A measured soil pH below a plant's sourced floor is worth saying out loud, with the shortfall and the direction of the fix. Above the ceiling too, but only for a plant that is documented to require acid soil — for everything else the ceiling is this corpus's own and fires nothing.
A good hunch · suboptimal if ignored → advise
How it works
Nutrient bioavailability is pH-dependent, so below the crop's optimum band phosphorus and molybdenum become less available; below about 5.5 the effect turns acute as aluminium and manganese dissolve fast enough to damage roots directly. Above an acid-requiring plant's range the same curve runs the other way: iron becomes unavailable and the plant goes chlorotic.
This bed's measured pH sits below what this plant wants. Near the floor that is a question of yield, not survival — the floor is the pH at which extension says to lime, not the pH at which the plant dies, and it is an estimated figure. Below about 5.5 it becomes a harder problem: aluminium and manganese dissolve and damage roots. Lime raises pH and sulfur lowers it, but we will not tell you how much of either, because the amount depends on your soil's buffering capacity and only a lab report's buffer pH measures that. Your extension soil lab is linked on the soil card.
The evidence
Evidence: confirmed against 3 sources in the research-based literature.
the 3 sources
UF/IFAS Extension HS1207, Liu & Hanlon, 'Soil pH Range for Optimum Commercial Vegetable Production', edis.ifas.ufl.edu/publication/HS1207 (fetched and read in-container 2026-07-25). The ACUTE half of the mechanism: 'At pH 5.5 or lower, the solubility of Al increases 1000-fold for every pH unit decrease. For example, at pH 5.0, Al solubility is only 0.05 ppm, but at pH 4.0, Al solubility increases to a toxic level of 51 ppm.' On phosphorus: 'Aluminum and phosphate precipitate in low-pH soil'; 'Because of the Al concentration increase, the bioavailability of P at pH 4.0 reduces to one thousandth of the concentration present at pH 5.0'. On roots: 'In the presence of toxic concentrations of elements such as Al at low pH, root growth and water uptake are inhibited'. WHAT THIS SOURCE DOES NOT SUPPORT, AND IT IS WHY THIS RULE IS ONLY A GOOD HUNCH: all of the above is stated for pH 5.5 and below, and 'metal toxicity occurs at soil pH lower than 5.0' — while the floors this rule fires on run 5.5 to 6.2. Over most of its firing range the same page says the opposite: 'in the pH range from 5.5 to 7.0, all the nutrients have favorable solubility for use by vegetable plants.' The two are reconcilable (UMD's threshold is about optimum yield, this one about nutrient failure) but not as a single promising claim. Its Figure 2 is the nutrient-bioavailability-versus-pH curve that WOULD support the mild half of the mechanism; it is an image and could not be read.
UMD Extension, Table B-1 'Target Soil pH Values for Vegetable Crops', extension.umd.edu/sites/extension.umd.edu/files/2021-03/B-1.pdf (fetched and read in-container 2026-07-25): the source of every per-crop floor this rule reads. Its second column is headed 'Target lime when pH falls below', which is why the ruling_text says the floor is an intervention point rather than a lethal limit — the table itself frames it that way.
FIGURE 2 IS A DEAD END, AND THIS CLOSES IT (2026-07-26). Earlier notes recorded that HS1207's Figure 2 - a nutrient-bioavailability-versus-pH curve - was the evidence that would restore this rule to promising, and that it could not be read because it is an image. The figure does not need to be seen. Its caption and alt text are recoverable from the page: 'Figure 2. The pH and bioavailability (%) of listed nutrients in soil solution. Credit: Finck (1976).' TWO THINGS FOLLOW. First, the publication states its OWN reading of the figure, and that reading is the objection rather than the rescue: the sentence that forced the B-to-C downgrade - 'in the pH range from 5.5 to 7.0, all the nutrients have favorable solubility for use by vegetable plants' - is explicitly cited to Figure 2. The figure is what the contradiction rests on. Second, it is not UF/IFAS data: it is credited to Finck (1976), a reproduced availability curve of the widely-copied schematic kind. SO NOBODY SHOULD CHASE THIS IMAGE AGAIN. The good-hunch tier is not waiting on it. [read]