Coral reef assemblages, trophic interactions, and food web stability were examined across a remote and densely populated atoll, in the Republic of the Marshall Islands (RMI). The biomass of sharks, large-bodied piscivores, and secondary invertebrate consumers was expectedly larger in the absence of major human populations. Less intuitively, we report a doubling in the density and biomass of small-bodied acanthurids with significant human presence, whereas large-bodied parrotfishes were halved. These trends provided evidence for prey release of small acanthurids, but also indicated a reduction in grazing function given that power-law relationships govern fish size and physiology. Path analyses supported the conjecture that apex predator biomass enhanced the abundance of large-bodied herbivores, with ensuing benefits to calcifying benthic substrates and coral diversity. Human populations, as low as 40 individuals, had opposing linked interactions. The study next used species abundances to depict food chain interaction strengths, and multivariate measures of heterogeneity to depict food chain diversity. The removal of few strong links along the gradient of investigation (i.e. Acropora corals and sharks), as well as numerous weak links, was associated with fish assemblages and benthic substrates that had reduced grazing and calcification potential, respectively. The results inferred that a reduction in the functional response time of trophic guilds will follow perturbations (i.e. reduce stability), synonymous to food webs with low return rates to their modeled, stable con figuration. This study provides a broader perspective for interpreting how humans and apex predators influence ecological stability across (RMI) coral reef ecosystems.

No abstract

Howland and Baker Islands are two small, isolated reef and sand islets located near the equator in the central Pacific Ocean that are situated approximately 60 km apart. In 2004 and 2006, species-level monitoring at multiple sites, coupled with towed-diver surveys in 2002, 2004, and 2006 on both of these federally protected islands, revealed diverse fish, coral, macroinvertebrate, and algal assemblages. This study examines interand intra-island spatial and temporal differences in community composition among sites and presents baseline biological community parameters for two of the least impacted reef systems in the world. Despite similarities in species composition, permutational multivariate analysis of variance (PERMANOVA) and multidimensional scaling ordinations (nMDS) suggest biological communities at the two islands are distinct with Baker Island containing a greater percent cover of branched Acroporid corals and turf algae and Howland Island containing a greater percent cover of crustose coralline red algae and small, compact genera of coral. Both islands also contained considerable cover of noninvasive macroalgae. PERMANOVA further revealed benthic and fish species composition to differ between forereef and reef shelf sites from different sides of each island. When islands were considered as a whole, temporal changes were not noted between 2004 and 2006; however, temporal changes at select sites did occur, with coral cover decreasing significantly along the west side of Baker Island from 2004 to 2006.

No abstract