SI-6-revised in proof

Supporting information

for

Nacre tablet thickness records formation temperature

in modern

and fossil shells

Pupa U.P.A Gilbert

1,2*

, Kristin D. Bergmann

3,4

, Corinne E. Myers

5,6

, Matthew A. Marcus

, Ross

T. DeVol

, Chang-Yu Sun

, Adam Z. Blonsky

, Erik

Tamre

2,3

, Jessica Zhao

, Elizabeth A.

Karan

, Nobumichi Tamura

, Sarah Lemer

, Anthony J. Giuffre

, Gonzalo Giribet

, John M.

Eiler

, Andrew H. Knoll

3,5

University of Wisconsin

–Madison, Departments of Physics, Chemistry, Geoscience,

Madison WI 53706 USA

Harvard University, Radcliffe Institute for Advanced Study, Fellowship Program,

Cambridge, MA 02138.

Harvard University, Department of Earth and Planetary Sciences, Cambridge, MA 02138.

Massachusetts Institute of Technology, Department of Earth, Atmospheric and Planetary

Sciences, Cambridge, MA 02139.

Harvard University, Museum of Comparative Zoology and Department of Organismic and

Evolutionary Biology, Cambridge, MA 02138.

Universi

ty of New Mexico, Department of Earth and Planetary Sciences, Albuquerque, NM

87131.

Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, CA, 94720,

USA.

California Institute of Technology, Division of Geological and Planetary Scienc

es, Pasadena,

CA 91125

corresponding author, pupa@physics.wisc.edu, previously publishing as Gelsomina De

Stasio

Pinnidae: an ideal family for this study

Specimens within the family Pinnidae, commonly called pen shells or pen clams,

were chosen for this study for several reasons. Pinnids are not uncommon in both the

modern and fossil records (Figure S1

). Thus, modern specimens enabled ground-truthing

of the nacre TT-based paleothermometer, and fossil specimens provided a unique window

into paleoclimate over 180 Ma of Earth history. Pen shells are large, fast growing bivalves

with a thick nacreous region ideal for both spectral and isotopic analyses. Furthermore, the

nacre layer in pinnids is eas

ily separated from the prismatic calcite layer. This facilitates

sample preparation and minimizes contamination of aragonite nacre with calcite prisms

for clumped isotope analysis. Finally, pinnids produce each shell transect in a relatively

short period o

f time, less than a year. Thus comparison with modern shells, grown at

known T, makes it possible to identify which shells were primarily deposited in cooler vs.

warmer seasons. Anecdote has it that Japanese pearl farmers harvest their pearls in winter,

the tablet thickness is lower, resulting in greater pearl luster

(Strack, 2001)

. This trend

in the Japanese pearl bivalve

Pinctada fucata

is similar to that observed in this work for

Pinnidae: greater TT at higher T and

vice versa

Seven fossil specimens were analyzed for both TT and T, drawn from three locations

reflecting warm geologic times and places (the Late Cretaceous Gulf Coastal Plain, ~ 65.5-

66 Ma

(Landman et al., 2004; Larina, 2015; Thibault and Gardin, 2006, 2007)

; the Early

Eocene Gulf Coastal Plain, ~ 52-54 Ma

(Agnini et al., 2007; Frederiksen et al., 1982; Sessa et

al., 2012)

; and the Middle Miocene Mid

-Atlantic Calvert Cliffs, ~ 12.7-13.2 Ma) (Kidwell,

1997). In addition, two Jurassic

Pinna

shells were analyzed for TT only, and T was deduced

from TT. Many more details are provided in the Pinnidae biology section below.

Pinnidae Biology

The family Pinnidae Leach, 1819 includes subtidal and coastal species (Dance, 2013)

found in tropical and temperate regions, both today and in the fossil record

. Pinnid

species

are important elements of certain marine ecosystems, including the sandy substrate of

seagrass beds, lagoons and coral reefs where they can sometimes aggregate in large

densities

(Rosewater, 1961)

. Pinnids are

sessile, suspension-feeding bivalves

found

partially buried

with their anterior end in sand or mud, such that only the posterior

prismatic portion of the shell is visible above the sediment (Aucoin and Himmelman,

2011b; Schultz and Huber, 2013; Turner a

nd Rosewater, 1958; Yonge, 1953

). The family

contains two extant genera and 55 accepted species (although there are more than 16

additional genera in the fossil record)

(Lemer et al., 2014; Rosewater, 1961; Schultz and

Huber, 2013; Turner and Rosewater, 1958). Pinnids are un

ited by an elongated subtrigonal

shell shape, heteromyarian adductor muscle scars (small in the anterior and large in the

posterior), a toothless hinge with primary and secondary ligament segments, thin to absent

periostracum, byssus, and generally large s

ize (commonly between 15-35cm; Modern

Pinna nobilis

may reach lengths over 1m (Richardson et al., 2004; Schultz and Huber, 2013;

Turner and Rosewater, 1958). Pinnidae is nested within the Order Pterioida

Pteriomorphia. However the exact position of Pinnidae and

the identity of its s

ister group

remains unclear. Earlier studies based on molecular data have suggested

a sister group

relationship to Pterioidea,

Ostreoidea,

or Mytiloidea

(Adamkewicz et al., 1997; Bieler et al.,

2014; Campbell, 2000; Giribet and Distel, 2003; Giribet and Wheeler, 2002; Matsumoto,

2003; Steiner and Hammer, 2000).

The biomineralized shells of the Pinnidae are composed of an outer calcitic

prismatic layer and an inner aragonitic nacreous layer. In pen shells the inner nacreous

region does not extend to the posterior margin of the prismatic outer shell, and instead

extends from the anterior margin approximately 1/3 to 2/3

the length of the shell (

Figure

4). The genera

Pinna

Linnaeus, 1758

and

Atrina

Gray, 1842 may be distinguished by the

shape and size of their nacreous layer. In species of the genus

Pinna

the internal nacreous

layer is divided by a sulcus into a dorsal and a ventral lobe. The position of the posterior

adductor muscle scar with respect to the margins of the nacreous layer is one of the major

taxonomic characters used to distinguish species. The posterior adductor muscle scars

(PAs in Figure 4) in

Pinna

are fully enclosed within the nacreous region (in the dorsal lobe),

whereas in

Atrina

PAs extend to the edge of the nacreous layer, and in some species even

beyond the margin into the prismatic portion of the shell (Rosewater, 1982; Schultz and

Huber, 2013; Turner a

nd Rosewater, 1958; Yonge, 1953

). Finally, in species of the

subgenus

Streptopinna

Martens, 1880 (considered a third extant genus

until Lemer et al.

2014

(Lemer et al., 2014)) the internal nacreous layer is reduced to a dorsal lobe.

