Hydrobiologia 517: 1–13, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
1
On the celestite-secreting Acantharia and their effect on seawater
strontium to calcium ratios
Patrick De Deckker
Department of Earth and Marine Sciences, The Australian National University, Canberra ACT 0200, Australia
E-mail: [email protected]
Received 27 April 2003; in revised form 16 August 2003; accepted 10 September 2003
Key words: Acantharia, sarcodine protozoan, strontium sulfate, Indian Ocean, coral palaeothermometry, seawater
chemistry
Abstract
Significant variations in the Sr/Ca in waters from the eastern Indian Ocean in the vicinity of Australia, both from the
surface and from shallow depth profiles, are documented. The strontium sulfate-secreting protozoans Acantharia,
which are common in the upper 400 m of the oceans, especially at low latitudes, contribute substantially to changes
in the Sr/Ca of oceanic waters by extracting Sr for the formation of their skeletons. Below such depths, these
organisms dissolve and the Sr/Ca of seawater regains its ‘conservative’ nature. This mechanism accounts for some
of the variability of the Sr/Ca near the surface of the ocean, but this may still be found at depth as surface waters
become entrained at greater depths. It is argued here that the noted Sr/Ca variations may explain discrepancies
between coral data sets from different parts of the oceans, and calls for caution when reconstructing sea-surface
temperatures from the Sr/Ca of corals.
Introduction
Acantharians are common inhabitants in all oceans.
This unique group of sarcodine protozoans secretes
a skeleton made of the mineral celestite, SrSO4 . Because seawater is grossly undersaturated with respect
to this mineral, the remains of these planktonic organisms readily dissolve and are very rarely encountered
as fossils. Hence, the study of acantharians has been
neglected by geologists and palaeoceanographers, although these organisms are among the most abundant
members of the plankton in the oceans (Beers et al.,
1975; Bishop et al., 1977; Michaels, 1988), frequently surpassing that of the calcareous test-forming
foraminiferans.
Acantharian numbers in the equatorial Atlantic
Ocean can reach up to 744 specimens m−3 , compared
to a value of 105 m−3 at high latitudes (Bottazzi, 1978;
Bottazzi & Andreoli, 1982). In the North Pacific Central Gyre, Michaels (1988) counted 16 specimens per
litre in the upper 20 m of the water column, but numbers dropped down to 10 between 40 m to 120 m.
In comparison, foraminifers amounted to less than 1
specimen per litre in the upper 80 m of these samples
(Michaels, 1988). Commonly, the maximum density of acantharians in the oceans is found between
50 and 200 m, but figures vary between studies, and
the type of mesh size for plankton sampling devices
can substantially affect results; even the nature of the
preservative used is important as some selectively dissolve acantharian spicules (Beers & Stewart, 1970).
Also, to undergo gametogenesis, some acantharians
migrate to depths between 300 and 400 m (Bottazzi,
1978). With respect to seasonal abundance, the pioneering taxonomic and ecological observations of
Schewiakoff (1926) for the Bay of Naples showed
that these organisms are especially abundant from late
spring to the beginning of autumn. In addition, studies
from the seas around Italy identified that salinity is
also a controlling factor in the abundance of acantharians (Bottazzi, 1978), with high specimen counts even
at salinities up to 38.3% (Bottazzi & Andreoli, 1978).
During the two Australian cruises in the eastern
Indian Ocean, on board the RV Franklin, numer-
2
ous acantharian specimens were recovered from surface plankton tows at the sites where water samples
and CTD profiles were obtained. The abundance
of the acantharians collected was on average 100
times higher than the live foraminifers and radiolarians; acantharian densities were consistent with other
oceans (Beers et al., 1975; Bishop et al., 1977;
Bottazzi & Andreoli, 1982; Michaels et al., 1995).
Acantharians were most abundant between latitudes
14◦ and 25◦ S.
The aim of this paper is to present chemical analyses of water samples collected during the 2 cruises
and also document the presence of acantharians collected at the same stations.
Material and methods
During two cruises with the RV Franklin in the eastern
Indian Ocean in late 1995 [Fr10/95] and early 1996
[Fr2/96], numerous surface plankton samples were
collected. The plankton samples were collected by
towing a circular plankton net, 40 cm in diameter and
with a 150 µm mesh, through the surface of the ocean
for approximately 1 km. About 4 m3 of water passed
through the plankton nets. No quantitative effort was
made to estimate the number of plankton species per
volume of water sampled, although it is estimated that
approximately 125 m3 of surface water were sampled.
