Wet shells and dry tales: the evolutionary 'Just-So' stories behind the structure-function of biominerals

Abstract

The ability of evolution to shape organic form involves the interactions of multiple systems of constraints, including fabrication, phylogeny and function. The tendency to place function above everything else has characterized some of the historical biological literature as a series of 'Just-So' stories that provided untested explanations for individual features of an organism. A similar tendency occurs in biomaterials research, where features for which a mechanical function can be postulated are treated as an adaptation. Moreover, functional adaptation of an entire structure is often discussed based on the local characterization of specimens kept in conditions that are far from those in which they evolved. In this work, environmental- and frequency-dependent mechanical characterization of the shells of two cephalopods, Nautilus pompilius and Argonauta argo, is used to demonstrate the importance of multi-scale environmentally controlled characterization of biogenic materials. We uncover two mechanistically independent strategies to achieve deformable, stiff, strong and tough highly mineralized structures. These results are then used to critique interpretations of adaptation in the literature. By integrating the hierarchical nature of biological structures and the environment in which they exist, biomaterials testing can be a powerful tool for generating functional hypotheses that should be informed by how these structures are fabricated and their evolutionary history.

Keywords: adaptation; biominerals; mechanical properties; structure–function.

Figure 1.

Overview of the two animal shells used in this study: the aragonitic shell of Nautilus pompilius (a–d) and the calcitic shell of Argonauta argo (e–h). The shell of N. pompilius (a) is composed of three primary layers (b): the outer prismatic layer that transitions into the nacre layer (c), which then transitions into the inner prismatic layer. EBSD of N. pompilius shows a clear increase in texture going from the homogeneous zone (top of the image) down to the columnar and finally the nacreous zones (d). The shell of A. argo (e), in contrast with N. pompilius, grows bi-directionally from a central organic layer (f). Most of the thickness of the shell is formed by acicular calcite crystals that grow in conical clusters. These clusters begin as spherulites in the organic layer (g). The conical crystal clusters that make up the shell of A. argo are visible within the EBSD map and show a co-orientation within the clusters, the blue/green clusters near the image centre (h). Much of the variation in orientation seen in the image is due to neighbouring clusters going into and out of the plane. The colour-coded inverse pole figures have their reference direction normal to the image plane.

Figure 2.

Mechanical characterization on the nano-scale. Reduced modulus (Er) and indentation hardness graphs were calculated from two indentation maps from Nautilus pompilius (a,b), covering an area of 1300 × 200 µm² and Argonauta argo (d,e) covering an area of 96 × 36 µm². The mean values presented in the graphs were made by averaging indentation results across a row of indents made at the same height. Shaded regions represent ±1 s.d. of the averaged data. NanoDMA experiments performed on a cross-section of the shell of N. pompilius (c) and A. argo (f) at a relative humidity of 90%.

Figure 3.

Mechanical characterization on the macro-scale. Storage modulus, loss modulus and loss factor versus frequency graphs for Nautilus pompilius (a–c) and Argonauta argo (d–f), respectively. Inverted triangles, diamonds and circles represent data obtained from N. pompilius nacre only, complete N. pompilius shell and complete A. argo shell, respectively. For the data of N. pompilius, the shaded regions represent ±1 s.d. of the averaged data. The error bars for the A. argo plots, while also calculated as ±1 s.d., show a greater spread compared to N. pompilius due to the geometric variation of the beams cut from the shell. In this regard, most of the ‘error’ for A. argo is due to geometric differences between the two sides of the same beam that were averaged together for each point.