Three of three · the field tool
What is this, and does it belong here?
Something turns up — in a shipment, on a boot, in a soil sample. Describe what you can see and the candidates narrow, checked against measured references for both invasive species and Hawaiian natives. No account, no upload, works on a phone.
Working now · every candidate photographed
The identification key
Measure what you have and the list narrows. Every species here has been photographed and measured by the seed laboratory, so each candidate comes with a picture you can hold your specimen against rather than a description copied from a book.
What did you measure?
Any one of these narrows the list. A ruler gives you length and width; a jeweller’s scale gives you mass, which separates species that look alike at the same size.
All 53 photographed species
every one has a reference photograph
Bidens macrocarpaAsteraceae · 10–13.5 mm · 3.9–7 mgEndemic · Orthodox
Bidens menziesii subsp. menziesiiAsteraceae · 11.2–13.1 mmEndemic · Orthodox
Bidens molokaiensisAsteraceae · 9–11.5 mmEndemic · Orthodox
Bidens pilosaAsteraceae · 10–14 mm · 1.35–1.7 mgInvasive · Orthodox
Bolboschoenus maritimus subsp. paludosusCyperaceae · 3–3.4 mm · 1.6–3.45 mgIndigenous · Intermediate
Chrysopogon aciculatusPoaceae · 2.65–3.05 mmIndigenous · Orthodox
Clermontia arborescens subsp. waikoluensisCampanulaceae · 1–1.15 mmEndemic · Intermediate
Clermontia clermontioides subsp. clermontioidesCampanulaceae · 0.85–0.95 mm · 0.07–0.135 mgEndemic · Intermediate
Cyrtandra kamoolaensisGesneriaceae · 0.3–0.4 mm · 0.01–0.022 mgEndemic · Not yet determined
Cyrtandra kealiae subsp. kealiaeGesneriaceae · 0.35–0.42 mm · 0.003–0.01 mgEndemic · Not yet determined
Cyrtandra kohalaeGesneriaceae · 0.3–0.4 mm · 0.002–0.009 mgEndemic · Intermediate
Cyrtandra lessonianaGesneriaceae · 0.5–0.6 mm · 0.015–0.03 mgEndemic · Orthodox
What this set is, and is not
These fifty-three species are the ones the seed laboratory has photographed and measured. They are almost all Hawaiian natives, so this answers could this be something that belongs here with real precision.
It does not yet answer which invasive is this, because the priority non-native species have never been imaged to this standard. Doing that for one hundred of them is the funded work, and it uses this same schema so the two halves are directly comparable on the day they are made.
How the imaging and the vision model fit together →What the funding builds
The reference library
The key narrows. Confirming an answer needs calibrated photographs of known seed to compare against, and measurements taken from that material rather than from a book. Those records already exist for Hawaiian native species — this is one of them.
Asteraceae · 10–13.5 mm · 3.9–7 mg
Endemic · orthodox
Asteraceae · 11.2–13.1 mm
Endemic · orthodox
Asteraceae · 9–11.5 mm
Endemic · orthodox
Asteraceae · 10–14 mm · 1.35–1.7 mg
Invasive · orthodox
Cyperaceae · 3–3.4 mm · 1.6–3.45 mg
Indigenous · intermediate
Poaceae · 2.65–3.05 mm
Indigenous · orthodox
Campanulaceae · 1–1.15 mm
Endemic · intermediate
Campanulaceae · 0.85–0.95 mm · 0.07–0.135 mg
Endemic · intermediate
Gesneriaceae · 0.3–0.4 mm · 0.01–0.022 mg
Endemic · uncertain
53 of 200 loaded records carry photographed seed. Three orientations each — close-up for surface texture, a pair showing different faces, a cluster for the range within one collection. Images © Lyon Arboretum Rare Plant Program Seeds Lab, served from SeedsOfHawaii.org.
The imaging standard
At least three focus-stacked orientations per species, plus ordinary phone-camera images representing what a user will actually photograph. The two are not interchangeable: stacked images are the reference a model learns from, and phone images are what it will be asked about at a dock in poor light.
Why focus stacking
A seed two millimetres across, photographed close enough to show its surface, has a depth of field thinner than the seed itself. One exposure is sharp across a band and blurred everywhere else. Stacking combines many exposures at different focal depths into a single sharp image — the difference between a photograph of a seed and a usable reference.
Native and non-native, side by side
Records like the one above already exist for Hawaiian natives. The funded work adds 100 priority invasive and high-risk species to the same standard, in the same vocabulary — so an unknown seed is checked against both sets at once, and an inspector can tell a native from an incursion rather than only ruling one in.
It matters that the schema matches. A reference record built to a different standard has to be reconciled later; one built to this standard is directly comparable on the day it is made. The key above asks its questions in exactly these words for the same reason.