nguyen et al - 2011

7/31/2019 Nguyen Et Al - 2011

1/22

(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attached

copy is furnished to the author for internal non-commercial research

and education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of the

article (e.g. in Word or Tex form) to their personal website or

institutional repository. Authors requiring further information

regarding Elseviers archiving and manuscript policies are

encouraged to visit:

http://www.elsevier.com/copyright
http://www.elsevier.com/copyrighthttp://www.elsevier.com/copyright

7/31/2019 Nguyen Et Al - 2011

2/22

Author's personal copy

Dissolution susceptibility of PaleoceneEocene planktic foraminifera: Implicationsfor palaeoceanographic reconstructions

Thi Minh Phuong Nguyen a,, Maria Rose Petrizzo b, Peter Stassen a, Robert P. Speijer a

a Department of Earth and Environmental Sciences, K.U.Leuven, Celestijnenlaan 200E, 3001 Leuven, Belgiumb Universit degli Studi di Milano, Dipartimento di Scienze della Terra, Ardito D esio, via Mangiagalli 34, 20133 Milano, Italy

a b s t r a c ta r t i c l e i n f o

Article history:

Received 11 January 2011

Received in revised form 24 June 2011

Accepted 8 July 2011

Available online xxxx

Keywords:

dissolution susceptibility

planktic foraminifera

North Pacific

Ocean Drilling Program

PaleoceneEocene

paleoceanography

paleoclimatology

taphonomy

We investigated shell characteristics and differential dissolution susceptibility of planktic foraminiferal

species derived from upper Paleocene and lower Eocene deep-sea sequences, Ocean Drilling Program (ODP)

Site 865 (Allison Guyot) and Sites 1209B, 1210B and 1212A (Shatsky Rise) in the North Paci fic Ocean. The

purposes of this study are: 1) assessing the effects of differential dissolution on upper Paleocenelower

Eocene planktic foraminiferal assemblages, at species level and within different biozones, to quantify

dissolution susceptibility of genera and species; 2) investigating the differences in shell characteristics; 3)

revealing the relationship between shell parameters and dissolution robustness of taxa, and 4) identifying the

key shell parameter(s) influencing the dissolution susceptibility of foraminiferal taxa.

Two independent experiments were carried out, one focusing on gradual qualitative deterioration of taxa by

dissolution and the other documenting the weight loss of taxa. Shell parameters such as wall thickness,

porosity and pore size were determined through Scanning Electron Microscopy (SEM) and image analysis

(JMicroVision).

We found that the large muricate Acarinina and Morozovella are most resistant, followed by the cancellate

Subbotina and the small muricate Igorina, confirming results of previous work. At species level, the thick-

walled Acarinina soldadoensis, Acarinina subsphaerica and the large Morozovella subbotinae are the most

resistant species. Most of the large Morozovella species such as Morozovella aequa, Morozovella formosa-

gracilis , Morozovella velascoensis and Morozovella pasionensis, together with Acarinina nitida show

intermediate dissolution resistance, whereas the small muricate Igorina species, the cancellate Subbotina

velascoensis and the thin-walled Morozovella acuta and Morozovella occlusa are the most vulnerable species.

We propose a formula for calculating the dissolution resistance of taxa based on their wall thickness and

size two key parameters in dissolution resistance of a species. Application of this formula reveals good

agreement between the calculated and measured dissolution resistance, indicating its robustness.

Furthermore, the agreement between our experimental results, in-situ experimental results on live

foraminifera and natural quantitative/qualitative records suggests that our experiments accurately mimic

natural dissolution processes. Consequently, these experimental results strongly bear on interpretations of

foraminiferal dissolution in natural environments, especially in studies on early Paleogene climatic events

that are often associated with dissolution phenomena. More generally, a proper assessment of taphonomic

alteration by dissolution should be part of every paleoenvironmental reconstruction based on quantitative

foraminiferal records.

2011 Elsevier B.V. All rights reserved.

1. Introduction

Planktic foraminifera are among the most abundant organisms in

the open ocean and their shells often form a major component of

calcareous sediments on the ocean's floors. The sensitivity of planktic

foraminifera to the properties of ambient sea water has been

extensively demonstrated (e.g. Berger, 1970, 1975; Coulbourn et al.,

1980; ari et al., 2005), making them excellent paleoenvironmental

recorders in Quaternary and deep-time studies (e.g. Berger, 1968;

Mix, 1989). More specifically, planktic foraminifera provide informa-

tion on upper ocean environments in which they calcify and play an

important role in interpretations of paleoenvironmental conditions at

the time of deposition (e.g. B, 1968; Berger, 1968; Herman and

Bogyo, 1980).

However, only a small proportion of all shells sinking to the seafloor are ultimately preserved in the sedimentary record, because of

carbonate dissolution in the water column, at the sea floor and within

the sediment. Dissolution of planktic foraminifera in the deep sea is a

Marine Micropaleontology 81 (2011) 121

Corresponding author. Tel.: +32 16 326452; fax: +32 16 322980.

E-mail addresses: [email protected],

[email protected] (T.M.P. Nguyen).

0377-8398/$ see front matter 2011 Elsevier B.V. All rights reserved.

doi:10.1016/j.marmicro.2011.07.001

Contents lists available at ScienceDirect

Marine Micropaleontology

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m a r m i c r o

7/31/2019 Nguyen Et Al - 2011

3/22


widespread phenomenon (e.g. Berger, 1970; Peterson and Prell, 1985;

Conan et al., 2002). In ocean cores, complete dissolution is easily

recognized by the absence of calcareous foraminifera in clay beds (e.g.

Spiegler and Jansen, 1989). Less pervasive dissolution is, however,

more difficult to recognize and is often ignored, although various

subtle indications, such as increased shell fragmentation, decreased

planktic/benthic (P/B) ratios and carbonate content are frequentlydocumented in the geological records (Melillo, 1988; Weaver and

Raymo, 1989; Iaccarino and Proto-Decima, 1990; Premoli Silva and

Spezzaferri, 1990; Chen and Farrell, 1991; Brunner, 1992; Basov,

1995; Sliter, 1995).

In paleoenvironmental studies, the basic assumption is that fossil

assemblages truthfully reflect the biocoenosis and underlying environ-

mental signals (e.g. van der Zwaan et al., 1990). Since differential

susceptibility of planktic foraminifera to dissolution will influence the

composition of fossil assemblages, the reliability of paleoenvironmental

reconstructions, based on planktic foraminiferal assemblages may be

significantly reduced. For this reason, an objective assessment of

potential taphonomic alteration by dissolution should be part of every

paleoenvironmental reconstruction based on quantitative foraminiferal

data, from deep-sea basins and continental margin settings.Most studies dealing with planktic foraminiferal dissolution are

restricted to recent settings and Quaternary sequences. Research carried

out on pre-Quaternary stratigraphic intervals is relatively rare and

focused on specific qualitative and quantitative aspects of dissolution

(e.g. Jenkins and Orr, 1971; Violanti et al., 1979; Malmgren, 1987;

Petrizzo et al., 2008). Most knowledge on dissolution susceptibility of

extant planktic foraminifera is derived from quantitative analysis, and in

rare cases, from experimental studies. A general outcome of these

studies is thatdissolutionsusceptibilitydependson thecharacteristicsof

the foraminiferal shell, such as surface/volume ratio, microstructural

features, or porosity (e.g. Sliter et al., 1975; Thunell, 1976; Corliss and

Honjo, 1981). A commonly used index for foraminiferal dissolution is

the percentage of planktic foraminiferal fragments in the assemblage

(Berger, 1973; Cullen and Prell, 1984; Peterson and Prell, 1985; Le and

Thunell, 1996). Yet, this index does not unequivocally characterize

dissolution levels (Petrizzo et al., 2008). For instance, fragment ratios

can drop rapidly in samples that are severely affected by dissolution. In

addition, the fragmentation index cannot be employed in lithified

sediments that require more intense processing than soft mud,

contributing to fragmentation.

This study is part of a larger effort dealing with the effects of

differential dissolution on benthic and planktic foraminiferal assem-

blages from stratigraphic intervals that record remarkable environmen-

tal changes. The general aims are to: 1) reveal the effects of dissolution

on the composition of foraminiferal assemblages; 2) determine which

factors control the dissolution susceptibility of taxa, and 3) improve the

reliabilityof paleoenvironmental reconstructions basedon foraminiferal

assemblages. Our previous experimental study (Nguyen et al., 2009)

was carried out on a lower Eocene foraminiferal assemblage (benthicand planktic) from the Esna Formation in the Dababiya section, the

Global boundary Stratotype Section and Point for the base of the Eocene

inEgypt (Dupuiset al., 2003). Theresults revealed,amongothers, that at

a generic level, Acarinina is slightly more dissolution resistant than

Morozovella, but more significantly, both these muricate, symbiont-

bearing, surface-dwelling taxa are much more resistant to dissolution

than the cancellate, asymbiotic, deep-dwelling Subbotina. Moreover,

morphology, test size and especially shell wall micro-texture seem to

play important roles in the dissolution susceptibility of planktic

foraminiferal taxa. These results motivated more detailed experimental

studies on upper Paleocene to lower Eocene planktic foraminiferal

species.

In this study, we carried out investigations on dissolution

susceptibility and shell parameters of taxa from upper Paleoceneand lower Eocene planktic foraminiferal assemblages from the Pacific

Ocean. This time slice was chosen because dissolution phenomena are

a recurrent problem in upper Paleocene to lower Eoceneforaminiferal

assemblages, especially in connection with the Paleocene/Eocene

Thermal Maximum (PETM, e.g. Ernst et al., 2006; Petrizzo, 2007;

Guasti and Speijer, 2007), the Early Late Paleocene Event (ELPE, Rhl

et al., 2004; Petrizzo, 2005) and the Eocene Thermal Maximum 2

(ETM 2 Shipboard Scientific Party, 2003). The selected sedimentary

sequences contain various indications suggesting that partial disso-

lution prevailed in the Pacific Ocean during the PaleoceneEocene

interval (Sager et al., 1993; Colosimo et al., 2006).