Once the towing was completed, the plankton sample
was immediately filtered through a small 10 cm diameter net (mesh 60 µm) and rinsed thoroughly with
distilled water to get rid of any remaining sea water
(to prevent a flocculant from forming once ethanol is
added to the sample, and also to prevent dissolution of
the acantharians as seawater is undersaturated with respect to celestite – see Discussion later). The samples
were then preserved in 100% ethanol.
Water samples were collected from both the ocean
surface and depth profiles. The water samples used
for chemical analyses were filtered through a 0.45 µm
Millipore membrane using a peristaltic pump, acidified with a few drops of ultrapure HNO3 . The seawater samples were all analysed on a Perkin Elmer
Elan 6000 ICP-MS, with the use of a Perkin Elmer
FIAS 2000 (used as an online auto-diluter) and a Perkin Elmer AS-90 auto-sampler. Two sets of samples
and standards were prepared. Both sets of samples
and standards were analysed for Ca42,43,44, Sr86,87,88,
Ba134,135, Sc45 and Y89 . The first set comprised a
1 in 10 dilution of the original samples and stand-
ards, the second set was prepared as the first, but
included spikes of internal standards of Sc45 and Y89 ,
resulting in end concentrations of 9 and 0.9 mg l−1 ,
respectively, in every sample and standard. A further
1 to 10 dilution was carried out for all samples and
standards using a Perkin Elmer FIAS 2000 running as
an online auto-diluter, feeding directly into a Ryton
Cross Flow Nebulizer/Spray Chamber attached to a
Perkin Elmer Elan 6000 ICP-MS. Artificial seawater
was used to make up the analytical standards; these
were also crossed-checked against natural seawaters
spiked with Ca, Sr and Ba standards to determine
analytical performance of the technique.
In the first set of samples, Sc and Y concentrations
proved to be low [less than 20 µg/l], but raw counts
were still used as minor corrections for Sc45 and Y89
concentrations in the results from the analysis of the
second set of samples. The results of the second set
of samples were used only, as the corrected internal
standards, Sc45 for Ca44 and Y89 for Sr88 , could be
used to correct for matrix interference, instrument drift
and short-term noise in the second analysis on the
analyses of interest. A further 8 sub-samples were analysed as duplicates, and 6 sub-samples were analysed
using Methods of Standard addition on both a Perkin
Elmer Elan 6000 ICP-MS and a Perkin Elmer 3100
Atomic Absorption Spectrometer, as cross-checks on
the original analysis. Considering ICP-MS analysis is
not the ‘favoured’ technique for the analysis of seawater, this technique showed considerable promise.
However, errors associated with large dilutions (this
was partially corrected for with the addition of internal
standards at the beginning of the dilution stages), and
the production of a suitable standard solution, matrixmatched to the samples, continued to cause small
problems. All samples with standards greater than 5%
were re-analysed.
Results
A total of 65 surface water samples was analysed from
the region between 108◦ E and 13◦ E and 12◦ S and
32◦ S and their location are shown in Fig. 1a. Note
that none of the surface samples were taken in less
than 50 km from the Australian coastline to prevent
possible ‘contamination’ from likely fluvial sources.
A broad range of salinities was encountered spanning
from S = 33.69 to S = 35.91 and those values are
schematically contoured in Fig. 1b. Variations were
also noted for Ca and Sr measurements, and again the
3
Figure 1. Maps of the eastern Indian Ocean adjacent to Western Australia showing the location of the sea-surface samples (dots) taken during
the two Franklin cruises in December 1995 (= Fr10/95) and February–March 1996 (= Fr2/96) (1a) and the location of the rivers which
contribute to continental outflow, especially during the monsoonal season. 1(b) shows the salinity contouring based on the data obtained during
the 2 cruises; 1(c) displays the contouring for Sr/Cl (as µg/g); 1(d) the measured molar Sr/Ca values in mmol/mol; 1(e) and 1(f) display the
contouring for Ca and Sr in ppm respectively. Chlorinity values used in 1c were obtained using the relationship of S() = 1.80655 Cl()
(Chester, 1990). Salinity values were obtained by conductivity measurements while at sea, and recalibrated later by CSIRO personnel. The
location of the 3 sites with CTD profiles displayed in Fig. 2 is shown by black crosses in the above map.
values are schematically contoured in Figs 1e, f. Ratios of Sr/Cl as well as Sr/Ca were also computed and
are displayed in the same fashion in Figs 1c, d. Water
samples from three CTD [conductivity-temperaturedepth] profiles were selected for chemical analyses
and the Sr/Ca results are presented against salinity
in Figs 2a–c. All data used for the present paper are
available on the Springer server.
