservation strategy. Its components call for re-evaluating land-use plans and zoning policies to clearly separate conservation zones from mining concessions; expanding the boundaries of the sanctuary and the bay under a ridge-to-reef, whole-of-island perspective; sustaining environmental protection initiatives that empower indigenous and local communities as conservation champions; providing ecological regeneration of mined-out areas grounded in scientific assessments of soil, water, and biodiversity; and reducing reliance on mining income through alternative and additional livelihoods. The framework further urges voters to prioritise candidates with firm anti-mining and environmental agendas, promotes multi-stakeholder collaboration involving universities, churches, and civil society, and emphasises environmental education in schools to cultivate a stewardship ethic among the young. The framework was refined through consultation with academics, priests, environmental lawyers, church leaders, and advocates, ensuring it reflects community and multidisciplinary perspectives rather than academic interpretation alone.

The political context is shifting rapidly. The provincial government has issued statements inclined toward the cessation or closure of the nickel mining operations and has formally requested their rescission through provincial resolution. A House Bill has been filed to declare Davao Oriental a mining-free zone, although its passage has since been halted. Internationally, environmental groups have repeatedly called for a complete halt to mining near the heritage site, and reporting on nickel mining damage near the sanctuary has stirred national outrage. The researchers argue that these overlapping pressures make sentiment analysis especially valuable: while social media data cannot deliver sector-specific conclusions, they reveal how socially visible participants frame the trade-offs between economic development and environmental protection, where public conflict originates, and where community expectations for accountability and livelihood transition are most clearly articulated.

The study’s broader message is that digital discourse deserves a formal seat in environmental governance. Guided by Social Representations Theory, the authors contend that collective meanings about mining are constructed, shared, and reinforced through online social interaction, and that these expressions can and should guide public policy and social action aimed at natural resource management and biodiversity conservation. In a province where indigenous cultural communities maintain longstanding material, cultural, and spiritual relationships with land and sea, ignoring public sentiment risks both ecological damage and social conflict. The team recommends institutionalising the RESPECT framework to re-evaluate and reimagine the province’s mining sector, insisting that transparency and genuine community engagement are prerequisites for any ecologically centred mining practice. As mining faces mounting scrutiny worldwide, this study suggests that the comments section, often dismissed as noise, may in fact be a rich and largely untapped data source for the science of conservation.

The historical backdrop helps explain why emotions run so high in these online exchanges. The Philippines industrialised its mineral sector during the American occupation, and in 1995 the Mining Act, Republic Act 7942, was enacted to stimulate extraction and economic growth. Yet the industry has long been criticised for habitat degradation, biodiversity loss, and pollution, with environmental impact assessments frequently functioning more as compliance exercises than genuine safeguards. In Banaybanay, Davao Oriental, an overflowing mining catchment pond released orange laterite sediment rich in chromium and nickel, threatening freshwater and marine life in nearby waterways.

The choice of Facebook as a data source reflects its dominance in the Philippines, where it is among the most widely used platforms and where users often express sharply polarised political opinions. This makes the platform a natural arena for contested environmental issues, though the researchers acknowledge important limitations: only publicly shared comments could be collected, and demographic details such as age, locality, and political background could not be verified, meaning the sample cannot be generalised to the wider population.

The authors also frame the work as a form of citizen science, in which online expressions of environmental concern can serve as evidence for monitoring ecological decline and preventing conflict. This matters in Mindanao, where mining grievances have intersected with insurgency, poverty, and social exclusion, and where the killing of indigenous peoples has fuelled resistance to corporate extraction. Because mining impacts are not experienced uniformly, but are shaped by gendered roles and inequalities, the researchers argue that sentiment analysis can reveal whose concerns are loudest and whose remain unheard, helping policymakers design responses that are inclusive, equitable, and culturally sensitive rather than purely economic in orientation.

