Quantification sinking velocities of individual coccoliths
will contribute to optimizing laboratory methods for separating coccoliths of
different sizes and species for geochemical analysis. The repeated
settling–decanting method was the earliest method proposed to separate
coccoliths from sediments and is still widely used. However, in the absence
of estimates of settling velocity for nonspherical coccoliths, previous
implementations have depended mainly on time-consuming empirical method
development by trial and error. In this study, the sinking velocities of
coccoliths belonging to different species were carefully measured in a series
of settling experiments for the first time. Settling velocities of modern
coccoliths range from 0.154 to 10.67 cm h
Coccolithophores are some of the most important phytoplankton in the ocean.
They can secrete calcareous plates called coccoliths, which contribute
significantly to discrete particulate inorganic carbon in the euphotic zone
and to
Temporal and spatial distribution of samples.
Both decanting and micro-filtering are widely used methods for coccolith
separation. The micro-filtering method separates coccoliths with a polycarbonate
micro-filter membrane (with pore sizes of 2, 3, 5, 8, 10 and
12
In the current study, we present a novel and rigorous estimation of sinking
velocity for 16 species of modern and Cenozoic coccoliths, carefully measured
in 0.2 % ammonia at 20
We measured the sinking velocity of 16 different species of coccoliths,
isolated from eight deep-sea sediment samples from the Pacific and Atlantic
oceans (Fig. 1, Table A1 in Appendix). Sample were principally of Quaternary
age but include two Neogene/Paleogene samples. In general, numbers of small
coccoliths, including
Schematic of settling experiments.
The sinking velocity measurement depends on absolute abundance estimation
(more details in Sect. 2.2.2). However, on microscope slides, larger
coccoliths and foraminifer fragments may cover smaller coccoliths, reducing
the accuracy of coccolith absolute numbers. Thus, before sinking experiments
were carried out, raw sediments were pretreated to purify the target
coccoliths to reduce errors in coccolith counting. The raw sediments were
disaggregated in 0.2 % ammonia and sieved through a 63
The shape parameters of vessels.
We are not aware of any prior direct determination of the sinking velocity of
individual coccoliths, although the sinking velocities of live
coccolithophores and other marine algal cells have been successfully measured
by the FlowCAM method (Bach et al., 2012) or a similar photography
technique (e.g., Miklasz and Denny, 2010). Here, we introduce a simple method
to measure the particle sinking speeds without special equipment.
After pretreatment, the coccolith suspensions were gently shaken and then
moved into comparison tubes which were vertically mounted on tube shelves. We
set the timer going and let the suspension settle for a specified period of
time, marked as sinking time or settling duration ( Thereafter, we removed the upper 15 mL supernatant into a 50 mL centrifuge
tube with a 10 mL pipette. This operation was performed slowly and gently to
avoid drawing lower suspensions upward. The absolute counting of coccolith was
achieved by using the “drop technique” to make quantitative microscope
sides (Koch and Young, 2007; Bordiga et al., 2015). In total, 0.3 mL mixed suspension
was extracted with pipettes onto a glass cover, and the slider was dried on a
hotplate. The lower suspension was than homogenized and another slider was
prepared as described above. The number of coccoliths in the upper and lower suspensions were carefully
counted on microscope at
To calculate the sinking velocities of coccoliths, we define a parameter
named the separation ratio (
The influence of temperature on sinking velocity. Density data are from Kell (1975), and viscosity data are from Kestin et al. (1978).
The separation ratio,
Sinking velocities of
There are still two issues to be explained. Firstly, to eliminate the shape
differences among vessels, all separation ratios have been transferred to
calibrated separation ratios (
Seven commonly used vessels were selected to detect the potential influence
of vessels (Fig. 3). Two of them are made of plastics (no. 2 and no. 3 in
Fig. 3), and all others are pyrex glass vessels. About 500 mg of sediment
from core KX21-2 were pretreated as described in Sect. 2.2.1 and suspended in about
500 mL diluted ammonia. After that, settling experiments were performed as
described in Sect. 2.2.2 using different vessels. In these experiments, only the
dominant species,
The sinking velocity and shape–velocity factor of different
coccolith species:
Temperature can change the density and viscosity of liquid. Generally
speaking, the higher the temperature is, the lower the density and viscosity
will become and the faster pellets will sink. Take water, for instance: if the
temperature increases from 15 to 30
The calibrated ratio (
The calibration of sinking velocity in high-concentration suspension has
been calculated by Richardson and Zaki (1954):
Coccolith sinking velocities and coccolith shape factors.