This study is aimed at: 1) assessing the effects of differential

dissolution on upper Paleocenelower Eocene planktic foraminiferal

assemblages, at species level and within different biozones, in order to

quantifydissolution susceptibility of genera and species; 2) investigating

differences in shell characteristics of planktic species; 3) revealing the

relationship between shell parameters and dissolution susceptibility;

4) identifying the key parameter(s)influencing dissolutionsusceptibility

of these planktic taxa. Results from these investigations form a basis for

developing a new dissolution index, which allows an objective and

quantitative evaluation of the degree of dissolution that has occurred in

planktic foraminiferal assemblages.

12121212

Fig. 1. Late Paleocene paleogeography (Hay et al., 1999) and locations of ODP Sites 865 (Allison Guyot, equatorial Pacific Ocean), 1209, 1210 and 1212 (Shatsky Rise, northwestern

Pacific Ocean).

2 T.M.P. Nguyen et al. / Marine Micropaleontology 81 (2011) 121

7/31/2019 Nguyen Et Al - 2011

4/22


Table 2

List of samples, taxa used in the weight loss experiment (experiment 2).

Samples Zone/subzone Taxa name Size fraction

(m)

125

330

330

490

143-865B-11H-4, 4850 P 6a (E3) Acarinina soldadoensis x

143-865B-11H-4, 485 0 P6a (E3) Morozovella subbotinae x x

143-865B-11H-4, 485 0 P6a (E3) Subbotina velascoensis x x

143-865B-12H-1, 130132 P5 (E2) Acarinina soldadoensis x x143-865B-12H-1, 130132 P5 (E2) Morozovella occlusa x x

143-865B-12H-1, 130132 P5 (E2) Morozovella aequa x x

143-865B-12H-1, 130132 P5 (E2) Subbotina velascoensis x x

143-865B-13H-1, 124125 P5 Acarinina nitida x x

143-865B-13H-1, 124125 P5 Igorina pusilla x

143-865B-13H-1, 124125 P5 Igorina tadjikistanensis x

143-865B-13H-1, 124125 P5 Morozovella acuta x x

143-865B-13H-1, 124125 P5 Subbotina velascoensis x x

143-13H-2, 5052 P4c Acarinina nitida x

143-13H-2, 5052 P4c Morozovella velascoensis x

143-865B-13H-3, 5052 P4c Acarinina nitida x x

143-865B-13H-3, 5052 P4c Igorina albeari x

143-865B-13H-3, 5052 P4c Igorina tadjikistanensis x

143-865B-13H-3, 5052 P4c Subbotina velascoensis x x

143-865B-13H-4, 5052 P4c Acarinina nitida x x

143-865B-13H-4, 5052 P4c Igorina pusilla x

143-865B-13H-4, 5052 P4c Igorina albeari x

143-865B-13H-4, 5052 P4c Igorina tadjikistanensis x

143-865B-13H-4, 5052 P4c Morozovella acuta x x

143-865B-13H-4, 5052 P4c Subbotina velascoensis x x

143-865B-14H4, 4749 P4a Igorina pusilla x

143-865B-14H4, 4749 P4a Igorina albeari x x

143-865B-14H4, 4749 P4a Igorina tadjikistanensis x

143-865B-14H4, 4749 P4a Morozovella acuta x

198-1209B-23H1, 6364 P4c Morozovella pasionensis x x

19 8-1 212 A-9 H-C C P6b ( E4 ) Acarinina soldadoensis x x

19 8-1 212 A-9 H-C C P6b ( E4 ) Acarinina subsphaerica x x

19 8-1 212 A-9 H-C C P6b ( E4 ) Igorina broedermanni x

19 8-1 212 A-9 H-C C P6b ( E4 ) Morozovella subbotinae x x

19 8-1 212 A-9 H-C C P6b ( E4 ) Morozovella formosa-

gracilis

x

19 8-1 212 A-9 H-C C P6b ( E4 ) Subbotina velascoensis x x

198-1210B-19H-CC P6a (E3) Acarinina soldadoensis x x

198-1210B-19H-CC P6a (E3) Acarinina subsphaerica x x

198-1210B-19H-CC P6a (E3) Igorina broedermanni x198-1210B-19H-CC P6a (E3) Morozovella subbotinae x

198-1210B-19H-CC P6a (E3) Subbotina velascoensis x x

Table 1

List of samples, taxa and size fractions used in the qualitative shell deterioration

experiment (experiment 1).

Sample Zone/

subzone

Taxa

name

Size fraction (m)

125

250

250

330

330

400

400

490

490

550

N550

143-865B-11H-4,

4850

P6a(E3)

Acarininasubsphaerica

x x

143-865B-

11H-4,

4850

P6a

(E3)

Morozovella

subbotinae

x x

143-865B-

11H-4,

4850

P6a

(E3)

Subbotina

velascoensis

x

143-865B-

12H-1,

130132

P5 (E2) Acarinina

soldadoensis

x x

143-865B-

12H-1,

130132

P5 (E2) Morozovella

occlusa

x x

143-865B-

12H-1,

130132

P5 (E2) Morozovella

aequa

x x

143-865B-

12H-1,

130132

P5 (E2) Subbotina

velascoensis

x x

143-865B-

13H-1,

124125

P5 Acarinina

soldadoensis

x x x

143-865B-

13H-1,

124125

P5 Igorina pusilla x

143-865B-

13H-1,

124125

P5 Morozovella

acuta

x x

143-865B-

13H-1,

124125

P5 Subbotina

velascoensis

x x

143-865B-

13H-2,5052

P4c Acarinina

nitida

x x x

143-865B-

13H-2,

5052

P4c Morozovella

velascoensis

x x x

143-865B-

13H-2,

5052

P4c Subbotina

velascoensis

x x x

143-865B-

13H-3,

5052

P4c Igorina pusilla x x

143-865B-

13H-3,

5052

P4c Subbotina

velascoensis

x

143-865B-

13H-4,

5052

P4c Acarinina

nitida

x x

143-865B-

13H-4,5052

P4c Igorina pusilla x

143-865B-

13H-4,

5052

P4c Morozovella

acuta

x

143-865B-

13H-4,

5052

P4c Subbotina

velascoensis

x x

143-865B-

14H4,

4749

P4a Igorina

tadjikistanensis

x x

143-865B-

14H-4,

4749

P4a Igorina pusilla x x

198-1209B-

23H-1,

6364

P4c Acarinina

subsphaerica

x

198-1209B-23H-1,

P4c Morozovellapasionensis

x x

(continued on next part of this page)

Table 1 (continued)

Sample Zone/

subzone

Taxa

name

Size fraction (m)

125

250

250

330

330

400

400

490

490

550

N550

6364

198-1209B-

23H-1,6364

P4c Subbotina

velascoensis

x

198-1209B-

23H-2,

101102

P4b Acarinina

subsphaerica

x x

198-1209B-

23H-2,

101102

P4b Morozovella

occlusa

x x

198-1209B-

23H-2,

101102

P4b Subbotina

velascoensis

x

198-1210B-

19H-CC

P6a

(E3)

Acarinina

soldadoensis

x x x

198-1210B-

19H-CC

P6a

(E3)

Morozovella

subbotinae

x x x

198-1210B-

19H-CC

P6a

(E3)

Subbotina

velascoensis

x x x

198-1212A-

9H-CC

P6b

(E4)

Acarinina

soldadoensis

x x x

198-1212A-

9H-CC

P6b

(E4)

Subbotina

velascoensis

x x x

3T.M.P. Nguyen et al. / Marine Micropaleontology 81 (2011) 121

7/31/2019 Nguyen Et Al - 2011

5/22


2. Material and methods

2.1. Material

The planktic foraminiferal taxa used in the experiments were

derived from ODP Leg 143, Hole 865B, Allison Guyot in the equatorial

Pacific Ocean and from ODP Leg 198, Holes 1209B, 1210 and 1212A,Shatsky Rise in the North Pacific Ocean (Fig. 1). The present water

depth for Hole 865B is 1518 m (Sager et al., 1993); for Hole 1209B is

2387 m (Takeda and Kaiho, 2007); for Hole 1210B is 2573 m

(Bralower et al., 2006) and for Hole 1212A is 2681 m (Bralower

et al., 2006).

At Hole 865B, 7 samples were retrieved from cores 11H-4, 12H-1,

13H-1, 13H-2, 13H-3, 13H-4, and 14H-4. The biozonation applied is

based on the data ofPetrizzo et al. (2005), adapted to the low-latitude

zonation scheme of Berggren and Pearson (2005). The samples

span planktic foraminiferal Subzone P4a to P6a (=E3, zonation by

Berggren and Pearson, 2005). At Hole 1209B, we used 2 samples from

cores 23H-1 and 23H-2 (Zone P4). At Hole 1210B, 1 sample from core

19H-CC (Subzone P6a=E3) and at Hole 1212A, another sample from

core 9H-CC (Subzone P6b=Zone E4) were analyzed. Species identifi

-cations (Tables 1 and 2, Plates 1 and 2) follow Olsson et al. (1999),

Petrizzo (2005), and Pearson et al. (2006).

The selected samples contain rich planktic foraminiferal

assemblages. The initial preservation state of these assemblages

is good, most specimens in these assemblages are intact and

generally free of infilling. However, the sedimentary successions

from which they were retrieved also show changes in fragmenta-

tion of planktic foraminifera and changes in carbonate content,

suggesting at least partial dissolution associated with the PETM

(Sager et al., 1993; Hancock and Dickens, 2005). Deposition at

these sites took place at lower bathyal to abyssal paleodepths of

13002000 m (Bralower et al., 1995; Sager et al., 1999; Shipboard

Scientific Party, 2002).

