How we build our plant profiles

1. Checked scientific names

Every profile starts from the scientific name, checked against Plants of the World Online (Kew) and World Flora Online. When a plant has been renamed, like the snake plant, now Dracaena trifasciata, we use the accepted name and list the old one as a synonym.

2. Facts before text

Before writing, we gather the species facts in a data sheet: light and water needs, temperature range, size, native range and toxicity. The scales, the quick summary and the text all come from the same data, so they do not contradict each other.

The scales use fixed criteria. For water, for example, 1 means letting the soil dry out completely, 2 means watering when the top layer is dry and 3 means keeping it slightly moist at all times.

3. Open, cited sources

We rely on botanical databases, university extension services, botanic gardens and, for animal toxicity, the ASPCA. Each profile lists its sources with a link to the species page and the date they were checked. When there is no reliable source, for example on toxicity, we say so instead of guessing.

4. Use of artificial intelligence

Texts are drafted with the help of artificial intelligence from the data sheet and the sources above, then go through automatic consistency checks (the text cannot contradict the facts) and sample-based review. Images labeled as illustrations are generated by AI from the botanical description of the species. For identification, we always recommend checking real photos, such as those on iNaturalist.

5. Reviews and corrections

Each profile shows the date of its last review. Species with conflicting information between sources, especially about toxicity, stay offline until they are reviewed.