Discussion
Surface waters
Surface water data from these Franklin cruises show
a typical patchiness with respect to all measured elements, either tabled singly or displayed as ratios such
as Sr/Ca or Cl/Sr (Figs 1a–f). The latter ratio was used
to determine whether a relationship could be found
between salinity and the concentration of Sr. This
could not be found. Patchiness of Sr values in upper
waters in the oceans had already been documented in
Renard’s (1985) review, and more recently for several
4
Figure 1(b). (Continued.)
sites adjacent to the Hawaiian islands where corals
grow (de Villiers et al., 1995).
Close examination of the location of the major
rivers bordering the Western Australian coast (Fig. 1a)
shows no obvious links between chemical and salinity variations found offshore (Figs 1b–f), despite the
fact that half of the water samples were collected
in December 1995 when three cyclonic depressions
occurred in the region with substantial rainfall, and
river discharge ensuing. Equally, no link was found
in February–March 1996 when the rest of the water
samples were taken.
Overall, salinity displays lower values at the lowest latitudes and highest salinities are recorded near
28◦ –30◦ S where evaporation is highest in the region.
Nevertheless, salinity values ≥ 34.6 are noticeable on
the northwest shelf and these correspond also to the
overall high Sr/Ca values.
A possible explanation being that very likely fine
needles of CaCO3 may precipitate due to the high
salinities and temperatures at the end of the ‘dry’
[=austral winter] season caused by some supersaturation in CaCO3 . However, this hypothesis requires
testing as a comparison between the plots for salinity (Fig. 1b) and Cl/Sr (Fig. 1c) show no obvious
similarity with Sr/Ca (Fig. 1d).
Depth profiles
Characteristically, in some profiles obtained using conductivity-temperature-depth [CTD] equipment, Sr/Ca ratios also show a systematic depletion
of Sr in the upper portion of the water column (see
5
Figure 1(c). (Continued.)
Figs 2a–c). This pattern is reminiscent of profiles
from the Atlantic (McKenzie, 1964; Brass & Turekian,
1972, 1974; Bernstein et al., 1987) and Pacific (Brass
and Turekian, 1973, 1974; Brass, 1980) Oceans, and
the same was recently found also by Müller & De
Deckker (2003) from a region near Tasmania and New
Zealand at the confluence of the Indian, Pacific and
Southern Oceans. More importantly, de Villiers (1999)
concluded from her work in the Atlantic and Pacific
Oceans that ‘seawater Sr and Sr/Ca exhibit spacial
gradients of 2–3% globally with the deep ocean enriched relative to the surface’. This is investigated
further below by attempting to determine the cause of
seawater Sr/Ca variability.
However, despite the lowest values recognized
near the surface in all three CTD profiles (for explanation see Discussion below), another phenomenon
showing a variation in the Sr/Ca with depth is noticeable. Examination of the salinity–temperature profiles
for the 3 CTD sites (Station Fr2/95-36 at 14◦ 48.68 S
114◦ 16.37 E; Station Fr10/95-39 at 28◦ 44.35 S
112◦ 47.89 E; station Fr10/95-28 at 20◦ 01.93 S
112◦ 43.37 E) help identify the various water masses
encountered in the region. Importantly are (1) the Indonesian Throughflow water (= ITF in Fig. 3) which is
characteristically of low salinity, (2) the Indian Central
Water (= ICW in Fig. 3) which is at the other extreme
with its highest salinity, and (3) the other two water
masses the Subtropical South Indian water (= SSIW in
Fig. 3) and the South Indian Central Water (= SICW in
Fig. 3). The latter two water masses are intermediary
between the ITF and the ICW. It is noteworthy that
the ICW, recognized in Fig. 2c at Station Fr10/95-28
between ∼200 and 300 m water depth, also has a high
6
Figure 1(d). (Continued.)
Sr/Ca as for the other 2 profiles below the surficial
Sr/Ca-‘depleted’ layer. The surprising feature, however, is that below the ICW lens at Station Fr10/95-28,
Sr/Ca return to low values; these waters belong to
watermasses (SSIW and SICW) that originated at the
surface at higher latitudes, and at the time of their
formation, may also have been ‘depleted’ in Sr/Ca and
remained as such.
