WHOI's healthy-reef soundscapes nearly doubled coral larval settlement, while selective breeding raised adult heat tolerance by about 1°C-week

Broadcasting healthy-reef soundscapes onto degraded reefs can substantially increase coral larval settlement. Fish, squid and crabs were already known to respond to reef sound; this work extends acoustic cues to coral larvae and tests playback directly on degraded reefs. News report: the team placed Porites astreoides larvae in cups that let in sound but blocked light and chemical cues, broadcast healthy-reef recordings through underwater speakers, and found larvae exposed to the recordings settled almost twice as often on average as those without the sound treatment; the report describes this as a field test begun in 2022.

Adult coral heat tolerance has a heritable genetic basis, and one generation of selective breeding can shift offspring heat tolerance. Previous heritability estimates for coral heat tolerance came largely from early life stages or adult bleaching responses without accounting for survival; this study is the first to estimate h² in selectively bred adult corals reared to reproductive maturity. Paper abstract and results: using Acropora digitifera and selecting parents under a 1-week +3.5°C and a 1-month +2.5°C simulated marine heatwave, narrow-sense heritability h² was estimated at 0.29 (±0.16 SE) and 0.23 (±0.16 SE); high-tolerance parents produced offspring that withstood about 0.8 and 0.9°C-weeks more heat stress than low-tolerance parents; the authors note that logistical constraints prevented population and generational replication, so genetic drift cannot be excluded.

Short-term and long-term heat stress tolerance may be under independent genetic control, so selecting for one may not improve the other. Some prior work suggested rapid thermal challenge responses predict moderate-term exposure responses; this study used a bivariate animal model and a selection experiment to test the genetic correlation between the two traits. Paper results: the phenotypic correlation between the two traits was weak but significantly positive, while the genetic correlation was rG = 0.06 (±0.66 SE, P > 0.05); offspring selected for short-term tolerance did not show enhanced long-term tolerance; the authors describe the genetic correlation estimate as highly uncertain and the conclusion as tentative.

Global coral restoration site selection favors accessible, more impacted and lower-diversity locations, and restoration success is poorly predicted by available ecological and environmental variables. The study combined data from 220 global restoration projects with ecological, environmental and climate data to quantify site selection and success, and projected future heat stress exposure at restored sites. Paper abstract and results: boosted regression tree models predicting site selection had an average true skill statistic of 0.48 ± 0.18; models predicting restoration success explained less than 5% of variance; 57% (170/299) of restored sites experienced at least one bleaching alert level I or worse event within five years, and 15% experienced at least one level II event; under SSP2-4.5, 99.6% of restored sites are projected to experience at least one mass mortality year by 2100, averaging 24.8 ± 14.9 events, with the first anomaly on average in 2063 (±16 years).

These results are aimed at coral restoration practitioners, conservation managers and policymakers: soundscape playback applies to local settings where larval settlement on degraded reefs needs a boost; selective breeding applies to species and facilities where brood stock can be reared and selected; the global restoration analysis applies to assessing site selection and long-term viability. The papers emphasize that these interventions are only meaningful alongside rapid emissions reduction and reduction of local human pressures.

The news story is a summary read and does not fully present the papers' figures and supplementary materials, so the sample size, replication design and statistical details of the soundscape playback experiment cannot be confirmed here; the selective breeding paper's authors note that genetic drift cannot be excluded and that the genetic correlation estimate between short- and long-term heat tolerance is highly uncertain; in the global restoration analysis, models of restoration success explained less than 5% of variance, which the authors partly attribute to monitoring durations ranging from 0.8 to 144 months, inconsistent definitions of success and insufficient data standardization.