Seawater Chemistry and Early Carbonate Biomineralization

See allHide authors and affiliations

Science  01 Jun 2007:
Vol. 316, Issue 5829, pp. 1302
DOI: 10.1126/science.1137284


The first appearances of aragonite and calcite skeletons in 18 animal clades that independently evolved mineralization during the late Ediacaran through the Ordovician (~550 to 444 million years ago) correspond to intervals when seawater chemistry favored aragonite and calcite precipitation, respectively. Skeletal mineralogies rarely changed once skeletons evolved, despite subsequent changes in seawater chemistry. Thus, the selection of carbonate skeletal minerals appears to have been dictated by seawater chemistry at the time a clade first acquired its mineralized skeleton.

Organisms use a wide variety of minerals to make their skeletons, including silica, apatite, and several polymorphs of carbonate, in particular aragonite and calcite. It is unclear, however, why different taxa evolved to use one mineral rather than another. Lineages rarely switched their mineralogy after acquiring skeletons, suggesting that, for most taxa, ambient seawater chemistry does not strongly influence skeletal mineralogy (1). However, seawater chemistry may have dictated the choice of skeletal mineralogy at the time skeletons first evolved in a clade (1, 2).

This hypothesis was tested by assessing the ancestral mineralogy (aragonite versus calcite) of 21 metazoan taxa that evolved mineralization during the late Ediacaran through the Ordovician (∼550 to 444 million years ago) and comparing the resulting pattern with independent constraints on coeval seawater chemistry (Fig. 1 and tables S1 and S2). Each taxon is thought to represent an independent acquisition of biomineralization (3) and therefore should offer an independent test of the hypothesis. Alternative taxonomic interpretations (table S1) affect the degree of support for the pattern presented here but do not change it. Four taxa known to be polyphyletic were excluded (table S1).

Fig. 1.

First appearances of carbonate skeletons in animals (A) and constraints on seawater chemistry (B) during the Ediacaran through the Ordovician Periods. Fluid inclusion data are listed in table S2. Indirect evidence for seawater chemistry, reviewed in (2), refers to constraints from nonskeletal marine precipitates and from models of mid-ocean ridge hydrothermal brine fluxes. Because most taxa are fossilized in abundance, it is assumed that first appearances of skeletons in the fossil record closely approximate actual first appearances of skeletons. “Calcite sea” and “aragonite sea” refer to seawater that favors calcite and aragonite precipitation, respectively. Boundary age constraints are from the International Commission on Stratigraphy International Stratigraphic Chart (2007) ( and (11). Although there is uncertainty regarding the position of stage boundaries relative to some first appearances of skeletons, the sequence of taxa is unaffected (table S2). Ma, million years; T, Tommotian Stage; A, Atdabanian Stage; Bot/Toy, Botomian and Toyonian Stages.

Determining the primary mineralogy of ancient carbonate fossils is not straightforward because aragonite recrystallizes to calcite at Earth's surface pressures. As a result, formerly aragonitic skeletons are often preserved as calcite. Five well-accepted criteria were used to infer primary mineralogy [Supporting Online Material (SOM) text]: (i) presence of original aragonite, (ii) phylogenetic distribution of mineralogies in extant members of a group, (iii) quality of preservation of original microstructures in calcite, (iv) morphology of crystals replicated by secondary minerals, and (v) magnesium content. Evidence for primary mineralogy was drawn from the literature (table S1) and critically evaluated with respect to these criteria.

For 18 of the 21 taxa, primary mineralogy is known with at least some confidence. Of these, eight are highly likely (N = 4) or likely (N = 4) to have been aragonitic, and 10 are highly likely to have been calcitic (either high- or low-magnesium calcite). Mapping inferred skeletal mineralogies on first appearances of mineralization (FAMs) results in a striking pattern (Fig. 1A): Aragonitic taxa systematically appeared earlier, with FAMs occuring in the Ediacaran Period from Nemakit-Daldynian Stage followed by calcitic taxa, which began mineralizing in the Tommotian Stage or later.

Fluid inclusions in marine evaporite deposits provide the most definitive constraints on the Mg2+/Ca2+ ratio of ancient seawater (4), which is thought to be the primary control on marine nonskeletal carbonate mineralogy (2, 5). These indicate that the Mg2+/Ca2+ ratio of late Ediacaran seawater favored aragonite precipitation (4, 6), whereas the Mg2+/Ca2+ ratio of Tommotian/Atdabanian through Toyonian seawater (4, 6, 7), as well as that of late Ordovician seawater (4, 8), favored calcite precipitation (Fig. 1B and table S2).

The close correspondence between the pattern of FAMs and the constraints on seawater chemistry (Fig. 1) suggests that selection of aragonite versus calcite was largely dictated by seawater chemistry at the time mineralized skeletons were first acquired in a clade. Mineralization of most animal skeletons is biologically controlled, occurring in an environment isolated from seawater (9). As a result, seawater chemistry does not have a direct influence on the mineralogy of most animal skeletons the way it does for biomineralizers that induce mineralization directly from seawater (9). However, seawater chemistry may have played an indirect role in determining skeletal mineralogy by affecting the physiological costs of biomineralization (1). Experimental and paleontological evidence suggests organisms have greater difficulty producing skeletons not favored by seawater chemistry (2, 10); thus, it is reasonable to posit that, when skeletons first evolved, natural selection favored the mineral easiest to precipitate (1). Once selected, however, taxa rarely switched mineralogies even though they may have been negatively affected when Mg2+/Ca2+ ratios changed (2). Although initially optimal, skeletal mineralogy may since have presented a considerable evolutionary constraint.

Supporting Online Material

SOM Text

Tables S1 and S2



View Abstract

Navigate This Article