2.2. Dissolution experiments

To construct a dissolution susceptibility ranking scheme for

planktic taxa, two independent experiments were carried out. Thefirst experiment focused on gradual qualitative deterioration of taxa

and the second on the weight loss of taxa.

2.2.1. Experiment 1 qualitative deterioration

The purpose of this experiment is to investigate the differential

dissolution resistance of planktic taxa, based on the deterioration of

their tests by dissolution. Samples were first soaked in water and

washed through a 38 m sieve, and then dried on a hot plate. In total

12 species belonging to the four most common upper Paleocene to

lower Eocene genera (Acarinina, Morozovella, Subbotina and Igorina)

were used in this experiment (Table 1 and Plates 1 and 2). Well-preserved empty specimens were picked from six different size

fractions of 125250 m, 250330 m, 330400 m, 400490 m,490550 m and N550 m. Wherepossible, 1020 intact specimens ofeach species and from every size fraction were used in the

experiment. Each group of specimens was stored in separate glass

containers. Prior to the experiment, the preservation state of these

specimens was determined by binocular microscopy. During the

experiment, specimens were exposed to 5 ml of dissolution media

(a buffered acetic acid solution with a pH of 6.6) in two hour

increments until the last specimen was fully dissolved or disinte-

grated. Every 2 h, specimens were cleaned through a 38 m sieve

using distilled water, dried, and their state of preservation was re-

assessed by binocular microscope.

A foraminiferal preservation scale modified from Boltovskoy andTotah (1992) was employed and scores were assigned according to

the Absolute Preservation Scores (APS), ranging from 1 (=test

destroyed) to 8 (=test intact). These values were then converted

into Relative Preservation Scores (RPS), which are calculated relative

to the initial APS (see also Nguyen et al., 2009):

RPS(tx)=APS(tx)/APS(t0), in which tx is the time (measured in

hours) of experimental exposure, APS(t0) is the APS prior to the

experiment.

We also calculated the Cumulative Relative Preservation Scores(CRPS) for each species within a size fraction, this is to express the

progressive dissolution characteristics of the taxa:

CRPS=RPS(tx), x ranges from 0 to the time (hours) when thetaxa are completely dissolved or disintegrated.

Because the CRPS of a taxon differs per size fraction, we calculated

the average CRPS for the combined size. This value enables us to

compare the differential dissolution between taxa.

Average CRPS= CRPS (size m)/nr. of size fractions, m=size

fraction of 125250 m,250330 m, 330400 m,400490 m, 490550 m and N550 m.

2.2.2. Experiment 2 weight loss

In the second experiment, we focused on investigating thedifferential weight loss of taxa resulting from dissolution. During

the first experiment, we observed that larger specimens of a species

possess thicker test walls than smaller specimens and thus contain

more calcite, so they are expected to take longer to dissolve. This

means that the relationship between size fraction and dissolution

susceptibility of taxa is most likely related to the original weight of

the specimens. Furthermore, calcite content in the test is probably

the key factor responsible for the differential dissolution between

specimens and species (Berger, 1967). Therefore, independently

from experiment 1, we examined the role of the original test weight

on the differential dissolution susceptibility of planktic foraminiferal

species.

Another set of about 13,000 specimens, composed of 15 species

belonging to the genera Acarinina, Igorina, Morozovella and Subbotina,

were picked from splits of the same samples used in experiment 1 and

were used to evaluate the weight loss (Table 2 and Plates 1 and 2). All

specimens used in this investigation were selected from two size

fractionsof 125330 m and 330490 m. Toassessthe general fateof

the species, all specimens encountered in the sub-samples (about

1000 specimens) were used in the experiment, without distinguish-

ing their preservationstate. The experiment wascarried out following

the same procedure used in experiment 1. Before and after the

experiment, the number of specimens of every species was counted

and their total weight was measured using the Micro Sartorius

scale. Then, the total weight was divided by the number of

specimens for each species to obtain the average weight of a

species at that moment. We also calculated the Relative Remaining

Weight (RRW) after every 2 h of acid exposure in comparison with

their original weight (100%) before the experiment. These valueswere then summed into the Cumulative Relative Remaining Weight

(CRRW) and used to compare the weight loss among species by

dissolution.

The calculations are as follows:

RRW(tx)=TW(tx)/TW(t0), where TWis the total weight of taxa, txis the time that taxa are exposed to acid treatment and t0 is the time

before taxa are exposed to acid treatment.

CRRW=RRW(tx), x ranges from t0 to the time (hours) that the

taxa are completely dissolved.

Because the CRRW of a single species somewhat varies between

samples, we calculated the mean value of theCRRW forthesesamples.

This is needed in order to compare the differential weight loss

between taxa.

Mean CRRW= (CRRWy Ny)/ Ny, in which CRRWy indicatesthe CRRW of taxa in sample y and Ny is the number of specimens of

taxa in sample y.


7/31/2019 Nguyen Et Al - 2011

6/22


Plate 1. SEM images illustrating the species concepts applied in this study. 1) Igorina pusilla, sample 865B-13H-1, 124

125 cm. 2) Morozovella subbotinae, sample 1210B-19H-CC.3) Igorina albeari, sample 1209B-22H-3, 111112 cm. 4) Igorina tadjikistanensis, sample 1209B-23H-2, 148149 cm. 5) Igorina broedermanni, sample 865B-8H-2, 5052 cm.

6) Morozovella velascoensis, sample 865B-13H-2, 5052 cm. 7) Morozovella aequa, sample 1209B-23H-2, 101102 cm. 8) Morozovella pasionensis, sample 1209B-23H-1,6364 cm. a,

umbilical view; b, lateral view; c, spiral view. Scale bars: 100 m.


7/31/2019 Nguyen Et Al - 2011

7/22


Plate 2. SEM images illustrating the species concepts applied in this study. 1) Morozovella occlusa, sample 1209B-23H-2, 101102 cm. 2) Acarinina nitida, sample 1209B-23H-2, 91

92 cm. 3) Acarinina subsphaerica, sample 1209B-23H-2, 101102 cm. 4) Acarinina soldadoensis, sample 865B-11H-4, 4850 cm. 5) Subbotina velascoensis, 1209B-23H-1, 6364 cm.

6) Acarinina soldadoensis, sample 1210B-19H-CC. 7) Morozovella formosa-gracilis, sample 1212A-9H-CC. 8) Morozovella formosa-gracilis, sample 1210B-19H-CC. a, umbilical view;

b, lateral view; c, spiral view. Scale bars: 100 m.


7/31/2019 Nguyen Et Al - 2011

8/22


2.3. Measurement of shell parameters

To determine data on shell parameters such as wall thickness,

porosity (% of surface area occupied by pores) and pore sizes (on the

external surface), we analyzed SEM images of taxa using JMicroVision

software (Nicolas, 2009). Measured specimens were randomly chosen

from samples in which the taxa were abundant. Wall thickness was

measured in the central part of the last chamber, or on the

penultimate chamber in specimens having a very thin wall in the

last chamber related to the reproductive cycle (gametogenic

chamber). We used the function 1 D measurement to measure the

wall thickness and Object extraction to measure the porosity

(% pore area of total surface area) and pore size (Fig. 2). Measure-

ments of these parameters for every taxon are listed in Appendices A

to F.

Our observations showed that some taxa (for instance Subbotina

velascoensis) show fewer chambers in the last whorl than others andwhen thelast chamber of this species dissolved, their size significantly

decreases. So it seems that the number of chambers in the last whorl

can also be an important factor affecting the dissolution resistance of

taxa. To examine whether this is true, we counted the number of

chambersin the last whorl of each species. This is carried out, together

with measuring the size (largest diameter) of individual species using

a binocular microscope.

2.4. Data analysis

All data on CRPS, CRRW and shell parameters were analyzed

statistically, using PAST software (Hammer et al., 2001). Correlation

analysis was used to examine the degree of correlation between these

variables. As our data are not normally distributed, we used the non-parametric correlation analysis (Spearman and Kendall correlation

measures; Gibbons, 1985) fora more powerful analysis that highlights

which parameters play the most important role in the dissolution

susceptibility of the species.

Subsequently, multiple regression analyses were performed, in

which shell parameters, showing a strong correlation with CRPS and

CRRW, were considered as independent variables while CRPS, CRRW

were considered as dependent variables. This is done in order to

quantify the effects of these parameters on the dissolution resistance

of species as well as to describe their relationship through a linear

function.

3. Results

3.1. Differential shell deterioration by dissolution (experiment 1)

3.1.1. Differential shell deterioration between genera

We investigated the variation in RPS of species belonging toAcarinina, Morozovella, Subbotina and Igorina within a sample. Data

from individual and combined size fractions show strong species

offsets as evidencedby RPS(Fig.3AD).Acarinina species are generally

most dissolution resistant, followed by Morozovella subbotinae,Morozovella aequa, Subbotina velascoensis, the other Morozovella

species and Igorina species, respectively. The differences in RPS

between taxa are prominent in larger size fractions and much less in

small size fractions(Fig. 3AC, Table 3A). When species withina genus

from different samples and different size fractions are merged

together, the RPS of the genera vary widely. The RPS of Igorina drops

rapidly, reaching zero after 22 h of exposure, whereas the RPS of

Subbotina, Morozovella and Acarinina took 36, 56 and 84 h, respec-

tively, to reach zero (Fig. 3E). The differences in dissolution damage

between these genera leads to dissimilar average CRPS betweengenera (Table 3B), whichis highest (11.5)inAcarinina, and isabout1.3

times higherthanthat ofMorozovella and2 timeshigher than theCRPS

1 D measurement

Value (m)

18.92385

Parameter

Line length

Whole image (image surface %)

Poligon 1(image surface %)

Poligon 1 count

Data viewer (Object extraction)

Class Class 1

54.45

17.17

758

Fig. 2. Measurements of wall thickness, shell porosity and pore size (external) using JMicroVision software ( Nicolas, 2009). To determine wall thickness we used the function 1 D

measurement and the result is indicated here as the value of Line length (18.92385) in the table on the left. To determine porosity and pore size, we used the function Object

extraction. Results of these measurements are indicated here in the upper table on the right. In this table, the value in Poligon 1 (17.17%) indicates the percentage of pore size on

the external surface of specimen (54.45% of the whole picture) and the value in Poligon 1 count (758) shows the number of pores on the surface of the specimens, as indicated in

this image. Area of these pores is listed in Appendix D.