The

present work focused on species within the extant genera

Atrina

and

Pinna

. These two

genera are common in the fossil record and thus provide the most robust history of

environmental conditions through time.

Very little is known about the reproductive strategies and the pelagic larval

duration of most pinnid species, except for some commercially important ones. Most

species are believed to be gonoc

horistic, to reproduce annually and to produce larvae with

a planktotrophic stage with trochophore and veliger stages like

Pinna atropurpurea

Atrina

pectinata

and

Atrina maura

(Beer and Southgate, 2006; Mendo et al., 2011). Because of

their potentially long larval pelagic phase, the dispersal capacity of pinnids

is expected to

be extensive

. This results in a cosmopolitan generic and family distribution, while habitat

specificity and past geographic isolation maintained a degree of species

-level endemism

(e.g.

P. nobilis

in the Mediterranean Sea). Once the veliger settles, the muscular foot is used

to bury in soft sediments. Anterior byssus threads then anchor the shell to hard substrate

or cobbles in the sediment, such as sea grass beds or coral boulders (Aucoin and

Himmelman, 2011a, b; Grave, 1911; Richardson et al., 2004

; Rosewater, 1982; Schultz and

Huber, 2013;

Turner and

Rosewater, 1958; Yonge, 1953

). Consequently, most species of

Pinnidae live within the subtidal photic zone (up to ~ 100

m)(Grave, 1911; Richardson et

al., 1999; Schultz and Huber, 2013)

. Some exceptions exist, e.g.,

P. carnea

and other species

found in the Pacific Ocean can be found intertidally, and a few species are only found in

deep water, down to 600m depth (Schul

tz and Huber, 2013; Yonge, 1953

). Pen shells are

fairly delicate, and may be damaged by wave action, currents, storms, or predation (Allen,

2011; Aucoin and Himmelman, 2011b; Grave, 1911; Rosewater, 1982; Turner and

Rosewater, 1958; Yonge, 1953

). Repair of the prismatic portion of the shell is quite efficient.

For instance modern

A. rigida

may repair up to 13mm/day (Grave, 1911), and fully

exhumed shells may re-bury

themselves. However, successful re-burial is not guaranteed,

and exhumation often results in death (Grave, 1911; Richa

rdson et al., 1999; Yonge, 1953

Growth Rates

.—Species of pen shells are some of the fastest growing bivalves known

and may grow upwards of 20cm or more radially in their first year of life (Aucoin and

Himmelman, 2011b; Butler and Brewster, 1979; Cendejas et al., 1985; Richardson et al.

1999; Richardson et al., 2004

). As in most mollusks, faster radial growth occurs early in life;

for pen shells this is generally in the first three years

(Butler and Brewster, 1979; Schöne,

2008). Individuals reach reproductive maturity in appr

oximately the second year, and have

been known to live up to 20 years in the wild

(Aucoin and Himmelman, 2011b; Hendriks et

al., 2012; Kožul et al., 2011; Richardson et al., 1999)

Slowdown and even stoppage in radial growth with age has been observed in many

mollusk species (e.g. (Schöne, 2008)

; for species of

Pinna

(Grave, 1911; Hendriks et al.,

2012)). Furthermore, both biotic and abiotic environmental factors appear to influence

rates of shell growth and stoppages. In addition to age, biotic factors affecting mollusk

radial growth rates include: ava

ilability of nutrients (food), predation pressure, production

of gametes, and spawning (Grave, 1911; Ivany, 2012; Joubert et al., 2014; Linard et al.,

2011; Lowenstam and Weiner, 1989; Schöne, 2008). Abiotic factors include: temperature,

salinity, lunar cycles and tides, seasonal variations (e.g

., day length, storms), turbidity,

ocean circulation patterns, sea level, and carbon (organic and inorganic), phosphate, and

nitrate concentrations in the water column (Ivany, 2012; Joubert et al., 2014; Linard et al.,

2011; Lowenstam and Weiner, 1989; Richardson et al., 1999; Schöne, 2008). Changes in

growth rates are commonly detected in bivalves by the deposition of a growth line in the

prismatic outer shell (Schöne, 2008)

. These growth lines, unfortunately, are not produced

in species of Pinnidae; however, the posterior adductor muscle scars (PAs in Figure 4) can

be used to identify the age of in

dividuals (Butler and Brewster, 1979; Richardson et al.

1999; Richa

rdson et al., 2004

). Richardson et al.

(Richardson et al., 1999) have documented

that PAs from the first year of growth are often absent, thus the age of

an individual may be

estimated as 1 + number of observed PAs.

Few previous studies have measured slowdowns or stoppages specifically in rates

of deposition of the nacreous layer, and none of these were conducted on pen shells. We do

not observe any disco

ntinuity in nacre formation for the specimens of

Pinna

and

Atrina

analyzed here (

Figure 6), however this does not preclude faster growth seasonally. The

most commonly investigated species are those of economic value in pearl aquaculture or

human consumption, such as the pearl oyster (

Pinctada margaritifera

) or abalone (e.g.,

Haliotis refuscens

). In these few studies, both temperature and nutrient concentrations

were observed to impact rates of nacre deposition and total nacre thickness in tank

experiments of

P. margaritifera

(Joubert et al., 2014; Linard et al., 2011)

. Additionally,

different suites of genes have been identified in

P. margaritifera

that produce different

regulatory proteins for prismatic (radial) vs. nacreous growth (Joubert et al., 2014; Marie et

al., 2012)

. It is very plausible that pen shells, which are closely related to the pearl oysters,

share this same genetic framework. Thus, rate of nacre deposition and TT in pen shells

may be influenced by abiotic and biotic environmental factors differently than prismatic

radial growth, or even continue at a regular rate throughout the life of these species, as the

shells of Pinnidae and

Pinctada

thicken significantly after the phase of rapid radial growth

(~3 years).

