Does Coral Reef Restoration Actually Work?

Coral restoration is neither a miracle nor a gimmick. The evidence says it works in a narrow, specific sense, and fails badly if you ask it to do more than that.

By Clarisa Strohmeyer, Founder | Updated 18 September 2026 | 14 min read

Yes, coral restoration can deliver measurable benefits at selected sites. A global review reported coral survival around 60 to 70% across restoration projects, although monitoring periods varied and were generally short. A 2025 meta-analysis of 764 marine restoration interventions found that about 64% of them met a threshold of at least 50% survival. But almost every project is small, most are monitored for less than two years, and the total area restored worldwide is trivial next to the area being lost. Restoration rebuilds particular reefs. It does not save coral reefs.

What “works” has to mean

Most arguments about coral restoration are really arguments about the question. “Does it work” can mean at least four different things, and the honest answer differs for each.

Do transplanted corals survive? Usually, yes. Does a restored patch of reef come to resemble a healthy reef? Sometimes, under good conditions, over years. Can restoration reverse the global decline of coral reefs? No, and the most recent analysis is blunt about it. Is restoration therefore worth doing? That is a judgement, not a measurement, and it depends on what you expect from it.

Anyone who answers the first question and implies they have answered the third is selling something. That includes organisations like ours, which is why this page separates the four.

What the survival data actually shows

The most comprehensive review of the field is Boström-Einarsson and colleagues, published in PLOS ONE in 2020. It gathered coral restoration case studies from around the world and looked at what was actually reported rather than what was claimed.

A global review reported coral survival around 60 to 70% across restoration projects, although monitoring periods varied and were generally short. These results demonstrate that corals can be grown and transplanted successfully, but do not by themselves establish lasting reef recovery.

The 2025 Nature Communications meta-analysis by Danovaro and colleagues looked at 764 marine restoration interventions across habitat types. About 64% of them met a threshold of at least 50% survival. That is a count of how many interventions cleared a bar, not an average survival rate, and it covers marine habitats generally rather than coral reefs alone. The same analysis found no significant decline in success as the area of an intervention increased, which matters for anyone arguing that restoration cannot be scaled up at all.

About 9% of interventions in that analysis failed outright, again across all the marine habitats studied rather than coral reefs alone. Roughly one in eleven producing nothing is what a normal applied science looks like. What matters when you read any survival figure is the denominator, the timeframe, and what the project counted as success.

Why the reported range is so wide

Published survival figures vary enormously between projects, and the variation is not random. Three things drive most of it.

Species choice. Around 59% of studies in the 2020 review focused on fast-growing branching corals, largely Acropora. These grow quickly and are easy to fragment, which makes projects look good on a two-year horizon. They are also among the first corals to die in a bleaching event. Branching corals do contribute real reef structure. A study of restored sites in Indonesia found their carbonate production tripling to match healthy reference reefs within four years. But a project reporting excellent survival on branching species has not shown that the same method works for the slow-growing massive corals that build a reef’s framework, and species choice should be judged against the site’s ecology, its diversity and what the project is actually trying to achieve.

Monitoring window. About 60% of projects reported under 18 months of monitoring. Coral survival curves are not flat. Losses cluster in the first months, then again during the first thermal stress event the site experiences. A project that reports 12-month survival and stops has measured the easy part.

What gets published. Projects that fail may be less likely to be written up. Danovaro and colleagues tested for publication bias and did not detect it. That does not rule out unreported failures, but it does mean the published average should not simply be assumed to be flattering.

What restoration cannot do

This is the section that is easiest to leave out, and the one worth reading closely.

The scale problem

In the 2020 review, the median restored area per project was 100 square metres. Not a hectare. Not a square kilometre. A square ten metres on a side.

Set that against the loss. A 2025 paper in Nature Ecology & Evolution by Mulà and colleagues estimated that about 11,700 square kilometres of coral reef degraded between 2009 and 2018. The arithmetic is not close. A hundred thousand median-sized projects would cover ten square kilometres, about 0.085% of what was lost in a single decade.

