The model, called the Global Plastic Distribution Model or GPDM, builds on a conceptual framework known as the Plastic Pathfinder, which was originally developed to simulate plastic transport in a single hypothetical basin. The research team scaled that framework up to planetary proportions by harmonizing global datasets on wind speed and direction from the Copernicus Climate Change Service, along with surface runoff, elevation and land cover from HydroSHEDS, and river networks from HydroRivers. All inputs were standardized to a common grid of 0.01 degrees, roughly one kilometer at the equator, and organized into 285 drainage basins defined by the HydroBASINS level-3 hierarchy. Eight basins, including the Caspian Sea and several small islands, were excluded because their shapes or characteristics made simulation impractical, but the discarded area represents less than half a percent of the total simulated surface.

At the heart of GPDM lies a deceptively simple physical principle borrowed from friction theory: plastic moves only when the driving forces of wind or surface runoff exceed the resistance offered by terrain and land cover. The researchers defined 61 threshold parameters governing mobilization across different land cover and slope classes. When wind or runoff surpasses a threshold, plastic is transferred to a neighboring grid cell along the wind direction or flow pathway; when thresholds are not exceeded, the plastic stays put. Each terrestrial cell receives one normalized unit of plastic at every monthly time step, a deliberate design choice that decouples environmental transport from the geography of waste generation. The result is a set of relative distribution coefficients rather than absolute mass fluxes, which is precisely what life cycle assessment practitioners need to link emissions to environmental fate.

The simulation ran for ten years at monthly resolution, long enough for the system to approach a quasi-steady state. Comparisons of outputs at three, five, seven and ten years showed limited variation after the initial period, giving the team confidence that the ten-year coefficients represent a stabilized distribution. Rivers emerged as the dominant aquatic pathway, receiving on average 20 percent of emitted plastic, with basin-level values ranging from zero to 52 percent. Lakes, by contrast, absorbed a negligible 0.19 percent on average. On land, grasslands and shrublands formed the largest sink at 24 percent, followed closely by forests at 21 percent, while agricultural and bare areas each retained about 14 percent. Urban areas, interestingly, retained almost nothing, a reflection of their smooth, impervious surfaces that offer little resistance to mobilized litter.

To test the robustness of these estimates, the team conducted a Monte Carlo sensitivity analysis across 15 basins chosen to span a range of climatic and land-use conditions. Each basin was run 100 times with lognormal perturbations of up to 20 percent applied to the threshold parameters. The results were reassuring: relative uncertainty was only 5 percent for riverine coefficients and 1 percent for terrestrial ones, though it rose to 18 percent for the small lake fraction, where tiny absolute amounts amplify proportional variability. The single most influential parameter turned out to be the surface runoff threshold governing grasslands, shrublands and agricultural plains, indicating that hydrologically driven transport, not wind, dominates large-scale plastic mobilization even though wind data were time-resolved across a decade of monthly averages.

The real test of any global model is what it reveals when confronted with real-world emission data. In a proof of concept, the researchers coupled their distribution coefficients with a comprehensive inventory of mismanaged plastic waste covering more than 50,000 municipalities worldwide. The coupled estimates suggest that of the plastic entering the environment globally each year, roughly 43 million metric tons accumulate on land, 7.17 million metric tons reach rivers, 0.31 million metric tons settle in a narrow two-kilometer coastal buffer, and only 0.12 million metric tons end up in lakes. Country-level rankings varied dramatically by compartment. India ranked first in most terrestrial categories and in lakes, but only third in grasslands and second in rivers. Indonesia topped the riverine transfer ranking, consistent with its vast network of river cells, while Nigeria led in grassland accumulation despite more modest showings elsewhere.

These compartment-specific rankings carry a powerful policy implication: the countries contributing most to terrestrial pollution are not necessarily the same ones driving riverine or coastal accumulation. A nation’s plastic footprint, in other words, cannot be captured by a single number. The authors argue that this argues for regionalized fate factors in environmental assessment rather than one-size-fits-all global averages. Their comparison with the published literature showed broad agreement for freshwater compartments but estimates one order of magnitude higher for terrestrial and coastal accumulation than most existing assessments, a divergence the researchers attribute to their explicit representation of environmental retention rather than a narrow focus on ocean inputs.

The model is candid about its limitations. It treats all macroplastic as a generic, undifferentiated unit, using mobilization thresholds originally derived from loose paper, which are more representative of lightweight items like bags and bottles than of heavier or denser objects. It excludes sinking, deposition, sedimentation, fragmentation and the conversion of macroplastics into microplastics. Narrow rivers smaller than the one-kilometer grid resolution are not explicitly represented, likely underestimating transfers to small streams. Extreme events such as floods and monsoons, known to dramatically amplify plastic mobilization, are not captured, nor are dams and reservoirs that act as retention structures. The outputs were not calibrated against field observations, partly because relative basin-level fractions cannot be directly compared with absolute point measurements. Coastal zones posed a particular challenge, with spatial bias detected near the Ganges estuary in Bangladesh and the Caspian coastline near Sumqayit in Azerbaijan, prompting the researchers to introduce the two-kilometer coastal buffer in their country-scale analysis.