Subject of Research: Social media sentiment analysis to inform community-based conservation in mining-affected landscapes near a UNESCO World Heritage Site in Davao Oriental, the Philippines

Article Title: Analysing social media sentiments to inform community-based conservation in mining-affected landscapes near a World Heritage Site in Davao Oriental, the Philippines

Article References: Villegas, J., Verzosa, R., Bauyot, M. F., Hemillan, J., Pilar, J., Ngoho, M. C. J. N., Rivera, E., Medina, M. N., Antonio, E., & Sumile, E. (2026). Analysing social media sentiments to inform community-based conservation in mining-affected landscapes near a World Heritage Site in Davao Oriental, the Philippines. Discover Conservation, 3(1), Article 39. https://doi.org/10.1007/s44353-026-00107-w

Image Credits: AI Generated

DOI: 10.1007/s44353-026-00107-w

Keywords: social media sentiment, community-based conservation, open-pit mining, UNESCO World Heritage Site, Mount Hamiguitan, Davao Oriental, biodiversity, nickel mining, thematic analysis, Facebook, indigenous communities, ecological regeneration

The premise rests on a simple pedagogical insight, famously captured by the Senegalese forester Baba Dioum in 1968: we will only conserve what we love, love what we understand, and understand what we are taught. Biological collections make that understanding tangible. Each specimen lot preserves multidimensional information, spanning historical, geographical, morphological and ecological dimensions, that abstract textbook diagrams simply cannot convey. When students handle a bivalve and trace the articulation of its valves, or spiral a gastropod shell in their fingers while considering its mode of life, they are engaging with empirical evidence, scientific methodology and the documentation practices of real research, all at once.

Yet the La Plata team is careful to stress that collections do not teach by themselves. Their educational value depends entirely on the intentional work of teachers who design activities around the specimens. That design principle is visible throughout the collection’s architecture. The researchers organised 254 specimens into five molluscan classes and assigned each lot to one of three purposes: teaching, marked DOC; outreach, marked EXT; and repository, marked REP, a reserve of replacement material for items damaged by handling. Colour-coded labels and a lot-naming system combining the first letters of the phylum and class allow staff and students to navigate the collection quickly, a practical but consequential detail for anyone who has watched a class of undergraduates wait while an instructor searches for the right jar.

The curation itself followed rigorous standards. Specimens were obtained during field surveys by the research group, mostly empty shells collected after natural death, using a minimal-impact sampling approach. Preservation relies on two methods: dry storage in jars and bags, and wet preservation in 70 percent ethanol. Taxonomic identifications were verified and updated against authoritative databases such as the Catalogue of Life and MolluscaBase, and the collection’s records were migrated into a spreadsheet structured according to the Darwin Core standard, the international data framework used by natural history collections worldwide. A data-entry protocol now governs how every field is completed, ensuring that information remains consistent as the collection grows.

The most eye-catching innovation, however, is what happened next: digitisation. Working with the Functional Digitisation Unit of the Museo de La Plata, the team generated three-dimensional models of specimens using two very different technologies. The first is deliberately low-tech in spirit. Using Polycam, a free smartphone application that builds 3D models through photogrammetry, the researchers created interactive models of two invasive bivalves, the golden mussel Limnoperna fortunei and the Asian clam Corbicula fluminea. The choice of a free mobile tool was intentional, so that any teacher with a phone or tablet could replicate the workflow and bring the same resources to their own students without institutional funding or specialised equipment.

The second technology operates at the other end of the precision spectrum. An Artec Micro II scanner, offering a resolution of five microns within a maximum scanning volume of 20 by 20 by 15 centimetres, was used to digitise seven non-native mollusc species. The result is the first three-dimensional Virtual Malacological Collection of invasive molluscs in Argentina, hosted online and freely explorable. In the rendered models, viewers can rotate a veined rapa whelk, Rapana venosa, and click on numbered points to see the anatomical names of individual shell parts, an experience that merges taxonomic instruction with the tactile exploration normally reserved for people standing in front of a museum drawer.

