A 2025 study in Nature Ecology & Evolution concluded that coral restoration cannot be scaled globally to save reefs. We agree — with current methods. So we're proposing a different method entirely.
Coral reefs provide ecosystem services valued at up to $9.9 trillion per year — coastal protection, fisheries, tourism, biodiversity, pharmaceutical compounds. Over one billion people depend on reef ecosystems for food or income. Reefs protect more than 150,000 km of shoreline from storm damage.
Since the 1950s, global coral coverage has declined by half. Between 2009 and 2018 alone, the world lost approximately 11,700 km² of coral — more than all the living coral in Australia. Between 2023 and 2025, over 83% of reef areas globally experienced bleaching-level heat stress.
The question is not whether we can afford to restore the ocean. It's whether we can afford the methods we're currently using.
The most comprehensive analysis of coral restoration ever published — Mulà, Bradshaw, and Strona's 2025 paper in Nature Ecology & Evolution — examined all documented restoration projects worldwide and reached a devastating conclusion: current restoration cannot be scaled to offset losses.
Even at the most optimistic cost estimates and assuming every project succeeds, restoring just 10% of degraded reef areas would require over $1 billion. At realistic costs, accounting for the high failure rates and expensive techniques involved, the figure exceeds $16 trillion. That's nearly four times the entire investment in coral restoration over the past decade.
The median project size was roughly 100 m². The median cost was around $400,000 per hectare. And the study found that most restoration sites are chosen based on accessibility to humans rather than ecological suitability — and that more than half of recently restored sites have already been hit by severe bleaching.
~100 m² median project size Bayraktarov et al.
$400,000/ha median cost Bayraktarov et al.
$558,000/ha median nursery-transplant Mulà et al.
60.9% median fragment survival Bayraktarov et al.
Genetic clones from fragmentation Harrison et al.
Specialist divers, labs, epoxy required
11,700 km² lost in one decade GCRMN
Cost/ha low enough for climate finance
Millions of hectares need habitat
Self-sustaining ecosystems that expand
Genetic diversity for adaptation Dela Cruz & Harrison
Community-operable with basic skills
Coral reefs, oyster beds, kelp forests, mussel banks, gorgonian communities — across every degraded marine ecosystem, the same dynamic plays out. Marine organisms need hard substrate to attach to. When that structure is destroyed — by trawling, dredging, dynamite, or simply by the death of the organisms that built it — the biology disappears. Without biology, no new structure forms. The system stalls: bare sediment, no recruitment, biological desert.
But the reproductive biology hasn't stopped. The water column is saturated with larvae, spores, and propagules: coral spawn drifting on the surface, oyster veligers searching for shell, kelp zoospores looking for rock, mussel larvae seeking any hard surface. Every spawning season, every tidal cycle, the ocean's organisms are trying to recolonise.
They don't need our expertise. They need a surface to grab onto.
This principle is universal. It governs scleractinian corals in Indonesia, Cladocora caespitosa in the Mediterranean, Mytilus edulis in the Baltic, Crassostrea virginica in the Chesapeake, Ostrea edulis in the North Sea, and gorgonians on every temperate shelf. The species change. The principle does not.
This universality is the leverage point. A single operational model — provide substrate, protect recruits through their vulnerable stage, deploy at scale — works across all of them. One method. Every ecosystem. Every coast.
Nothing we propose is novel. Every piece of this approach has been demonstrated independently, often at large scale. The innovation is assembly: combining proven elements into a single, scalable, replicable system.
In September 2025, the Chesapeake Bay Program completed the world's largest oyster reef restoration — 10 tributaries, over 1,500 acres of constructed reef, 7.19 billion hatchery-spawned oysters planted in Maryland alone by end of 2024. Cost: approximately $93 million. Harris Creek, the first completed site, hit 98% of reefs meeting target density. The method: produce substrate, seed with spat, deploy, let nature grow.