Despite these caveats, the significance of the work is hard to overstate. Life cycle assessment, the standard tool for evaluating the environmental footprint of products, has long suffered from a blind spot: while characterization factors exist for marine microplastics and for macroplastic entanglement and ingestion at sea, comparable factors for terrestrial and freshwater macroplastic impacts have been missing entirely. Most existing fate models simply assume that macroplastics do not move after being emitted to land. GPDM demolishes that assumption by quantifying, basin by basin, the probability of cross-compartment transfer. The model’s code and raw spatial outputs are openly available on Zenodo, allowing other researchers to couple the coefficients with their own emission inventories or process-based models.

Perhaps the deepest message of the study is conceptual. By demonstrating that land is the primary global sink for macroplastic waste, GPDM shifts the analytical center of gravity away from the ocean-centric narrative that has defined the field since the earliest global estimates of plastic inputs to the sea. Plastic that lingers in soils, forests and grasslands is not harmless; it fragments, leaches chemicals, enters food webs and degrades ecosystems in place, and it can be remobilized by floods or storms years or decades later. As negotiations toward a global plastics treaty continue, the study suggests that waste management interventions on land, particularly in low-income countries where uncollected waste and open dumpsites drive emissions, may deliver far greater environmental returns than previously appreciated. The missing plastic, it turns out, was never really missing. It was hiding in plain sight, scattered across the landscapes we live in.

Subject of Research: Global spatially explicit modeling of macroplastic distribution from land into terrestrial, riverine and lake compartments

Article Title: A Global Basin-Delineated Model of Macroplastic Distribution from Land to Aquatic Environments

Article References: Deschênes, C. E., Dorber, M., & Verones, F. (2026). A Global Basin-Delineated Model of Macroplastic Distribution from Land to Aquatic Environments. Environmental Challenges, 25, Article 101672. https://doi.org/10.1016/j.envc.2026.101672

Image Credits: AI Generated

DOI: Not provided

Keywords: plastic pollution, macroplastics, global model, river basins, terrestrial sink, life cycle assessment, hydrology, land cover, mismanaged plastic waste, environmental fate, freshwater, sustainability

The central problem, the authors contend, is fragmentation. Nanopesticide studies typically report pest mortality or pathogen suppression, nanofertiliser work emphasises nutrient uptake and physiological responses, RNA interference experiments measure gene silencing, and carrier-based formulations are judged by loading capacity, release kinetics, stability or adhesion. Each endpoint is scientifically legitimate in its own context, but the studies differ so widely in dose metrics, exposure routes, target organisms and biological organisation that their results cannot be compared directly. A formulation that kills 90 percent of fall armyworm larvae in a leaf-dip assay tells a farmer almost nothing about whether the same product will protect a maize crop through a rainy season, because the dose that reaches the insect’s gut in the field may be orders of magnitude lower than what the assay delivered.

To resolve this, the review proposes an environmental exposure–response–translation framework that treats formulation design, environmental fate, exposure, biological response and agronomic outcome as a single causal chain rather than separate disciplines. Material attributes such as particle size, surface chemistry, loading and release mechanism are modified after application by sunlight, rainfall, soil pH, organic matter, microbial activity, plant-surface properties and the behaviour of the target pest. These interactions determine whether a formulation stays dispersed, deposits on foliage, survives weathering, releases its active ingredient at the right moment and reaches the intended biological interface. Crucially, the framework distinguishes five dose levels: the nominal dose applied, the dose loaded onto the carrier, the dose released under field conditions, the dose present at the exposure site, and the biologically delivered dose that actually reaches the site of action. The authors argue that biologically delivered dose will predict both pest suppression and non-target risk far more reliably than the nominal concentrations that dominate current reporting.

The framework also classifies nano-enabled agro-inputs along two axes: the mechanism of the active component and the delivery function of the nanoscale system. Five primary classes emerge: intrinsically active nanomaterials such as metal and metal-oxide nanoparticles; small-molecule pesticidal payload systems, including botanical actives like azadirachtin encapsulated in chitosan, mesoporous silica, cyclodextrins or lipid nanoparticles; nucleic-acid payload systems that deliver double-stranded RNA to silence pest genes; biological-agent payload systems that protect and deliver viruses, fungi or bacteria; and nutrient, plant-signal and defence-eliciting systems such as nano-silicon or zinc-based inputs. The classification deliberately separates what produces the biological response from how the response is enabled, so that a chitosan carrier, for example, can be recognised as a pesticide vehicle, an RNA delivery system or a plant-defence stimulant depending on its payload and target rather than its material identity.

The empirical weight of the review comes from paired laboratory-to-field comparisons, drawn heavily from research on fall armyworm, one of the world’s most destructive maize pests and the best-represented target in the nano-enabled crop-protection literature. The reported potency range is staggering: laboratory LC50 values against fall armyworm span from roughly 0.04 milligrams per litre for emamectin benzoate nanoformulations to nearly 10,000 milligrams per litre for silica nanoparticles alone, a difference of about 250,000-fold. Copper oxide nanostructures required around 119 to 135 parts per million, while zinc oxide and silicon dioxide nanoparticles needed 343 and 727 parts per million respectively. The lesson, the authors stress, is that nano-enabled agro-inputs cannot be treated as a single potency class; toxicity depends on the active component, the formulation function, the exposure route and the developmental stage of the pest.

