The scale of the problem the researchers set out to address is staggering. More than ten million contaminated sites are estimated to exist worldwide, with mining, foundries, industrial waste, petrochemical production, and fossil fuel combustion among the leading sources of potentially toxic elements, or PTEs, in soil. Traditional remediation has often meant excavating contaminated earth and dumping it in landfills, an approach that is costly, disruptive, and simply moves the problem elsewhere. Phytoremediation offers a gentler alternative: selected plant species extract, stabilize, or sequester pollutants through their roots and tissues. Some hyperaccumulator plants can absorb large quantities of metals without suffering physiological damage, making them living cleanup machines. But phytoremediation has a persistent weakness, namely that it occupies land for years while generating little direct revenue, which has limited its adoption on marginal and contaminated sites.

The Italian team’s answer is to stack a second land use on top of the first. Agrivoltaics, the co-location of solar panels and crops, has gained momentum as a way to squeeze more value from agricultural land, and previous studies have reported land equivalent ratios above 1.2, meaning more than 20 percent higher combined productivity than growing food or generating power separately on equivalent land. By pairing photovoltaic arrays with phytoremediation species, the researchers argue, contaminated land can be cleaned while simultaneously producing renewable energy, and the electricity revenues can make remediation economically viable. The panels may even help the plants, reducing evapotranspiration and soil evaporation through partial shading, a benefit worth an estimated 20 to 30 percent reduction in irrigation requirements in water-scarce Mediterranean environments.

Choosing the right plants was the first challenge. The team screened roughly 100 peer-reviewed publications drawn from Scopus, Web of Science, and Google Scholar, prioritizing studies from Mediterranean, semi-arid, or industrially contaminated settings comparable to Augusta. Candidate species were evaluated using standard phytoremediation performance indicators, including the bioconcentration factor, which measures a plant’s ability to absorb metals from soil; the translocation factor, which describes how efficiently metals move from roots to shoots; and the remediation factor, which captures the share of total soil contaminants extracted over a growing season. Species with a translocation factor above 1 are classified as translocators, suited to phytoextraction, while low values favor phytostabilization, where contaminants are locked in place around the roots. Soil pH, electrical conductivity, organic matter, microbial activity, and contaminant bioavailability all entered the selection calculus, since plants can only absorb metals present in bioavailable forms.

Three species emerged as the most promising candidates, each representing a complementary strategy. Arundo donax, giant reed, is a hardy C3 perennial with high biomass productivity and documented phytoextraction capacity for zinc, alongside phytostabilization of chromium and lead that reduces leaching and groundwater risk. Chrysopogon zizanioides, vetiver grass, is a C4 species with a deep, fibrous root system, exceptional drought tolerance, and proven performance on saline and alkaline soils. Cannabis sativa, industrial hemp, is a moderately shade-tolerant C3 crop with a dense root system that improves soil structure and a high economic value rooted in its fiber, seed, and secondary metabolite applications. The researchers stress that no single species is universally optimal; rather, the trio collectively satisfies the agronomic, physiological, environmental, and economic criteria of the framework.

Contamination tolerance was assessed by comparing biomass production in polluted soils against uncontaminated controls, using experimental data from prior studies. Giant reed showed yield losses of up to 40 percent only at the highest chromium concentration tested, 600 milligrams per kilogram, and remained a strong zinc extractor. Hemp, specifically the late-flowering variety Futura 75, maintained biomass production across all tested levels of cadmium, lead, and nickel contamination, with the steepest decline, up to 50 percent, occurring at the highest nickel concentration of 1,500 milligrams per kilogram. Vetiver tolerated moderate lead and cadmium treatments with limited biomass reduction, though critical combined concentrations of zinc, lead, and cadmium cut its yield in half. Shading effects were estimated separately from meta-analytical response curves for C3 and C4 crop groups, which revealed a non-linear relationship between biomass and solar radiation reduction. C3 cereals showed less-than-proportional yield declines up to roughly 15 percent shading, while C4 species proved markedly more sensitive, a distinction that shaped which plant was matched to which panel technology.