The acantharian skeleton and seawater Sr/Ca
The acantharian skeleton is made of fine SrSO4
needles/spicules and rapidly becomes dismantled
upon death once the protoplasm decomposes, falls
through the water column [probably fairly rapidly due
to its high (3.96) specific density] and consequently
dissolves readily (Beers et al., 1975). Acantharians
are rare or non existent below 1500 m as recognised
from trap studies (Beers & Stewart, 1970). This explains the change in the Sr composition of seawater
as acantharians utilise a substantial amount of Sr in
the upper parts of the water column where they live
and secrete their test, but this element is returned into
solution once spicules dissolve while they sink. Several studies have alluded to the substantial uptake of Sr
by acantharians from their ambient waters. Estimates
for one acantharian vary between 20.6 µg cm−2 yr−1
(Bernstein et al., 1987) and 22 to 40 µg cm−2 yr−1
(Brass, 1980). Again, estimates vary between the
amount of Sr acantharian skeletons contain; 0.2 nmol
(Bishop et al., 1977) and between 5–49 nmol (Bernstein et al., 1987) based on mean weights. Assuming a
mean value of 20 nmol of Sr, an acantharian specimen
would require 0.76 µg of Sr to fabricate its acicu-
7
Figure 1(e). (Continued.)
lar skeleton; for an average number of 10 specimens
per litre at anyone time, up to 7.6 µg of Sr would
be taken from the ambient seawater. As an example,
with 2 acantharian generations per year, and an average residence time of 25 years for the upper waters
in the ocean, an acantharian population would have
extracted 380 µg of Sr from 1 l of seawater over those
years. By this process, the Sr can be extracted from
the original water sample, provided there is no mixing
with underlying waters, and explains the 5% areal and
vertical variations of Sr/Ca in the upper 400 m of the
water column.
de Villiers (1999) discussed the characteristic depletion of Sr variations in the upper few hundred
metres in her samples form the Atlantic and Pacific
Oceans with respect to deeper horizons below. She
concluded that the dissolution of acantharians contrib-
ute to a significant increase in the upper part of the
ocean [below the horizon near the surface which is
depleted in Sr as a result of Sr uptake from acantharians], but that also a secondary influence is caused
by calcium carbonate cycling. It is suggested here
that different upper-ocean water masses may, in fact,
also have Sr/Ca values as illustrated in Fig. 2c for the
profile which displays an obvious correlation between
salinity and Sr/Ca. It may be that partially-evaporated
seawater may have chemically evolved differently,
such that aragonite needles may have precipitated at
high salinities [and high temperatures], thus forcing a
‘partial’ depletion in Sr values. Such a phenomenon
could be expected on the northwest shelf of Australia
where intense evaporation does frequently occur and
high salinities are generated.
8
Plate 1. SEM photos of acantharians collected in plankton tows during the RV Franklin cruises in the eastern Indian Ocean. Collection stations
with coordinates are listed after the identification. (A) Acanthocolla cruciata Haeckel with spicules partly broken (Fr2/96–19; 15◦ 00.12 S
110◦ 29.80 E); B, F: Acanthostaurus purpurascens Haeckel (Fr2/96-26; 21◦ 59.76 S 107◦ 42.71 E); C: Phyllostaurus siculus Haeckel with
shrunk protoplasm still visible (Fr2/96-18); D: Amphibelone hydrotomica Haeckel with juvenile planktonic foraminifers stuck in between the
spicules (Fr2/96-18); E: Coleapsis vaginata Haeckel (Fr2/96-26). Note scale bars for all specimens represent 50 µm and the background which
consisted of carbon tape has been artificially erased.
9
Plate 2. SEM photos of acantharians collected in plankton tows during the RV Franklin cruises in the eastern Indian Ocean. Same remarks about
collection stations as for Plate 1. (A), (B) Diploconus aff. fasces Haeckel (note juvenile planktonic foraminifers for relative scale) (Fr2/96-19);
(C) Amphibelone anomala Haeckel (Fr2/96-19; 26◦ 59.68 S 109◦ 29.68 E); D: Acanthospira spiralis Haeckel with possible radiolarian stuck
among the spicules and with some shrunk protoplasm still visible (Fr2/96-36; 14◦ 48.68 S 114◦ 16.37 E); (E) Phyllostaurus siculus Haeckel
with shrunk protoplasm still visible and a juvenile planktonic foraminifer adhering to some of the spicules (Fr2/96-36); (F) possible early
stage of Pleuraspis costata Müller with some protoplasm also visible (Fr2/96-19). Note scale bars for all specimens represent 50 µm and the
background which consisted of carbon tape has been artificially erased.
