Agricultural fertilisers contribute substantially to microplastic concentrations in UK soils

The paper investigates how microplastic concentrations in agricultural soils have changed over time and whether agricultural fertilisers contribute to these trends. Plastics are widely used across sectors, including agriculture, and degrade via chemical, physical, and biological pathways, producing persistent microplastics that can be transported through various environmental media. Agricultural soils receive plastic residues directly (e.g., mulch films, polymer-coated fertilisers, silage films) and indirectly (e.g., biosolids, wastewater irrigation, manure, atmospheric deposition). Microplastics and their additives can negatively affect soil properties, soil biota, and crop yield, but their long-term concentrations and trends in agricultural soils remain largely unknown. While temporal records in urban environments show increasing microplastics since the 1950s, such temporal evidence has been lacking for agricultural soils. Using archived samples from the Broadbalk long-term winter wheat experiment, the study aims to reconstruct a temporal record to test whether microplastics have increased and to assess the contribution of organic and inorganic fertilisers to soil microplastic loads.

The authors summarize existing knowledge that plastics in agriculture yield short-term agronomic benefits but may cause long-term soil degradation. They outline definitions of plastic size classes and typical shapes found in agricultural soils (films, fragments, fibers, foams, pellets) and review multiple agricultural and non-agricultural sources delivering microplastics to soils, including polymer-coated agrochemicals, farm inputs and equipment, biosolids, wastewater, tyre wear, and atmospheric deposition. Prior studies report negative impacts of microplastics and plastic-associated additives on soil organisms, soil physical and chemical properties, and crop yield, with many experiments using elevated microplastic concentrations. Temporal records have shown proliferation of microplastics in urban sediments since the 1950s, but comparable time-series evidence in agricultural soils was previously absent. The study positions itself to fill this gap using a historical archive.

Study site and sampling: Soil samples (0–23 cm, plough layer) were taken from the Broadbalk winter wheat experiment at Rothamsted Research (Harpenden, UK; 51°48′N, 0°22′W), a long-term conventionally managed experiment on a Chromic luvisol (clay loam to silty clay loam). Samples were collected from three treatments across 18 time points (1846–2022): (a) FYM—farmyard manure at 35 t ha⁻¹; (b) Nil—no soil amendments; (c) N3(P)KMg—144 kg N ha⁻¹, 35 kg P ha⁻¹ triple superphosphate until 2000, 90 kg K ha⁻¹ potassium sulphate, and 12 kg Mg ha⁻¹ Kieserite (35 kg Mg every third year between 1974–2000). Samples were milled to 2 mm and stored in sealed glass bottles or card boxes. For each treatment and time point, 1.5 g subsamples were analyzed. Three procedural blanks were run. Sample preparation and extraction: Organic matter was removed by adding H2O2 and heating to 60 °C. After effervescence subsided and visible organic matter was removed, samples were cooled to 40 °C and 5 ml of 0.05 M Fe(II)SO4 added, then reheated to 60 °C and covered for 24 h to decompose residual H2O2 and flocculate clays. Density separation was performed using 26% w/v NaCl (600 ml), mixed and settled for 24 h. Supernatants were filtered through 0.45 µm glass fibre filters; beakers were rinsed with HPLC water and washings filtered. Filters were stained with 3 ml of 0.5% Nile Red in n-hexane, then rinsed with 3 ml n-hexane, dried, mounted on slides, covered with glass slips, and wrapped in foil. Contamination control: Samples were handled in sterilised aluminum trays within a sterilised fume cupboard. Glassware, instruments, and chemicals were leached with acetone and baked at 400 °C for 4 h. Cotton lab coats and non-synthetic clothing were used. Samples were analyzed in triplicate with blanks. Recovery rates were measured and satisfactory. Microscopy and identification: A Leica MZFLIII stereo fluorescence microscope with integrated digital camera (GXCapture software) was used. Particles were examined at Ex:Em 425:480 nm, 475:535 nm, and 510:560 nm (Nile Red typically fluoresces Ex 450–490 nm; Em 515–565 nm). Selection criteria for microplastics included: clearly visible, well-defined edges; three-dimensional synthetic-like shape; size >10 µm; absence of internal organic structures; clear green-yellow fluorescence; visible across all Ex:Em combinations. A single analyst performed identification and quantification. Statistics: Simple linear regressions with group analysis using dummy coding of multi-categorical predictors were fitted to data from 1966 onward (when microplastics were detected in all plots). Significance threshold was P < 0.05. Analyses were performed in SPSS (IBM Statistics for Windows, Version 28.0).