7/31/2019 Nguyen Et Al - 2011

9/22


of Subbotina. The small muricate Igorina shows the lowest average

CRPS, indicating that it is most dissolution prone.

3.1.2. Differential shell deterioration among species

Dissolution susceptibility of species within genera was determined

for specimens from all samples merged together, for the different size

fractions (Table 3A) and for the combined size fraction (Fig. 4 and

Table 3B).

3.1.2.1. Acarinina species. Three Acarinina species were used in the

investigation of shell deterioration: A. nitida, Acarinina soldadoensis

and Acarinina subsphaerica. In examinations of individual sizefractions and the combined size fraction, changes in the RPS of

these species with exposure time show considerable differences in

dissolution susceptibility. The mean RPS and average CRPS values ofA. nitida are always lower than those of A. subsphaerica and

A. soldadoensis. For A. nitida, the RPS reached zero after 42 h of exposure.

A. soldadoensis lastedtwiceas long (84 h) andA. subsphaerica in between

(56 h) (Fig. 4A).These patternsconsistentlyindicate, bothin thedifferent

size fractions and in the combined size fraction, that A. nitida is most

susceptibleAcarinina, followedbyA. subsphaericaandthenA. soldadoensis

(Table 3A3B).

3.1.2.2. Igorina species. Two Igorina species were investigated: I. pusilla

and Igorina tadjikistanensis. The RPS values of these species are nearly

identical throughout the experiment, reaching zero after 22 h ofexposure (Fig. 4B). The similarity in their average CRPS for both size

fractions of 125250 m and 250330 m as well as at the combined

Acarinina Igorina

Morozovella Subbotina

0 4 8 12 1 6 20 2 4 28 32 3 6 40 4 4 48 52 5 6 60 64 6 8 72 7 6 80 84

1

0.8

0.6

0.4

0.2

0

RPS

Exposure time (hours)

E

0

0.2

0.4

0.6

0.8

1

0 4 8 12 1 6 20 2 4 28 32 36 40 4 4 48 5 2 56 60 64 6 8 72 7 6 80 84

RPS

RPS

A. soldadoensis400-490 m

M. subbotinae400-490 m

S. velascoensis400-490 m

A



C

0 2 4 6 8 10 12 14 16 18 200

0.2

0.4

0.6

0.8

1

RPS

I. pusilla125-250 mA. nitida125-250 m

S. velascoensis125-250 m


0

0.2

0.4

0.6

0.8

1

0 4 8 12 16 2 0 24 2 8 32 3 6 40 44 48 52 56 6 0 64 6 8 72 7 6 80 84

RPS

D

0

0.2

0.4

0.6

0.8

1

0 4 8 12 16 20 24 28 32 36 40

B


A. soldadoensis330-400 m M. aequa330-400 m

M. occlusa330-400 m S. velascoensis330-400 m

A.soldadoensis M.subbotinae S.velascoensis

Fig. 3. AC: Examples of the differential reduction in Relative Preservation Score (RPS) with exposure time for some common Paleogene species in the same size fraction, from the

same sample. Aspecies from sample 1210B-19H-CC (Shatsky Rise), size fraction of 400490 m; Bspecies from sample 865B-12H-1, 130132 cm (Allison Guyot), size fraction of

330400 m; Cspecies from sample 865B-13H-4, 5052 cm (Allison Guyot), size fraction of 125250 m; Ddifferential reduction in Relative Preservation Score between species

from sample 1210B-19H-CC when all size fractions are combined together and Eidem for genera (all species within each genus are merged) from different samples, different size

fractions combined.


7/31/2019 Nguyen Et Al - 2011

10/22


size fraction (Table 3A3B) demonstrate that these taxa are equally

susceptible to dissolution.

3.1.2.3. Morozovella species. Six Morozovella species were examined:M. acuta, M. aequa, Morozovella occlusa, Morozovella pasionensis,

M. subbotinae and Morozovella velascoensis. The results indicate a

significant distinction in the solution susceptibility between

M. subbotinae and the other Morozovella species. During the

experiment, the RPS of M. subbotinae is always higher than that of

theother species and reached zero after 56 h (Fig. 4C). Thedecrease in

RPS is very fast in M. acuta and M. occlusa, yet slower in

M. velascoensis, M. pasionensis and M. aequa. From 26 to 34 h, the

RPS of the following species successively dropped to zero: M. acuta,M. pasionensis, M. aequa, M. occlusa and M. velascoensis (Fig. 4C).

The observed difference in dissolution resistance between

Morozovella species is also expressed by their average CRPS values

at the individual size fractions and combined size fraction. In both

cases, this value is highest in M. subbotinae, followed by M. aequa

(Table 3A3B). For the other taxa, comparisons between size fractions

show small deviations in their rank, with respect to the average CRPS

values (Table 3A). The combined size fraction indicates that among

Morozovella species, M. subbotinae is the most dissolution resistant,

followed by M. aequa, M. velascoensis, M. pasionensis, M. occlusa and M.acuta, respectively (Table 3B).

3.1.2.4. Intraspecific differential dissolution between samples and

locations. Our results also highlight intraspecific dissolution suscep-

tibility between locations (Fig. 5). For instance: A. soldadoensis

specimens from sample 1210B-19H-CC are the most resistant,

followed by specimens from sample 1212A-9H-CC, 865B-12H-1,

130132 cm and 865B-13H-1, 124125 cm, respectively (Fig. 5A).

The distinct difference in dissolution resistance between locations is

also observed in M. subbotinae and in S. velascoensis (Fig. 5B and C). In

all cases, the taxa from Site 865 are less dissolution resistant than the

taxa from the Shatsky Rise sites.

3.2. Differential weight loss from dissolution (experiment 2)

3.2.1. Average genus and species weights

Our measurements show a considerable difference in average

initial weight between genera and species. In the size fractions

N125 m Acarinina is the heaviest genus, with an average weight of13.6 g/specimen, followed by Morozovella (12.2 g), Subbotina

(7.9 g) and Igorina (6.2 g) (Table 4).Among Acarinina, A. soldadoensis is the heaviest species (average

weight 16.5 g), followed by A. subsphaerica (13.4 g) and A. nitida

(11.3 g). Igorina albeari is the heaviest Igorina (7.1 g), followed byI. pusilla (6.3 g), I. tadjikistanensis (5.7 g) and I. broedermanni (5.5 g).The morozovellids can be divided into two groups: a heavy group,

composed of the large species M. formosa-gracilis, M. subbotinae,

M. aequa, M. velascoensis and M. pasionensis (average weights

11.417.1 g) and a second lighter group consisting of M. occlusa andM. acuta (7.68.6 g/specimen; Table 4 and Appendix A).

3.2.2. Weight loss by dissolution

The weight loss experiments show that dissolution causesdifferential decreases in weight among genera (Fig. 6). The weight

ofAcarinina reached zero after 58 h of exposurewhile for Morozovella,

Table 3A

List of species ranked in decreasing order of dissolution resistance, based on the average CRPS from the start until the end of experiment 1, in different size fractions. Note that the

data presented here are a combination of different samples.

Size

(m)

Acarinina species Igorina species Morozovella species Subbotina species

Taxa name CRPS Taxa name CRPS Taxa name CRPS Taxa name CRPS

125250 A. nitida 4.0 I. pusilla 3.5 S. velascoensis 3.4

I. tadjikistanensis 3.4

250330 A. nitida 6.0 I. pusilla 5.1 S. velascoensis 4.9

A. subsphaerica 8.0 I. tadjikistanensis 5.1

A. soldadoensis 9.8

330400 A. nitida 8.6 M. acuta 4.6 S. velascoensis 6.8

A. subsphaerica 10.6 M. pasionensis 5.7

A. soldadoensis 11.8 M. occlusa 5.7

M. aequa 8.0

M. subbotinae 8.9

400490 A. nitida 11.5 M. velascoensis 6.6 S. velascoensis 8.7

A. subsphaerica 15.4 M. acuta 7.1

A. soldadoensis 16.1 M. occlusa 7.4

M. pasionensis 8.0

M. aequa 8.7

M. subbotinae 10.3

490550 A. soldadoensis 20.6 M. velascoensis 7.7

M. subbotinae 15.8

N550 M. velascoensis 9.0

Table 3B

List of genera and species ranked in decreasing order of dissolution resistance, based on

the average CRPS from the start until the end of experiment 1. Note that the data

presented here are a combination of different size fractions from different samples.

Taxa name Number of specimens Time (hours) taxa persist Average CRPS

Acarinina 303 84 11.5

Morozovella 279 78 8.8

Subbotina 284 36 6.0

Igorina 90 22 4.3

A. soldadoensis 172 84 14.6

A. subsphaerica 61 56 11.4

M. subbotinae 97 56 11.7

A. nitida 70 42 7.5

M. aequa 30 32 8.4

M. velascoensis 52 34 7.8

M. pasionensis 30 30 6.8

M. occlusa 40 34 6.6

S. velascoensis 284 36 6.0

M. acuta 30 26 5.9I. pusilla 60 22 4.3

I. tadjikistanensis 30 22 4.2


7/31/2019 Nguyen Et Al - 2011

11/22


Subbotina and Igorina it took only 36, 34 and28 h, respectively. During

the experimental process, the relative remaining weight (RRW) and

the mean CRRW of Acarinina are highest, followed by those of

Morozovella, Subbotina and Igorina, respectively (Fig. 6 and Table 4).