Australia's National Sea Simulator raises approximately one million young corals per season. Mediterranean Cladocora caespitosa micro-fragmentation achieves 89.8% survival after one year. CSIRO modelling confirms larval reseeding scales at low cost per colony. These are laboratory and small-nursery results — our approach uses the same biology with simpler, larger infrastructure.
Every artificial reef ever deployed — from ancient Japanese fishermen dropping boulders, to Mediterranean stone structures dating to antiquity, to modern concrete modules — demonstrates the same thing: put hard substrate on bare seabed, and marine life colonises it. This is not a hypothesis. It is the oldest, most replicated finding in marine ecology.
In Virginia, 38 million seagrass seeds were distributed from small boats over 11 years, restoring 1,700 hectares of seagrass habitat — the most successful marine vegetation restoration ever documented. Australia's RRAP now targets billions of coral larvae across multi-kilometre reef areas using similar dispersal logic.
The dominant coral restoration method — cutting fragments from donor colonies and replanting them — produces genetic clones. An entire restored site can be populated by fragments from just a handful of parent colonies. This creates monocultures: genetically identical populations vulnerable to the same diseases, the same thermal thresholds, the same pathogens.
Research consistently shows that sexually propagated corals — grown from larvae produced by spawning events — have greater genetic diversity and higher adaptive potential than clonal fragments. Each larval recruit is genetically unique, carrying a novel combination of parental traits.
Genetic copies of donor colonies. Limited adaptive capacity. All individuals share the same thermal tolerance threshold. One disease can kill an entire restored site.
Requires careful genotype tracking and mixing of 30+ donors to partially mitigate — rarely achieved in practice.
Every recruit is genetically unique. Full genetic shuffle of sexual reproduction preserved. Natural variation in heat tolerance, disease resistance, and growth rate.
The genotypes best adapted to local and future conditions survive and reproduce. The reef adapts itself.
In a warming ocean, this genetic diversity is not a luxury — it is the difference between a restored site that bleaches and dies in the next marine heatwave and one that persists and adapts. Our nursery approach uses natural spawning and larval settlement, preserving the full evolutionary toolkit.
Engineered substrates — 3D-printed ceramic tiles, manufactured coral skeletons, purpose-built concrete modules — all work as settlement surfaces. But they introduce cost, manufacturing complexity, and supply chain dependencies that limit scale.
Its mineral composition — calcium carbonate in limestone, silicates in volcanic rock — is chemically compatible with marine organism settlement. Its surface irregularity at every scale, from micro-pits to ridges to pores, provides vastly more settlement area and shelter than any moulded surface. It accumulates biofilm and coralline algae naturally, creating the chemical cues that trigger larval settlement behaviour.
Rock is also heavy, irregular, and self-stabilising on the seabed. Boulders interlock, resist wave energy, and create the complex three-dimensional architecture — caves, overhangs, crevices, channels — that reef ecosystems require for fish shelter, predator avoidance, and microhabitat diversity. No design iteration. No manufacturing. No supply chain. Just rock.
The Chesapeake Bay oyster restoration used recycled shell and simple substrate bases. Japan has used natural boulders for reef construction for millennia. The Mediterranean has stone structures underwater that are centuries old and fully colonised. The precedent is ancient, global, and consistent.
When your primary input is the cheapest bulk material on Earth, available in every country, requiring no manufacturing — your cost-per-hectare drops to a range that climate finance, development banks, and national environmental budgets can actually fund.
Mulà et al. (2025) concluded that current coral restoration methods cannot be scaled globally — the costs are too high, the projects too small, the sites chosen for convenience rather than ecology, and most restored reefs face severe bleaching risk within decades.
We agree completely. Current methods can't scale. That's not a reason to stop — it's a reason to change the method.
What if restoration didn't require specialist divers? What if the substrate cost approached zero? What if the method produced genetically diverse populations instead of clones? What if it worked for coral, oysters, mussels, kelp, and gorgonians alike? What if the nurseries could be built and operated by any coastal community with basic aquaculture skills?
What if we just gave the ocean a rock?