How to Run a Biodiversity Survey Project, Step by Step
Running a biodiversity project usually means juggling a stack of tools: one for surveys, one for maps, one for training, a spreadsheet for scores and a separate job to keep the website up to date. Our 60-second walkthrough shows one project running on CitizenClimate from set-up to live data, in ten steps.
The film uses a fictional project, Riverside Forest Restoration, with three communities. Below we go through each step and explain what's happening behind the screens.
Part one: setting up the project
The project developer does the set-up once, on a phone or laptop.
1. Create the project: name it, add the communities and map the site boundary.
2. Invite the community: share a QR code or invite code and members join in seconds.
3. Add an AI plant ID survey: photo, AI species suggestions, plant health and GPS location in one form.
4. Track SDG 15, Life on Land: pick the template and scoring runs automatically.
5. Teach local plants: short lessons with infographics and a scored quiz.
6. Open the forum and grievances: a place to share sightings, plus a formal channel to raise concerns.
Part two: the field takes over
7. Members collect data: snap a plant and AI suggests the species. In the film the top suggestion is Prunus africana, the African cherry, listed as Vulnerable by the IUCN because of bark harvesting.
8. Records land on the live map: every record is geotagged and clustered, so coverage gaps are easy to spot.
9. Scored automatically: biodiversity indices and an SDG 15 score update as records arrive.
10. Live on your website via API: your project page updates itself.
What the biodiversity scores mean
The film shows three standard indices from the CitizenClimate demo account. The Shannon-Wiener index (H′ 3.20) combines how many species you've recorded with how evenly they're spread: higher means richer, more balanced biodiversity. Simpson's index (D 0.98) is the probability that two records picked at random are different species, so values near 1 mean high diversity. Pielou's evenness (J 0.97) shows how evenly individuals are spread across species, where 1 is perfectly even.
​
None of these tells the whole story on its own, which is why the dashboard shows them side by side and tracks them over time. A rising Shannon index in restored plots, against a flat one in comparison plots, is the kind of evidence funders and verifiers look for.
Why AI plant ID changes the economics
Traditional botanical surveys depend on scarce specialists. AI suggestions with confidence scores let trained community members record species reliably, while low-confidence results go to an ecologist who can review and override them. You get far more records from far more places without losing quality control. Every photo is GPS-tagged and timestamped, giving a clear chain of custody from field to report.
Getting the most from a biodiversity survey
A few habits make a big difference. Fix your sampling locations so you compare like with like each season. Keep forms short so people finish them. Use the lessons to teach the species that matter most in your landscape, especially threatened ones. And use the forum: sightings posted and discussed by the community often pick up things a form never would.
Sharing results with funders
Funders and partners increasingly want to see progress between reports, not just at the end of a grant. With the API feeding your project page, they can watch observations, species counts and scores rise as the community works, while personal details stay anonymous.
CSV and JSON exports cover formal reporting and registry submissions, and the same records can support certification under standards such as Plan Vivo. One set of field data, used three ways, with no copying between systems.
On every screen, in 11 languages
Join the CitizenClimate movement to track and push forward monitoring progress in your community:
-
Download the app – Get started collecting data using questionnaires built for your needs
-
Access the dashboards – Use our API to visualise and compare SDG scores
-
Share your impact – Show everyone what your community's doing for sustainability
Frequently Asked Questions
How does AI plant identification work in CitizenClimate?
A member photographs a plant and the app returns species suggestions with confidence scores. Results below a set threshold are flagged so an ecologist can review and override them.
​
What biodiversity indices does the dashboard show?
The Shannon-Wiener index, Simpson's diversity index and Pielou's evenness, alongside species counts and an SDG 15 score, all updated as new records come in.
​
Can biodiversity data appear on our own website?
Yes. Through the CitizenClimate API, verified data can feed your project's website so the figures update automatically, with personal details kept anonymous.
​
Can community members raise concerns about the project?
Yes. Projects can switch on a grievance mechanism alongside the community forum, so members can raise issues formally and see them resolved.
​
Do community members need botanical training?
No formal training is needed to start. AI suggests the species, confidence scores show what needs checking, and the app's lessons teach the plants that matter most locally. Expert review catches the rest.
​