As Acarinina has the highest original starting weight this finding is

expected.

Among Acarinina, the weight ofA. nitida dropped to zero after just

36 h of the experiment (Fig. 6). The weight of A. subsphaerica andA. soldadoensis decreased more slowly and they fully fragmented after

44 h and 58 h, respectively. This results in the low mean CRRW of

A. nitida in comparison with the other Acarinina species (Table 4). TheIgorina species dissolved at similar rates (Fig. 6), resulting in nearly

identical CRRW values (5.35.6), with I. albeari being slightly less

susceptible (Table 4). In contrast, Morozovella species show large

differences in weight loss. M. subbotinae was the most robust of the

Morozovella species investigated, taking longest to dissolve and with

highest RRW throughout the experiment. M. occlusa and M. acuta

dissolve most rapidly and have the lowest RRW. Consequently,


0 10 20 30 40 500

0,2

0,4

0,6

0,8

1

RPS

A. soldadoensis330-400 m, 198-1210B-19H-CC

A. soldadoensis330-400 m, 198-1212A-9H-CC

A. soldadoensis330-400 m, 143-865B-12H-1, 130-132

A. soldadoensis330-400 m, 143-865B-13H-1, 124-125

A

0 4 8 12 16 20 24 28 320

0,2

0,4

0,6

0,8

1

RPS

S. velascoensis330-400 m, 143-865B-13H-2, 50-52

S. velascoensis330-400 m, 198-1209B-23H-1, 63-64


C


B

0 4 8 12 16 20 24 28 32 36 40 44 48 52 560

0,2

0,4

0,6

0,8

1

RPS

M. subbotinae490-550 m, 198-1210B-19H-CC

M. subbotinae490-550 m, 143-865B-11H-4, 48-50

M. subbotinae400-490 m, 198-1210B-19H-CC

M. subbotinae400-490 m, 143-865B-11H-4, 48-50

Fig. 5. Examples of differential dissolution susceptibility of species (same size fraction)

among samples from different locations. In general, taxa from Site 865 show higher

susceptibility than the same taxa from Sites 1209, 1210 and 1212. A

A. soldadoensis insize fraction of330400 m fromZones E2-E3.BM. subbotinae in size fractionsof 400

490 m and 490550 m from Zone E3 (subzone P6a). CS. velascoensis in size fraction

of 330400 m from Subzone P4c.

0 2 4 6 8 10 12 14 16 18 20 22


0

0,2

0,4

0,6

0,8

1

I. pusilla I. tadjkistanensis

B


0 10 20 30 40 50 60 70 80 900

0,2

0,4

0,6

0,8

1

RP

S

RPS

RPS

A. nitidaA. soldadoensis A. subsphaerica

A

M. acuta

M. pasionensis

M. aequa M. occlusa

M. subbotinae M. velascoensis

0 10 20 30 40 50 600

0,2

0,4

0,6

0,8

1


C

Fig. 4. Differential reduction in the mean Relative Preservation Score with exposure

time between species that belong to the same genus. Note that values presented here

are combined from all size fractions and all samples for each taxon.


7/31/2019 Nguyen Et Al - 2011

12/22


M. occlusa and M. acuta have the lowest mean CRRW (4.7 and 5.2,

respectively), followed by M. pasionensis, M. velascoensis, M. formosa-

gracilis and M. aequa (6.7 to 7.9). M. subbotinae has the highest mean

CRRW (9.0), being twice as robustas themost susceptible M. occlusaand

M. acuta (Table 4 and Fig. 6).

3.3. Shell characteristics

3.3.1. Size

Results on the variation in size (largest diameter) between species

(Fig. 7A, Appendix B) show that most of the large species belong to the

genus Morozovella of which M. formosa-gracilis is the largest (average

size 480 m), followed by M. velascoensis (440 m) and M. subbotinae

(430 m). The mid-size group of taxa (300400 m) consists of theremaining Morozovella species, theAcarinina species and S. velascoensis.Igorina yields the smallest taxa, with average size ranges from 200 to

250 m.

3.3.2. Wall thickness

Ourmeasurements on wall thicknessof the last normalchamber of

each specimen reveal significant differences between the taxa

studied, both at genus and species level (Fig. 7B, Appendix C). Based

on the average thickness, these taxa can be divided into two main

groups: 1) the thick-walled group that exclusively consists of

Acarinina species, of which A. soldadoensis has wall thickness of

31 m (as well as the larger variation); A. subsphaerica and A. nitida

showing thickness of 24 m; 2) and a thin-walled group, of which

large Morozovella species such as M. aequa, M. subbotinae, andM. formosa-gracilis yield an average wall thicknesses of 1419 m.

The test walls of Igorina spp., S. velascoensis, and of the other

morozovellids are slightly thinner (913 m).

3.3.3. Shell porosity and pore size

Results indicate considerable dissimilarity in the proportion of

porosity on the external test surface (% pores) and pore sizes

between species (Fig. 7CD). S. velascoensis has the highest porosity

(23%) and pore size (N7 m2), but it also shows the largest variation.Acarinina species exhibit quite low and equable porosities, from 11%

in A. subsphaerica to 12% in A. nitida. The pore sizes of this group are

intermediate, which range around 5 m2. The difference in porosity

between Igorina species is also minor, varying from 13% in I. pusilla

to 15% in I. albeari and pore sizes range from 5 to 6 m2.Morozovellids yield porosities ranging from 9% (M. occlusa) to 16%

(M. velascoensis) and have the smallest pores (34 m2) (Fig. 7C and

Appendices D and E).

Table 4

List of genera and species ranked in decreasing order of dissolution resistance, based on

the mean CRRW during experiment 2.

Taxon Number of

specimens

Average original

weight (g)

Time (hours)

taxa persist

CRRW

Acarinina 2969 13.6 58 9.4

Morozovella 2696 12.2 36 7.0Subbotina 2912 7.9 34 6.0

Igorina 3263 6.2 28 5.5

A. soldadoensis 884 16.5 58 11.3

A. subsphaerica 1024 13.4 44 9.3

M. subbotinae 1097 15.2 36 9.0

M. aequa 68 14.7 32 7.9

A. nitida 1061 11.3 36 7.7

M. formosa-gracilis 59 17.1 34 7.1

M. velascoensis 269 14.6 30 6.9

M. pasionensis 45 11.4 28 6.7

S. velascoensis 2912 7.9 34 6.0

I. albeari 1209 7.1 28 5.6

I. pusilla 661 6.3 24 5.4

I. tadjikistanensis 710 5.7 22 5.4

I. broedermanni 683 5.5 20 5.3

M. acuta 1000 8.6 20 5.2

M. occlusa 158 7.6 18 4.7

%totalC

aCO

3

SUM

ofspe

cim

ens

M.occlu

sa

M.acuta

I.ta

djiki

stanen

sis

I.br

oede

rmanni

I.pu

silla

I.albe

ari

S.vel

asco

ensis

M.vel

asco

ensis

M.formos

a-gr

acilis

A.nitid

a

M.aeq

ua

M.subbo

tinae

A.sub

spha

erica

A.solda

doen

sis

M.pasio

nensis

50% lost of total CaCO3 and total number

of specimens of each species

0

4

8

12

16

20

24

28

32

36

40

44

48

52

56

60

0 5 0 10 00 5 0 10 00 5 0 1 000 5 0 1000 50 1 000 5 0 10 00 5 0 10 00 5 0 1 000 5 0 10 00 50 1 000 50 1 000 5 0 10 00 5 0 1 000 5 0 1000 5 0 10 00 7500150000 5 0 100

Time(hours)taxaexposureintoacidtreatment

Fig. 6. Differential weight loss among species by dissolution. Horizontal axis of thefi

rst graph indicates the percentage of total CaCO3 of all species and the second graphs shows thenumber of all specimens used in the experiment. Forthe othergraphs, horizontal axis shows the weight percentage of species. Data presented here resulted from the combination of

all size fractions of species, from different samples. Species are arranged from left to right in increasing order of Cumulative Relative Remaining Weight (CRRW). The subhorizontal

line connects the levels at which 50% of the CaCO3 and number of specimens are lost.


7/31/2019 Nguyen Et Al - 2011

13/22


3.3.4. Number of chambers in the last whorl

M. pasionensis has the highest number (on average 7.7 chambers) andS. velascoensis the least (3.5 chambers) (Fig. 7E, Appendix F). Igorina

species andM. occlusa, M. velascoensis, M. formosa-gracilis have more or less

the same number of chambers in the last whorl, ranging from 5.9 to 6.4.

Acarinina species and M. acuta, M. subbotinae and M. aequa have 4 to 5.1chambers in the last whorl. Thelargest variation in numberof chambers is

observed in M. velascoensis and M. occlusa, ranging from 5 to 8 chambers.