Fossil Record

.—The Pinnidae are observed in t

he fossil record as far back as the

Silurian Period (~ 444 Ma)

(En

-Zhi et al., 1986; Zhang, 1988)

. A PaleoBiology DataBase

(PBDB;

paleobiodb.org/

) search on 5/25/15 recovered 926 fossil occurrences of

Pinna

and

176 fossil occurrences of

Atrina

, with global distribution (see maps from time periods of

interest in Figure S1

Pinna

first occurs in Middle Mississippian sediments (age range: 345

Ma – present)(Cash, 1882; Wheelton, 1905)

;

Atrina

is also first observed in the

Carboniferous (age range: ~ 359 Ma – present) (Rosewater, 1961; Ruedemann, 1916,

1918). Although their delicate shells result in difficulty preserving whole body fossils,

fragments and partial specimens are not uncommon in warm water, near

-shore to offshore

paleo

-sediments, where their unique morphology makes for easy identification. Further,

when preserved in situ, individuals of Pinnidae are

often quite abundant in “thickets”, akin

to modern Mytilidae

(Butler and Brewster, 1979; Idris et al., 2008)

DETAILED

METHODS

Sample acquisition

We apply a specimen naming convention where the first letter designates the genus,

second letter designates the species, and the number indicates the order in which the

specimen was received and analyzed. Each sample is further identified by a hyphen and

either a “1” (fragment analyzed using PEEM) or “2” (fragment analyzed using clumped

isotopes), or greater numbers for additional samples. For example, the

Atrina rigida

specimen was received 5

, thus the PEEM-analyzed sample is labeled: Ar5-1, and the

clumped isotope-analyzed sample is labeled Ar5-2. In all figures in this work, however, the

last hyphen and number are omitted for simplicity. Associated specimen information is

archived in a nacre sample compendium on:

http://home.physics.wisc.edu/gilbert/nacre/sample_compendium.html. Table S1

contains

a summary of specimen information and Figures 8, 9, S11 provide sample collection

location and temperature data measured by nearby buoys or weather stations from the

modern samples over the lifetime of the organisms (Pc2, Ar5, and Pn1).

Ar5:

Modern

Atrina rigida

(size: 29 cm along the vertical in Figure 4, S4). Purchased from

Gulf Specimen Marine Laboratory, Panacea, FL, USA. The living animal was collected at the

beginning of September 2014 from St. Joseph Bay, Gulf county, FL (29° 43' 15" N / 85° 19'

39" W) from a collection depth of between 0.5 and 2.0m, kept in a tank for 3-4 weeks,

shipped live to Madison,

WI and sacrificed on Sept. 30th, 2014. The specimen grew in

water temperatures ranging between ~10°C and ~32°C (buoy T data in Figure S11

). PEEM

sample Ar5-1 was cut from the right valve (RV) of shell Ar5, as shown in Figure S4

Clumped isotope analysis was performed on the left valve (LV) and named Ar_5_a,

Ar_5_a_6_10, Ar_5_c. The clumped isotope data from Ar5 were not included because there

was major contamination during the analysis, and the results are off by 1000°C.

Ar3

: Modern

Atrina rigida

(size: 21 cm). Purchased the same day as Ar5 from Gulf

Specimen Marine Laboratory, Panacea, FL, USA. In all other respects identical to Ar5 (see

Figure S11

for location and buoy T data). The LV was analyzed here only using SEM in

Figure 5, and is therefor

e not included in

Table S1

, or in the sample counting.

Pc2:

Modern

Pinna carnea

(size: 22.5 cm). Specimen from the Malacology collection of the

Museum of Comparative Zoology (MCZ), Harvard University, Cambridge, MA, USA. Catalog

number MCZ 382622: collecte

d by Sarah Lemer on March 15, 2015 in Boca del Drago

(9°24'17" N / 82°19'26" W), Isla Colón, Archipelago of Bocas del Toro, Panama (Figure 8),

at a depth of 1

-1.5m. Before sacrificing it, the specimen was kept for six days in an outdoor,

shaded tank at the

Smithsonian Tropical Research Institute (STRI) Marine station under

constant flow of seawater from the coastline at its natural temperature. The specimen grew

in water temperatures ranging from ~26°C to ~31°C

(Figure 8). The LV was used for both

PEEM (Pc2-1) and clumpled isotope (Pc2-2) analyses.

Pn1:

Modern

Pinna nobilis

(size: 28 cm). Also from the Malacology collection of the MCZ,

Harvard University, Cambridge, MA, USA. Catalog number MCZ 371544: collected by Juan

Giribet in Mallorca, Spain, 1991 (Figur

e 9). The temperature record from 1990-1991 is

unavailable. Temperature records in Figure 9 show the locations where the buoy T data

were collected in recent years. The specimen likely grew in water temperatures ranging

from ~10°C to ~30°C

(Figure 9). Th

e LV was used for both PEEM (Pn1

-1) and clumped

isotopes (Pn1-2) analyses.

Ah2:

Miocene

Atrina harrisii

(fragment size: 7 cm). Courtesy of Robert Hazen and John

Nance, Calvert Marine Museum Invertebrates Collection, Solomons

, MD USA. Catalogue

number CMM-I-237: collected from Bed 19, Boston Cliffs Member, Choptank Formation,

Chesapeake Group (Middle Miocene Epoch ~12 Ma). The sample appearance and

preparation are shown in

Figure S7. Fragment size prohibited identification of RV or LV for

PEEM (Ah2-1) and clumped isotope (Ah2-2) analyses but both analyses were conducted on

the same fragment.

Ah3:

Miocene

Atrina harrisii

(fragment size: 4.5 cm). Courtesy of Susan Butts and Jessica

Utrup, Yale Peabody Museum (YPM), Yale University, New Haven, CT, USA. Catalogue

number IP 527493: collected by S. M. Kidwell in Saint Marys Co, MD in 1979. Specimen

collected from Unit 5 of the Drumcliff Member, Choptank Formation (Middle Miocene ~13

Ma). For more information, see: http://peabody.yale.edu/collections/search

-collections?ip

A single fragment (RV/LV unknown)

was used for PEEM (Ah3-1) and clumped isotope

(Ah3-2) analyses. Species designation determined by the authors

and confirmed by John

Nance (Calvert Marine Museum of Invertebrates) based on absence of interior sulcus or

exterior keel, and comparison to other age

-equivalent specimens previously described

from the Drumcliffs Member (Glenn, 1904; Kidwell, 1982).

Ah4

: Miocene

Atrina harrisii

(fragment size: 5.5 cm). Courtesy of Susan Butts and Jessica

Utrup, YPM, Yale University, New Haven, CT, USA. Catalogue number IP 527512: collected

by S. M. Kidwell in Saint Marys Co, MD in 1978. Specimen collected from the top of Unit 1,

Drumcliff Shell Bed, Choptank Formation (Middle Miocene ~13 Ma). For more information,

see:

http://peabody.yale.edu/collections/search

-collections?ip

. A single fragment

(RV/LV?) was used for PEEM (Ah4-1) and clumped isotope (Ah4-2) analyses. Species

designation determined by the authors and confirmed by John Nance (Calvert Marine

Museum of Invertebrates) based on absence of interior sulcus or exterior keel, and

comparison to other age-equivalent specimens previously described from the Drumcliffs

Member (Glenn, 1904; Kidwell, 1982).

Px1:

Eocene cf.

Pinna

sp., (fragment size: 3.5 cm). Courtesy of Susan Butts and Jessica Utrup,

YPM, Yale University, New Haven, CT, USA. Catalogue number IP 527489: collected by C. O.

