Cobalt fertiliser needs a soil-specific reassessment

Cobalt fertiliser cannot yet be treated as a routine agricultural input, according to a critical review of evidence on the metal’s plant requirements, phytotoxic effects and consequences for soil health. The authors conclude that the distance between reported beneficial and harmful exposure can be narrow, while cobalt availability varies with pH, organic matter, texture, iron and manganese status, and other soil properties. They argue that deliberate additions should therefore follow crop- and soil-specific assessment rather than the assumption that more cobalt is inherently beneficial.

The review synthesises recent reviews and primary studies spanning agricultural soils, cobalt inputs, plant physiology, soil enzymes and microbial communities. It is not a new field trial or meta-analysis, and the authors note that exposure systems, dose metrics, endpoints and soil conditions often cannot be compared directly. Cobalt occurs naturally in the terrestrial environment and can enter soils through geological weathering, mining, industrial emissions, sewage sludge and agricultural materials.

Its concentration is closely tied to parent material: basic and ultrabasic rocks may contain around 100 mg kg−1, basaltic rocks 40–50 mg kg−1, and acidic rocks such as granite only 1–10 mg kg−1. The global average crustal concentration is generally estimated at 20–25 mg kg−1, but local soil concentrations can depart substantially from that value. European geochemical data illustrate the resulting variability.

The LUCAS topsoil dataset contained 21,682 samples with a mean cobalt concentration of 6.35 mg kg−1 and a range from 0.32 to 91.89 mg kg−1. Agricultural and grazing lands had a mean of 8.9 mg kg−1 in another European survey. At two Greek sites, values were strongly linked to local geology: soils in the Triada area of Euboea ranged from 40 to 208 mg kg−1 because of ultrabasic and nickeliferous rocks, while a Mediterranean site on Megara had a median of 19.0 mg kg−1, about 2.4 times the European median of 7.8 mg kg−1.

These geographic contrasts show how parent material can produce widely different baseline exposures; they do not establish a universal upward or downward trend over time. Several numerical benchmarks appear in the literature, but they answer different questions. Values around 0.5–3.0 mg kg−1 have been described as deficiency ranges, with approximately 3 mg kg−1 often cited as a deficiency threshold.

Natural-background estimates commonly fall around 20–25 mg kg−1, while regulatory and ecological values include 20 mg kg−1 as a limit value, 50 mg kg−1 as a warning value, 240 mg kg−1 as a critical value, 40 mg kg−1 as an agricultural-use guideline, and approximately 13 mg kg−1 as an ecological screening level. None is a universal plant-toxicity threshold. The reviewed evidence was insufficient to define a general cobalt sufficiency range for crops because soil-based dose–response data linking total soil cobalt, plant-available cobalt and crop response remain limited.

Cobalt’s clearest plant function is indirect. In legumes, rhizobial microorganisms use it in symbiotic nitrogen fixation, supporting nodulation, metabolism, growth and yield. Some studies also report improved growth, stress responses or altered alkaloid production in non-legumes, but cobalt is not established as essential for all higher plants.

Deficiency does not consistently depress growth under natural conditions, its physiological role outside nitrogen-fixing symbioses remains incompletely resolved, and reported tissue concentrations may reflect environmental exposure or specialised accumulation rather than nutritional need. Most plants, including legumes, typically contain 0.1–10 mg kg−1 dry weight, although a larger synthesis spanning more than 140 non-hyperaccumulating species reported a much wider range. Such values cannot by themselves distinguish cobalt presence, accumulation, a beneficial response and a true physiological requirement.

Plant responses also depend strongly on how cobalt is delivered. Nutrient solutions and hydroponic media generally expose roots to cobalt more immediately than soil, where sorption, complexation and mineral reactions reduce availability. In solution-based studies, growth depression appeared at 5 µM in mung bean, while chickpea germination remained comparable with the control at 50 µM but declined above that concentration.

Tomato and cauliflower showed growth depression at 50 µM, with severe necrosis and biomass collapse at 500 µM. By contrast, a soil-based assessment placed an EC20, corresponding to a 20% reduction in growth metrics, at 30.6 mg kg−1 for alfalfa, barley, radish, perennial rye and Brassica rapa. Solution concentrations in micromoles, soil concentrations in milligrams per kilogram and foliar rates in grams per hectare describe fundamentally different exposure systems, so these figures cannot be transferred directly between soils or crops.

The review identifies a biphasic pattern in several selected studies: low exposure was associated with stimulation, whereas inhibition appeared after a comparatively small increase. The transition from the concentration associated with benefit to the first inhibitory or toxic concentration was 1.25 times higher in onion, 1.5 times in Paulownia seedlings and field-treated sugarcane, and 2 times in maize, radish and solution-grown sugarcane. These unitless ratios compare concentrations within individual studies and do not convert unlike units into a shared application threshold.

