Volatility in coral cover erodes niche structure, but not diversity, in reef fish assemblages

Archived file name: “Tsai_etal_2022_ScienceAdvances_archive.RData” 1. “Tsai_etal_sumzdata”: Reef fish community structure variables and environmental variables across 40 reefs are stored as R data.frame object. Within this data.frame object, the reef IDs (“reefID”), latitude (“lat”), cross-shelf position (“shelf”), mean coral cover (“mC”), and standard deviation (“sdC”) and log-standard deviation (“logsdC”) in coral cover, (time-averaged) Poisson-lognormal estimated richness (“plnRichness”), (time-averaged) Poisson-lognormal estimated unevenness (“plnUnevenness”), and VPRSA variance components due to persistent species differences (“Vr”), stochastic fluctuations (“Ve”), and overdispersion (demographic/sampling stochasticity) (“Vd”) are recorded. A transformed shelf position, 1 - “shelf”, was used for the piecewise SEM analysis in the MS. 2. “AIMSLTMPfishcomm.list” and “AIMSLTMPfishcomm.list.reefID”: We extracted and compiled the community dynamics data of LTMP from 1994 to 2004 (11 years) for analyses. We pooled fish community and coral cover data at the scale of the entire reef, summing species abundances over all 15 transects laid on each reef. Percentage living hard coral cover is averaged across transects and sites within reefs. We adopted this approach to reduce stochastic sampling error, thereby obtaining more precise estimates of the community structure statistics that are of interest. Because the small-sized fish taxa (mainly Pomacentridae) were surveyed in narrower transects than other, larger fish taxa, we used subsampling (Poisson samples) to rescale the abundances of large-sized species to standardize sampling effort. Each fish counted on the wider transects was given a 20% probability of appearing in the sub-sample (because the small-fish transects cover only 20% of the area of large-fish transects). Sub-sampled data were used for the analyses.