10
Figure 3. Temperature-salinity profiles obtained with CTD equipment for the 3 stations discussed in the text (36 = Fr2/96-36; 39 =
Fr10/95-39; 28 = Fr10/95-28]. The various watermasses of relevance to the discussion in this paper are labeled on these profiles.
These are: ITF for Indonesian Throughflow; SSIW for Subtropical
South Indian Water; SICW for South Indian Central Water; ICW for
Indian Central Water (nomenclature after Fieux et al., 1996).
Implications of seawater Sr/Ca
Figure 2. Continuous salinity profiles measured by CTD (bold line)
and Sr/Ca (in mmol mol−1 ) (thin line with dots showing the depths
of sampling using Niskin bottles) taken for 3 selected sites during
the two Franklin cruises (see text for more details) to show the
characteristic depletion in Sr in at least the upper 200 m of the
ocean, and the increase at various levels down to 1000 m where
acantharians would completely dissolve. Note the broad range of
salinities resulting from the sampling of different water masses, and
the different depths displayed for those profiles. In particular, note
the relationship in (2c) between salinity and the Sr/Ca measured at
station Fr10/95-28.
The residence times for calcium and strontium in
today’s oceans have been estimated to be of the order
of 1.1 × 106 and 4.9 × 106 years, respectively (Martin
& Whitfield, 1983; Chester, 1990) and, consequently,
the ratio of these two elements has been considered by
many to be constant for the entirety of the Quaternary.
This assumption led to the use of Sr/Ca fluctuation detected in the skeleton of modem and fossil corals as a
proxy for sea-surface temperature changes in tropical
oceans (Smith et al., 1979; Beck et al. 1992; McCulloch et al., 1994; Min et al., 1995; Beck et al., 1997;
Alibert & McCulloch, 1997; McCulloch et al., 1999;
Schrag, 1999). Nevertheless, there have been ample
published accounts documenting significant variations
in either Sr, or in the Sr/Ca and Sr/Cl ratios of seawater
in the open oceans (Andersen et al., 1970; Brass &
Turekian, 1972, 1973, 1974; Bernstein et al. 1987;
Bernstein et al., 1992; de Villiers et al., 1994; de
Villiers, 1999). These findings have been ignored by
most of those using the coral Sr/Ca as a palaeothermometer. These contradicting views need resolving,
especially since there are now several claims for substantial sea-surface temperature differences between
the present and previous episodes of the Late Quaternary (e.g., for the Last Glacial Maximum) (Guilderson
et al., 1994), or younger episodes (Beck et al., 1992,
11
1997) that are in conflict with sea-surface temperature reconstructions using other proxy methods such
as foraminiferal faunal assemblages (CLIMAP, 1981;
Anderson et al., 1989; Martinez et al., 1997) or Uk37
analyses from the remains of calcareous nannoplankton (Bard et al., 1997). Here, additional chemical
analyses of seawater samples from the eastern Indian
Ocean are presented and these identify that the Sr/Ca
of waters at the surface and through profiles is again
not constant throughout in an oceanic region where
corals live.
Seawater Sr/Ca and coral palaeothermomet
Plate 3. SEM photos of acantharians collected in plankton tows
during the RV Franklin cruises in the eastern Indian Ocean. Same
remarks about collection stations as for Plate 1. (A) Acanthostaurus
purpurascens Haeckel with some shrunk protoplasm visible and a
smaller acantharian adhering to some of its spicules (Fr2/96-19); (B)
Amphibelone hydrotomica Haeckel (Fr2/96-15; 23◦ 44.23 S 108◦
31.91 E); C: Amphilithium clavarium var. lanceolatum Schewiakoff
with several planktonic foraminifers and one radiolarian adhering
to numerous other ancantharian spicule debris; one dinoflagellate is also visible at the opposite end of the debris amalgamate
(Fr2/96-36). Note scale bars for all specimens represent 50 µm and
the background which consisted of carbon tape has been artificially
erased.
The analytical data presented here seriously question
the validity of using the Sr/Ca of fossil corals in reconstructions of past sea-surface temperatures, because
calculated temperature shifts (Beck et al., 1992; Guilderson et al., 1994; Min et al., 1995; Beck et al., 1997;
Schrag, 1999; McCulloch et al., 1999) all assume a
constant or near-constant Sr/Ca for the water in which
the corals grew. The present study, in addition to others
mentioned above, shows that the Sr/Ca of seawater,
especially in the realm where corals grow, cannot be
accepted to be constant through time and space.
It is noteworthy that the several sites and periods
for which fossil corals have been analysed represent
locations – and periods of time – during which salinity
may have changed as a result of reduced precipitation.
