Human, Oceanographic and Habitat Drivers of Central and Western Pacific Coral Reef Fish Assemblages

Abstract

Coral reefs around US- and US-affiliated Pacific islands and atolls span wide oceanographic gradients and levels of human impact. Here we examine the relative influence of these factors on coral reef fish biomass, using data from a consistent large-scale ecosystem monitoring program conducted by scientific divers over the course of >2,000 hours of underwater observation at 1,934 sites, across ~40 islands and atolls. Consistent with previous smaller-scale studies, our results show sharp declines in reef fish biomass at relatively low human population density, followed by more gradual declines as human population density increased further. Adjusting for other factors, the highest levels of oceanic productivity among our study locations were associated with more than double the biomass of reef fishes (including ~4 times the biomass of planktivores and piscivores) compared to islands with lowest oceanic productivity. Our results emphasize that coral reef areas do not all have equal ability to sustain large reef fish stocks, and that what is natural varies significantly amongst locations. Comparisons of biomass estimates derived from visual surveys with predicted biomass in the absence of humans indicated that total reef fish biomass was depleted by 61% to 69% at populated islands in the Mariana Archipelago; by 20% to 78% in the Main Hawaiian islands; and by 21% to 56% in American Samoa.

Fishing and environmental influences on estimates of unfished herbivorous fish biomass across the Hawaiian Archipelago

“…Coral reef fisheries in Hawaii exploit multiple species but especially herbivorous fishes such as parrotfishes and surgeon-fishes (Friedlander & Parrish 1997). While there are multiple regulations governing the harvest of reef fish in Hawaii, such as size restrictions and spatial closures, the difficulty of monitoring fishing effort has resulted in inconsistent fisheries-dependent data, creating a need to explore fishery-independent data (Nadon et al 2015).…”

“…In 2012 and 2013, divers estimated the maximum vertical relief within their SPC cylinder and the proportion of their cylinder in different relief bins (< 0.2, 0.2−0.5, 0.5−1.0, 1.0−1.5, >1.5 m from substrate), and these data were used to generate a mean vertical relief value in each cylinder. In order to analyze fish biomass data associated with the 2 different complexity metrics, we used a data calibration published in Williams et al (2015). Williams et al (2015) collected both types of complexity data and generated an equation for predicting mean vertical relief from the 5-point habitat complexity scale.…”

“…In order to analyze fish biomass data associated with the 2 different complexity metrics, we used a data calibration published in Williams et al (2015). Williams et al (2015) collected both types of complexity data and generated an equation for predicting mean vertical relief from the 5-point habitat complexity scale. We used their linear regression formula to generate estimates of mean vertical relief at each site surveyed in 2010 and 2011.…”

“…Instead, first-order approximations of B unfished are typically generated from current biomass estimates at unfished locations such as remote, unpopulated areas or long established marine protected areas where the effects of fishing are minimal (McClanahan et al 2007, Williams et al 2011, Edwards et al 2014, but see Ault et al 2008, Nadon et al 2015 for alternatives using length-based methods). Consequently, past studies of fishing effects on coral reef fishes have generally had 2 key ingredients for depletion estimates: (1) current estimates of fish biomass in populated, fished areas, and (2) current estimates of fish biomass in unpopulated areas or regions where fishing is prohibited.…”

Coral reef soundscapes : spatiotemporal variability and links to species assemblages

“…Coral reefs vary considerably in their species assemblages in both space and time (Parravicini et al, 2013;Williams et al, 2015). Accordingly, a large body of ecological literature has focused on trying to identify the drivers of this variability.…”