Our editorial standards

Every plant in GardeningLab is a record: its names, where it is native, the conditions it prefers, what it does for wildlife, and how to grow it. These are the standards each record is held to, and what we will not do. Every source the records cite, with how many records cite it, is on Our sources; who builds and checks the records is on How we work.

1.A plant's role in its ecosystem is part of the record

We treat what a plant does for living things as core fields of the record, alongside how to grow it: which caterpillars it feeds, which pollinators use it, what wildlife it supports, and how much deer favour it. A record that covers only horticulture does not meet our standard.

2.Sources are ranked, and each fact names the one it came from

We draw on three tiers of source, and use the highest tier that covers a fact. A citation names the source actually used for that fact, with its tier. The plant sheet shows them under "Sources". These are all the sources our plant records cite, by tier:

Tier 1: government databases, universities and their extension services, and the research-based native-plant centres

Tier 2: botanic gardens, native-plant societies, and peer horticultural references

  • Calscape, California Native Plant Society
  • Plant Finder, Missouri Botanical Garden
  • Chileflora, the encyclopedia of Chilean plants
  • iNaturalist, research-grade observations, for when a plant flowers
  • The published plant lists of the botanical gardens and public gardens on Gardens of the World, each cited by the garden's own name

Tier 3: general horticultural and reference sources

A nursery's catalogue tells us what can be bought, never what a plant does in the wild. Photo identification in the app suggests names from Pl@ntNet; the name you choose is what is saved, and it is not cited as a fact.

3.Before we compare a source to a question, we check what the source measures

A conservation-agronomy database, a trade association and a nursery each answer real questions, but different ones. We use a source only for what it measures, and we have withdrawn our own findings when they did not meet that standard.

4.If no source we trust states it, we leave it blank

An empty field is better than a guessed one. Where no source we trust covers a fact, the app says so: "No source we trust covers this yet." We never fill in a default sun requirement or bloom season. A bloom window longer than nine months must name the source that lists every month. Sun is given as shade, part shade or sun, never as a number of hours we did not measure.

5.What a plant prefers is not what your garden measured

A plant record describes the conditions the plant prefers. Your bed's record describes the conditions you observed. The two are kept as separate records. When they are compared, it is to rank plants for a bed, and the ranking is a suggestion: nothing is hidden, and every plant stays yours to choose.

6."Native" always means native to a named place

A Native badge refers to an EPA Level I ecoregion. Where a plant is native to the ecoregion but its wild range does not reach your state, the record says so. A record for a whole genus never claims nativity, because a genus can include species from other continents. Whether a plant is a keystone is worked out each time from the National Wildlife Federation's lists for your ecoregion, never typed in by hand.

7.Nothing is presented as reviewed until a person has reviewed it

Records are drafted from their sources with the help of AI. Every record carries the label "AI-drafted, pending review" until a person has checked it field by field. A plant you look up yourself in the app is drafted by a language model, has no citations, and wears the same label. Where our records do not yet meet these standards, we say so on How we work.

8.Corrections

If you find a wrong fact, write to us at the address on How we work. Include the page address and what is wrong. A person reads every correction.

Last reviewed: 30 October 2026.