Explaining Spatial Variation in Coral Size Structure in American Samoa.

Abstract

Coral size structure distributions (i.e. the distribution of individual colony sizes within a population) have been shown to vary between and among populations exposed to different environmental regimes and disturbance histories. Subsequently, assessing size structure spatiotemporal variability in relation to biogeophysical factors can provide insight into underlying mechanisms driving spatial patterns observed in coral populations. A total of 22 species of coral are now listed as threatened under the U.S. Endangered Species Act (ESA); however species level data on demographic processes and responses of corals to threats needed for effective management is deficient. This includes: 1) quantitative assessments of population status, and 2) identification of potential environmental drivers that influence the status of a given species. Here, analysis of a rare ESA-listed species, Isopora crateriformis (Isopora spp. used as a proxy), and an abundant species, Montastrea curta, from a 2015 NOAA Ecosystem Sciences Division (ESD) survey across five islands in American Samoa (within 0 – 30m depth range) was used to determine spatial variation in population size structure patterns across two spatial scales (siteand strata-level resolution). Using co-located data available, a range of environmental (e.g. benthic geomorphology), oceanographic (e.g. temperature, wave energy), biological (e.g. benthic cover), and anthropogenic impact covariates were collated and synthesized at comparable spatial scales to the coral population data. Generalized modeling and multi-model inference were used to evaluate the strength and magnitude of the relationship between biogeophysical/anthropogenic covariates (explanatory covariates) and size distribution parameter estimates (response variable) for each coral species. Due to its versatility and effectiveness, the Weibull distribution was used to characterize the observed size spectra and the distribution shape parameter, k, was used as the size spectra response metric in statistical modeling. Analyses reveal that i) size structure spatially varied among and between species and ii) modeled biogeophysical relationships varied significantly between species. Mean net carbonate accretion rate and net carbonate accretion rate variability (i.e. net carbonate accretion rate coefficient of variation), in addition to geomorphological slope and slope variability, accounted for a large proportion of spatial variation in the Isopora spp. site-level size spectra (R2 = 58%). For Isopora spp. strata-level analysis, wave energy and mean accretion rate explained a large proportion of spatial variation (R2 = 46%). In contrast, for Montastrea curta site-level analysis, irradiance (photosynthetically active radiation), percent coral cover, wave energy, mean depth, M. curta juvenile abundance, and SST explained a large proportion of spatial variation in the size spectra (R2 = 45%). For M. curta strata-level analysis, irradiance, mean accretion rate, SST, and wave energy explained a large proportion of spatial variation (R2 = 57%). Our results suggest that the Weibull shape parameter, k, is a reliable metric that captures variability in the coral size distribution and that species-specific biogeophysical factors explain coral size structure variability across space.