For example, the ‘postulated’ substantial temperature
drop (6.5 ◦ C) at Vanuatu (Beck et al., 1992, 1997)
appears to coincide with the occurrence of a significant period of aridity registered in northern Australia
at a similar latitude (De Deckker et al., 1991), implying a possible reduction of runoff from the island
that could cause a change in the Sr/Ca but not necessarily a change of sea-surface temperature (SST).
Foraminifer assemblages from nearby ODP core 828
(Martinez et al., 1997), detect a shift of summer SST
of no more that 1 ◦ C over the last 20 Ka, challenging
the coral Sr/Ca evidence (Beck et al., 1997).
Similarly, for the Barbados coral (Guilderson et al.,
1994) which grew during the Last Glacial Maximum
(= LGM), a reconstructed SST (e.g., a drop of 5 ◦ C
compared to today) may be misleading because of
much more arid conditions in the region. So far, no
lacustrine cores have returned LGM sediments from
the island, very likely due to the extreme dry conditions there during the LGM. The postulated decrease
in precipitation is confirmed for adjacent areas in the
Caribbean (Bradbury et al. 1981) and could have
12
caused the increase in the salinity and δ18 O of the ambient waters that was originally interpreted as resulting
from an SST drop (Guilderson et al., 1994). In addition, at the LGM 1.2 must be added to the salinity to
adjust for the fall in global sea level. Similarly, an increase of the Sr/Ca of seawater by 3% (due to a change
in the supply of local Sr-depleted continental waters)
near the coast where corals usually grow, or a lack
of local acantharian productivity, for example, would
have registered as a 5 ◦ C temperature drop by the coral
Sr/Ca palaeothermometer. Modelling studies (Stoll &
Schrag, 1998) on the effect of sea level changes during the Quaternary also postulate that Sr in seawater
would have changed during periods of low sea levels.
No quantitative measurements have so far been made
to verify these models but they support the argument
that the basis of coral Sr/Ca thermometry for parts of
the Quaternary ought to be looked at more closely.
Conclusion
This paper argues that acantharians, despite their small
size, play an important role in the Sr depletion in
the upper few hundred metres of the ocean’s water
column. It is argued also that if such ‘Sr-depleted’ waters eventually ‘sink’ or become covered by the water
masses, the Sr depletion may remain at greater depths.
Such a phenomenon may perhaps be used to actually
trace the origin of a water mass.
Despite the lack of uniformity in the Sr/Ca of sea
water near the surface as shown in this paper, the
coral Sr/Ca thermometer remains a very powerful tool
for demonstrating secular variations and for recent
times’ reconstructions (e.g., the last millennium) as
oceanic conditions for that time frame would have
changed little. Nevertheless, several studies have identified that the temperature equation using the Sr/Ca in
the same species of coral varies from place to place
(de Villiers et al., 1994; Beck et al., 1997; Alibert &
McCulloch, 1997), showing that oceanic Sr/Ca varies between sites. However, in order to extend past
SST reconstructions for periods of the record which
registered significantly different conditions – the extreme being during the LGM – such as salinity changes
(which could affect acantharian productivity levels (effectively changing the Sr/Ca of seawater locally), light
intensity, depth of the nutricline, wind-induced turbulence and mixing of the surface layer of the ocean)
caution is definitely warranted. A number of proxies, used together at the same site – such as corals
and other methods/organisms – are essential before
inferring substantial sea-surface temperature changes
at low latitudes during the LGM.
Acknowledgements
I am thankful to the captains and crew of the Australian National Facility RV Franklin for facilitating
the collection of all samples in the eastern Indian
Ocean, and the numerous colleagues who helped while
at sea, and for several ARC grants to support this work.
I benefited from the insight of Dr D. Nürnberg (Geomar, Kiel) who suggested that acantharians play a role
in the uptake of Sr in the oceans. He also provided
key references and commented on an earlier draft of
the manuscript. Mr F. Krikowa of the University of
Canberra performed the analyses presented here. Dr
T. Corrège provided some cross-checking analyses
by ICP-MS. Very profitable discussions with Michael
Shelley during the early stages of writing of this paper
helped improve the clarity of the arguments and text.
Mr C. Hilliker drew the figures and Mr A. Sadekov
helped with the photographs. Dr R. Barwick was also
provided much valuable advice with the illustrations.
I am also grateful to Dr E. Gliozzi who managed to obtain an original of the publication and
plates of the important and outstanding work of W.
Schewiakoff from the Zoological Station in Naples
through negotiations with the librarian there.
