Turbidity gave no clear heat buffer for annual coral cover change in four Atlantic and Pacific samples in 1998-2019

We tested whether site mean turbidity changed the link between survey-date heat and annual hard-coral cover change. HeatCRD version 1 gave 1,110 matched annual pairs at 425 sites. We matched site, source, month and depth. Four ocean and period models controlled first-survey heat, its turbidity interaction, initial cover, depth, year, month and source. The second-heat slope changed by -3.69, +1.82, +4.99 and +0.89 cover points per four TSA degree-heating weeks per turbidity doubling. All four corrected intervals included zero. Only 66 pairs had second-survey heat of at least four; 14 had that heat and moderate turbidity. Country clusters, initial-cover removal, a turbidity limit and a linear turbidity scale did not establish a common buffer. These selected records do not show that turbidity has no effect. Sparse heat support, site mean optics, survey-date heat, source imbalance and missing visits limit inference. The package gives source hashes, matched data, model matrices, code and independent checks.

coral cover · marine heat stress · turbidity · observation bias · annual change · null result