Coral cover at St. John, VI, and Mo'orea LTER sites, 1992-2019

<p><strong>Overview</strong></p>

<p>The reefs in Mo'orea and St. John have been monitored for decades, and recent publications from these efforts provide the ecological background to fully interpret the present results. Coral cover (pooled among taxa) for all three habitats in Mo'orea, as well as St. John, are presented, with these data originating from photoquadrats (0.5 × 0.5 m) that have been recorded annually at permanent locations from 2005 to 2019. In brief, in Mo'orea, the three habitats were sampled at six sites around the island, with 40 photoquadrats/site at 10-m and 17-m depth on the fore reef, and 20 photoquadrats/bommie, and five bommies/site, in the back reef. In 2005, photoquadrats were randomly positioned along a ~ 50 m transect at 10-m and 17-m depth at each site on the fore reef, and thereafter were sampled in the same positions. Five photoquadrats were randomly positioned along four, 5-m transects along cardinal axes at each bommie every year. In St. John, the fringing reefs (at 7-9-m depth) were sampled at six sites, with 40 photoquadrats randomly positioned every year along a single 40-m transect at each site.</p>

<p>Photoquadrats from the fore reef and from St. John were analyzed using CoralNet Software, with 200 dots randomly located on each image, and the substratum beneath each dot identified. The photoquadrats in the back reef were analyzed with a coarser resolution in which the dominant substratum is each of 25 sub-squares in each photoquadrat was identified (i.e., providing 4% resolution). The cover of scleractinians (pooled among taxa) is presented using sites as replicates, and mean cover by year is used in tests of association with concurrent coral recruitment.</p>

<p>To test the hypotheses guiding the present study, coral recruitment was measured in Mo'orea and St. John using settlement tiles deployed from 2005 to 2019, and loggers were used to measure temperature. The immersion times of settlement tiles slightly differed between regions because the two sampling programs originally were designed for different purposes.</p>