Gains and losses of coral skeletal porosity changes with ocean acidification acclimation

Abstract

Ocean acidification is predicted to impact ecosystems reliant on calcifying organisms, potentially reducing the socioeconomic benefits these habitats provide. Here we investigate the acclimation potential of stony corals living along a pH gradient caused by a Mediterranean CO2 vent that serves as a natural long-term experimental setting. We show that in response to reduced skeletal mineralization at lower pH, corals increase their skeletal macroporosity (features >10 μm) in order to maintain constant linear extension rate, an important criterion for reproductive output. At the nanoscale, the coral skeleton's structural features are not altered. However, higher skeletal porosity, and reduced bulk density and stiffness may contribute to reduce population density and increase damage susceptibility under low pH conditions. Based on these observations, the almost universally employed measure of coral biomineralization, the rate of linear extension, might not be a reliable metric for assessing coral health and resilience in a warming and acidifying ocean.

Figure 1. Study location.

Located off the southwestern coast of Italy (a), near Panarea Island (b), there are underwater volcanic vents releasing persistent gaseous emissions (98–99% CO₂ without instrumentally detectable toxic compounds), resulting in a stable pH gradient. Four sites at various distances from the primary vent were initially selected for study. No temperature difference exists among the four study sites throughout the year. (c) Ranges of measured pH_TS (number of observations [n]=103–110 per site) and Ω_arag (n=96–104 per site) values at the four sites, showing consistent increases in both pH_TS and Ω_arag from the vent crater to its periphery. (d) Bathymetric profile of the sites with associated mean pH_TS and Ω_arag values. No corals were found at Site 4, characterized by the lowest pH (mean pH_TS 7.4). Living specimens of Balanophyllia europaea, photographed at night with expanded tentacles (e) and during the day with contracted tentacles (f); marker 5 mm.

Figure 2. Skeletal morphology of Balanophyllia europaea growing under different pH conditions from the macroscale to the nanoscale.

Each row in the figure corresponds to a different study site and sample age is 9–11 years. Images are representative of all observed skeletons. (a,e,i) Low magnification SEM images of coral skeletons, marker 5 mm. (b,f,j) Internal sections of corallites from μCT images, marker 5 mm. (c,g,k) SEM images of entire skeletal fibres from fractured septae, marker 10 μm. (d,h,l) AFM images of mineral grains on the skeletal fibre surfaces, marker 50 nm.

Figure 3. Scatterplots of skeletal parameters, and correlation analysis between porosity and net calcification rate.

Site 1=blue, Site 2=green, Site 3=red. Straight lines represent the best-fit linear regression (mean, solid line), 25% quantile and 75% quantile (dashed lines). (a–e) Skeletal parameters (y-axes) plotted against pH_TS. (f) Scatterplot of porosity (P_A) versus net calcification rate in corals from Sites 1 to 3. For a–d,f N=44; for e n=122. ***P<0.001; **P<0.01; *P<0.05, robust t-statistics test.

Figure 4. Summary of responses in Balanophyllia europaea skeletal parameters as a function of pH from the ocean to the nanoscale.

The significant decline in population density with pH was measured in previously published research. Net calcification rate and bulk density decrease with decreasing pH, whereas porosity (P_A) increases, preserving the linear extension rate and corallite shape (biometry and interseptal volume fraction as measured by μCT do not correlate with pH). The increase in porosity is associated with a decrease in skeletal stiffness. At the nanoscale, the ‘building blocks' (the fundamental structural components of the coral skeleton, aragonite fibre bundles and their constituent mineral grains) produced by the biomineralization process are substantially unaffected by increased acidity. Green boxes denote parameters found to have a direct relationship with pH, blue boxes denote parameters that have an inverse relationship and red boxes denote parameters found to have no relationship with pH. The significances of the regression of each parameter (dependent variable) with pH (independent variable) are indicated; ***P<0.001; **P<0.01; NS indicates no significance. Micro and nanoscale structure observations by SEM and AFM represent qualitative data.