Calcification response of a key phytoplankton family to millennial-scale environmental change

Abstract

Coccolithophores are single-celled photosynthesizing marine algae, responsible for half of the calcification in the surface ocean, and exert a strong influence on the distribution of carbon among global reservoirs, and thus Earth's climate. Calcification in the surface ocean decreases the buffering capacity of seawater for CO₂, whilst photosynthetic carbon fixation has the opposite effect. Experiments in culture have suggested that coccolithophore calcification decreases under high CO₂ concentrations ([CO₂(aq)]) constituting a negative feedback. However, the extent to which these results are representative of natural populations, and of the response over more than a few hundred generations is unclear. Here we describe and apply a novel rationale for size-normalizing the mass of the calcite plates produced by the most abundant family of coccolithophores, the Noëlaerhabdaceae. On average, ancient populations subjected to coupled gradual increases in [CO₂(aq)] and temperature over a few million generations in a natural environment become relatively more highly calcified, implying a positive climatic feedback. We hypothesize that this is the result of selection manifest in natural populations over millennial timescales, so has necessarily eluded laboratory experiments.

Figure 1. Size-normalisation of coccolith mass.

(A) Schematic representation of coccolithophore cell, with variables defined. (B) Regression of molar PIC:POC ratio against volumetric ratios of calcite to organic material (Eq. 1). (C) Regression of coccosphere radius against the square root of coccolith area. The red dashed line represents an independently derived relationship between coccolith size and coccosphere size. (D) Regression of coccosphere thickness against coccolith thickness. (E) Regression of molar PIC:POC ratio against coccolith aspect ratio (; Eq. 3). The dark region around each regression line represents the 1σ confidence interval of the regression, whilst the lighter region with the dashed border represents the 1σ prediction interval of the regression. Error bars on individual points represent the 1σ confidence interval of each measurement. SEM images courtesy of Jeremy Young, used and adapted with permission (http://ina.tmsoc.org/Nannotax3/).

Figure 2. Calcification response of the Noëlaerhabdaceae to environmental changes over two glacial-interglacial cycles.

(A) benthic δ¹⁸O, and 5kyr interval average sea-surface temperature (from Mg/Ca ratios in planktic forams; filled triangles over TII represent an alternative size fraction - see methods) and [CO₂(aq)] (calculated from CO₂ atm and SST) at ODP site 1123 in the southern Pacific Ocean. (B–D) Coccolith morphometrics: (B) Mass, (C) Area and (D) Aspect ratio. Raw data are displayed as frequency-density contour plots. Emiliania huxley-affiliated morphotaxa (MG_Emi) are hollow squares, Gephyrocapsa spp.-affiliated morphotaxa (MG_Geo) are hollow diamonds and filled circles (population mean) represent the mean of the Noëlaerhabdaceae-affiliated morphotaxa in each sample. Arithmetic coefficient of variation (CV) is a measure of the mean-independent variance of log-normal data, and is shown in grey. Map was made using GMT 5.2.1 (http://gmt.soest.hawaii.edu/).

Figure 3. Fractional change in PIC:POC of natural heterogeneous populations over glacial terminations (left), compared with that of monoclonal strains subject to an equivalent CO₂ change in culture (right).

(A) The response of MG_Emi in nature is within the range of uncertainty and of plasticity of E. huxleyi observed in the laboratory. (B,C) The PIC:POC of MG_Geo in nature increases across both terminations, which is opposite to the equivalent plastic response as inferred from culture manipulations in G. oceanica. Error bars represent the 1σ uncertainty, which for the down-core response, is calculated using the prediction interval of Eq. 3. For culture results, details of the regressions are given in the Supplementary material.

Figure 4. Instantaneous effect of biomass production on seawater carbonate chemistry as a function of PIC:POC (ρ).

Solid line isocontours represent [CO₂] (μmol/kg), and dashed isocontours represent pH. Solid arrows represent the instantaneous effect of biogenic matter formation, and shaded arrows CO₂ exchange with an atmospheric carbon pool that is large relative to the perturbed sample of seawater. Depending on its ρ, biogenic material may form an instantaneous sink or source of CO₂. For conditions typical of the modern ocean, when ρ is <or > 1.42, pH initially increases or decreases respectively, and when ρ is < or > 1.86 [CO₂] initially decreases or increases respectively. These critical values depend on the carbonate chemistry of the surface ocean.