Taxa nameMean % pore

Standard deviation

C

0

5

10

15

20

25

30

Taxa name

Mean shell wall thickness

Standard deviation

B

0

10

20

30

40

Mean size of species

Standard deviation

0

100

200

300

400

500

600

A.solda

doen

sis

A.subs

phae

rica

A.nitid

a

M.aequ

a

M.subb

otina

e

M.pasio

nensis

M.form

osa-gra

cilis

M.velas

coen

sis

S.velas

coen

sis

I.albe

ari

I.tadjik

istan

ensis

I.pusilla

I.broed

erma

nni

M.occlu

sa

M.acuta

Taxa name

A

Mean pore area

Standard deviation

Taxa name

D

0

2

4

6

8

10

Taxa name

Mean number of chambers

Standard deviation

0

2

4

6

8E

SizeofSp

ecies(m)

Shellwallthick

ness(m)

Porosity(%pore)

Porearea(m

2)

Numberofchambers

A.solda

doen

sis

A.subs

phae

rica

A.nitid

a

M.aequ

a

M.subb

otina

e

M.pasion

ensis

M.velas

coen

sis

S.velas

coen

sis

I.albe

ari

I.tadjikist

anen

sis

I.pu

silla

I.br

oede

rman

ni

M.occlu

sa

M.acuta

M.form

osa-gracilis

A.solda

doen

sis

A.subs

phae

rica

A.nitid

a

M.aequ

a

M.subb

otina

e

M.pasio

nensis

M.form

osa-gracilis

M.velas

coen

sis

S.vela

scoensis

I.alb

eari

I.tad

jikist

anen

sis

I.pusill

a

I.br

oede

rman

ni

M.occlu

sa

M.acuta

A.solda

doen

sis

A.subs

phae

rica

A.nitid

a

M.aequ

a

M.subb

otina

e

M.pasio

nensis

M.form

osa-gracilis

M.velas

coen

sis

S.velas

coen

sis

I.alb

eari

I.tadjikist

anen

sis

I.pusilla

I.br

oede

rman

ni

M.occlu

sa

M.acuta

A.solda

doen

sis

A.subs

phae

rica

A.nitid

a

M.aequ

a

M.subb

otina

e

M.pasio

nensis

M.form

osa-g

racilis

M.velas

coen

sis

S.velas

coen

sis

I.albe

ari

I.tadjikist

anen

sis

I.pusilla

I.broe

derm

anni

M.occlu

sa

M.acuta

Fig. 7. Shell characteristics among species: Athesize of species is measured as thelargest diameter of individual specimens;Bthewall thicknessis measured on thelast preserved

chamber of specimens; Cshell porosity is the percentage of total pore area on the external surface of the specimens (% pore); Dthe average size of a pore on the outer surface is

calculated as the total surface area of pores divided by the number of the pores and Ethe number of chambers in the last whorl.


7/31/2019 Nguyen Et Al - 2011

14/22


3.4. Relationships between weight and other shell parameters

Our correlation analysis reveals a high value (0.87) between the

weight and the size of the species (Table 5). For weight and wall

thickness this value is slightly lower (0.74). The correlation values

between the original weight and the number of chambers in the last

whorl, the average pore area and the porosity are negative and much

lower (0.43, 0.31 and 0.12, respectively). When the originalweights of thespecies areplotted vs.average size andwall thickness it

is shown that the larger species are the heavier ones. In most cases,

these are also the ones with the thicker test walls of the last chambers

(Fig. 8AB).

A regression analysis for 15 species was carried out (Table 6) in

which the original weight of species was considered as a

dependent variable, while the shell wall thickness and the mean

size were considered as independent variables. The results indicate

a coefficient of 0.234 between the weight and the shell wall

thickness and of 0.036 between the weight and the size. Based on

these values, the regression for the weight of species and their shell

wall thickness and size can be formulated as below. R2 for this

regression analysis is very high (0.923), indicating that more than

Table 5

Correlation between CRRW, CRPS and shell parameters. Correlation values are given in the lower triangle of the matrix, and the probabilities that these parameters are uncorrelated

to each other are given in the upper part.

CRRW CRPS Wall thickness Original weight Size Porosity Pore area Number of chambers

CRRW 0.00 0.00 0.00 0.04 0.51 0.68 0.13

CRPS 0.90 0.00 0.00 0.00 0.33 0.46 0.19

Wall thickness 0.95 0.83 0.01 0.12 0.38 0.56 0.08Original weight 0.85 0.94 0.74 0.00 0.71 0.33 0.16

Size 0.60 0.76 0.47 0.87 0.53 0.22 0.30

Porosity 0.21 0.31 0.28 0.12 0.20 0.44 0.76

Pore area 0.13 0.24 0.19 0.31 0.38 0.24 0.27

Number of chambers 0.46 0.41 0.52 0.43 0.33 0.10 0.35

Original weight

Linear (Original weight)

0

5

10

15

20

Originalweight(g)

y = 0.0419x - 3.0491

R2= 0.8075

100 200 300 400 500

A

Average size of species (m)

Originalweight(g)

Shell wall thickness (m)

Original weight

Linear (Original weight)

B

y = 0.3991x + 4.8851

R2= 0.4146

5 10 15 20 25 30 3500

5

10

15

20

Fig. 8. The relationship between the original weight of species and their average size

(A), and wall thickness (B). R

2

value is very high in A and much lower in B,demonstrating that a larger part of thevariation in theweight of species is explained by

the linear relation with their size and a smaller part of this variation is explained by the

linear relation with wall thickness.

Table 6

Multiple-regression results.

1. Dependent variable: original weight.

Independent variables: wall thickness, size.

N: 15

Multiple R: 0.966

Multiple R2: 0.933

Multiple R2 adjusted: 0.923

ANOVA:

F: 83.986

p: 8.788E-08

Coeff. Std. err. t p

Constant 4.449 1.226 3.630 0.003Wall thickness 0.234 0.049 4.786 0.000

Size 0.036 0.004 9.659 0.000

Weight=4.449+(0.234 wall thickness)+(0.036 size)

2. Dependent variable: CRPS.

Independent variables: wall thickness, size

N: 12

Multiple R: 0.927

Multiple R2: 0.860


ANOVA

F: 27.436

p: 1.480E-04

Coeff. Std.err. t p

Constant 2.609 1.742 1.497 0.169

Wall thickness 0.294 0.058 5.088 0.001

Size 0.018 0.005 3.324 0.009

CRPS=2.609+(0.294 wall thickness)+ (0.018 size)

3. Dependent variable: CRRW.

Independent variables: wall thickness, size

N: 15

Multiple R: 0.954

Multiple R2: 0.910


ANOVA

F: 60.259

p: 5.514E-07

Coeff. Std. err. t p

Constant 1.513 0.644 2.348 0.037

Wall thickness 0.227 0.026 8.807 0.000Size 0.006 0.002 3.120 0.009

CRRW=1.513+(0.227 wall thickness)+(0.006 size).


7/31/2019 Nguyen Et Al - 2011

15/22


90% of the variation in the original weight of taxa is explained by

this linear regression.

Weight of taxa = 4:449 + 0:234shell wall thickness

+ 0:036size

3.5. Relationships between dissolution susceptibility and shell parameters

In order to examine the relationship between dissolutionsusceptibility and shell parameters, we analyzed the correlation

between CRPS, CRRW and the shell parameters (Table 5). Results

show that CRPS and CRRW are closely correlated (0.90). Wall

thickness has a strong bearing on both CRPS and CRRW as indicated

by the high correlation values of 0.83 for CRPS and 0.95 for CRRW.

Indeed, the distinct linear relationship between wall thickness and

both preservation scores is shown in Fig. 9AB.

The original weight of the taxa also shows a strong correlation

with both CRRW, CRPS (0.85 and 0.94, respectively). In all cases, the

heavy species are most dissolution resistant, as indicated in Fig. 9CD.

In comparison with wall thickness and original weight, the average

size of taxa expresses a slightly weaker relationship with CRRW and

CRPS (0.60 and 0.76 respectively). Even though, smaller species inmost cases are more susceptible to dissolution than the larger species

(Fig. 9EF).

CRPS

A

Shellwallthic

kness(m)

2 4 6 8 10 12 14 16 18

0

10

20

30

40

y = 1.891x + 0.8423

R2

= 0.6717

Mean wall thickness

Linear (mean wall thickness)

Standard deviation

CRRW

B

0

10

20

30

40

Shellwallthickness(m)

4 6 8 10 12

y = 3.2854x - 7.7508

R2

= 0.8361

Mean wall thickness

Linear (mean wall thickness)

Standard deviation

4 6 8 10 12 14 16 18

CRPS

0

5

10

15

20

Originalweight(g)

Originalweight(g)

y = 1.0393x + 2.8589

R2

= 0.7532

2

C

y = 1.7945x - 1.5467

R2

= 0.6493

4 6 8 10 12

CRRW

0

5

10

15

20D

2 4 6 8 10 12 14 16

CRPS

0

100

200

300

400

500

600

Sizeofspecies(m

)

Sizeofspecies(m

)

E

y = 16.939x + 200.83

R2

= 0.4561

Mean size

Linear (mean size)

Standard deviation

y = 27,269x + 143,1

R2

= 0,3262

4 6 8 10 12

CRRW

0

100

200

300

400

500

600F

Mean size

Linear (mean size)

Standard deviation

Fig. 9. Cumulative Relative Preservation Scores (CRPS) and Cumulative Relative Remaining Weight (CRRW) plotted vs. wall thickness (figures A and B), initial weight (figures C and

D) and size (figures E and F). R2

values are very high in figures A, B, C, D and lower in figures E and F, indicating that a major part of variation in CRPS and CRRW is explained by thelinear relations with wall thickness and weight while a smaller part of the dissolution indicators is explained by the linear relations with size.


7/31/2019 Nguyen Et Al - 2011

16/22


The number of chambers in the final whorl presents a weak,

negative correlation with the CRPS and the CRRW (0.41 and0.46,

respectively). Similarly, test porosity and pore size indicate a very

weak and negative relationship, yet with high un-correlation values

for the CRPS and the CRRW (upper part of Table 5). There are no

significant differences in these shell parameters between resistant

species and vulnerable species, as indicated in Fig. 10AF.