The virtual collection is already reshaping teacher training. In the Invertebrate Biology course at La Plata’s Faculty of Humanities and Education Sciences, students who are themselves future teachers are asked to design lesson plans about invasive mollusc species using the virtual collection as the central resource. The assignment is deliberately open-ended, and the results have ranged from illustrated fact sheets with QR codes linking to the 3D models, to guided discussion questions built on scientific articles and newspaper reports, to a Kahoot! quiz and a full board game. Classroom observations suggest that manipulating digital specimens without physical constraints helps students connect theoretical knowledge with socioscientific issues such as biological invasion, a problem the team notes is among the most pressing global threats to native species and ecosystem function.

The physical collection, meanwhile, travels beyond the university. Through a university outreach project, native marine molluscs from the Buenos Aires coast, including species such as the ribbed mussel Aulacomya atra, the blue mussel Mytilus edulis and the violet clam Eucallista purpurata, form the basis of workshops delivered in secondary schools and teacher training institutes across the province. Students work with dichotomous keys and malacological fact sheets accessed through QR codes, comparing shell articulation, umbo position and coiling patterns to infer how each animal fed and where it lived. Workshops close with debates on conservation strategies and the ethical responsibilities of citizens. At public fairs, visitors encounter a black ‘mystery box’ containing the same specimens; selecting one triggers questions about identification, habitat and human uses. The team reports that the activity routinely corrects misconceptions, for example the spherical, transparent egg capsules of Pachycymbiola brasiliana, which beachgoers frequently mistake for turtle eggs.

The researchers argue that these experiences address a documented gap. Earlier surveys cited in the study found that between 71 and 81 percent of students in Argentine teacher training programmes valued biological collections as educational resources, yet most did not know how to construct or use them, and between 33 and 38 percent reported being unable to use them in secondary schools for lack of material. Strikingly, between 69 and 81 percent of trainee teachers were unfamiliar with virtual collections, while 81 percent said they would use them if they could. Collections remain underutilised in classrooms, the authors note, because of structural limits such as scarce materials, restricted institutional access and insufficient teacher training, problems that digitisation can materially ease by removing temporal and spatial barriers to access.

The La Plata team’s conclusions are measured but forward-looking. A catalogued, organised and regularly updated physical and virtual collection streamlines the selection of teaching materials, reduces the wear and deterioration that handling inflicts on specimens, and gives teachers and students ready access to resources they could never build alone. But access alone, they caution, guarantees nothing; the pedagogical payoff depends on teachers who know these resources exist and plan deliberately around them. If the model spreads, the humble mollusc, ancient, abundant and endlessly varied, could become an unlikely ambassador for biodiversity literacy, its shells preserved in ethanol and resin, and now in pixel-perfect three dimensions, teaching a generation of students why the living world is worth understanding, and worth keeping.

The choice of molluscs as the foundation for this collection was no accident. Mollusca is one of the largest phyla in the animal kingdom, and its shells are durable, visually striking and easy to handle, making them unusually well suited to classroom work. The researchers also point to the phylum’s deep entanglement with human society, from food and ornament to economic and cultural uses, which gives teachers multiple entry points for connecting biology to students’ everyday experience.

The project sits within a broader institutional effort by the Biodiversity Education Study Group, known as GEEBio, whose Invertebrate Biology Collection Group has taken on the long-term administration of the collection. This matters because sustaining a teaching collection is not a one-off project but an ongoing commitment to curation, reconditioning and documentation, work that often goes unrecognised in academic settings.

The study also situates biological collections within ex situ conservation strategies, alongside seed banks and arboretums, framing them as primary repositories of biodiversity knowledge rather than mere teaching props. From this perspective, a well-maintained collection serves research, conservation management and public engagement simultaneously, with exhibitions and outreach activities fostering the public participation that conservation ultimately depends upon.

For teachers wondering how to begin, the researchers draw on prior work identifying three practical pathways: searching data portals to build customised lessons, collaborating with researchers to develop or adopt an existing resource such as the virtual collection, and contributing to citizen-science portals that expand collections with new data. Each route lowers the barrier that unfamiliarity has created.

