Contaminant Uptake in Hemp – What Moves from Polluted Soil into the Plant

Growing on a contaminated site raises two questions: what the plant pulls out of the soil, and where that substance sits once the season ends. Both have been measured, though on thinner ground than the popular story of soil-cleaning hemp suggests. The evidence rests on a single peer-reviewed field trial on PFAS, alongside pot trials run by a German state authority and older work on heavy metals. What follows from that for growers concerns the uptake pathway, the distribution inside the plant, and the fate of the harvest.

Why soil is the difficult case

PFAS are a class of substances built around unusually stable bonds between carbon and fluorine, in use since the 1940s in stain- and water-repellent coatings and in firefighting foams. In November 2023 the International Agency for Research on Cancer classified PFOA as carcinogenic to humans and PFOS as possibly carcinogenic. Several removal methods exist for water. For soil, the standard answer is still excavation: dig it out, haul it away, landfill it.

That is where phytoextraction comes in, the idea of pulling contaminants out through a plant and its harvest. The typical sites are military and fire training grounds sprayed with fluorinated foams over decades. Several traits make industrial hemp an obvious candidate there: it is annual, grows fast, draws a lot of water, is rarely browsed and returns little leaf mass to the soil.

What the field trial in Maine measured

The only peer-reviewed field trial so far ran in 2022 on the former fire training area of a closed military base in northern Maine, across five plots of 1.2 × 6 m, with four hemp cultivars. Only one of them grew normally and reached 1.2 m; the other three stalled at roughly 0.3 m because of photoperiod. PFOS was the dominant contaminant in every soil sample, at around 107 ng per gram near the drainage area and around 7.5 ng per gram on the berm.

The balance for one growing season: roughly 1.4 mg of PFAS mass removed above ground in total, about 85 % of it in the leaves. Per unit area, the cultivar that grew well removed 0.21 mg per square metre in the heavily contaminated section and 0.09 mg per square metre in the lightly contaminated one, which corresponds to about 0.2 and 2.0 % of the PFAS present in the rooted zone there. The headline figure of 2 % therefore comes from the lightly contaminated section, not from the centre of the damage. Between spring and autumn sampling the trial found no statistically supported decline in soil concentrations, with all paired comparisons at p greater than or equal to 0.05. A measurable drop in soil contamination is not documented at this site.

Two limits belong with those numbers: only the top 15 cm of soil were sampled, and one of the authors is connected to the supplier of part of the seed, which the paper declares as a conflict of interest.

Chain length and acid type decide, not the species

The bioaccumulation factor fell with increasing chain length in the field trial. At equal carbon number, carboxylic acids accumulated more strongly than sulfonic acids, in the field as well as in the greenhouse comparison, where the difference ranged from 0.4-fold to 360-fold. Short-chain compounds move preferentially into above-ground tissue, long-chain ones such as PFOS barely do: they stay mostly in the root. A single short-chain compound, perfluoropentanoic acid, accounted for 56 % of the total mass removed.

For growers this means the shape of the molecule governs the outcome rather than the choice of crop. The roots themselves were not analysed, since a hemp harvest leaves them in the ground anyway; because they reach deeper than the sampled 15 cm, the depth the absorbed fraction came from is also unresolved. Whether planting additionally binds PFAS in place, acting as phytostabilisation rather than removal, has not been studied for hemp specifically. Four of seven only tentatively identified compounds did fall by more than 20 % in the hemp plot and not in the unplanted control; the authors suspect microbial breakdown of precursor compounds in the rhizosphere, but with two replicates they could not test it statistically.

Where the substances end up inside the plant

The leaf is the loaded fraction. For heavy metals, a German field trial found the highest concentrations of every metal tested in the leaf as well, with nickel ahead of lead and cadmium, and considered the material’s use as a marketable raw material limited for that reason. The stem carries less, which makes fibre use conceivable; how PFAS behave during industrial fibre processing has not been examined. One greenhouse trial found over 45 % of the absorbed load in the pollen, at a bioaccumulation factor above 20.8 in that fraction. No field measurement exists to match it, so the transfer to real stands remains open.

The decisive step is the last one. Plants and their associated bacteria generally do not break the bond between carbon and fluorine: what is taken up has been relocated, not destroyed. Treating the harvested material by hydrothermal liquefaction at 300 °C for two hours broke down most of the perfluorocarboxylic acids. For the sulfonic acids the effect stayed limited, and PFOS mass in the product even rose, which the authors attribute to precursor compounds converting. German practice shows the same problem at the disposal end: a biogas plant does not degrade absorbed PFAS, and they remain in the digestate.

What the German situation means for growing

The pot and field trials run by the agricultural technology centre in Augustenberg show that almost exclusively short-chain PFAS pass into above-ground plant parts, and that crops accumulate them in flower, seed and fruit as well, to a degree that varies by species. Growing on contaminated land is therefore a question about the harvested product, not only about the soil. Even minimal concentrations in irrigation water lead to transfer into edible parts in sensitive crops, which puts contaminated water alongside soil as a pathway of its own. In the Mittelbaden and Mannheim areas roughly 1700 hectares of farmland count as contaminated, caused by compost mixed with paper sludge; because the load there consists mainly of long-chain compounds, the authority regards plant-based remediation as severely limited. By its own assessment, precise predictions of uptake are barely possible even when soil concentrations are known.

The legal yardstick is missing. Since 1 August 2023 the German federal soil protection ordinance has carried trigger values for seven individual PFAS, but only for the soil-to-groundwater pathway. For the soil-to-crop pathway there are none, because the data behind the transfer factor is considered insufficient. No remediation target for PFAS in solid soil is set. On the cultivation side, hemp in Germany is tied to certified seed, reporting duties towards the federal agricultural agency and official controls, which makes even a remediation planting a notifiable operation.