Dunbar on Feb. 4, 1966 in Butler Co, AL (Coll. 3). Specimen collected from the Bashi Shell

Marl, lower Hatchetigbee Formation (Early Eocene ~54 Ma). For more information, see:

http:

//peabody.yale.edu/collections/search

-collections?ip

. A single fragment (RV/LV?)

was used for PEEM (Px1

-1) and clumped isotope (Px1-2) analyses.

Px2:

Eocene cf.

Pinna

sp. (fragment size: 6 cm). Courtesy of Susan Butts and Jessica Utrup,

YPM, Yale University, New Haven, CT, USA. Catalogue number IP 527490: collected by C. O.

Dunbar on Feb. 4, 1966 in Butler Co, AL (Coll. 3). Specimen collected from the Bashi Shell

Marl, lower Hatchetigbee Formation (Early Eocene ~54 Ma). For more information, see:

http://peabody.yale.edu/collections/search

-collections?ip

. A single fragment (RV/LV?)

was used for PEEM (Px2

-1) and clumped isotope (Px2-2) analyses.

Ps5:

Late Cretaceous

(Maastrichtian)

Pinna

sp. (fragment size: 8 cm). Courtesy of Neil

Landman and Bushra M. Hussaini, American Museum of Natural History (AMNH), New

York, NY, USA. Catalogue number 99982: collected by Neil Landman, Susan Klofak, Matt

Garb, Remy Rovelli, and C

orinne Myers on May 28, 2010 in Tippah, Co, MS. Specimen

collected from the Owl Creek Fm, Selma Group (Maastrichtian ~66 Ma). Accessioned

specimen consists of many shell fragments; two separate fragments were used for Ps5 and

Ps6 below. A single fragment (RV/LV?) was used for PEEM (Ps5-1), clumped isotope (Ps5-

2a, Ps5-2c), and EPMA (Ps5-4) analyses. By comparison with

Pinna laqueata

specimens

found at this site at the same time and those previously observed in the Owl Creek of

Missouri

(Stephenson, 1957)

, it is possible that this shell species was

Pinna laqueata

Ps6:

Late Cretaceous (Maastrichtian)

Pinna

sp. (fragment size: 4 cm). Courtesy of Neil

Landman and Bushra M. Hussaini, AMNH, New York, NY, USA. Specimens from same

catalogue number, locality, and stratigraphic interval as Ps5 above. A single fragment

(RV/LV?) was used for PEEM (Ps6

-1) and clumped isotope (Ps6-2a, Ps6-2c) analyses. As for

Ps5, it is possible that Ps6 is

Pinna laqueata

, as other

P. laqueata

specimens were found at

this site

(Stephenson,

1957).

Pfo1:

Early Jurassic (Pliensbachian)

Pinna folium

(remaining shell size: 12 cm, estimated

total size including missing umbo: 15 cm), from Blockley, Gloucestershire, UK, extracted

from the Lower Lias. Courtesy of Steven Davies, Dinosaurland Fossil Museum collection,

Lyme Regis, Dorset, United Kingdom. The shell shows excellent iridescence, and does not

have any calcite prismatic layer in the areas analyzed with SEM. It does not have the umbo

anymore. The two valves are closed and filled with calcite, identified by PEEM analysis. A

1cm thick slice was embedded and cut as close to the umbo as possible (~3 cm from it),

embedded, polished, and analyzed with SEM. The best region (position 3) was cut, re

embedded and polished for PEEM analysis (Pfo1

-3)(Figure

S8). No clumped isotope

analysis was done as this specimen occluded extensive calcite crystals between the two

valves (see top of transects in

Figure 6), and the nacre layer was too thin (300 μm at most)

for safe, uncontaminated separation.

Ps8:

Earl

y Jurassic (Pliensbachian)

Pinna folium

(remaining shell size: 5.6 cm, estimated

total size including missing umbo: 7 cm), from Blockley, Gloucestershire, UK, extracted

around the year 2000 from the Northcott Brick works, Ibex zone,

Beanicerus luridum

Courtesy of Christopher Andrew, Lyme Regis Museum, Lyme Regis, Dorset, United

Kingdom. The shell shows limited iridescence only in one region, and from SEM and PEEM

analysis does not have any calcite prismatic layer preserved. It does not have the umbo

any

more. The two valves are closed and filled with polycrystalline calcite (from PEEM

analysis), the entire specimen was embedded, cut, polished, and analyzed with SEM. The

best region (position 2) was cut, re

-embedded and polished for PEEM analysis (Ps8

2)(Figure

S8). No clumped isotope analysis was done as this specimen had extensive calcite

crystals between the two valves, outside of the nacre layer, in between the nacre layer, and

even percolating through some of the nacre tablets (see Figure 2F). The nacr

e layer was

even thinner than in the Pfo sample, with 90 μm maximum thickness.

Sample preparation

Three modern and seven fossil specimens were acquired from the sources

described above. The three

modern shells were rinsed in ethanol, air

-dried, and cut with a

jeweler’s saw along the bisector line. A second cut, shown in Figure S4b

, isolated a shell

fragment ~1cm-wide including the thickest nacre (red arrows in

Figures 4, S4, S7

) at its

center. This sample was then embedded in epoxy to expose the cross

-section in the thickest

nacre region, polished, and coated. The thickest nacre region was chosen because it

provides the most sample to measure, and was deposited over the longest period of time.

Consequently, experiments in this region maximize the environmental information stored

in each specimen.

Fossil Pinnidae shells are extremely fragile, as shown by the shell fragments at the

right hand side of

Figure S7a,b

for specimen Ah2. These fragments were cut off for clumped

isotope analysis using a razor blad

e gently pressed through the soft, flaky nacre.

Subsequently the remaining specimen was embedded in epoxy, then cut using a diamond

saw as shown in

Figures S6 and S7c,d

. Fossil specimens were then re-embedded to expose

a shell cross

-section in the region of thickest nacre (missing cuboid in Figure S6 and S7c,d

polished, and coated similar to modern specimens.

The embedding epoxy in all cases was EpoFix (EMS, Hatfiled, PA, USA), poured

around the shell fragment in 1-inch round molds, and cured for 13 hours. Before

embedding, extreme care was taken to coarsely polish each shell to obtain a flat cross

section perpendicular to the shell inner surface. The specimen was then mounted on

double

-stick tape to the bottom of the embedding mold in order to minimize any

orientation error introduced during the embedding stage. The finished shell mount

contained nacre layers perpendicular to the polished surface within an angle of

5°, thus

the mounting error on TT measurements was negligible.

