Ecological forecasts are used to assess the vulnerability of ocean habitats. But these important predictions of biological response are often based on a single environmental threshold, such as a temperature or pH limit, without accounting for how organisms or populations may differ in their tolerance or resilience. Incorporating biological variation into our early warning thresholds could produce better forecasts and better decisions downstream. We are launching microgrants for researchers developing such "Living Thresholds", to refine environmental forecasting with biological data and test whether they improve predictions.
Submission Deadline
Oct 16, 2026
Which coral reef is most at risk of bleaching? Where should conservation resources be directed? Ecosystem managers today use the Coral Bleaching Heat Stress Alert index, setting a temperature threshold of 1°C above the local summer maximum to guide these decisions. But in practice, some high-risk reefs remain resilient, while others suffer catastrophic damage due to differences in the underlying biology, where thermal tolerance varies by community composition, symbiont species pairings, and historical exposure.
Many forecasting management tools work in the same way for other marine ecosystems, using one physical measurement (e.g. temperature, salinity, pH) to establish a threshold and then broadly assuming an even response from the complex ecosystems they are measuring. Because organisms and populations vary widely in how they respond at these thresholds, a single index can’t adequately predict ecosystem level outcomes.
This limitation of physical thresholds poses challenges across a range of ocean ecosystems, for example:
Mangrove forest vulnerability is assessed using a threshold rate of relative sea-level rise, but some forests are able to keep up while others are not.
Seagrass management sets a minimum light requirement, but different species and sites differ significantly in how much light they need.
Coastal dead zones are defined by where oxygen drops below 2 mg/L, but different species of fish and shellfish fare differently on either side of that line.
In several cases, building biological differences into an index has improved forecasts. However, in most systems, well-documented variation has not yet made it into the tools managers use to predict ecosystem outcomes.
This call seeks out living thresholds to complement and improve ecosystem forecasting. These microgrants support researchers asking which biological variation makes an existing index predict outcomes better, and how much detail it takes to improve predictive power.
Strong proposals will:
Start from an index or threshold in operational or common use
Add a documented source of biological variation, and explain why it should matter
Test the original and revised index against observed outcomes
Define how you will measure whether the revised index predicts better
Seek to understand how much biological resolution (e.g. regional vs. species level) is needed to improve prediction
Surprise us!
Because these are microgrants, we expect projects to use existing data, such as long-term monitoring records, satellite observations, or published tolerance experiments, rather than new fieldwork. Proposals should be scoped to a focused question that can be completed within one year.
Who should apply
We welcome ecologists who know a system well and modelers applying machine learning to real-world forecasting.
This call is limited to ocean and coastal systems, including coral reefs, kelp, coastal forests, seagrass meadows, shellfish beds, tidal wetlands such as salt marshes and mangroves, and the open ocean. Freshwater and fully terrestrial systems are outside its scope.
How to apply:
Eligibility:
The program is open to:
Funding Considerations
Please review our FAQs for more information and details.
Grants are open to independent researchers, university labs, tribes, NGOs, individuals, and dedicated amateurs. We're unable to fund for-profit companies or startups.
Start a project on Experiment and submit it to this grant. Full guidance on writing a strong submission is in the Experiment researcher guide
Yes, with some limitations. International grant funds are distributed through PayPal, so you'll need a PayPal account that can both receive and withdraw funds in your country of residence — PayPal's supported features vary widely by country, and some locations only support sending money, not receiving or withdrawing it. We also can't fund applicants based in a country currently subject to comprehensive U.S. sanctions.
Yes. If your team applies together, one member should be designated as the lead contact with the Experiment Foundation and PI who also completes the survey.
No. Independent researchers and dedicated amateurs are encouraged to apply.
No, existing project ideas can be submitted as long as they fit the Experiment protocol. Please note that all projects on Experiment commit to openly sharing their results, data, and progress with an open-source mindset.
We are not able to fund for-profit companies, startups, or LLCs through this grant.
Endorsements should come from a suitable peer, advisor, or institutional contact — see the endorsement guide for who qualifies. If you're unable to find an eligible endorser, reach out to [email protected] for help.
Submissions are accepted on a rolling basis until the posted deadline. Applications in progress as of the deadline are reviewed at the Field Builder's discretion.
Projects are reviewed for eligibility, then evaluated by a Program Officer for scientific merit, fit with the grant's goals, and alignment with the program's mission. Some elements of chance may be used in project selection. Projects selected for funding then go through a final administrative and legal review before funds are released.
Reviews are conducted on a rolling basis starting one week after the application opens; most applicants hear back within 2-4 weeks after the submission deadline.
Selected projects will be invited to join the Field Building Community. Beyond funding, each Grant Officer provides feedback and support throughout your project, and experiment staff will review your project write-up and offer formatting guidance before launch.
Funds are added once your project launches on experiment. Typical launch phases run about 30 days, with a 2–4 week payout period after the project closes. Learn more in the Experiment payout guide.
You'll need to crowdfund the remaining amount to receive any payout, since experiment projects run on an all-or-nothing funding model. You can also use Stretch Goals to raise more than your minimum target.
Yes — every project is also eligible for crowdfunding on Experiment. You can also use Stretch Goals to raise additional funds.
Grants will be administered to projects up to $10,000. The goal of this small, fast grant program is to get these projects off the ground quickly. If selected, the grant can support an ongoing or new experiment project.
Project leads are encouraged to try to raise additional support from the crowdfunding process, as granted amounts may only cover part of the project funds.
In addition to funding, grantees are invited to join a team of researchers working together across the field.
This grant program is open to non-profits and individuals, and is not able to support for-profit companies or activities. Please review the eligibility criteria and FAQ for additional details.