Adopting standards like Darwin Core also connects a modest university collection to a global infrastructure of natural history data, meaning that local specimens can eventually inform research on species distributions, climate change and biological invasions far beyond Argentina. The open access publication of the study itself reflects this ethos of widening participation, allowing educators anywhere to examine the methods and adapt the model to their own institutional collections and regional fauna.

Subject of Research: Use of physical and virtual malacological collections as interactive educational resources for biodiversity learning

Article Title: Malacological collections as educational resources for interactive visual learning and biodiversity exploration

Article References: Custodio, H., Motta, J., Amoia, A., Arregui Longui, M., Vilches, A., & Darrigran, G. (2026). Malacological collections as educational resources for interactive visual learning and biodiversity exploration. Discover Ecology, 2(1), Article 19. https://doi.org/10.1007/s44396-026-00037-w

Image Credits: AI Generated

DOI: 10.1007/s44396-026-00037-w

Keywords: malacology, biological collections, biodiversity education, 3D digitisation, virtual collection, invasive species, environmental education, teacher training, Darwin Core, photogrammetry, science outreach, molluscs

The team built a stock-based dynamic material flow analysis model, a technique that tracks how materials accumulate in use, how long they stay there, and when they flow out as waste. Historical stock development, material intensities and average lifetimes for Dutch macro infrastructure between 1950 and 2023 were drawn from the SUBLIME database, and the survival of infrastructure assets was modeled with a Weibull distribution, the statistical function most widely recommended for long-lived structures. Crucially, the researchers did not treat infrastructure as a single undifferentiated mass. They distinguished highways and railways, water works, oil and gas networks, electricity infrastructure and municipal utilities, because each category obeys its own lifespan dynamics, demand drivers and material intensities.

To capture uncertainty about the future, the team developed three scenarios running to 2070. A Reference scenario follows central projections from infrastructure asset owners and official outlooks, including Rijkswaterstaat traffic elasticity estimates, CBS population forecasts, the Dutch Delta Programme and KNMI climate scenarios. A High scenario pairs rapid energy transition and economic growth with aggressive end-of-life recovery mandates and advanced recycling infrastructure, while a Low scenario assumes weak policy support and slow technological diffusion. Demand was quantified through the stock-flow-service nexus, multiplying projected population, which may grow from about 17.9 million to between 19 and 24 million inhabitants by 2070, by per-capita infrastructure requirements and material intensities.

The headline finding is a dramatic divergence between materials. By 2070, annual demand for steel is projected to increase by 94 to 104 percent compared with 2024 levels, and concrete demand by 58 to 94 percent, while asphalt demand remains essentially flat, growing by only 1 to 2 percent. The driver of the surge is not road building but the energy transition. Offshore and onshore wind turbines, solar installations and the reshaping of oil and gas pipelines dominate future inflows, with energy infrastructure expected to account for roughly 55 to 76 percent of total steel demand and 48 to 75 percent of concrete demand by 2070. Total steel stocks are projected to roughly double from 22 million tonnes in 2000 to between 42 and 50 million tonnes, and concrete stocks to grow from 43 to between 62 and 86 million tonnes, while asphalt plateaus at about 380 to 400 million tonnes because the road network is mature and demand comes mainly from resurfacing rather than expansion.

On the supply side, the model projects a steadily growing stream of end-of-life materials. Steel outflows rise from 0.2 million tonnes per year in 2000 to 0.9 to 1.3 million tonnes per year by 2070, and concrete outflows grow from 0.5 to between 1.4 and 2.2 million tonnes per year. Applying material circularity strategies of reuse, remanufacturing, repurposing and recycling, the researchers estimate that between 0.51 and 1.1 million tonnes of steel, 0.6 and 1.75 million tonnes of concrete, and 19 and 22 million tonnes of asphalt could be recovered annually between 2024 and 2070. In theory this secondary supply could cover 74 to 86 percent of steel demand, 71 to 83 percent of concrete demand, and an extraordinary 92 to 101 percent of asphalt demand across the scenarios.