Shell mounts were polished with coarse grit, followed by 300

-nm Al

nanoparticles, followed by 50

-nm Al

nanoparticles (MasterPrep, Buehler, Lake Bluff, IL,

USA) suspensions. Before use, both polishing suspensions were dialyzed against 22g/L

in DD water for 24h with three

solution changes. The polished samples

were rinsed in ethanol, air

-dried, covered with a mask in the area to be analyzed (purple

square in Figure S4d

, transparent square in

Figure S7e

,) and coated with 40nm Pt. This

produced the high

-reflectivity re

gion all around the square, (the white region in

Figure S4d

and black region in

Figure S7e)

. The mask was then removed, and the entire sample surface

was coated again with 1nm Pt, while rotating the sample. One nm is sufficient to ensure

good conductivity,

but is less than the ~5 nm depth below the sample surface from which

the secondary electrons detected in a PEEM experiments originate (Frazer et al., 2003)

Hence the majority of the detected signal comes from the sample, not the coating. One nm

Pt, however, is not enough to make good electrical contact, motivating the thicker coating

surrounding the area of interest. One nm coating must be done using a high-precision

sputter coater that enables slow, precise coating, during which the sample is tilted and

spun (208HR High Resolution Sputter Coater, Cressington, UK, and Ted Pella, Inc., USA).

The differential-thickness resulting from two rounds of Pt coating makes it possible to

perform photoemission experiments on shells, minerals, rocks, or any other insulators, and

was introduced by our group

(De Stasio et al., 2003; DeVol et al., 2014)

. It also prevents any

charging phenomena or artifacts (Gilbert et al., 2000)

. Figure S7e

shows a polished,

trimmed, and coated sample, ready for PEEM analysis.

Sample powders were prepared for clumped isotope analyses from the epoxy

embedded shell fragments surrounding the PEEM sub-sample for Ah3-2, Ah4-2, Px1-2 and

Px2

-2. A bulk sample of nacre was removed from the epoxy and powdered using a mortar

and pestle. For Ah2-2, a subsample from the fragments produced while cutting the sample

with a razorblade was powdered with a mortar and pestle. Unembedded samples of Ar5,

Pc 1-2, Pn 2-2, Pl

-5-2 and Pl-6-2 allowed separation of the prismatic calcite layer and the

nacre using a razor blade. Both were powdered using a mortar and pestle prior to clumped

isotope analysis.

XRD Analysis

X-ray diffraction patterns (XRD) were collected on beamline 12.3.2 at the Berkeley

Advanced Light Source, as described pr

eviously, using a DECTRIS Pilatus 1 M area detector

(Gilbert et al., 2008; Tamura and Gilbert, 2013; Yang et al., 2011). A ~1-μm spot, 9 keV

monochromatic beam illuminated the sample surface in the locations shown in Figure S2

Each pattern of reflections was indexed using the XMAS software (Tamura, 2014) , with

excellent calcite and aragonite recognition.

SEM Analysis

For scanning electron microscopy (SEM) analysis, the anterior regions of three modern

shells were cut, rinsed with ethanol, air

-dried, and coated with 20

-nm Pt. The Hitachi S-

3400N scanning electron microscope in the UW-Madison Department of Geoscience was

used to produce the images in Figure 5

, under the secondary

-electron mode and an

acc

elerating voltage of 15 kV.

EPMA analysis

Electron Probe Micro

-Analysis (EPMA) was conducted to obtain quantitative elemental

spot analysis and elemental mapping on the shell Ps 5

-4 to assess trace metal variability

across the shell. EPMA was done on the JEOL JXA-8200 Electron Microprobe at the

lifornia Institute of Technology. For all quantitative results, the accelerating voltage was

15 kV, the beam current was 20 nA, and the beam size was 1 μm. The CITZAF method was

used for matrix correction. Sample standards for the five chemical elements

analyzed,

included: dolomite for Mg, siderite for Fe, rhodochrosite for Mn, strontianite for Sr, and

anhydrite for S. Mg had an average detection limit of 0.01% Fe–272 ppm, Mn–388 ppm, Sr

–

438 ppm, and S–0.02%. EPMA results are presented in Figure S3

and reported in Table S3

The Late Cretaceous shell Ps5 is also extremely well preserved chemically. EPMA

measurements, both maps and transects, indicate that Fe and Mn trace metal

concentrations are extremely low across the shell (aragonite [Fe] = 110 ± 12 ppm and [Mn]

= 90 ± 10 ppm; calcite [Fe] = 202 ± 42 ppm and [Mn] = 93 ± 27 ppm, mean ± std. error of

the mean). As both metals tend to incorporate into calcite and aragonite in reducing

environments below the sediment-water interface, this suggests minimal

diagenesis

(Brand and Veizer, 1980)

. Sr and Mg are consistent with primary precipitation of the two

phases from seawater (aragonite [Sr] = 2567 ± 71 ppm and [Mg] = 134 ± 33 ppm; calcite

[Sr] = 986 ± 38 ppm and [Mg] = 3909 ± 187 ppm) (Figure S3, Table S3

Detailed description of digital ruler measurements

Each PIC

-map was opened in Adobe Photoshop

, then immediately duplicated into a

second “Photoshop layer”. This second layer was then rotated until the nacre

tablets were

horizontal. Tilt in the other direction was prevented with accurate sample mounting,

perpendicular to the polishing and imaging plane. A vertical un-rotated digital ruler with

ticks and numbers was pasted into a third “Photoshop layer”. For all images, the digital

ruler was arranged such that the “0” tick was located precisely at the bottom of the un-

rotated PIC

-map. All other ticks were then moved so that each coincided with the boundary

between two nacre tablets. In

Figure S9

the three “Photo

shop layers” are displayed for each

PIC-map: at the bottom is the un-rotated original PIC

-map layer, on top of it is the rotated

PIC-map, and on top of both is the digital ruler, with its white ticks and numbers. Note that

“0” is at the bottom of the field

of view, and all other ticks and numbers were shifted up or

down, one at a time, such that they coincided with nacre tablet boundaries. The vertical-

FoV was 21.5 μm in all panels of Figure S9

. Note that in Figure S9

we displayed a larger

field of view to

include all of the un

-rotated images. During the measurement the file size

was not (and should not be) increased, so the FoV measurement remained the same, even

when a Photoshop layer was rotated. This avoids the introduction of any quantitative

errors. To obtain the “average TT” in each PIC-map (Figure S10

) we divided the vertical-

FoV in each image by the number of tablets counted in that image using the digital ruler.

All PIC

-maps were measured twice, by two co-authors, recording the results in two

separate columns of a Microsoft Excel

spreadsheet. The averages and standard deviations

were calculated in Excel

, all the plots were produced in Kaleidagraph

4.5.2 for Mac.