References
Alibert, C. & M. T. McCulloch, 1997. Strontium/calcium ratios in
modern Porites corals from the Great Barrier Reef as a proxy
for sea-surface temperature: calibration of the thermometer and
monitoring ENSO. Paleoceanography 12: 345–363.
Andersen, N. R., J. D. Gassaway & W. E. Maloney, 1970. The relationship of the strontium: chlorinity ratio of water masses in
the tropical Atlantic Ocean and Carribean Sea. Limnology and
Oceanography 15: 467–472.
Anderson, D. M., W. L. Prell & N. J. Barratt, 1989. Estimates of sea
surface temperature in the Coral Sea at the last glacial maximum.
Paleoceanography 4: 615–627.
Bard E., F. Rostek & C. Sonzogni, 1997. Interhemispheric synchrony of the last deglaciation inferred from alkenone palaeothermometry. Nature 385: 707–710.
Beck, J. W., R. L. Edwards, E. Ito, F. W. Taylor, J. Récy, F. Rougerie,
P. Joannot & C. Henin, 1992. Sea-surface temperature from coral
skeletal strontium/calcium ratios. Science 257: 644–647.
Beck, W. T. , J. Récy, F. Taylor, R. L. Edwards & G. Cabioch,
G., 1997. Abrupt changes in early Holocene tropical sea surface
temperature derived from coral records. Nature 385: 705–707.
13
Beers, J. R., F. M. H. Reid & G. L. Stewart, 1975. Microplankton of
the North Pacific Central Gyre. Population structure and abundance, June 1973. Internationale Revue Gesamtes Hydrobiologia
60: 607–638.
Beers, J. R. & G. L. Stewart, 1970. The preservation of acantharians in fixed plankton samples. Limnology and Oceanography 15:
825–827.
Bernstein, R. E., P. R. Betzer, R. A. Feely, R. H. Byrne, M. F.
Lamb & A. F. Michaels, 1987. Acantharian fluxes and strontium to chlorinity ratios in the North Pacific Ocean. Science 237:
1490–1494.
Bernstein, R. E., R. H. Byrne, P. R. Betzer & A. M. Greco,
1992. Morphologies and transformations of celestite in seawater:
the role of acantharians in strontium and barium geochemistry.
Geochimica et Cosmochimica Acta 56: 3272–3279.
Bishop, J. K. B., J. M. Edmond, D. R. Ketten, M. P. Bacon & W. B.
Silker, 1977. The chemistry, biology, and vertical flux of particulate matter from the upper 400 m of the equatorial Atlantic
Ocean. Deep-Sea Research 24: 511–548.
Bottazzi, E. M., 1978. Systematic-ecological aspects of Radiolaria
with special reference to Acantharia. Bolletino Zoologica 45:
133–144.
Bottazzi, E. M. & M. G. Andreoli, 1978. Distribuzione stagionale
degli Acantari e dei Radiolari (Protozoa, Sarcodina) in diverse zone costiere dei mari italiani. L’ Ateneo Parmense Acta
Naturalia 14: 477–500.
Bottazzi, E. M. & M. G. Andreoli, 1982. Distribution of adult and
juvenile Acantharia (Protozoa, Sarcodina) in the Atlantic Ocean.
Journal of Plankton Research 4: 757–777.
Bradbury, J. P., B. Leyden, M. Salgado-Laboriau, C. Lewis, C.
Schubert, M. Binford, D. Frey, D. Whitehead & F. Weibezahn,
1981. Late Quaternary environmental history of Lake Valencia.
Science 214: 1299–1305.
Brass, G. W., 1980. Trace elements in acantharian skeletons.
Limnology and Oceanography 25: 146–149.
Brass, G. W. & K. K. Turekian, 1972. Strontium distributions in sea
water profiles from the GEOSECS I (Pacific) and GEOSECS II
(Atlantic) test stations. Earth and Planetary Science Letters 16:
117–121.
Brass, G. W. & K. K. Turekian, 1973. Stontium and alkalinity variations in the South Pacific. In Fraser, R. (ed.), Oceanography of
the South Pacific 1972, New Zealand National Commission for
UNESCO, Wellington: 15–18.
Brass, G. W. & K. K. Turekian, 1974. Strontium distribution in
GEOSECS oceanic profiles. Earth and Planetary Science Letters
23: 141–148.
Chester, R., 1990. Marine Geochemistry. Unwin Hyman, Boston.
CLIMAP Project Members, 1981. Seasonal reconstruction of the
earth’s surface during the last glacial maximum. Map and Chart
Series, No. 36. Geological Society of America, Boulder.