The photovoltaic side of the framework was evaluated through PVsyst simulations for one-hectare plots under Augusta’s real conditions, with panel heights of 4 to 4.5 meters to accommodate tall crops and allow mechanized harvesting. Three configurations were modeled. The first paired giant reed with conventional monofacial modules, 450-watt panels arranged in 11 north-south rows spaced 10 meters apart, totaling 1,026 panels. The second combined vetiver with vertically mounted bifacial modules, 440-watt panels oriented east-west in 11 rows, a layout whose more homogeneous light distribution suits the higher irradiance demands of C4 photosynthesis. The third matched hemp with semi-transparent modules, 528 panels in north-south rows, whose partial light transmittance mitigates shading for the moderately tolerant C3 crop. All simulated energy outputs were conservatively reduced by 20 percent to account for overall system losses, including soiling and air pollution effects.

The results reveal a clear trade-off landscape. The monofacial giant reed configuration delivered the highest electricity production at 568,660 kilowatt-hours per year, with biomass of 26.3 tonnes per hectare annually under shading and a net economic return of 2,906 euros per hectare per year. The bifacial vetiver system produced 297,987 kilowatt-hours annually but achieved the highest biomass yield, 60.0 tonnes per hectare per year, and the best net return, 6,730 euros per hectare per year. The semi-transparent hemp configuration generated 285,638 kilowatt-hours with 14.0 tonnes of biomass and a net return of 1,543 euros per hectare per year. The researchers conclude that the optimal configuration depends on the system’s objectives: monofacial panels maximize energy, semi-transparent panels protect biomass in shade-tolerant crops, and bifacial vertical arrays offer the most balanced compromise, combining satisfactory energy production with the highest net revenue among the three options.

The economic analysis extended to what happens after harvest. The team modeled biomass valorization through thermochemical gasification, assuming a lower heating value of about 18 megajoules per kilogram for lignocellulosic biomass, gasification efficiency near 60 percent, and electrical conversion efficiency of roughly 30 percent through gas engines, at an average European electricity price of about 0.215 euros per kilowatt-hour. Gross revenues ranged from 8,762 euros per hectare per year for vetiver under optimal conditions down to 995 euros for hemp under severe nickel stress. But the ledger darkens when management costs are counted: mechanical harvesting at 20 to 30 euros per tonne, transport at 5 to 15 euros per tonne, pretreatment and chipping at 10 to 15 euros per tonne, and, critically, the disposal of contaminated ashes at 100 to 200 euros per tonne, since burned biomass concentrates the extracted metals. Net revenues can fall by 30 to 50 percent, underscoring the importance of optimizing logistics and exploring alternative pathways such as biochar production or biocomposite manufacturing.

Augusta itself provides a sobering backdrop for the framework. Located in eastern Sicily within the Augusta-Priolo-Melilli petrochemical district, the area hosts refineries and chemical plants whose emissions have released lead, cadmium, mercury, and arsenic, along with polycyclic aromatic hydrocarbons, PCBs, and dioxins, into soil, water, and the degraded marine environment of Augusta Bay. The calcareous soils are structurally deteriorated, clay-rich layers promote water stagnation, and coastal salinity adds another stress factor for crops. The Mediterranean climate, with hot dry summers that can exceed 35 degrees Celsius and scarce rainfall concentrated in autumn and winter, makes water management essential. The researchers are careful to note that their framework remains a modeling exercise: field trials are still needed to determine whether the selected crops can withstand the simultaneous exposure to two major stressors, shading and soil contamination. Still, as energy demand grows and contaminated land multiplies, the vision of solar arrays standing over fields of metal-drinking grass offers a compelling glimpse of remediation that pays for itself.

