Back

Active late Holocene accretion of mesophotic coral-rubble fields in the northern Red Sea

Eyal, G.; Weinstein, D. K.; Hammerman, N. M.; Shaked, Y.; Torfstein, A.

2026-10-09 ecology
10.64898/2026.10.07.757296 bioRxiv
Show abstract

Mesophotic coral ecosystems are commonly interpreted as thin biological veneers over relict shallow-water reef structures, but their capacity to generate late Holocene carbonate deposits remains poorly quantified. Here we combine diver excavations, rubble-bed thickness surveys, radiocarbon age-depth modelling, sediment grain-size analyses, coral-rubble assemblages, trace-element geochemistry, and modern Alveopora allingi growth measurements from two 40-65 m deep coral-rubble fields in the northern Gulf of Eilat/Aqaba, Red Sea. Rubble deposits are at least 1.0-1.5 m thick and accumulated at mean rates of 2.21 {+/-} 0.52 mm yr-1 at the Dekel site and 1.33 {+/-} 0.13 mm yr-1 at the IUI site, implying persistent carbonate accumulation over centuries to >1 kyr. Age-depth models indicate quasi-continuous deposition despite pronounced faunal changes, including a shift from A. allingi-dominated rubble to upper layers enriched in Euphyllia paradivisa at Dekel. The dominant modern rubble-forming coral, A. allingi, extends at 3.04 {+/-} 0.87 mm yr-1 and has skeletal density of 1.17 {+/-} 0.27 g cm-3. First-order cover-based accretion estimates at present coral cover fall well below the radiocarbon-derived rates, indicating that long-term mesophotic rubble accumulation integrates historical cover, secondary calcification and sediment retention rather than present live coral cover alone. These data demonstrate that mesophotic rubble fields can act as active carbonate factories and palaeoecological archives, challenging the assumption that modern mesophotic habitats are geomorphologically passive.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.