Angle spread measurement

The angle spread of c’

-axes (the projections of c

-axes onto the polarization plane of the

illuminating radiation) was measured using the “Polarization Analysis Package” of GG

Macros

(GG–Macros, 2016)

. For every PIC-mapped area, we made a polarization stack of 19

images, with linearly polarized illumination covering a range of 90° with a 5° step. By

fitting the intensity of a pixel as a function of p

olarization to

Equation 1 (see PIC-mapping),

we then determined the values of

c’

, the c’-axis angle, and

, the amplitude of polarization-

dependent intensity, for every pixel in a PIC

-mapped area.

To remove artifacts before measurement, we first masked off pixels with extreme

values of

, namely those falling into bins populated by less than 10

pixels on a histogram

of all

-values from the PIC

-map. In particular, this removes artifacts with low polarization-

dependence and correspondingly low values of

We then placed the

c’

-values of all remaining pixels on a histogram and measured

the range of these c’

-axis angle bins that were each occupied by at least 10

pixels. This

somewhat arbitrary occurrence cut-off was used to further ensure that remaining c’-axis

angles, still representing the majority of the 10

pixels in the image, are not artifactual.

Great care was taken to confirm that no real nacre tablets were completely excluded from

the AS analysis by this cut-off. The measured angle spread in each PIC

-map was always

between 10° and 30° for all areas analyzed in

modern and fossil nacre, with the only six

exceptions due to diagenesis and shown in the six panels of

Figure 2

Using the GG Macros (GG–Macros, 2016)

, this analysis of a given PIC

-map can be

done by following a few simple steps:

1) Having opened the experiment file, stacked the 19 images, and produced a PIC

-map, now

click on the “Analyze output” button in the “Polarization Panel”.

2) The “Polarization Result Analysis” panel appears. In that panel, press “Create Masked

PICmap”. When you do that, make sure that box “From POL B” is ch

ecked in the “Masking”

section of the panel: that masks off pixels with extreme

-values, with <10

frequency, as

described above.

3) A histogram of

c’

-values of unmasked pixels appears. Check that the vertical cut-offs are

placed at the 10

level; if nece

ssary, move them manually.

4) Click on “Extract Masked AS”. The angle spread is measured and displayed in the

command panel.

The average angle spread thus obtained for each shell is shown in Figure S13

, and, although

extremely noisy, it appears to be anti

-correlated with T. It is possible that this is a sampling

artifact: at greater T tablet thickness is greater, hence there are fewer tablets in the field of

view and their possible orientations are under

-sampled, resulting in smaller angle spreads.

This is confirmed by the shape of

c’

-value histograms, which significantly departs from

Gaussian at higher T, and is generally closer to Gaussian at lower T. Rather than the angle

spread in each PIC

-map, a measurement of all orientation angles over the entire she

ll might

be more informative, but is beyond the scope of this work.

2.7 Abiotic aragonite growth experiment

Aragonite growth

.—Synthetic aragonite particles were grown simultaneously in sealed

Pyrex bowls placed in bio

-culture rooms or incubators with different temperatures. The

growth solution recipe follows

(Liu et al., 2009)

. Four

22 mm

22 mm glass coverslips

were each placed in a 35mm plastic Petri dish containing 3 ml of 10mM CaCl

+ 10mM

MgCl

. Petri dishes were covered and sealed with Al foil surrounded by Parafilm. Four holes

were poked in the Al foil of each dish, using a needle. Each of four covered Petri dishes was

placed in a separate sealable 950 ml Pyrex bowl. A chunk of ammonium carbonate (~8-9 g)

was placed on a 35 mM Petri dish with no cover, and put in each of the Pyrex bowls. The

Pyrex bowls were then sealed with plastic covers, and placed in a bowl at one of four

different temperatures: 4°C, 15°C, 22°C, and 30

°C (4°C and 15

°C = “cold” room; 22°C = lab

room temperature; 30

°C = bacteria plate incubator set to 30°C.). Twenty hours later

, each

glass coverslip was removed from the growth solution with forceps and rinsed in 10mM

Tris, pH 11 for a few seconds, rinsed in 100% ethanol for about 5 seconds, and placed in

clean 35mm Petri dishes at an angle resting on the lip of the dish to dry for 1 hour. Dry

coverslips were then placed in labeled 50ml screw cap tubes containing Kimwipes to hold

the coverslips in place, and transported for imaging.

Particle size measurement

.—Each coverslip with aragonite particles wa

s imaged using

crossed polarizers in reflection mode on a Nikon MM400 visible light microscope. The

larger particles grown at 15, 22, 30°C were imaged using the smallest magnification (5

objective); the 4°C

-particles were imaged with the highest magnification (100

objective).

Fifteen or more image files were acquired and saved from each sample. Aragonite crystal

diameters were measured in Image J

using the “straight line selection” tool, after setting

the scale to the appropriate magnification. The green lines shown in Figure 10 were drawn

in Adobe Photoshop

in locations similar to those selected for measurement in Image J

Supporting Information References

Adamkewicz, S.L., Harasewych, M., Blake, J., Saudek, D., Bult, C.J., 1997. A molecular

phylogeny of the bivalve mollusks. Molec biol Evol 14, 619

-629.

Agnini, C., Fornaciari, E., Raffi, I., Rio, D., Röhl, U., Westerhold, T., 2007. High-resolution

nannofossil biochronology of middle Paleocene to early Eocene at ODP Site 1262:

implications for calcareous nannoplankton evolution. Marine Micropaleont 64, 215

-248.

Allen, J.A., 2011. On the functional morphology of Pinna and Atrina larvae (Bivalvia:

Pinnidae) from the Atlantic. J Marine Biol Assoc UK 91, 823-829.

Aucoin, S., Himmelman, J., 2011a. Factors determining the abundance, distribution and

population size

–structure of the penshell Pinna carnea. Journal of the Marine Biological

Association of the United Kingdom 91, 593-606.

Aucoin, S., Himmelman, J., 2011b. Factors determining the abundance, distribution and

population size

–structure of the penshell Pinna carnea. J Marine Biol Assoc UK 91, 593-606.

Beer, A.C., Southgate, P.C., 2006. Spat collection, growth and meat yield of Pinna bicolor

(Gmelin) in suspended culture in northern Aust

ralia. Aquaculture 258, 424-429.

Bieler, R., Mikkelsen, P.M., Collins, T.M., Glover, E.A., González, V.L., Graf, D.L., Harper, E.M.,

Healy, J., Kawauchi, G.Y., Sharma, P.P., 2014. Investigating the Bivalve Tree of Life–an

exemplar

-based approach combining molecular and novel morphological characters. Invert

Systemat 28, 32-115.

Brand, U., Veizer, J., 1980. Chemical diagenesis of a multicomponent carbonate system--1:

Trace elements. J Sedim Res 50, 1219-1236.

Butler, A., Brewster, F., 1979. Size distributions and growth of the fan

-shell

Pinna bicolor

Gmelin (Mollusca: Eulamellibranchia) in South Australia. Marine Freshwater Res 30, 25-39.