The sinking velocities of
However, our experiments were premised on the basis that the concentration of
suspension was equal among different vessels. This means that large vessels
can treat more sediment at one time, but if we choose a larger vessel, we need to spend more time in pumping suspensions, and it often costs more time in terms of sinking (often
due to longer sinking distance). Assuming that the sediment is composed of
50 % calcite (with a density of 2.7 g cm
The selection of separation velocities: the sinking velocities of
three main coccolith species in sample from core KX21-2 were calculated by
the length distribution and velocity factors in Table 2. The yellow dots
represent sinking velocities of coccoliths with mean length. The edge of
boxes show the sinking velocities of coccoliths within 1 standard deviation
of length (
We measured the separation ratios of different coccoliths in comparison tubes
at 20
Generally speaking, the sinking velocities of coccoliths increase with distal shield length (Fig. 5a), as expected from the increase in volume to sectional area for a given geometry as length increases. Our data imply that the sinking velocity has a power function relationship with distal shield length.
We propose that the sinking velocity of coccoliths might have a quadratic
relationship with distal shield length as described by Stokes' law (Fig. 6a).
If we use data for all species except
To improve coccolith separation by settling methods, we measured sinking
velocities of different coccoliths by gravity. Sinking velocities in this
study varied from 0.154 to 10.61 cm h Measure the length of coccoliths in your target assemblage under the
microscope and regress the length distribution by the assumption of a normal
distribution (details are in Appendix C). Estimate sinking velocities for each important species. For species
whose sinking speed has been directly measured, we can use the length–velocity
factor directly ( Calculate the separation time for the main species. For example, in KX21-2 there
are three main coccoliths ( Homogenize the sediment suspension and let coccoliths settle for the period
calculated in Step 3. After that, pump out the upper part of the suspension. In
the upper part, we exclusively have the smaller of the main coccoliths.
However, the column will still contain some smaller ones. So this step (settling
and pumping) should be repeated until the lower part no longer has any significant contribution from the smaller coccoliths. This step has been described well in previous studies, and more details can be found in Stoll and
Ziveri (2002) and Bolton et al. (2012).
We find that, if we use the general formula, a closed central area coccolith will
sink faster than predicted (
The sinking velocities and coccolith length results can be found in Table 2.
Sample selections. SCS represents the South China Sea; W. P. represents the western Pacific; N.A. represents the northern Atlantic.
Imaged of coccoliths measured in this study:
To measure the distal shield length of coccoliths, pictures were taken at a
magnification of 1250
Size distribution of coccoliths measured in the present study. The abbreviations of coccolith names follow Table A1.
The classification of coccoliths by length was supported by mixture analysis
in PAST (Hammer et al., 2001), such as
The classical classification of
The short axis and long axis length distribution of coccoliths in Fig. 6d.
In this part, the derivation of the equation will be explained in detail including proofs of several assumptions mentioned in the Materials and methods section.
When the well-mixed sediment begins to sink, the decrease in coccolith number in the upper suspension (
Through integrating Eq. (D1), we can get the variation in coccolith
number in the upper column over time:
Combining the Eqs. (D1), (D2), (D3) and (D4), we obtain the relationship
between the separation ratio,
The comparison tubes used in our experiments have the same
The coccoliths' lengths in the sediment have some variations. So what we
measured is actually the bulk settling velocity of the whole coccolith
population. We also offer a test for the assumption that the average sinking
velocity of all coccoliths can be treated as the sinking velocity of
coccoliths with the average length. Here, we used the data of
For
The simulations of coccoliths settling with different lengths:
The errors of the measured separation ratio (
The error distribution with different
This study was conceived by HZ and CL. Measurements and calculations were conducted by HZ. HZ, HS and CB wrote the paper with the help from XJ and CL.
The authors declare that they have no conflict of interest.
This study was supported by grants from the Chinese National Science Foundation (91428310, 91428309 and 41530964, to Chuanlian Liu) and ETH Zurich (to Heather Stoll). It was also supported by Chinese Scholarship Council (CSC) scholarship to Hongrui Zhang. We thank the Integrated Ocean Drilling Program (IODP) for providing the samples. The IODP is sponsored by the US National Science Foundation and participating countries under management of IODP Management International, Inc (IODP-MI). We thank Zhimin Jian for providing the sample of the core ODP 807. We thank three anonymous reviewers as well as the editor for their comments and suggestions, which helped us to improve the original version of the paper. Edited by: Lennart de Nooijer Reviewed by: three anonymous referees