1 min read

A Healing Touch Revives Coral

The same rescue that saves one bleached coral turns hostile the moment the donor is a genetic stranger. And the fastest part of the recovery went unexplained.

In an aquarium in California, a pale scrap of coral, stripped of the algae that feed it, sat nuzzled against a healthy neighbor. Within weeks it had regained its color. Fresh algae had crossed over through a bridge of living tissue.

Corals depend on single-celled algae living inside their tissue, which photosynthesize and provide the bulk of their food. Heat stress makes corals expel those algae. That is bleaching, and a bleached coral can starve or get smothered by other algae.

Researchers at California's Hopkins Marine Station heat-bleached aquarium-kept pieces of Pocillopora damicornis, each between 1 and 1.5 centimeters (0.4 and 0.6 inch) long. Each pale piece was then glued in place touching a healthy one that was its genetic clone. All five regained full color within eight weeks, one in just four. Pieces kept 5 to 7 millimeters (about 0.2 to 0.3 inch) away stayed mostly white, grew a shaggy coat of algae and in two cases died.

Touch mattered because the pieces fused. When the team pulled pairs apart, both halves showed matching tears where color first returned, the remains of tissue bridges. One week of contact supplied enough algae for a bleached piece to finish recovering alone. By week two, the algae inside were doubling roughly every four days.

There is a firm condition. Paired with a healthy piece from a different genetic individual, a bleached coral met stinging filaments, and the healthy coral grew over it.

The study could not explain why the transferred algae multiplied so fast, or show whether any of this would work on a reef rather than in a tank.

Every successful donor was, genetically speaking, the patient. Every other donor treated the patient as real estate.

Read the full story at Science Advances, October 14, 2026


Hot Take: Coral nurseries keeping cuttings from every colony should file them as spare parts, sorted by genotype. Mixing up the labels is, on this evidence, not recommended.

Subscribe to our newsletter.

Be the first to know - subscribe today