To quantify the effects of shell parameters on the dissolutionsusceptibility of taxa, as well as to describe the relationship between

them, we carried out multi-regression analyses, in which the CRPS

and the CRRW are considered as dependent variables. Wall thickness

and average size are the largest controls on the CRPS and CRRW and

are considered as independent variables. We do not take the original

weight into this analysis since this parameter is derived from the

combination of wall thickness and size, as mentioned above. The

results show that 83% of CRPS and 90% CRRW are explained by the

combination between these two structural parameters. Regression

coefficient values between the shell wall thickness and the CRPS,CRRW are 0.294 and 0.227, respectively. The corresponding values

between the mean size and the solution preservation indicators are

y = -0.2959x + 15.551

R2= 0.0691

2 4 6 8 10 12 14 16

CRPS

0

5

10

15

20

25

30

Porosity(%

pore)

Porosity(%

pore)

A

Mean porosity

Linear (mean porosity)

Standard deviation

0

5

10

15

20

25

30

CRRW

4 6 8 10 12

B

y = -0.3797x + 15.816

R2= 0.0493

Mean porosity

Linear (mean porosity)

Standard deviation

Poressize(m

2)

Poressize(m

2)

2 4 6 8 10 12 14 16

CRPS

10

C

0

2

4

6

8

y = -0.0826x + 4.8789

R2= 0.0337

Mean pores size

Linear (mean pores size

Standard deviation

CRRW

D

4 6 8 10 120

2

4

6

8

10

y = -0.0684x + 4.7184

R2

= 0.0088

Mean pores size

Linear (mean pores size)

Standard deviation

2 4 6 8 10 12 14 16

CRPS

0

2

4

6

8

Numberofchambers

E

y = -0.1614x + 6.5061

R2

= 0.1793

y = -0,2744x + 7,272

R2

= 0.2113

4 6 8 10 12

CRRW

0

2

4

6

8

Numberofchambers

F


Linear (mean number of chambers)

Standard deviation


Linear (mean number of chambers)

Standard deviation

Fig. 10. Dissolution indicators (CRPS and CRRW) vs. porosity (figures A and B), pore size (figures C and D) and number of chambers in the final whorl (figures E and F). At all graphs,

regression lines are nearly parallel to X-axis and R2 values are very low, demonstrating that there are no general relations between the dissolution susceptibility of species and their

porosity, pore size or number of chambers.


7/31/2019 Nguyen Et Al - 2011

17/22


0.018 and 0.006 (Table 6). Linear regression formulas between theCRPS, CRRW and these shell parameters can be expressed as follows:

a/ CRPS=2.609+ (0.294 wall thickness)+ (0.018 size)

b/ CRRW=1.513+(0.227 wall thickness)+(0.006 size)

4. Discussion

4.1. Relative ranking of dissolution susceptibility

We have constructed a dissolution ranking for some common late

Paleocene to early Eocene planktic foraminiferal species. Similarity in

the ranking schemes derived from both experiments, shown by the

high correlation and R2 values between CRPS and CRRW (0.90, Table 5

and 0.89, Fig. 11), highlights the reliability of our proposed scheme.

Accordingly, the large and muricateAcarinina is most robust, followed

by Morozovella. Thecancellate Subbotina is intermediate in dissolution

susceptibility and the small muricate Igorina is the most dissolution

vulnerable genus.

At species level, in both experiments, Acarinina species show a

consistent ranking: in ascendingorderof dissolutionsusceptibility these

are:A. soldadoensis,A. subsphaerica andA. nitida. For Igorina, both results

show similar susceptibility for I. pusilla and I. tadjikistanensis. AlthoughI. albeari and I. broedermanni were not assessed in the first experiment,

their similar CRRW values indicate that there is a little difference in

dissolution susceptibility between the Igorina species.

Within Morozovella, we observed an overall similar dissolution

ranking in both experiments. Accordingly, M. subbotinae and M. aequa

are the most resistant species. Morozovella formosa-gracilis was not

examined in the shell destruction experiment, nevertheless theresemblance in CRRW between this species, M. pasionensis and M.

velascoensis suggests that these three species have the same interme-

diate dissolution robustness. The two smaller taxa, M. acuta and M.

occlusa show higher dissolution susceptibility in both assessments.

Among the species combined, the thick-shelled A. soldadoensis, A.

subsphaerica and the strongly muricate M. subbotinae indicate high

dissolution robustness. All the other large Morozovella species,

together with A. nitida show intermediate dissolution susceptibility.

The cancellate S. velascoensis, the small muricate Igorina species and

the two small, thin-walled M. acuta and M. occlusa reveal high

dissolution vulnerability in both assessments.

4.2. Relationship between dissolution resistance and weight

Our results show that lighter species are dissolved more rapidly

than heavier ones (Fig. 9CD and Table 4). As heavier species consist

of more calcite, it could be expected that dissolution media need a

longer time to break down the shells. This demonstrates that

progressive dissolution results in the relative enrichment of heavier

species (and specimens) within the assemblage.

The observed relationship between dissolution resistance of taxa

and their weight is in good agreement with in-situ studies on

selective dissolution of modern planktic foraminifera (Berger,1967). It is also a reasonable explanation for the relative enrichment

of Acarinina and Morozovella in partially dissolved Paleogene

planktic foraminiferal assemblages in records from various locations

around the Tethys (e.g. Canudo et al., 1995; Arenillas and Molina,

1996; Berggren and Ouda, 2003; Luciani et al., 2007) and ODP sites

(e.g. Basov, 1995; Kelly, 2002; Lyle et al., 2002). The direct

relationship between the weight of taxa and their dissolution

susceptibility in our controlled laboratory environment is clear. In

natural environments, in which the time that taxa are exposed to

dissolution agents is not controlled, it can be expected that this

connection could even be stronger. Generally, dissolution of

foraminiferal tests is considered to be a taphonomic process (e.g.

Martin and Liddell, 1991), but shell loss from dissolution already

starts in the water column from partial resorption of the plankticshells and is mediated through biodegradation at the sea floor (e.g.,

Ruddy, 1997). Because lighter shells (or taxa) generally descend

more slowly to the seafloor, they are exposed longer in the water

column and thus, are more prone to early dissolution (e.g. Martin

et al., 1995; Schiebel et al., 2007). This process provides a

taphonomic filter and leads to a relative increase of heavy speci-

mens and taxa in sediments.

4.3. Wall thickness and size dependency

Our experiments demonstrate that the dissolution susceptibility of

planktic foraminiferal species strongly depends on their wall thickness

(Table 5 and Fig. 9AB), and that planktic assemblages becoming

relatively enriched in thick-walled individualsas dissolutionprogresses.

Within a species, larger specimens are often heavier and possess

thicker test walls and as a result, they are the more dissolution

resistant. This is not directly applicable between species, because

species can be larger, but also thinner than others. In this case, the

overwhelming role of shell wall thickness on the dissolution

resistance of taxa will be weakened and size can become a more

dominant control on dissolution susceptibility. For example the shells

ofM. subbotinae and M. aequa are thinner than A. nitida but these two

Morozovella species are more dissolution resistant. The difference in

size between these three taxa is a reasonable explanation for this

discrepancy as the Morozovella species are generally larger thanA. nitida (Fig. 7A). This resulted in less original weight of A. nitida

(Table 4), and consequently, less dissolution resistance of A. nitida in

comparison with M. subbotinae and M. aequa.

The close link between wall thickness, size fraction and solutionsusceptibility of taxa in our experiment is supported by previous

laboratory studies (Nguyen et al., 2009) as well as by in-situ studies,

based on both modern planktic (Berger, 1967; Berger, 1970; Conan

et al., 2002) and benthic (Corliss and Honjo, 1981) foraminiferal

assemblages. Our findings are also in line with quantitative distribu-

tion data from ancient sediments (e.g. Sliter, 1995) and modern

deposits (e.g., Sliter, 1975; Conan et al., 2002). In those records, thin-

walled and small planktic and benthic foraminifera were documented

as more susceptible to carbonate dissolution than other thick-walled

and larger taxa.

In addition, the size and shell-thickness of species can vary in time

and space, depending on the surrounding environmental parameters,

such as water depth, chemical composition of water column [CO32],

temperature and light, (B and Hemleben, 1970; Berger, 1971;Anderson, 1975; Schiebel et al., 2007). This variation leads to

differences in the weight of species between different stratigraphic

CRRW

Linear (CRRW)

CRRW

CRPS

4

6

8

10

12

4 8 12 14 16 182 6 10

y = 0.5958x + 2.4045

R2= 0.8879

Fig. 11.CumulativeRelative Remaining Weight(CRRW) plotted vs. Cumulative Relative

Preservation Score (CRPS). R2 value is very high, indicating a high agreement in

dissolution ranking for species within the two experiments.


7/31/2019 Nguyen Et Al - 2011

18/22

7/31/2019 Nguyen Et Al - 2011

19/22


observed that the ELPE assemblages are made up mainly of heavily

overgrown igorinids whereas the other genera are almost absent in

the assemblages. Outside the ELPE, igorinids are less common and

show no shell thickening. The unusual dominance of thick-walled

igorinids within the ELPE might reflect their facility of adapting to

changes in water masses circulation (change in deep water circulationand/or variation in surface water productivity) and in carbonate

saturation (see Shipboard Scientific Party, 2002; Petrizzo, 2005). This

record indirectly emphasizes the importance of assessing the wall

thickness in the determination of the dissolution susceptibility of taxa,

as observed in our experiments.

4.6. Implications for dissolution estimates in natural environments

4.6.1. Experimental vs. natural dissolution

In natural environments, dissolution of foraminiferal assemblages is

controlled by two groups of factors: 1) factors concerning the shell of

taxa (surface/volume ratio, wall thickness, size, porosity, morphology),

and 2) factors concerning the surrounding environment prior to, during

and after deposition (physico-chemical parameters of sea-water,interstitial water or ground water, biological productivity and activity,

sedimentation conditions; e.g. Conan et al., 2002; Herrero and Canales,

2002). A complex interaction between these factors results in

differences between natural dissolution and dissolution in a controlled

experimental setup, in which only shell parameters and exposure time

are considered. Yet, the close agreement between our experimental

results and in-situ experimental results (e.g. Berger, 1967, 1970) aswell

as natural quantitative/qualitative records indicative of dissolution

(Sliter, 1975; Petrizzo et al., 2008) suggest that our experimentsaccurately mimic a combination of natural processes, and thus the

experimental results have a strong bearing on the interpretation of

foraminiferal dissolution in natural environments.

