Ghorbel, Saïdi, Aldilami, Mechichi and Alsherif — the group that previously characterized the barley-growth-promoting strain Arthrobacter sp. MD36 — report in Scientific Reports (2026) a new isolate, Arthrobacter sp. MD55, from the rhizosphere of the salt-tolerant grass Diplachne fusca growing beside a man-made sewage dumping lake (the paper does not state the site’s country; voucher 1579, University of Jeddah Herbarium). The site soil is heavily contaminated: Table 1 gives Ni 856, Cu 724, Cr 526, Co 369 and Pb 265 mg kg⁻¹ (n = 4). The headline result is eye-catching — barley grains that show 0% germination in 100 mM metal solutions germinate at up to 86–89% when primed with MD55 — and the paper frames it as a bioinoculant for “microbe-assisted phytoremediation.” Here is what the data do and do not support.
What reproduces from the paper’s own tables
MD55 is multiply metal-resistant in plate assays (growth at 500 mM NiCl₂ and 500 mM CuSO₄; GenBank PX758636; MALDI-TOF log score ≈ 2.32, best match A. creatinolyticus, so genus-level designation only). Its germination rescue is consistent across Table 4: at 50 mM, inoculated grains reach 53.7% (Ni), 89.0% (Co), 80.0% (Cr), 86.0% (Cu), 85.0% (Pb) and 64.0% (five-metal mixture) versus 20, 19, 0, 0, 10 and 0% uninoculated; at 100 and 150 mM every uninoculated treatment is 0%, while inoculated grains still germinate (Pb 86.7% at 100 mM — one of the two cells, with Cu at 50 mM, that the abstract’s “86% vs 0%” headline is cherry-picked from). In the 30-day pot experiment the strain increased root length by up to 320% and shoot length by up to 120% under the mixed-metal stress. IAA production, phosphate solubilization and 1.5 M NaCl tolerance are documented traits, and the halotolerance is genuinely interesting: it suggests salt and metal co-selection in the sewage-lake habitat.
Four problems a reviewer should not miss
The soil numbers contradict themselves. The Results text says the site soil has “average values fluctuating around 20 mg/kg for Cu, and 40 mg/kg for Ni.” Table 1 reports Cu 724 and Ni 856 mg kg⁻¹ — twenty-fold higher — and the text’s own ordering of metals (Ni > Cu > Cr > Co > Pb) matches Table 1, not the 20/40 sentence. The Table 1 standard deviations (Ni 856.00 ± 0.37; Pb 265.00 ± 0.07; Mg 123.00 ± 0.01) are not just implausibly tight for field soil replicates — they sit below the precision of the ICP-MS instrument itself — and the table cannot even decide whether they are SDs (column header) or SEs (caption). One of the two descriptions of the isolation soil is wrong, and the paper does not notice.
The abstract claims nitrogen fixation; the data say otherwise. The abstract lists “nitrogen fixation” among MD55’s traits, but Table 3 shows a negative result and the body states twice that the strain “could not fix nitrogen” — with an explicit recommendation to confirm by nifH or acetylene-reduction assays before claiming the trait. An abstract claim contradicted by its own Table 3 is exactly the kind of sentence that propagates through citation chains.
The assay doses are acute-toxicity theater. The germination assay is filter paper saturated with 50–150 mM metal salts — 50 mM Ni²⁺ is ≈ 2.9 g L⁻¹, 150 mM ≈ 8.8 g L⁻¹. Even the study’s own heavily contaminated isolation soil carries 856 mg kg⁻¹ total Ni; soil porewater Ni in contaminated field soils is micromolar, three orders of magnitude below the lowest treatment. Uninoculated barley at 0% germination is not a plant failing on a contaminated soil — it is a seedling facing a near-lethal salt-shock. Surviving that is a real and repeatable phenotype, but it is an acute-stress rescue, and the paper’s leap from Petri-dish rescue to “phytoremediation in contaminated soils” is not bridged by any measurement of metal removal: no plant-tissue metal concentrations, no biomass, no post-experiment soil or solution chemistry. Smaller slips reinforce the carelessness: the Pb salt is printed as Pb(NO₂)₃ (lead(II) nitrate is Pb(NO₃)₂), one Table 2 header reads 10,000 in the Co column (as printed, a 10 M CoSO₄ test — not a preparable solution), the methods say three replicates of 25 grains while the Table 4 caption says n = 4, the statistics section names the LSD test while the table captions say Tukey’s HSD, and one cell prints “80.00 d ± 7” with the significance letter inside the number.
The germination arithmetic cannot be right. With replicates of 25 grains, every replicate percentage must be a multiple of 4 — so means of three or four such replicates can never yield the two-decimal values Table 4 reports (53.67, 60.50, 49.50, 68.50, 86.67, 70.67…). Either the design description or the table is wrong, and since the same letters column declares 53.67% (“a”) and 20.00% (“a”) not significantly different in the very column the text calls a significant rescue, Table 4’s inferential layer cannot be read at face value either. The direction of the effect — inoculated grains germinate where uninoculated ones do not — is visible in the raw pattern (0% across the board without MD55 at 100–150 mM) and is plausible for an IAA-producing, metal-resistant rhizobacterium. But the table as printed cannot be the data of the experiment as described.
What it genuinely means
Strip the overclaim and there is a defensible core: a metal- and salt-tolerant Arthrobacter that measurably protects barley — the world’s fourth-largest cereal — through germination and early seedling establishment under acute stress, via IAA and phosphate solubilization. For phytostabilization of degraded, sewage-affected land, getting a crop to establish at all is the binding constraint, and a halotolerant inoculant co-selected in exactly such an environment is a plausible, cheap intervention. The paper’s own limitations section says the honest version: qualitative traits only, no biosorption quantification, no field trials — and point (iv) of its future-work list proposes integration “with phytoremediation techniques using hyperaccumulator plants.” That is the right instinct, because the endpoint these assays cannot reach — actual metal removal — is exactly what hyperaccumulator biology is for: our nickel hub records what field-grade Ni removal requires, and our alpine pennycress record shows the difference between a strain that helps a crop survive metal and a plant that moves metal into harvestable shoot tissue at percent-level concentrations. MD55 is a promising germination insurance policy; it is not yet, and should not yet be cited as, phytoextraction.
Source: full open-access publisher HTML and all five tables read in full; every number verified verbatim against the source by three independent reviewers. Selection, sourcing and verification methodology: /methodology/.