Campbell, D.C., 2000. Molecular evidence on the evolution of the Bivalvia. Geol Soc Lond

Spec Publ 177, 31

-46.

Cash, W., 1882. Yorkshire Fossil Mollusca, Proceedings of the Yorkshire Geological and

Polytechnic Society. Geological Society of London, pp. 77

-88.

Cendejas, J., Carvallo, M., Juárez, L., 1985. Experimental spat collection and early growth of

the pen shell, Pinna

rugosa (Pelecypoda: Pinnidae), from the Gulf of California. Aquaculture

48, 331-336.

Dance, S.P., 2013. The encyclopedia of shells.

De Stasio, G., Frazer, B.H., Gilbert, B., Richter, K.L., Valley, J.W., 2003. Compensation of

charging in X

-PEEM: a successfu

l test on mineral inclusions in 4.4 Ga old zircon.

Ultramicroscopy 98, 57-62.

Dennis, K.J., Affek, H.P., Passey, B.H., Schrag, D.P., Eiler, J.M., 2011. Defining an absolute

reference frame for ‘clumped’isotope studies of CO 2. Geochim Cosmochim Acta 75, 7117-

7131.

DeVol, R.T., Metzler, R.A., Kabalah

-Amitai, L., Pokroy, B., Politi, Y., Gal, A., Addadi, L., Weiner,

S., Fernandez-Martinez, A., Demichelis, R., Gale, J.D., Ihli, J., Meldrum, F.C., Blonsky, A.Z.,

Killian, C.E., Salling, C.B., Young, A.T., Marcus, M.A., Scholl, A., Doran, A., Jenkins, C., Bechtel,

H.A., Gilbert, P.U.P.A., 2014. Oxygen spectroscopy and Polarization-dependent Imaging

Contrast (PIC)-mapping of calcium carbonate minerals and biominerals. J Phys Chem B 118,

8449-8457.

En-Zhi, M., Boucot

, A., Xu, C., Jia-Yu, R., 1986. Correlation of the Silurian rocks of China. Geol

Soc Amer Special Papers 202, 1-81.

Frazer, B.H., Gilbert, B., Sonderegger, B.R., De Stasio, G., 2003. The probing depth of total

electron yield in the sub keV range: TEY

-XAS and X-PEEM. Surf Sci 537, 161

-167.

Frederiksen, N.O., Gibson, T.G., Bybell, L.M., 1982. Paleocene

-Eocene boundary in the

eastern Gulf Coast. Gulf Coast Assocn Geol Soc Trans 32, 289-294.

GG–Macros, 2016.

http://home.physics.wisc.edu/gilbert/GG-macros.

Gilbert, B., Andres, R., Perfetti, P., Margaritondo, G., Rempfer, G., De Stasio, G., 2000.

Charging phenomena in PEEM imaging and spectroscopy. Ultramicroscopy 83, 129-139.

Gilbert, P.U.P.A., Metzler, R.A., Zhou, D., Scholl, A., Doran, A., Young, A., Kunz, M., Tamura, N.,

Coppersmith, S.N., 2008. Gradual Ordering in Red Abalone Nacre. J Am Chem Soc 130,

17519-17527.

Giribet, G., Distel, D.L., 2003. Bivalve phylogeny and molecular data. Molec System

Phylogeogr Moll, 45-90.

Giribet, G., Wheeler, W., 2002. On bivalve phylogeny: a high

‐

level analysis of the Bivalvia

(Mollusca) based on combined morphology and DNA sequence data. Invert Biol 121, 271-

324.

Glenn, L., 1904. Mollusca, Pelecypoda. System Paleontol Miocene: Maryland Geological

Survey Miocene Volume, 274-401.

Grave, B.H., 1911. Anatomy and physiology of the wing-shell

Atrina rigida

. Govt. print. off.

Hendriks, I.E., Basso, L., Deudero, S., Cabanellas

-Reboredo, M., Álvarez, E., 2012. Relative

growth rates of the noble pen shell Pinna nobilis throughout ontogeny around the Balearic

Islands (Western Mediterranean, Spain). J Shellfish Res 31, 749

-756.

Idris, M., Arshad, A., Bujang, J., Ghaffar, M., Daud, S., 2008. Biodiversity and distribution of

pen shells (Bivalvia: Pinnidae) from the seagrass beds of Sungai Pulai, Peninsular Malaysia.

Res J Fisheries Hydrobiol 3, 54

-62.

Ivany, L.C., 2012. Reconstructing paleoseasonality from accretionary skeletal carbonates

—

challenges and opportunities, Reconstr

ucting Earth's Deep

-Time Climate—The State of the

Art in 2012, pp. 133-165.

Joubert, C., Linard, C., Le Moullac, G., Soyez, C., Saulnier, D., Teaniniuraitemoana, V., Ky, C.L.,

Gueguen, Y., 2014. Temperature and Food Influence Shell Growth and Mantle Gene

Expression of Shell Matrix Proteins in the Pearl Oyster Pinctada margaritifera. PloS one 9,

e103944.

Kidwell, S.M., 1982. Stratigraphy, in Vertebrate Taphonomy and Depositional History of the

Miocene Calvert and Choptank Formations, Atlantic Coastal Plain. Yale University.

Kidwell, S.M., 1997. Anatomy of extremely thin marine sequences landward of a passive-

margin hinge zone: Neogene Calvert Cliffs succession, Maryland, USA. J Sedim Res 67, 322-

340.

Kožul, V., Glavić, N., Bolotin, J., Antolović, N., 2011. The experimental rearing of fan mussel

Pinna nobilis (Linnaeus, 1758), Procs 46th Croatian and 6th Internat Symp on Agriculture,

Opatija, Croatia, p. 806.

Landman, N.H., Johnson, R.O., Edwards, L.E., 2004. Cephalopods from the

Cretaceous/Tertiary boundary

interval on the Atlantic Coastal Plain, with a description of

the highest ammonite zones in North America. Part 2. Northeastern Monmouth County,

New Jersey. Bull Am Mus Nat Hist 287, 1-107.

Larina, E.A., 2015. Upper Maastrictian ammonite biostratigraphy of

the Gulf Coastal Plain,

Geology. Brooklyn College, p. 65.

Lemer, S., Buge, B., Bemis, A., Giribet, G., 2014. First molecular phylogeny of the

circumtropical bivalve family Pinnidae (Mollusca, Bivalvia): evidence for high levels of

cryptic species diversit

y. Molec phylog evol 75, 11

-23.

Linard, C., Gueguen, Y., Moriceau, J., Soyez, C., Hui, B., Raoux, A., Cuif, J.P., Cochard, J.-C., Le

Pennec, M., Le Moullac, G., 2011. Calcein staining of calcified structures in pearl oyster

Pinctada margaritifera and the effect of food resource level on shell growth. Aquaculture

313, 149-155.