Subject of Research: Integrating phytoremediation with agrivoltaic systems to decontaminate soils polluted by potentially toxic elements at an industrial site in Augusta, Italy

Article Title: Phytoremediation-agrivoltaic systems for PTEs decontamination: a case study from an industrial site in Augusta, Italy

Article References: Di Agosto, M., Armaro, C., Di Novo, S., Salmeri, F., Randazzo, N., & Sergi, F. (2026). Phytoremediation-agrivoltaic systems for PTEs decontamination: a case study from an industrial site in Augusta, Italy. Environmental Science and Pollution Research, 33(28), 14301-14322. https://doi.org/10.1007/s11356-026-38149-1

Image Credits: AI Generated

DOI: 10.1007/s11356-026-38149-1

Keywords: phytoremediation, agrivoltaics, soil contamination, potentially toxic elements, photovoltaics, heavy metals, Augusta Sicily, industrial hemp, vetiver grass, giant reed, renewable energy, biomass valorization

The study, carried out by L. A. S. P. Jayasinghe of the Tea Research Institute of Sri Lanka and Simplicio M. Medina of the University of the Philippines Los Baños, was published in the journal Discover Soil. It addresses a gap that has long frustrated both agronomists and environmental scientists. While soil loss from tea lands has been quantified repeatedly, far less is known about how much dissolved nitrogen, phosphorus, and potassium slip away in runoff water itself, and whether the physical barriers farmers already build can meaningfully stem that flow.

The experimental site sits at mid elevation, between 580 and 600 meters above sea level, in the wet zone mid country agro-ecological region where annual rainfall exceeds 3,200 millimeters with a 75 percent probability. The terrain is unforgiving: slopes in the experimental area ranged from 60 to 72 percent. The soil belongs to the Maskeliya series of the Red Yellow Podzolic group, a strongly acidic, sandy clay loam derived from feldspar-rich rock. Initial sampling revealed conditions already below optimal for tea, with soil organic carbon at just 0.7 percent against a recommended 2.2 to 3.0 percent, total nitrogen at 0.15 percent, and every exchangeable cation measured, including potassium at 62.1 micrograms per gram, falling short of favorable levels. In other words, the field was already nutrient-poor before a single drop of runoff was collected, which makes every kilogram of lost fertilizer all the more consequential.

To measure the losses directly, the researchers established twelve erosion plots, each 10 meters long and 3 meters wide, laid out along the contours and arranged in a randomized complete block design with three replicates. Aluminum sheets buried along plot boundaries prevented water from entering or escaping sideways, and 250-liter plastic barrels at the lower end of each plot captured every liter of runoff. Four treatments were compared: a stone terrace built level with the soil surface, the same stone terrace planted with vetiver grass along its inner edge at 15-centimeter spacing, a raised terrace wall 30 centimeters high, and a control plot of tea with no conservation structure whatsoever. The tea cultivar was TRI 2023, fertilized twice a year with VP/UM 910 at 200 kilograms per hectare, following the estate’s standard schedule.

The monitoring window, from October 2020 to November 2021, captured a punishing 5,970 millimeters of rainfall across 26 sampling events. Runoff followed rainfall closely, spiking during continuous high-rainfall periods when the soil became saturated. Over the full study period, the untreated control plots shed 1,681.8 cubic meters of water per hectare per year. Stone terraces alone cut that volume by 55 percent, terrace walls by 68 percent, and the stone terrace with vetiver grass by a remarkable 80 percent, leaving just 332.9 cubic meters per hectare to escape. The differences among the three conservation measures were not statistically significant, but each one dramatically outperformed the unprotected control.

When the researchers analyzed the chemistry of that runoff, an important nuance emerged. The concentrations of ammonium-nitrogen and nitrate-nitrogen in the water, ranging from 3.2 to 3.9 and 2.1 to 2.3 milligrams per liter respectively, did not differ significantly among treatments, and dissolved phosphorus concentrations, between 0.88 and 1.26 milligrams per liter, were likewise statistically indistinguishable. What mattered was volume. Because the conservation structures slashed the sheer quantity of water leaving the plots, the total mass of nutrients exported fell sharply even where concentrations stayed similar. Dissolved ammonium-nitrogen losses in the vetiver treatment dropped to 1.3 kilograms per hectare per year and nitrate-nitrogen to 0.8 kilograms, both significantly lower than the control. For context, tea fields in China under 240 kilograms of nitrogen per hectare have shown runoff nitrogen concentrations of 4.4 to 17.4 milligrams per liter, and Japanese tea catchments have recorded nitrate levels near 50 milligrams per liter in rivers and reservoirs, underscoring how much worse unmanaged systems can become.

