Bryophytes are the quiet workhorses of metal biomonitoring: cheap, ubiquitous, and famous for passively soaking up whatever the atmosphere or soil offers. But how does a non-vascular plant actually handle a metal load once it arrives through the soil rather than the air? Cimerman, Grandljič, Žnidaršič and colleagues answer that for the model liverwort Marchantia polymorpha subsp. ruderalis in an open-access paper in Plant and Soil (published 22 September 2026), and the answer is a genuine mechanism — plus a calibration warning for everyone who interprets moss-bag data.
The gradient
The team collected metalliferous soil near the former Žerjav lead–zinc smelter in Slovenia’s Mežica valley — pseudototal (aqua regia) concentrations of 61.9 mg kg⁻¹ cadmium, 9,287 mg kg⁻¹ lead and 631 mg kg⁻¹ zinc — and built a four-step dilution series against matched non-metalliferous control soil (native soil, 2/3 and 1/3 mixtures, control). Eight genetically distinct Slovenian genets (four female, four male) were cloned from gemmae, 320 plants in total, and harvested after 35 days. The soils were realistic: pH ~7, and the plants were exposed through the substrate only, with no aerial deposition.
The cost, and the hiding place
Even the 1/3 mixture cost the liverwort measurable growth. By day 35, control thalli averaged 14.3 ± 2.92 cm² in surface area against 4.00 ± 1.32 cm² on smelter soil; biomass fell four- to fivefold (the paper’s own framing of 466 ± 149 vs 86.3 ± 34.1 mg fresh weight), chlorophyll dropped from 24.4 ± 3.75 to 17.82 ± 2.71 µg mL⁻¹, and the ionome shifted — K, Mn and P fell as Zn, Cd, Pb, Ca, Fe and S rose.
The mechanism that keeps the gametophyte alive is spatial, not chemical: micro-PIXE mapping showed Zn, Cd and Pb strongly enriched in the lower parenchyma and ventral epidermis — the non-photosynthetic underside of the thallus — reaching mean local concentrations of 591, 399 and 1,175 mg kg⁻¹ respectively. Transmission electron microscopy adds the structural correlates: a ventral cuticle more than twice as thick on smelter soil (34.6 ± 14.5 nm vs 14.4 ± 4.6 nm), electron-dense cytoplasmic vesicles clustered near chloroplasts, and stressed plastids loaded with starch and plastoglobuli. Sequestration works — but the plant pays for it in biomass, and the ventral values are local tissue hotspots, not whole-thallus concentrations; the authors themselves note the PIXE cross-sections differ substantially from bulk ICP-MS numbers.
The non-obvious number: cadmium saturates
Bulk thallus maxima were 180 ± 26.8 mg kg⁻¹ Zn, 136 ± 22.8 mg kg⁻¹ Pb — and only 56.7 ± 15.2 mg kg⁻¹ Cd, well under the 100 mg kg⁻¹ dry-weight threshold the field uses for vascular-plant Cd hyperaccumulation. The dose response shapes differ cleanly: Zn and Pb rise approximately linearly with bioavailable soil concentration, while Cd follows a saturating exponential that plateaus around 53.0 ± 10.9 mg kg⁻¹ in planta. The practical reading is the one the authors draw: M. polymorpha thallus concentrations can track soil Zn and Pb for biomonitoring, but not Cd — at a high-Cd site the liverwort’s thallus simply stops rising, flattening the very gradient a monitor wants to measure. That nonlinearity, plus an earlier moss-bag report of Pb up to 2,276 µg g⁻¹ that this soil-based experiment could not replicate, should temper airborne-exposure claims about this species.
Why this is not a phytoextraction paper
Concentration is not removal. Total metal content per thallus was dominated by the biomass penalty: Zn content per plant was actually highest in controls (1.7 vs 1.3 µg) because smelter-soil plants were four to five times smaller, and even on smelter soil a thallus held just 0.41 µg Cd and ~1 µg Pb. A mat of stressed liverwort is a dosimeter, not a harvestable sink. The honest opportunity is the monitoring one — a cheap, clonable sentinel whose Zn and Pb read linearly and whose saturation behaviour now has a measured ceiling — together with a clean model for studying ventral sequestration in a species with superb molecular tools. The honest risk is misreading it: a fourth caveat the authors flag is that male gametophytes accumulated 1.3–1.5× more Cd and Pb than females (n = 4 per sex, and opposite to one published sex-tolerance result), so sentinel networks that ignore sex could be biased.
Bounds
This is one soil system, one 35-day greenhouse exposure, eight genets, and dose–response fits over only four soil levels that the authors call descriptive trends rather than predictions. A consensus hyperaccumulation threshold for bryophytes does not yet exist, so the “accumulator, not hyperaccumulator” verdict is relative to a vascular-plant yardstick.
Provenance: source read in full from the open-access publisher version; all figures above verified against the primary text. Method: /methodology/.