The cost problem

The same paper put published restoration costs at anywhere from about US$6,000 to US$143 million per hectare, depending on method and setting. Ten per cent of that degraded area is 117,000 hectares. At the lowest published cost of about US$6,000 per hectare that comes to roughly US$702 million, and the paper’s own estimate runs higher than that. These are highly variable, technique-dependent extrapolations rather than a single price tag. Total global investment in coral restoration over the past decade was around US$258 million.

Their conclusion is worth quoting directly: “restoration alone is not a practical or affordable solution to counteract the global decline of coral reefs.”

The warming problem

A restored reef is not protected from the thing that killed the original. For the restored localities it analysed, under the intermediate SSP2-4.5 emissions scenario, the same 2025 paper projected that 99.6% would experience at least one bleaching event by 2100. Planting coral into water that will keep getting warmer does not stop it getting warmer.

This is the hard limit. Coral restoration is a local recovery tool operating under a growing global climate threat. Its lasting benefits depend on reducing local pressures alongside rapid cuts in greenhouse gas emissions. It can improve a site. It cannot out-plant warming.

So why do it at all?

Three defensible reasons survive the above, and one common reason does not.

Specific places are worth saving on their own terms. A reef that feeds a particular village, or shelters a particular coastline, matters to the people who depend on it whether or not the global picture improves. Restoration at that scale is not a gesture. It is infrastructure repair.

Reef function can come back, and quickly, at a site. Lange and colleagues, writing in Current Biology in 2024, measured restored reefs in South Sulawesi, Indonesia, where rubble fields had been consolidated with coated steel structures and planted. Within four years the net carbonate budget at those sites had tripled, matching healthy reference reefs. The authors are careful about what that does and does not mean: branching corals were preferentially transplanted, so community composition differed from a natural reef, which may affect habitat for some species and resilience to future heatwaves. It is evidence that an important reef-building function can be restored, not evidence that a whole reef ecosystem has been.

Local pressure reduction genuinely works. A global analysis by Selig and Bruno, published in PLOS ONE in 2010, examined 8,534 coral cover surveys across 310 protected areas recorded between 1969 and 2006. Coral cover inside them stayed roughly constant while cover on unprotected reefs declined. That is an association between protection and maintained coral cover, not proof that active transplanting works, and it is a smaller claim than “recovery”. Holding ground while conditions worsen is still a real result.

You cannot manage what you have never measured. Many reefs in the Coral Triangle have never had a baseline survey, and without one no claim about improvement can be checked. Restoration projects that document properly produce data that outlives the project.

The reason that does not survive: restoration as a substitute for emissions reduction, or as reassurance that the problem is handled. It is not, and it does not.

One nuance worth adding, because it cuts the other way. The 2025 Danovaro meta-analysis found that successful restorations can begin before every stressor has been removed, which contradicts a common assumption that intervention is pointless until conditions are perfect. Waiting for ideal conditions is not the cautious choice. It is just a different way of doing nothing.

What credible restoration looks like

If you are assessing a project, whether as a donor, a journalist or a student, these are the things that separate serious work from theatre.

A baseline before anything is planted. Without a survey of the site as it was, no claim about improvement can be checked. It is an easy step to skip, because it produces nothing photogenic.

Monitoring past the first thermal event. Twelve-month survival is the industry default and it is not long enough. The meaningful question is what survives the first serious heat stress the site experiences.

Species beyond the fast growers. A project working only with branching Acropora should be able to explain that choice against the site’s ecology and its own objectives, not just against what grows fastest.

Published failures. An organisation that has never reported a loss is either extraordinarily lucky or not telling you everything. Mortality is normal, and reporting it is what makes the survival figures worth anything.

Local employment and local consent. Restoration done by visiting teams on someone else’s reef, without the people who fish it, tends not to outlast the funding.

Honest framing about climate. If the language suggests the reef is being saved, rather than a patch of it being rebuilt under worsening conditions, treat the rest of the claims with care.

