Next-year heat gave no clear bleaching link in four matched Atlantic and Pacific coral samples in 2003-2018

We tested next-year heat stress as a check on coral bleaching associations. We matched the same site and calendar month in consecutive years within 2003-2010 and 2011-2019. Four Atlantic and Pacific models used 840 monthly records and controlled current heat, depth, site, year and month. Future heat slopes were +5.01, -0.85, -0.61 and +1.60 bleaching percentage points per four TSA degree-heating weeks. All four country-cluster intervals included zero after adjustment for four comparisons. Ecological-region, year and site-by-month checks also gave no clear future link. Only 840 of 7,820 source months met the match and repeat rules. Few clusters, broad intervals, heat persistence and selection on future survey reports limit the test. A null future slope does not show that current heat estimates are free from bias. This study does not estimate pollution effects, species loss or causal climate effects. The package includes code, source hashes, raw-data checks, matched evidence and offline reproduction instructions.

coral bleaching · heat stress · future exposure · observation bias · negative control · regional replication · null result