Mean fish share distance was 0.10 higher at matched coral cover than between early years at five Moorea sites, 2006–2017

We tested whether recorded fish shares returned to early values when nearby live stony coral cover returned. We used open MCR LTER fish and coral data. Five of six fixed fore-reef sites had a first coral-cover match within 20% of their 2006–2007 mean in 2013–2018. The match years were 2014–2017. Two fixed fish transects had the same observers throughout. At the five matches, mean Bray–Curtis distance from the two early fish surveys was 0.4538. Distance between the early years was 0.3533. The mean excess was 0.1005, with four positive sites. Leaving out the north shore reduced it to 0.0183. Observer, taxon-group, cover-window, and equal-count checks retained a positive mean. At 200 fish per survey vector, the mean excess was 0.0903; the central 95% sampling range was 0.0545–0.1245. This range conditions on visual counts and is not an ecological confidence interval. A second return year gave a positive mean diagnostic. The result shows different recorded fish shares at similar coral cover in this fixed sample. It does not separate disturbance effects from background change, seasonal survey shifts, or detection error. The package has unchanged source data, provenance, code, figures, all site results, and independent raw-CSV checks. A clean run reproduced 28 result and image files byte for byte.

coral reef · Moorea · fish composition · coral cover · temporal change · reproducibility