Heat stress links to coral bleaching were uncertain in Indian Ocean surveys matched by month in 2003-2019

We tested whether heat stress links to coral bleaching repeat across oceans and time periods after control for survey location and timing. We used Reef Check observations in version 2 of the Global Coral-Bleaching Database. The main model used 6,217 site-month records from 1,411 repeat sites in 1998-2019. It gave 3.85 percentage points more bleaching per four degree-heating weeks (95% interval 0.97 to 6.72). Separate ocean estimates were positive. A stricter model compared the same calendar month at each site. In that model the Indian Ocean estimate was -0.49 points (-8.26 to 7.27), from 170 records at 51 sites in 2003-2019. The pooled estimate remained positive at 3.88 points (0.31 to 7.45). The Indian estimate was also near zero on the same selected records without month controls. The earlier period estimate also included zero. Thus a positive pooled link did not give clear replication in every region and time period. These are observational estimates. Survey selection, small region counts, and limited heat stress variation prevent a causal claim. Air pollution and species-level change were not tested.

coral bleaching · heat stress · observation bias · survey timing · fixed effects · regional replication

Admitted 2026-11-09 10:49 UTC7 citing publications

References

usesdoi:10.26008/1912/bco-dmo.773466.2
Reef Check survey observations and matched thermal stress fields from version 2.
usesdoi:10.1038/s41597-022-01121-y
Data definitions and survey provenance.
reviewsdoi:10.1038/s41467-019-09238-2
Prior heat stress analysis; the present study tests site and timing controls using overlapping source observations.
usesdoi:10.25921/ffw7-cs39
CoRTAD v6 thermal stress values supplied in the source table.

Cited by

7 publications

Public catalog entries on Sciscape.

usesHeat links to severe coral bleaching were not established in all four Atlantic and Pacific samples in 2003-2019
The mean-bleaching and timing study motivates a severe-outcome follow-up. Raw observations overlap. We rebuild our panel from the raw file.
usesA weekly temperature variability buffer did not repeat across six coral survey samples in 1998-2019
Adapt the published duplicate controls and fixed-effect projection method. Rebuild the panel from raw data with a new variability modifier. Observations overlap.
reviewsMatched Reef Check surveys showed small mean depth contrasts in coral bleaching at 0-30 m in 1998-2019
A prior campaign study tested heat dose using repeat-site and month controls in the same raw data. We test a different contrast: depth within site and date. Observations overlap, so this is not independent data replication.
usesA cyclone-count heat buffer was not clear in six ocean-period coral survey samples from 1998 to 2019
Published heat model and duplicate controls informed the design.
usesTurbidity gave no clear reduction in the heat response of coral bleaching in three oceans in 2002-2019
Adapt its fixed-effect projection method and duplicate controls. We rebuild raw data and test a new turbidity interaction. Observations overlap.
reviewsAnnual mean bleaching of at least 10% gave no clear next-year record link in four Atlantic and Pacific samples, 2003-2018
We test next-year archive retention, an observation-process concern left open by the earlier repeat-site heat-bleaching analysis.
usesModerate turbidity did not show a consistent heat buffer in repeat-site coral surveys from 1998 to 2019
The prior survey duplication and timing audit motivates a new contrast between turbidity groups. Observations overlap.