- Across all three treatments (FYM, N3(P)KMg, Nil), microplastic concentrations increased significantly from 1966 to 2022 (R² = 0.546, F(1,28) = 33.607, p < 0.001), consistent with other archive and sediment analyses. - No microplastics were detected in samples from 1846–1914, aligning with the emergence of modern synthetic plastics circa early 20th century. Particles <10 µm of unknown composition were present in all samples, attributed to unavoidable contamination during milling, storage, collection, and analysis. - For 1966–2022, microplastic concentrations in FYM and N3(P)KMg treatments were significantly higher than the Nil treatment (model R² = 0.8, F(3,26) = 4.6, p < 0.001 for both comparisons), with no significant difference between FYM and N3(P)KMg (p = 0.441), indicating both organic and inorganic fertilisers directly contribute to soil microplastic loads beyond baseline. - FYM treatment: gradual increase from 1966 to 2010, then an apparent plateau after 2010 despite increases in the other treatments; the plateau may reflect vertical transport of microplastics below the plough layer exceeding inputs. - N3(P)KMg treatment: steep increases observed between 1997–2005 (+350%) and 2010–2022 (+183%), reflecting an overall increase over the time series. - Post-decoupling from Nil, microplastic loads in Nil likely reflect tyre wear from farm machinery and non-agricultural inputs (wind-driven redistribution, runoff, atmospheric deposition). - The overall proliferation across all treatments likely reflects both treatment-specific inputs and increasing global plastic use, especially in the last decade. Fertilisers and additives are estimated to release ~22,500 tonnes of microplastics annually. - The analysis confirms agricultural soils act as receptors and reservoirs of microplastic pollution, and fertiliser applications are a substantial contributor to increasing soil microplastic concentrations over time.

The study addresses the knowledge gap on temporal changes in agricultural soil microplastics by leveraging the Broadbalk sample archive. Findings show a clear increase in microplastic concentrations since 1966 across all treatments, with significantly higher levels in fertiliser-amended plots compared to the unamended control. This implicates both organic (FYM) and inorganic (polymer-coated and associated) fertilisers as important sources of microplastics to soils beyond baseline environmental deposition. The temporal patterns, including sharp increases in the inorganic fertiliser treatment and a plateau in FYM after 2010, likely reflect changes in agricultural plastic usage, degradation dynamics of polymer coatings, and soil processes such as vertical transport, bioturbation, and hydrological events that can redistribute microplastics. The decoupling from the Nil treatment underlines the contribution of agricultural inputs relative to background sources like atmospheric deposition and tyre wear. These results substantiate the view that agricultural soils are accumulating microplastics over time and act as long-term reservoirs, with implications for soil health, ecosystem functioning, and food security. Given the persistence of plastics and the poor reversibility of their accumulation, there is an urgent need to better understand the fate, transport, and effects of realistic microplastic loads in field conditions.

This work provides the first temporal record demonstrating increasing microplastic concentrations in a typical UK agricultural soil and shows that both organic and inorganic fertilisers significantly contribute to soil microplastic loads beyond baseline levels. The evidence indicates agricultural soils are receptors and reservoirs of microplastic pollution, with a growing legacy driven by agricultural practices and broader societal plastic use. Given the largely negative reported interactions of microplastics with soil biota, soil properties, and crop production, these trends pose risks to agricultural productivity and food security. Future research should prioritize long-term field trials at environmentally representative concentrations, improved source attribution (including fertilizer coatings and manure pathways), vertical and lateral transport processes, and mitigation strategies to reduce inputs and mobilization. A reassessment of plastic use in agriculture, alongside development of sustainable alternatives and management practices, is critical.

- Detection threshold: Only particles >10 µm were confidently identified; smaller particles (<10 µm) of unknown composition were observed and attributed to unavoidable contamination, potentially underestimating total microplastic loads and affecting specificity at the smallest sizes. - Potential contamination: Despite procedural blanks and controls, lab and handling environments are known sources of airborne microplastics (especially fibers), which may introduce background noise. - Methodological constraints: Nile Red fluorescence with visual identification can misclassify particles without polymer-specific confirmation; stringent selection criteria were used to minimize this risk. - Scope: Analyses focused on the plough layer (0–23 cm) and a single long-term experimental site, which may limit generalizability to other soil types, climates, and management systems.