De Deckker, P., T. Corrège & J. Head, 1991. Late Pleistocene record
of eolian activity from tropical northeastern Australia suggesting
the Younger Dryas is not an unusual climatic event. Geology 19:
602–605.
de Villiers, S., 1999. Seawater strontium and Sr/Ca variability in the
Atlantic and pacific oceans. Earth and Planetary Science Letters
171: 623–634.
de Villiers, S., G. T. Shen & B. K. Nelson, 1994. The Sr/Ca temperature relationship in coralline aragonite: influence of variability
in (Sr/Ca) seawater and skeletal growth parameters. Geochimica
et Cosmochimica Acta 58: 197–208.
de Villiers, S., B. K. Nelson & A. R. Chivas, 1995. Biological
controls on coral Sr/Ca and δ18 O reconstructions of sea surface
temperatures. Science 269: 1247–1249.
Fieux, M., C. Andrié, E. Charriaud, A. G. Ilahude, N. Metzl, R.
Molcard & J. C. Swallow, 1996. Hydrological and chlorofluoromethane measurements of the Indonesian Throughflow entering
the Indian Ocean. Journal of Geophysical Research 101 (C5):
12,433–12,454.
Guilderson, T. P., R. G. Fairbanks & J. L. Rubenstone, 1994. Tropical temperature variations since 20,000 years ago: modulating
interhemispheric climate change. Science 263: 663–665.
Martin, J. H. & M. Whitfield, 1983. The significance of river inputs
of chemical elements to the ocean system. In Wong, W. S., E.
Boyle, K. W. Bruland, J. D. Burton & E. D. Godberg (eds), Trace
Metals in Sea Water. Plenum, New York: 265–296.
McCulloch, M. T., M. K. Gagan, G. Mortimer, A. R. Chivas & P. J.
Isdale, 1994. A high resolution Sr/Ca and δ18 O coral record from
the Great Barrier Reef, Australia, and the 1982–1983 El Niño.
Geochimica et Cosmochimica Acta 58: 2747–2754.
McCulloch, M. T., A. W. Tudhope, T. M. Esat, G. E. Mortimer, J. Chappell, B. Pillans, A. R. Chivas & A. Omura,
1999. Coral record of equatorial sea-surface temperatures during the penultimate deglaciation at Huon Peninsula. Science 283:
202–204.
McKenzie, F. T., 1964. Strontium content and variable strontium
chlorinity relationship of Sargasso sea water. Science 146: 517–
518.
Martinez, J. I., P. De Deckker & A. R. Chivas, 1997. New estimates
for salinity changes in the Western Pacific Warm Pool during
the Last Glacial Maximum: oxygen-isotope evidence. Marine
Micropaleontology 32: 311–340.
Michaels, A. F., 1988. Vertical distribution and abundance of
Acantharia and their symbionts. Marine Biology 97: 559–569.
Michaels, A. F., D. A. Caron, N. R. Swanberg, F. A. Howse &
C. M. Michaels, 1995. Planktonic sarcodines (Acantharia, Radiolaria, Foraminifera) in surface waters near Bermuda: abundance, biomass and vertical flux. Journal of Plankton Research 17:
131–163.
Min, G. R., R. L. Edwards, F. W. Taylor, J. Recy, C. D. Gallup
& J. W. Beck, 1995. Annual cycles of U/Ca in coral skeletons
and U/Ca thermometry. Geochimica et Cosmochimica Acta 59:
2025–2042.
Müller, A. & P. De Deckker, 2003. Magnesium, calcium and strontium in waters of the southern Tasman Sea at the confluence
of the Indian, Pacific and Southern Oceans. Marine Freshwater
Research 53 (7): 1115–1128.
Renard, M., 1985. Géochimie des carbonates pélagiques. Documents du BRGM 85: 1–650.
Schewiakoff, W. 1926. Die Acantharia. Fauna e Flora del Golfo di
Napoli 37: 1–755.
Schrag, D. P., 1999. Rapid analysis of high-precision Sr/Ca ratios in
corals and other marine carbonates. Paleoceanography 14: 97–
102.
Smith, S. V., Buddemeir, R. W., Redalje, R. C. & Houck, J. E., 1979.
Strontium-Calcium thermometry in coral skeletons. Science 204:
404–407.
Stoll, H. M. & D. P. Schrag, 1998. Effects of Quaternary sea level
cycles on strontium in seawater. Geochimica et Cosmochimica
Acta 62: 1107–1118.