Accumulating data regarding the increasingly degraded coral reefs worldwide have directed focus to the unique mesophotic coral ecosystem (MCE) as a potential refuge for shallow-water endangered species. Sponges play a crucial role in coral-reef functioning but are often overlooked in benthic surveys. This knowledge gap is especially true for the Red Sea, where data on sponge abundance, coverage and species composition are scarce. Moreover, no study to date has examined Red Sea mesophotic sponges.

Here we compared sponge diversity metrics between the shallow and mesophotic reefs in the Gulf of Aqaba, Northern Red Sea. We also examined the role of biotic and abiotic parameters in determining sponge spatial variability. Sponge diversity metrics significantly varied with increasing depth and between sites. Sponge species composition also significantly differed between depth and sites. However, while no significant evidence supported a sponge bottom-up control, the findings indicated that local factors, such as site characters and biotic interactions, might play essential roles in determining sponge diversity and community composition. The sponges’ spatial differences found in this study highlight the importance of considering variability in benthic communities and the factors controlling this, when designing management tools for coral reefs, particularly for the Northern Red Sea.

Data was collected during suryes at the shallow (5, 10, 20 m) and mesophotic (50, 60, 70 m) reef at the Gulf of Aqaba/Eilat northern Red Sea at two sites the IUI and OJ. At each site and depth the quantitative survey comprised three transects of 3.5 m² each (total 10.5² per site*depth). Each transect was created from 25-30 photo-quadrats, processed using ImageJ [40] to create image collections of equal size (~70X50 cm). The MCE survey was conducted from the R/V Sam Rothberg using a Remotely Operated Vehicle (ROV ECA-Robotics H800) equipped with FHD camera, four video lights (4000 lumens) and two parallel laser beams for scale (7.5 cm apart). A similar methodology was performed at the shallow reefs, using SCUBA diving, with photo-quadrats taken ~1 m above the reef using a CANON G12/16 camera mounted on a tray with two video lights (3500 lumens) and two parallel laser beams for scale. At both habitats, additional close-up pictures of the surveyed area (in each photo-quadrat) were taken in order to facilitate identification of the living organisms on the substrate. For sponge coverage calculation, 100 sampling points were overlaid on each image (50x70 cm) in a stratified random manner using CoralNet online software. Benthos components beneath each point were classified into one of three categories: sponge, coral or algae, and each component’s coverage was calculated in relation to the others (relative abundance). To assess sponge diversity at each site/depth (e.g., richness, abundance, Shannon, evenness), the sponge species were counted and identified in each image to the lowest possible taxonomic level. Where feasible, sponge species were collected during the survey, facilitating identification in the photo-quadrats. However, when taxonomic identification and/or sampling were not possible, sponges were defined according to their morphology and referred to as "morphospecies". Finally, the average sponge diversity per site or depth was calculated based on the averaged values per transect. Seawater was sampled adjacent to the reef (<50 cm) at 50 and 5 m depth at the two survey sites, using horizontal 12 L Niskin bottle (Fig. S1). Samples were collected during winter (Nov 2018-Mar 2019) and summer (Jun-Sep) of 2019 and processed for: Total Organic Carbon (TOC), Particulate Organic Carbon (POC) and inorganic nutrients (NOx^-, SiO₂, PO₄^3-), and fluorescence-activated cell-sorting (FACS) was used to sort the cells.

All statistical analyses and data visualizations were performed with RStudio 2021.09.0+351