4.6.2. Weight and dissolution ranking scheme for planktic foraminiferal

assemblages

Since it is our overall aim to develop a dissolution index that can

easilybe employed or adapted to anyquantitative foraminiferal study,

we need parameters that satisfy the following requirements: 1) a

strict relationship with the differential dissolution susceptibility of

taxa, and 2) applicability in a routine procedure of quantitative

foraminiferal analysis. The close connection between dissolution

resistance of species and weight, wall thickness and size indicatesthatthese are prime parameters for the evaluation of dissolution in

foraminiferal assemblages.

We apply the formulas presented in Sections 3.4 and 3.5 to the

measured shell wall thickness and size of early Paleogene species, to

see how the weight, CRPS and CRRW of species calculated from these

formulas fit with measured data. Despite the very good agreement

between CRPS and CRRW, we prefer to use the formula developed for

CRRW (3.5.b) rather than that for CRPS (3.5.a), because CRRW is

gained from the experiment in which all the specimens of each

species were used, therefore it is more representative for the

composition of the assemblage than the CRPS. In these formulas,

only wall thickness and size are included. The reason for this selection

is that these parameters satisfy our two requirements listed above.

Even though the weight of species directly controls the dissolution

resistance of taxa, we do not include this parameter in the regression

formulas of CRPS and CRRW because the weight of taxa is directly

derived from the combination of wall thickness and size (Tables 5, 6

and Fig. 8). Consequently, if weight would be included in these

formulas, the role of wall thickness and size would be doubled.

Results from the application of the regression formulas 3.4 and

3.5.b indicate good agreement between the measured and the

calculated weight and the CRRW (Table 7 and Fig. 12). However, a

discrepancy is observed for S. velascoensis, as the calculated weight is

much higher than the measured weight (Fig. 12A). This result is

expected considering the size and the wall thickness ofS. velascoensis

as discussed in Section 4.4. The measured values for CRRW are in line

with the calculated values, except for M. subbotinae and A. nitida

(Fig. 12B). The calculated CRRW in M. subbotinae is about 15% lower

than the measured CRRW and vice versa in A. nitida. The average sizeofA. nitida is 285 m, only slightly bigger than the average size of thesmallest group (Igorina species) but it has a thick wall (~24 m,

(Appendices B and C). These values in M. subbotinae are 428 m and16.5 m, respectively. It seems with those two species, that the sizeand the wall thickness affect the CRRW in a way that differs from the

other species and our regression formula could not accurately take

this into account. Our regression formula seems to provide an

overestimation for the role of wall thickness and an underestimation

for size in these particular examples.

The negative constants in formulas (see Sections 3.4 and 3.5.a)

demonstrate that our regression formulas can be applied only within

certain boundary conditions. At the condition that values in weight,

CRPS and CRRW are 0 and that wall thickness is at least 1 m, we

find that our regression formulas arevalid when thesize of a species is125 m. This is also the size fraction that we used in ourexperiments.

A

A.solda

doen

sis

A.subs

phae

rica

A.nitid

a

M.aequ

a

M.subb

otina

e

M.pasio

nensis

M.form

osa-gra

cilis

M.velas

coen

sis

S.velas

coen

sis

I.albe

ari

I.tadjikist

anen

sis

I.pusilla

I.broed

erma

nni

M.occlu

sa

M.acuta

A.solda

doen

sis

A.subs

phae

rica

A.nitid

a

M.aequ

a

M.subb

otina

e

M.pasio

nensis

M.form

osa-gracilis

M.velas

coen

sis

S.velas

coen

sis

I.albe

ari

I.tadjikist

anen

sis

I.pusilla

I.broe

derm

anni

M.occlu

sa

M.acuta

Weight(g)

Taxa nameMeasured weight

Calculated weight

0

5

10

15

20

Measured CRRW

Calculated CRRW

Taxa name

CRRW

B

0

4

8

12

Fig. 12. Measured data vs. calculated data on weight (A) and Cumulative Relative

Remaining Weight (B). Species are arranged from left to right in decreasing order of

measured wall thickness. The difference between the measured data and the calculated

data is generally small, pointing to the robustness and reliability of our regression

formulas on the weight and the dissolution ranking schemes of species.


7/31/2019 Nguyen Et Al - 2011

20/22


The discrepancies between the measured and the calculated data

in weight and CRRW underline the fact that our formulas for

calculating weight and the dissolution ranking of taxa should be

employed only after understanding the shell parameters of taxa.

Overall, however, the excellent agreement between the calculated

and the measured values in weight and CRRW indicates that the

dissolution susceptibility ranking of foraminifera can be determinedusing the experimentally developed regression formulas.

4.6.3. Application of dissolution ranking formula

So far, most of our knowledge on the dissolution susceptibility of

planktic foraminifera is restricted to recent taxa (e.g. Berger, 1967,

1970; Sliter et al., 1975). Understanding the dissolution resistance of

extinct planktic species in older, especially lithified sediments is

limited (Malmgren, 1987; Petrizzo et al., 2008). Our findings can be

applied to elucidate the taphonomic overprint related to partial

dissolution of Paleogene planktic foraminiferal assemblages. For

instance, common observations across the Paleocene/Eocene bound-

ary at ODP sites, such as increased relative abundance of species ofAcarinina and Morozovella, increased fragmentation of Subbotina

together with low P/B ratio, low foraminiferal numbers, and lowrelative abundances ofSubbotina species (e.g. Kelly, 2002; Kaiho et al.,

2006; Petrizzo, 2007) likely indicate partial dissolution. Thesefindings

may be extrapolated to different environmental settings or time

intervals characterized by different taxa with similar test

characteristics.

In most quantitative foraminiferal studies, wall thickness and size

of the common taxa are not routinely determined. Yet, as these

parameters fulfill our requirements of easy collection and robust

statistical relationships, a dissolution susceptibility ranking for

benthic and planktic foraminiferal assemblages from different

stratigraphic intervals can be achieved in relatively limited time.

This enables a dissolution quantification based on changes in relative

abundance of dissolution resistant taxa (having thick-walled and

large size) in any quantitative foraminiferal study.

5. Conclusions

We propose a dissolution susceptibility ranking for common

PaleoceneEocene planktic foraminiferal species, based on the

differential destruction and weight loss of these taxa through

dissolution. Among these species, the thick-walled A. soldadoensis

and A. subsphaerica and the large M. subbotinae are the most resistant

species. Most of the large Morozovella species such as M. aequa,

M. formosa-gracilis, M. velascoensis and M. pasionensis, together withA. nitida have intermediate dissolution susceptibility. Small, muricate

Igorina species, together with the cancellate S. velascoensis and the

thin-walled M. acuta and M. occlusa are the most dissolution

vulnerable species.We found a strict dependency of dissolution resistance on weight,

wall thickness and size. The agreement between our experimental

results and in-situ experimental results on recent foraminifera as well

as natural quantitative/qualitative records suggests that our experi-

ments well reflect natural dissolution processes. Consequently, our

results provide importantinsight into the impact of partial dissolution

on foraminiferal assemblages in natural environments, especially

during Paleogene hyperthermal events that are often associated with

dissolution phenomena.

A formula for assessing dissolution resistance of individual taxa is

proposed. Application of this formula shows a good agreement

between the calculated and the measured dissolution resistance,

indicating its robustness and reliability. With this approach, based on

the wall thickness and size, dissolution susceptibility assessments forforaminiferal assemblages from different areas and stratigraphic

intervals can be achieved. A proper assessment of taphonomic

alteration by dissolution should be part of every paleoenvironmental

reconstruction based on quantitative foraminiferal records.

Acknowledgments

The paper greatly benefited from constructive comments by two

anonymous reviewers and useful suggestions by the Editor FransJorissen. This research used samples provided by the Ocean Drilling

Program (ODP) sponsored by the U. S. National Science Foundation

(NSF) and participating countries under management of Joint

Oceanographic Institutions (JOI), Inc. We thank the Research

Foundation Flanders (FWO G.0422.10) and the K.U.Leuven

Research Fund forfinancial support to R.P. Speijer. Financial support of

MIUR-PRIN 2007 to M.R. Petrizzo is acknowledged.

Appendix A. Supplementary data

Supplementary data to this article can be found online at doi:10.

1016/j.marmicro.2011.07.001.

ReferencesAgnini,C., Macri,P., Backman, J.,Brinkhuis, H.,Fornaciari,E., Giusberti, L.,Luciani,V., Rio,

D., Sluijs, A., Speranza, F., 2009. An early Eocene carbon cycle perturbation at52.5 Ma in the Southern Alps: chronology and biotic response. Paleoceanography24 (PA2209). doi:10.1029/2008PA001649.

Anderson, J.B., 1975. Factors controlling CaCO3 dissolution in the Weddell Sea fromforaminiferal distribution patterns. Marine Geology 19 (5), 315332.

Arenillas, I., Molina, E., 1996. Biostratigrafa y evolucon de las asociaciones deforaminiferos planctnicos del transito PaleocenoEoceno en Alamedilla (CordillerasBtias). Revista Espanola de Micropaleontologica 18 (1), 7596.

Basov,I.A., 1995. Paleogene planktonicforaminifer biostratigraphy of Sites 883 and 884,Detroit Seamount (subartic Pacific). Proceedings of the Ocean Drilling Program,Scientific Results, 145 Available from World Wide Web:http://www-odp.

nguyen et al - 2011

Documents