When laboratory potency is compared with practical application rates, a sobering pattern emerges. Some nanoformulations, including DSPE-EB and EMPP/CD@PEG, were applied at approximately 87 and 25 times their laboratory LC50 values, consistent with enormous losses between the tank mix and the delivered dose caused by incomplete deposition, canopy heterogeneity, weathering, degradation and limited ingestion. Not every encapsulation improves matters: polymeric neem formulations performed worse than commercial neem oil at the same azadirachtin concentration, and botanical silver nanoparticle formulations showed weaker responses under field conditions than in the laboratory. The authors interpret these findings as evidence that controlled release or improved loading constitutes a genuine advantage only when it coincides with the pest’s susceptible life stage and maintains adequate exposure under realistic conditions.

Yet the review also documents genuine successes, and they share a common signature: equal or better field control at the same or lower active-ingredient rate. Mesoporous silica carriers known as CLAP@MSNs improved control at essentially the same emamectin benzoate rate as the conventional product. Nano-formulated thiocyclam and chlorantraniliprole retained strong control at one-tenth of the conventional rates. A system combining a pesticide, a star polycation carrier and double-stranded RNA targeting the Nrf2 gene improved field performance while reducing both carrier and dsRNA concentrations relative to laboratory testing. These cases demonstrate delivery, persistence or component-specific optimisation under practical conditions, and the authors argue that they are far more persuasive than potency numbers alone.

For RNA interference technologies, the framework demands an unusually complete evidence chain, because gene knockdown is only an intermediate step between delivery and crop protection. A double-stranded RNA payload must survive gut nucleases, enter cells, escape endosomes, engage the Dicer and Argonaute machinery, suppress the target transcript and its protein, and ultimately produce a phenotype such as reduced feeding, delayed moulting, lowered fecundity or mortality. Substantial knockdown can occur without rapid protein depletion, and protein reduction can fail to produce a phenotype when the gene is redundant or weakly expressed in the exposed tissue. The review therefore calls for time-resolved transcript and protein measurements, matched organism-level phenotypes, and controls including naked nucleic acid, carrier alone and non-target sequences, along with more than one independent RNA sequence where feasible.

Environmental safety receives equally rigorous treatment, framed not as an afterthought but as an integral determinant of efficacy and risk. The same properties that improve delivery can extend residence time, alter mobility and increase contact with non-target organisms. The framework distinguishes environmentally relevant exposure entities, which may be intact particles, aggregates, released active ingredients, dissolved ions or transformation products, and shows that chemically related materials behave very differently: zinc oxide dissolves readily, especially in acidic soils, whereas titanium dioxide tends to persist as intact, soil-associated particles. Biosafety testing is organised by exposure route, from pollinators and natural enemies for foliar sprays to soil microbes, earthworms and aquatic communities for soil, seed and runoff pathways. The authors also address regulation, noting that most jurisdictions handle nano-enabled products through existing pesticide, fertiliser and biostimulant law, and that authorisation of a conventional active ingredient should not automatically extend to a nanoformulation that changes release, persistence, residue behaviour or non-target exposure.

The review closes with a set of five translation-readiness gates and six testable propositions that together turn evaluation into a decision process: delivery adequacy, biological efficacy, comparative advantage over an appropriate conventional product, environmental safety and practical, regulatory readiness. Systems that fail a gate require reformulation, further evidence or rejection rather than optimistic extrapolation. The authors argue that future progress depends less on generating yet more formulations than on testing causal relationships, quantifying biologically delivered dose, aligning release with the target’s exposure window and conducting matched, multi-season field trials that measure agronomic performance, environmental fate and non-target effects concurrently. If adopted, the framework could finally separate nano-enabled technologies that deliver reproducible, efficient crop protection from those whose laboratory sparkle dissolves at the field’s edge.

Subject of Research: An integrated evaluation framework linking environmental fate, exposure and field translation for nano-enabled crop-protection agro-inputs

Article Title: An integrated framework for evaluating nano-enabled agro-inputs: linking environmental fate, exposure and field translation for sustainable crop protection

Article References: Awan, S., Oyege, I., Mwebesa, N., & Kansiime, E. M. (2026). An integrated framework for evaluating nano-enabled agro-inputs: linking environmental fate, exposure and field translation for sustainable crop protection. Journal of Nanoparticle Research, 28(10), Article 248. https://doi.org/10.1007/s11051-026-06747-2

Image Credits: AI Generated

DOI: 10.1007/s11051-026-06747-2

Keywords: nano-enabled agro-inputs, nanopesticides, environmental fate, biologically delivered dose, fall armyworm, crop protection, RNA interference, field translation, sustainable agriculture, environmental exposure, nanocarriers, regulatory readiness