How Reef Without Borders approaches this

We have not yet run a funded restoration mission, so we have no survival data of our own and nothing to claim. What we can describe is the plan and the reasoning behind it.

Our first mission is scheduled for Q4 2026 at Kontiki, Mactan, Cebu. In July and August 2026 our founder spent six weeks in Cebu and personally photographed and filmed the site, which gives us a visual record of the reef before any work begins. That is a photographic record, not a quantitative ecological baseline, and the difference matters. Before we plant anything at Kontiki we intend to run repeatable surveys, define in advance how we will measure survival and growth, and where it is feasible leave an untreated comparison area, so that what we report afterwards can be told apart from changes the site would have seen anyway.

Field operations are led by Dindo C. Paquibot, a PADI Instructor since 2007 with twenty years of professional diving in Philippine waters, including coral nursery and transplantation work in 2019 with a Philippine reef restoration non-profit that has since dissolved. Our technical and scientific content is reviewed by Kyle Ali, Marine Biologist.

We intend to publish what we find, including losses. Given everything above, the most useful thing a small organisation can contribute is not another optimistic number. It is a well-documented account of one site, including the parts that do not work.

Frequently asked questions

Does coral restoration actually work?

At the scale of a site, yes. Restored corals typically survive at rates around 60 to 70%, and coral restoration has been shown to rebuild specific reef functions, including carbonate production at restored Indonesian sites within four years. At the scale of the global reef crisis, no. The area restored worldwide is a tiny fraction of the area being lost.

How long does a restored reef take to recover?

There is no single figure, and be sceptical of anyone who offers one. Recovery depends on the disturbance, the species, the site and what happens to water temperature in the meantime. Most published projects have not monitored long enough to answer this properly.

Is coral restoration a solution to climate change?

No. It is a local recovery tool operating under a growing global climate threat. Projections suggest almost all restored sites will experience bleaching this century. Restoration can help a specific reef; reducing emissions is the only thing that addresses the cause.

Why is restoration so expensive?

Published costs run from roughly US$6,000 to US$143 million per hectare depending on method and setting. The work is done underwater, by trained divers, on a site that needs repeat visits over years. Costs are driven largely by labour, boat time and the number of return visits, and they vary widely by method and setting.

What should I ask a coral restoration charity before donating?

Ask what they surveyed before planting, how long they monitor, what their survival rate is and how they measured it, and whether they have ever published a failure. The answers tell you more than any photograph.

Sources and further reading

1. Coral restoration: a systematic review of current methods, successes, failures and future directions — Boström-Einarsson et al., PLOS ONE, 2020. Source for typical survival of 60 to 70%, median restored area of 100 m², and monitoring durations.
2. Assessing the success of marine ecosystem restoration using meta-analysis — Danovaro et al., Nature Communications, 2025. Source for 764 interventions, the ~64% meeting a threshold of at least 50% survival, the ~9% complete failure rate across marine habitats, and the finding of no significant decline in success as intervention area increased.
3. Restoration cannot be scaled up globally to save reefs from loss and degradation — Mulà et al., Nature Ecology & Evolution, 2025. Source for 11,700 km² degraded, cost per hectare, and bleaching projections.
4. A Global Analysis of the Effectiveness of Marine Protected Areas in Preventing Coral Loss — Selig & Bruno, PLOS ONE, 2010. Source for the marine protected area findings.
5. Coral Reef Restoration Monitoring Guide — NOAA National Centers for Coastal Ocean Science, 2020.
6. Monitoring Working Group — Coral Restoration Consortium.
7. Coral restoration can drive rapid reef carbonate budget recovery — Lange et al., Current Biology, 2024. Source for the recovery of net carbonate budgets at restored Indonesian sites within four years, and the caveats about community composition.
8. The cost and feasibility of marine coastal restoration — Bayraktarov et al., Ecological Applications, 2016.

Related reading

• Coral Propagation Explained Simply
• What Is a Coral Nursery?
• What Successful Reef Restoration Looks Like
• Coral Reef Statistics 2026

Learn about our work →