The ecological methods are described in detail in Edmunds et al. (2024, doi:10.1007/s00442-024-05517-y), and are briefly summarized below.

The study utilized the time series of the Moorea Coral Reef LTER, as they relate to coral community dynamics on the north shore fore reef. Annual measurements of coral cover, the density of coral settlers, and the density of small corals were used together with records of the environmental conditions to which they were exposed. Analyses focused on 2008–2021, which captured the final years of the last population outbreak of the crown of thorns (COTs) sea star, the coral population recovery that took place between 2010 and 2019, and coral mortality attributed to bleaching in 2019. Biological data came from two sites (LTER1 and LTER2) that are ~ 3 km apart, with environmental data from the same or similar sites (temperature), one of the two sites (flow at LTER1), or from 4.5 km resolution remote sensing data (Chlorophyll a as described below).

Coral cover was measured annually (April except for 2020 [August] and 2021 [May]) at 10-m depth along a 50 m, permanently marked transect at LTER 1 and LTER 2. Along each transect, 40 photoquadrats (0.5 × 0.5 m) were photographed at positions that were randomly selected in 2005, but fixed thereafter. Pictures were illuminated with strobes, and analyzed using CPCe or CoralNET software with manual annotation of 200 randomly located points on each image. Substrata beneath the points were categorized to coral genus, and the percentage cover for all corals (scleractinians and Millepora) and Pocillopora spp., is reported. The changes in cover of corals (scleractinians and Millepora) provided a holistic summary of the coral community consistent with how we have described it elsewhere and how it is described in the broader scientific literature on coral reefs. The separate summary for Pocillopora spp. provided a measure of coral cover that is the product of the most abundant coral settlers found on tiles deployed in the same habitat (i.e., pocilloporids). The density of small corals (≤ 4-cm diameter) was quantified in the field annually, shortly after the photoquadrats were recorded (but not in 2020 due to COVID-19), and was completed using quadrats (0.5 × 0.5 m) placed in the same positions as the photoquadrats. The benthos, including beneath branching corals, was inspected for small corals that were recorded to genus, and the densities of all corals and Pocillopora spp. are reported in units of corals 0.25 m^-2.

The density of coral settlers was measured using unglazed terracotta tiles (15 × 15 × 1 cm), seasoned (~ 6 months) in seawater beneath the marine laboratory dock, and then immersed on the reef at 10 m depth for ~ 6 months. Tiles were deployed from August/September to January/February and from January/February to August/September of each year at LTER1 and LTER2. Each tile was deployed independently and horizontally using a stainless steel stud with a ~ 1 cm gap beneath. Fifteen tiles were deployed at 10 m depth at each site, with tiles separated by a few centimeters to ~ 1 m. Upon retrieval, tiles were cleaned in dilute bleach, dried, and microscopically inspected (40 x magnification) for coral recruits that were identified to family. The top, bottom, and sides of the tiles were inspected, and densities of settlers for all corals and pocilloporids are reported. Because ~ 82% of the settlers was found on the lower surface of the tiles, densities (summed among surfaces) were expressed per 225 cm² of tile (i.e., the lower surface) and scaled linearly to settlers 0.25 m^-2. This assumption resulted in a slight upwardly-biased estimate in the density of recruits (versus a downwardly-biased estimate through consideration of the upper and lower surface at 450 cm²), but it did not affect interpretation of settlement tiles as an assay for the density of settling corals. For each site, mean densities from both tile immersions each year were summed to estimate annual settlement.

* See "Related Datasets" section for access to related dataset pages which include dataset-specific methodology.

site_locations.csv