Phosphorus told a different story. Losses of dissolved phosphorus ranged from 0.42 to 2.03 kilograms per hectare per year across treatments, but the differences were not statistically significant. Against an applied phosphorus rate of 35 kilograms per hectare per year, the conservation treatments lost only 1.1 to 2.7 percent of the input, while the control lost 6.7 percent. The explanation lies in phosphorus chemistry: the element binds strongly to soil minerals and organic matter in acidic tea soils, so most of what is applied stays in the field, either fixed in place or taken up by the bushes. A study in Taiwan similarly found that less than 5 percent of applied phosphorus left the field in storm runoff. The environmental stakes remain real, since phosphorus in waterways drives eutrophication, but on these slopes the primary lever for protecting waterways is keeping soil particles in place rather than managing dissolved losses.

Potassium, often overlooked because it is not directly linked to water quality impairment in the way nitrogen and phosphorus are, proved to be the biggest casualty of unchecked runoff. Dissolved potassium removal ranged from 1.27 to 9.5 kilograms per hectare per year, and the differences among treatments were highly significant. The vetiver-and-terrace combination lost just 1.27 kilograms per hectare per year, compared with 9.09 kilograms from the unprotected control, with the stone terrace alone at 3.37 and the terrace wall at 2.09 kilograms. The result echoes findings from India, where vetiver planted on bunds significantly improved nutrient conservation, and from Brazilian cropping trials where contour planting reduced potassium losses from 3.2 to 0.5 kilograms per hectare relative to downslope planting.

The mechanism behind the vetiver system’s success is a marriage of engineering and biology. The stone terrace acts as a physical barrier that slows the flow of water down the slope, giving it time to infiltrate, while the dense vetiver hedges act as living filters that trap fine soil particles suspended in the runoff. Neither measure alone performed as well as the two together, and although the vetiver treatment was numerically the best performer, the authors note that its advantage over the other conservation structures was not statistically significant. What was unambiguous was the contrast with doing nothing: every conservation measure substantially reduced both water and nutrient export, and nutrient removal broadly tracked runoff volume across the board.

For an industry in which fertilizer is among the most expensive management inputs, the economics are as compelling as the ecology. Every kilogram of nitrogen or potassium washed downhill must be replaced at cost, and the lost nutrients also degrade streams and reservoirs downstream of the plantations. With climate change expected to intensify rainfall extremes across tea-growing regions, the pressure on steep, thinly covered soils will only grow. The study’s message is that farmers on mid-elevation slopes do not need exotic technology to defend their soil; a well-built stone terrace lined with a strip of vetiver grass, a plant already familiar across the tropical world, can retain the majority of the nutrients that rain would otherwise carry away, protecting both the productivity of the land and the health of the waters below it.

Subject of Research: Effects of soil conservation measures on dissolved nutrient loss through surface runoff in mid-elevation tea lands of Sri Lanka

Article Title: Effects of soil conservation practices on nutrient loss through surface runoff in mid elevation tea lands of Sri Lanka

Article References: Jayasinghe, L. A. S. P., & Medina, S. M. (2026). Effects of soil conservation practices on nutrient loss through surface runoff in mid elevation tea lands of Sri Lanka. Discover Soil, 3(1), Article 122. https://doi.org/10.1007/s44378-026-00280-3

Image Credits: AI Generated

DOI: 10.1007/s44378-026-00280-3

Keywords: tea cultivation, soil conservation, surface runoff, nutrient loss, vetiver grass, stone terraces, soil erosion, Sri Lanka, dissolved nitrogen, dissolved potassium, dissolved phosphorus, sustainable agriculture