Fig 1.
Shell growth patterns and microstructure of Arctica islandica.
(A) Sketch of the left valve of a juvenile specimen. Cutting axis is indicated as orange line. (B) Radial shell section of the valve. The shell is divided into an outer shell layer (OSL; white) and inner shell layer (ISL; gray), separated by the pallial myostracum (blue line). (C) Magnified sketch of the hinge plate showing annual growth lines (dashed) and axis of maximum growth (black line with arrow). (D) Sketch of the microstructures of the ventral shell portion. The outer portion of the outer shell layer (oOSL) consists of homogeneous (HOM) microstructure, which gradually merge into crossed-acicular (CA) microstructure toward the inner portion of the outer shell layer (iOSL). Transitional fine complex crossed-lamellar (FCCL) and CA microstructures are formed in the inner shell layer (ISL). Annual growth lines (GL) and pallial myostracum (My.) consist of irregular simple prismatic (ISP) microstructure. (E) Sketch of the microstructures in the hinge plate. Growth increments and lines are composed of CA and ISP microstructures, respectively. Dashed black lines represent annual growth lines. Boxes in B and C show the extent of the microstructure sketches portrayed in D and E, respectively.
Table 1.
Shells of Arctica islandica used in the present study.
Fig 2.
Image segmentation of hinge plate SEM backscatter images.
(A-C) Morphological analyses were conducted on SEM backscatter images. BMU size and coverage were analyzed in the same image (A,C), taken after oxidation of the organic matrix by immersion in 10.5 vol% H2O2 for 20 minutes (image A magnified for visual clarity). Images for pore morphometry (B), in contrast, were taken in a polished and chemically untreated shell slab. (D-F) Binary images (objects of interest = white; remainder = black) used for the automated image segmentation process. Individual BMUs (D; here shown in various colors to allow discrimination of the individual entities) and pores (E, white) were recognized by the machine learning–based image segmentation process and separated from the remainder of the images (black). (F) For the calculation of BMU coverage a threshold based on the average gray value of the image series was applied. Values above this threshold were assigned to the crystalline (white) phase, those below the threshold to the inter-crystalline phase (black).
Fig 3.
Growth patterns, Mn/Ca values and microstructures in the hinge plate of the studied Arctica islandica.
(A) Mutvei-stained hinge plate of specimen Nioz-TC-12-A2R when viewed under a stereomicroscope with sectoral dark-field illumination. LA-ICPMS analysis was conducted prior to the immersion in Mutvei’s solution. Black dashed lines denote annual growth lines. Red solid lines = disturbance lines. These lines reflect handling stress (collection in the Baltic Sea, transferal to NIOZ, transferal to experimental tanks) and delimit the growth band that formed during the 4-week acclimatization phase. (B+C) Magnifications of (A): light microscopy (B) and SEM (C). (B) Portions formed during the acclimatization phase showed faint microgrowth patterns (the most defined ones indicated in yellow) possibly corresponding to circatidal or circalunidian increments and lines. Portion formed in tank was devoid of annual or disturbance lines. (C) Shell microstructures. Shell portions formed in the Baltic Sea consisted predominantly of crossed-acicular (CA) microstructure and exhibited variable sizes of the BMUs. The beginning of the experimental period is marked by a prominent spherulitic prismatic (SphP) disturbance line (red) followed by a zone of fine complex crossed-lamellar (FCCL) microstructure during the acclimatization phase. After another disturbance line, uniform CA microstructure is visible which formed in the experimental aquaria. (D) Ontogenetic changes of shell Mn/Ca ratios. Gray lines denote the different studied specimens; boxplots show distribution of manganese in shell portions formed in nature and during artificial tank environments.
Fig 4.
Microstructures in the hinge plate of the studied Arctica islandica specimens.
(A) Hinge portions formed in the Baltic Sea predominantly exhibited crossed-acicular (CA) microstructure. Annual growth lines (black line) consisted of irregular simple prisms (ISP), often followed by 10–20 μm-broad fine complex crossed-lamellar (FCCL) fringes, as shown here. Numerous faintly visible disturbance lines (red line) consisted of fibrous prismatic (FP) or spherulitic prismatic (SphP) microstructures. (B) Magnified image of the CA microstructures formed in the Baltic Sea. BMUs consist of a few elongated crystallites, which merged with each other and were loosely aligned in two predominant directions (open angles in direction of growth). (C) Microstructures of shell portions formed in the Baltic Sea (B.S.; leftmost image portion) and the early acclimatization phase (rightmost image portion). A prominent disturbance line (red) composed of SphP microstructure delimits the two shell portions. During the acclimatization phase, FCCL microstructures were produced. (D) Magnified image of the FCCL microstructures formed during the acclimatization phase. Compared to CA microstructures, the FCCL BMUs are even more elongated, strictly aligned in two dip directions and intersect with their neighboring units. (E-F) The CA microstructures formed under controlled temperature in the laboratory exhibited a rather uniform appearance along the growth axis. Dip directions of the needle shaped BMUs varied more strongly than those formed in the Baltic Sea, leading to a looser arrangement and a less organized appearance. All microstructures–except ISP–contained nm- to μm-sized pores (yellow arrows). Black arrows indicate direction of growth.
Fig 5.
Pores of Arctica islandica shells.
(A-C) Pores viewed on an untreated, fractured surface of a hinge plate. Insets in (A) and (C) depict an overview of the fractured shell. Yellow crosses within the insets denote the location of the respective images. (B,C) Higher magnification reveals the presence of shriveled organic envelopes (Org.) surrounding the partially void pores. (D-G) Pores viewed in a polished hinge surface after oxidation of the organic matter by immersion in 10.5% H2O2 for 20 minutes. Direction of growth is to the right. (D) Pores were aligned with their longest axis parallel to the growth front, as shown here near an annual growth line (black dashed line). (E) A pore filled with co-aligned, fibrous particles facing the same direction. (F,G) Pores partially filled with granular, spherical particles and remnants of organics.
Fig 6.
Empirical cumulative distribution functions of (A) BMU size and (B) pore size. The majority of BMUs and pores incorporated in the shells were small, whereas only few large entities existed. Small insets depict magnifications of the smallest 1% of the values and the large insets depict the largest 2% of the values (dashed lines). Ranges of the 15 largest values (solid lines) are shown as horizontal lines. The maximum size of pores and BMUs increased with culturing temperature. Detection limit of the analytical method (2 pixels) is indicated represented by vertical gray line.
Fig 7.
Quantitative microstructural data of hinge plate grown under temperature-controlled conditions.
(A) The size of the 15 largest BMUs of each temperature treatment and (B) the space occupied by crystalline phase (BMU coverage) displayed a statistically significant linear increase with culturing temperature. Note, the relationship between BMU size and temperature was stronger than that of the BMU coverage values (r2 = 0.67 vs. 0.54). Large inter-individual variability was observed among the BMU coverage values of specimen cultured at 3°C. (C) BMU elongation did not exhibit any significant link with culturing temperature. (D) The size of the largest 15 pores increased exponentially with rising temperature. Shaded areas represent ± 2 standard errors of each parameter of the respective models. Model equations and r2 goodness-of-fit values are annotated in the graphs. For justification of the choice of sample size (15 largest values) and type of model see S2 File.