Agricultural
evidence.
A traceable view of interventions studied for each crop and livestock enterprise — including effect size, conditions, trade-offs and evidence quality.
Records are displayed separately from recommendations. Missing evidence is shown as missing, not inferred.
Loading evidence…
Factors shown to change saleable yield, with the comparison and conditions attached.
Grouped by how they act
One row per documented factor. Open a row for comparator, dose, interactions and source.
| Factor | Direction | Documented effect | Evidence | Prairie | Studies |
|---|
No records match these filters
A missing record is not evidence of no effect.
- Do not stack. Combined treatments stay combined. Light × CO₂ is not LED plus CO₂.
- Saleable is not biomass. Tipburn, BER, cracking and pack-out count only when the paper measured marketable output.
- Keep the system label. Open-field or indoor results stay tagged. They are not greenhouse defaults.
What the crop was worth
Point-of-observation prices pulled from published series. Each price type answers a different question, so they are never averaged together.
Browse observations and provenance
| Commodity | Market / region | Date | As published | Normalized | Basis | Publisher |
|---|
Quarantined rows
Rows the pipeline refused to guess at, grouped by the reason. A quarantined row is a stated gap, not a dropped one.
| Reason | Rows | Detail |
|---|
Sources
| Source | Price type | Status | Obs. | Quar. | Document | SHA-256 | Note |
|---|
- Do not average across types. An insurance reference price is not a price anyone was paid, and a terminal bid is not a farm gate.
- Cite the basis. Where a unit was converted, the conversion is stated per row.
- Gaps stay visible. Quarantined rows are published rather than silently dropped.