Latitude gave no clear common heat-slope change in four Atlantic and Pacific coral samples in 2003-2019

We tested whether absolute latitude changes the within-site link between heat stress and coral bleaching. Four Atlantic and Pacific samples in 2003-2010 and 2011-2019 contain 5,336 Reef Check site-months. Models control site, year, calendar month and depth. All four primary country-cluster intervals included zero after correction for four tests. Heat-slope changes per ten degrees of latitude were -3.05, +9.52, -3.64 and -2.54 bleaching points per four TSA degree-heating weeks. A squared-heat term reduced the last estimate to -0.84. Equal site weights gave a negative adjusted interval only in the late Pacific sample. Site clusters gave opposite clear signs in the two late samples. Thus no common direction repeated across all four samples. This test does not reject a prior nonlinear latitude main effect. Selected reports, few country clusters, shared heat events and unmeasured coral composition limit inference. Code, source hashes, aggregate results, numerical checks and a figure are included. External raw data remain outside the bundle. No air pollution or causal climate-change effect is tested.

coral bleaching · latitude · thermal stress · observation bias · regional replication · fixed effects