Articles

Findings from the database, not from the internet. We only publish an article if it rests on a number our own data produces — and we link the query so you can reproduce it.

• Rain gardens: only 20 plants survive both the flood and the drought 11 September 2026 — a rain garden floods, then dries out. USDA rates those tolerances separately, and 95% of the wet-soil plants you would be offered fail the second test.
• 41 plants sold in nurseries that suppress whatever you plant beneath them 11 September 2026 — black walnut is famous for it. Sugar maple, four junipers and two lawn grasses do it too, and nothing on the label says so.
• 528 native plants no nursery sells 11 September 2026 — 30% of documented North American natives have no commercial source. The bottleneck on native planting is supply, not knowledge.
• Dry shade gets the attention. Shade over clay is the harder problem. 11 September 2026 — waterlogged in March, concrete in August, in light too low to recover in. 264 natives cope; 145 are ones you can buy.
• Dry shade: 179 plants can take it, 87 you can actually buy 17 August 2026 — half the plants that suit the hardest site in the garden are not sold by anyone. The gap is the story.

How we decide what to publish

One test governs the queue: could this article exist without our database? If yes, we don't write it. "Ten best shade plants" has ten thousand better-ranked versions and needs no data to produce. A count of how many plants pass two USDA tolerance thresholds and are commercially available does.

So every article here contains at least one claim only our data produces, with the query linked. Every factual statement traces to USDA PLANTS or a named authority — the same rule that governs the plant profiles. Numbers are dated and tied to a catalogue version, because re-ingesting the source data changes them. And where the data cannot support a claim, we say so in the article rather than rounding up to a confident sentence.

See data & sources for what we use and how.