Within-reef heat links to bleaching depth contrasts were uncertain in four Atlantic and Pacific samples in 2002-2019

We tested whether surface heat changed deep minus shallow coral bleaching at repeatedly surveyed reefs. We formed matched site-date pairs in Reef Check records from BCO-DMO 773466 version 2. Four Atlantic and Pacific period models used 726 pairs. Models controlled site, year, month and depth gap. Slopes ranged from -5.26 to +5.79 bleaching percentage points per four TSA degree-heating weeks. All four country-cluster intervals included zero after correction for four tests. The later Pacific estimate was negative but uncertain. In the later Atlantic model, one country supplied 99.3% of residual heat variation. Depth-gap, date and same-month checks also gave intervals that included zero. These records do not establish a common heat buffer at depth. Matching reduces differences in survey location, date and source. Few heated pairs, selected repeat visits, missing outcomes, changing depth endpoints and unknown coral communities limit inference. The result is not a causal depth effect, a climate trend or an air pollution test. The package includes paired evidence, source hashes, code and numerical checks.

coral bleaching · marine heat stress · depth refuge · matched surveys · observation bias · regional replication · null result