Liu, R., Xu, X., Cai, Y., Cai, A., Pan, H., Tang, R., Cho, K., 2009. Preparation of calcite and

aragonite complex layer materials inspired from biomineralization. Cryst Growth Des 9,

3095-3099.

Lowenstam, H.A., Weiner, S., 1989. On biomineralization. Oxford University Press.

Marie, B., Joubert, C., Tayalé, A., Zanella-Cléon, I., Belliard, C., Piquemal, D., Cochennec-

Laureau, N., Marin, F., Gueguen, Y., Montagnani, C., 2012. Different se

cretory repertoires

control the biomineralization processes of prism and nacre deposition of the pearl oyster

shell. Proc Natl Acad Sci USA 109, 20986-20991.

Matsumoto, M., 2003. Phylogenetic analysis of the subclass Pteriomorphia (Bivalvia) from

mtDNA COI sequences. Molec Phylogen Evol 27, 429-440.

Mendo, T., Koch, V., Wolff, M., Sínsel, F., Ruiz

-Verdugo, C., 2011. Feasibility of intertidal

bottom culture of the penshell Atrina maura in Bahia Magdalena, Baja California Sur,

Mexico. Aquaculture 314, 252-260.

Richardson, C., Kennedy, H., Duarte, C., Kennedy, D., Proud, S., 1999. Age and growth of the

fan mussel

Pinna nobilis

from south-east Spanish Mediterranean seagrass (

Posidonia

oceanica

) meadows. Marine Biol 133, 205-212.

Richardson, C., Peharda, M., Kenn

edy, H.

, Kennedy, P., Onofri, V., 2004

. Age, growth rate and

season of recruitment of

Pinna nobilis

(L) in the Croatian Adriatic determined from Mg: Ca

and Sr: Ca shell profiles. J Exper Marine Biol Ecol 299, 1

-16.

Rosewater, J., 1961. The family Pinnidae in the Indo

-Pacific. Indo

-Pacif Mollusca 1, 501.

Rosewater, J., 1982. Review of Hawaiian Pinnidae and revalidation of Pinna exquisita Dall,

Bartsch, and Rehder, 1938 (Bivalvia: Mytiloida).

Ruedemann, R., 1916. The Paleontology of Arrested Evolution. New Yo

rk State Mus Bull

196, 107-134.

Ruedemann, R., 1918. The Paleontology of Arrested Evolution: Address by the President of

the Paleontological Society, Albany, December 1916. University of the State of New York.

Schöne, B.R., 2008. The curse of physiology

—challenges and opportunities in the

interpretation of geochemical data from mollusk shells. Geo-Marine Lett 28, 269

-285.

Schultz, P.W., Huber, M., 2013. Revision of the worldwide recent Pinnidae and some

remarks on fossil European Pinnidae. ConchBooks, Öhringen, Germany.

Sessa, J.A., Ivany, L.C., Schlossnagle, T.H., Samson, S.D., Schellenberg, S.A., 2012. The fidelity

of oxygen and strontium isotope values from shallow shelf settings: Implications for

temperature and age reconstructions. Palaeogeogr Palaeocli

matol Palaeoecol 342, 27

-39.

Steiner, G., Hammer, S., 2000. Molecular phylogeny of the Bivalvia inferred from 18S rDNA

sequences with particular reference to the Pteriomorphia. Geol Soc Lond Spec Publ 177,

11-29.

Stephenson, L.W., 1957. A shorter contribution to general geology

. Geol Survey Profess

Paper 274, 97.

Strack, E., 2001. Pearls. Rühle-Diebener Verlag, Stuttgart. ISBN 3

-00-008636-6.

Tamura, N., 2014. XMAS: A versatile tool for analyzing synchrotron X-ray microdiffraction

data. Strain and Dislocat

ion Gradients from Diffraction. Spatially-Resolved Local Structure

and Defects; Barabash, R, 125

-155.

Tamura, N., Gilbert, P.U.P.A., 2013. X-ray Microdiffraction of Biominerals, in: De Yoreo, J.J.

(Ed.), Research Methods in Biomineralization Science Academic Press, Burlington, MA, pp.

501-532.

Thibault, N., Gardin, S., 2006. Maastrichtian calcareous nannofossil biostratigraphy and

paleoecology in the Equatorial Atlantic (Demerara Rise, ODP Leg 207 Hole 1258A). Rev

Micropaléontol 49, 199

-214.

Thibault, N., Gardin, S., 2007. The late Maastrichtia

n nannofossil record of climate change in

the South Atlantic DSDP Hole 525A. Marine Micropaleontol 65, 163-184.

Turner, R.D., Rosewater, J., 1958. The family Pinnidae in the western Atlantic. Johnsonia 3,

285-326.

Wheelton, H., 1905. A monograph of the British Carboniferous Lamellibranchiata Volume II.

Palaeontographical Society.

Yang, L., Killian, C.E., Kunz, M., Tamura, N., Gilbert, P.U.P.A., 2011. Biomineral nanoparticles

are space-filling. RSC

-Nanoscale 3, 603

-609.

Yonge, C.M., 1953. Form and habit in Pinna carnea Gmelin. Philosophical Transactions of

the Royal Society B: Biological Sciences 237, 335

-374.

Zhang, R., 1988. Stratigraphic and paleobiogeographic aspects of Devonian bivalves of

China.

Figure S1

. Fossil distribution of

Atrina

and

Pinna

during the Miocene, Eocene, and Late

Cretaceous epochs. Fossil occurrence data downloaded from the Paleobiology Database on

5/25/15.

Figure S2

. X-ray diffraction patterns, obtained from the red and green spots

corresponding

ly numbered on the photographs. Exterior prismatic layer (top) and interior

nacreous layer (bottom) photographs are shown. Indexing demonstrates that in red spot

locations the mineral is calcite, and in green spots, aragonite.

Ps5 and Ps6 are the two

Late Cretaceous

samples in this study, and the only specimens

preserving calcite. All other fossil samples contained only aragonite nacre, which was

identified as unaltered using spectroscopy at the O K

-edge (as in

Figure 1

) or at the Ca L

edge

(DeVol et al., 2014)

Figure S3

. EPMA trace metal maps (with warmer colors indicating higher relative

concentration

) and quantitative transects (white trace) across one shell sample from the

Late Cretaceous (Ps5

-4).

(A) Mg concentration in units of weight %. (B, C, D) Sr, Fe, Mn

concentrations in units of parts per million (ppm). Notice in all images the calcite prismatic

layer on top,

which appears in yellow, as it is Mg

-rich, in panel A, and the rest of the image

is nacre with occasional horizontal cracks but otherwise homogeneous elemental

distributions.