Aim

Conservation efforts have traditionally addressed the threat of biodiversity loss by prioritizing regions and habitats with high endemic species richness. However, species-poor habitats often harbor distinct, valuable, and/or functionally unique species that contribute to regional diversity. In the tropical marine realm, the “mangrove-seagrass-coral reef continuum” has dominated both the scientific literature and marine spatial planning. Bare sediment habitats are also part of this continuum and often comprise the majority of bottom area, as they do in most regions worldwide, but are understudied and rarely considered in conservation. To address this information gap, we partition the relative contributions of five habitats in the coastal mosaic (forereefs, patch reefs, seagrass, mangroves, and bare sand) to fish diversity across a tropical seascape.

Location

Carrie Bow Cay, Belize Barrier Reef

Methods

Using diver visual censuses, we assessed fish community composition at three replicate sites from each of the five habitats annually for five years. We then partitioned diversity into within (α) and between (β) habitat contributions, and further partitioned β-diversity into components attributed to species replacement and richness difference. Finally, we determined the local contributions to β-diversity (LCBD) by each habitat and identified the key individual species behind these contributions.

Main Results
Tropical sand flats were unique habitats, contributing disproportionately to total fish species diversity across the seascape, despite having the fewest species overall. Sand flats also had higher-than-expected local contributions to β-diversity (LCBD) and had the highest percentage of species not found in other habitats.

Main Conclusions

Our results demonstrate that previously understudied sand habitats harbor a relatively unique assemblage of species compared with other nearby habitats.  Marine spatial planning aimed at maximizing preservation of total biodiversity (including unique species) in the tropics should consider sand habitats.

Data were collected by diver visual census using the Reef Life Survey protocol: https://reeflifesurvey.com/methods/. In brief, divers swam 10m wide x 5m high x 50m long transects working in teams of two, each diver surveying a 5m wide transect block. Divers counted and sized all swimming fish using Reef Life Survey "Method 1." Then, in a second pass ("Method 2"), each diver surveyed a 1m wide block for cryptic fish and mobile invertebrates that reach an adult size of at least 2.5cm. Fish are placed into size bins, and invertebrates are counted but not sized.

Three permanent sites were surveyed for each habitat across all five years. 

Two seagrass sites, "Blueground Seagrass" and "Tobacco Seagrass" were each only surveyed in one year each.

Taxonomic names were validated and classifications were assigned using the 'wormsbymatchnames' function within the 'worms' R package, which scrapes the world register of marine organisms database.

Each transect contains two 'blocks' (per Reef Life Survey methods) which are surveyed separately. We pooled the blocks for all analyses, as they are adjacent to one another and could not be treated as independent replicates. The 'scientific_name' column refers to the validated name used for analysis. The 'orig_scientific_name' column refers to the original name recorded by the observer, which in some cases was a misspelling or unaccepted name that was changed during the taxon validation process. All numbered columns refer to size classes, and data are entered as a count of individuals within each size class, then totaled in the column 'total_count.'  

This information and more can be found in the ReadMe file attached.