They nevertheless indicate that a modest rise in local cobalt availability could move a crop from a stimulatory response into toxicity, especially where fertiliser distribution is uneven or soil accumulates the element. In common bean, for example, stimulation was reported at 0.05–0.10 mg Co kg−1 soil, while adverse effects emerged at 0.25 mg kg−1, illustrating that even a nitrogen-fixing species can have a narrow response window under the tested conditions. Excess cobalt disrupts several connected physiological processes.

It has been associated with lower chlorophyll, carotenoid and gas-exchange performance, as well as chloroplast damage and iron-deficiency-like chlorosis. Reactive oxygen species can accumulate, antioxidant enzymes may first rise and then become inadequate, and membrane injury can follow. Cobalt also interferes with nutrient balance, particularly through effects on iron transport, and can restrict the movement of phosphorus, sulphur, manganese and zinc.

In legumes, cobalt stress may compromise nodule formation and nitrogen fixation; in one bacterial system, reducing expression of the dmeF gene in Sinorhizobium meliloti lowered nodule formation without necessarily disabling nitrogenase activity in the nodules that did form. Across the reviewed literature, sensitivity is therefore a multidimensional trait involving growth inhibition, pigment degradation, oxidative injury, nutrient imbalance and metabolic disruption rather than a single universal ranking of plant species. Soil responses can provide evidence of biological disruption, but they vary with the measured function and the soil itself.

Experimental soils showed measurable inhibition of dehydrogenases and urease from 25 mg kg−1 cobalt, and clay-and-sand incubations revealed reduced cellulose decomposition and lower cellulase and β-glucosidase activity, with stronger effects in sandy soil. Cobalt can affect catalase, peroxidase, phosphatase, arylsulphatase and β-glucosidase as well. Microbial responses vary: some taxa may proliferate under low exposure, while Azotobacter and cellulose-decomposing organisms can be sensitive, and resistant communities may gain a relative advantage.

Evidence linking cobalt around 22 mg kg−1 in urban soil to antibiotic-resistance genes adds a wider public-health dimension, although that association does not by itself demonstrate a pathway of human exposure. Cobalt also has documented effects on the bacterial-fungal balance, nitrogen cycling and other soil functions. Farm inputs make blanket additions especially difficult to justify.

Phosphate fertilisers commonly contain 1–12 mg kg−1 cobalt, nitrogen fertilisers 5–12 mg kg−1, lime materials 0.4–3.0 mg kg−1, and manures a broader 0.3–24 mg kg−1. Deliberately enriched products can supply cobalt as sulfate, nitrate or chloride salts through soil application, foliar sprays or seed coatings. In the European Union, cobalt has been regulated among seven micronutrients eligible for declaration in fertilising products, and current rules specify minimum rather than maximum cobalt contents for many categories.

Labels require micronutrient fertilisers to be used only where there is a recognised need, but the review notes that this general provision does not create a cobalt-specific operational requirement for prior soil testing or background verification. That regulatory gap matters because intentional inputs occur alongside repeated incidental ones. Compound nitrogen, phosphorus and potassium fertilisers can carry cobalt through their raw materials, and applying them at hundreds of kilograms per hectare may constitute a recurring loading pathway.

Cobalt is also an incidental impurity in some pesticide formulations, although it is not listed as an active substance in plant-protection products. The authors’ proposed direction is cautious rather than prohibitive: future studies should use crop- and soil-specific designs under realistic field conditions, with consistent reporting of cobalt dose, speciation or bioavailability, plant endpoints, soil enzymes and microbial responses. Until such evidence can identify reliable transfer and toxicity thresholds, they say, cobalt should not be treated as a routine fertiliser constituent.

Subject of Research: Critical assessment of cobalt fertilisation, plant requirements and phytotoxic risk in agricultural soils

Article Title: Re-evaluating cobalt in agricultural soils: plant requirement and phytotoxic risk

Article References: Tsourgianni, E., Konstantinou, M., Stefanou, S., & Giannakoula, A. (2026). Re-evaluating cobalt in agricultural soils: plant requirement and phytotoxic risk. Plant and Soil. https://doi.org/10.1007/s11104-026-09167-2

Image Credits: AI Generated

DOI: 10.1007/s11104-026-09167-2

Keywords: cobalt fertilisation, agricultural soils, phytotoxicity, plant nutrition, soil bioavailability, soil enzymes, microbial communities, legume nitrogen fixation, soil health, fertiliser regulation, Re-evaluating, cobalt