Yet the study’s most sobering message lies in the gap between theory and reality. When technological, regulatory and financial constraints are factored in, realized closed-loop circularity, meaning recovery of materials at their original functionality and value, is estimated at only around 31 to 40 percent for steel, a mere 3 to 23 percent for concrete, and 50 to 85 percent for asphalt over the projection period. Fully recovering end-of-life stocks would reduce annual virgin material demand by 16 to 49 percent between 2024 and 2070, but the Netherlands currently falls far short of that potential. The country proudly recycles over 98 percent of its construction and demolition waste, an EU-leading figure, but the authors show this is largely open-loop downcycling: crushed concrete ends up as road foundation and filler rather than returning to new structural concrete.

Concrete is the weakest link. Only about 3 percent of end-of-life concrete is genuinely recycled each year, with roughly 93 percent downcycled and 4 percent landfilled, reflecting strict quality standards, contamination problems and the lack of cost-effective technologies to recover high-value sand, coarse aggregates and cement paste. Steel faces a different bottleneck: most Dutch scrap is exported to Spain, Turkey, Finland and Germany because the country lacks domestic electric arc furnace capacity for structural steel, and its primary producer focuses on high-grade flat products requiring tightly controlled scrap. Recycled steel content in Dutch macro infrastructure manufacturing is consequently limited to 25 to 31 percent. Contamination adds another layer of difficulty, as an estimated 1 to 3 percent of Dutch steel scrap contains hexavalent chromium from legacy paints and is classified as hazardous waste, discouraging domestic processing. Regulation matters too: the NEN-EN 1090 standard, introduced in 2014, requires costly certification that older structural steel components cannot easily obtain, inadvertently restricting reuse, although the new NTA 8713 procedure of 2023 offers a path toward safe component-level reuse.

Asphalt, by contrast, is a circularity success story. The Dutch industry already produces new pavement with a 50 percent reclaimed asphalt pavement mix, and reclaimed supply closely tracks demand, allowing closed-loop recovery to meet up to 85 percent of demand under the High scenario. The main complications are logistical and chemical. Surplus reclaimed material must be stockpiled in a densely urbanized country where space is scarce, and pavements laid before the 1991 tar ban contain carcinogenic polycyclic aromatic hydrocarbons, so this contaminated granulate must be thermally cleaned at specialized facilities before its aggregates can re-enter the market.

The authors argue that the greatest leverage for improving national circularity lies in the booming energy sector, which will account for 37 to 50 percent of total steel and 23 to 40 percent of total concrete demand by 2070, meaning even modest improvements in design and recovery there yield disproportionate system-wide benefits. They call for commercializing high-grade concrete upcycling technologies, pivoting toward direct component-level reuse of structural steel, as pioneered by the national bridge parts bank, and embedding design for disassembly, modularity and digital material passports in new assets. Policy interventions such as end-of-life recovery mandates, secondary material procurement incentives and the Environmental Cost Indicator could push closed-loop circularity up to 13 to 35 percent higher than current trends. With the Netherlands committed to cutting emissions 55 percent by 2030 and achieving net zero by 2050, the study concludes that closing the infrastructure materials loop is not merely an environmental aspiration but a strategic necessity, one whose lessons apply equally to Germany, the United Kingdom and the United States as their own post-war infrastructure ages into scrap.

Subject of Research: Scenario-based dynamic material flow analysis of future material demand and circularity potential of Dutch non-building macro infrastructure stocks

Article Title: Scenario based dynamic material flow analysis of Dutch macro infrastructure stocks: assessing future material demand and circularity potentials

Article References: Tareq, M. F., Berrill, P., & Tukker, A. (2026). Scenario based dynamic material flow analysis of Dutch macro infrastructure stocks: assessing future material demand and circularity potentials. Journal of Industrial Ecology. https://doi.org/10.1007/s44498-026-00176-z

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00176-z

Keywords: dynamic material flow analysis, macro infrastructure, circular economy, steel recycling, concrete downcycling, asphalt pavement, secondary raw materials, closed-loop circularity, Netherlands, end-of-life recovery, net-zero 2050, energy infrastructure