The biogeography of major diatom taxa in Southern Ocean sediments

Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93 – 126 www.elsevier.com/locate/palaeo

The biogeography of major diatom taxa in Southern Ocean sediments: 1. Sea ice related species Leanne K. Armanda,T, Xavier Crostab, Oscar Romeroc, Jean-Jacques Pichonb,F b

a School of Earth Sciences, University of Tasmania, Private Bag 79, Hobart, Tasmania 7001, Australia DGO, UMR-CNRS 5805 EPOC, Universite´ de Bordeaux I, Avenue des Faculte´s, 33405 Talence Cedex, France c Department of Geosciences, Bremen University, P.O. Box 33 04 40, 28334 Bremen, Germany

Received 18 November 2004; accepted 24 February 2005

Abstract Diatoms from 228 Southern Ocean core-top sediment samples were examined to determine the geographic distributions of 32 major diatom species/taxa preserved in the sediments of three zonally-distinct regions; Sea Ice, Open Ocean and the Tropical/Subtropical. In the first of three papers, 14 species/taxa occurring in the region where sea ice covers the ocean surface on an annual basis are geographically documented. Comparisons are drawn between the diatom abundances on the sea floor, sea ice parameters (annual duration and concentration in February and September) and February sea-surface temperature. Such parameters are commonly used in reconstructing past oceanographic conditions in the Sea Ice and Open Ocean zones. Analysis of the geographic patterns and sea-surface parameter correlations reveals species-specific distributions regulated primarily by sea ice coverage and sea-surface temperature, which support the use of diatom remains for the estimation of these past seasurface environmental parameters. Comparison with reliable accounts of the 14 species from the sediments or plankton also provides the first glimpses into species-specific ecology and habitat linkages. D 2005 Elsevier B.V. All rights reserved. Keywords: Diatom; Bacillariophyceae; Biogeography; Sea ice; Southern Ocean; Sediments

1. Introduction T Corresponding author. Current address: Centre d’Oce´anologie de Marseille, Laboratoire d’Oce´anologie et de Bioge´ochimie, Campus de Luminy, case 901, F-13288 Marseille Cedex 09, France. Tel.: +33 4 91 82 93 46; fax: +33 4 91 82 19 91. E-mail addresses: [email protected] (L.K. Armand), [email protected] (X. Crosta), [email protected] (O. Romero). F Deceased. 0031-0182/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2005.02.015

Diatoms of the Southern Ocean are microscopic algae that dominate the primary productivity cycle around Antarctica. Their remains are preserved more readily in the sediments beneath the Antarctic Circumpolar Current than other calcareous microorganisms used for palaeontological reconstructions.

94

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

Diatoms have been shown to be important for reconstructions of past sea-surface temperatures (Pichon et al., 1992a; Labeyrie et al., 1996) and more recently to past sea ice estimations in the lower latitudes (Armand, 1997, 2000; Crosta et al., 1998a,b; Armand and Leventer, 2003; Gersonde et al., 2005). It is acknowledged that sediment trap data from the Southern Ocean have shown that the major diatom flux from the surface waters to the sediments occurs over summer and that the presence of sea ice provides a barrier to production and sedimentation (Abelmann and Gersonde, 1991; Leventer, 2003). For this reason, making relationships, and ultimately deriving estimates, between diatom sedimentary distributions and that of various seasonal parameters should be made in reference to summer conditions. Here, we examine such summer relationships, but also incorporate comparisons with the winter parameters on the basis that summer/winter relationships occur among the parameters. The winter parameters here provide the contrasting environmental extreme expected on an average annual scale at a particular sample site. The role of the total habitat seasonality should be expected to play a role in the surface distribution of particular species or taxa. This includes the variation in winter parameters, and in some cases, it is expected to be linked to specific species distributions (e.g. such as Fragilariopsis cylindrus and Fragilariopsis curta; Gersonde, 1984; Garrison, 1991; Leventer and Dunbar, 1996). Large diatom data sets have recently been compiled in both the South Atlantic and Indian Oceans for transfer function work principally focused on sea-surface temperature (SST) estimates (Pichon et al., 1992a; Zielinski and Gersonde, 1997). Yet our understanding of the distributions throughout the Southern Ocean remains relatively patchy and influenced by early summaries provided by Kozlova (1966), Donahue (1973), Semina (1979), DeFelice and Wise (1981) and more recently by Zielinski and Gersonde (1997). The relevance of mapping these distributions is important since one of the basic assumptions of transfer function analysis is that the species located in the sediments reflects the distribution of physical parameters at the surface of the ocean. This work investigates the sedimentary distributions of 14 sea ice related diatom species or

groups of species with relative abundances greater than 2% in 228 core-top surface sediment samples from the Southern Ocean. We consider the abundance relationships of each of these species to three physical parameters (sea ice duration, sea ice concentration and SST), the species’ geography and provide useful details on previous sedimentary and plankton reports, which outline or provide glimpses to their environments of preference. We define sea ice related species as those confined southward of the Polar Front generally observed within the Sea Ice Zone (Tre´guer and Jacques, 1992) and with a known presence within, on or under sea ice or in the water column surrounding sea ice. Where species have previously been attributed or suspected of being bipolar, we have not cited comparisons to their environmental preferences or biogeographic range in the northern hemisphere. Aside from providing a basis to the Southern Ocean surface water hydrology signature of species in palaeo-environmental reconstructions, this paper represents one in a series of three papers describing diatom distributions in surface sediments from wide areas of the Southern Ocean (Tropical/Subtropical— Romero et al., this volume; Open-Ocean—Crosta et al., this volume) in which we provide the basis for future ecological and habitat research that will inevitably be required to link in with future biogeochemical and ecosystem modeling of our Earth system.

2. Materials and methods 2.1. Diatom data set Diatom analysis, sediment treatment and slide preparation followed the technique described in Rathburn et al. (1997), Armand (1997) and Romero and Hensen (2002). Diatom counts largely followed Schrader and Gersonde (1978) and Laws (1983). Generally, a minimum of 300 diatom valves was counted in each sample at a magnification of 1000. Diatoms were identified to species or species group level. Relative abundances of each diatom species or species group in a given surface sample were calculated as the fraction of the diatom species against the total diatom abundance in the

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

sample. Girdle bands of the diatom species Dactyliosolen antarctica and silicoflagellate specimens were recognized and counted but not used in the calculation of the diatom relative abundances. Distribution tendencies of these species using the Pichon database have been plotted and analyzed elsewhere (Armand, 1997). This analysis reports only on the diatom species or species groups currently employed in quantitative reconstructions for the Southern Ocean (e.g. Crosta et al., 2004). Many species observed in the Southern Ocean do not reach the required relative abundance of 2% in at least one sample and they are not included in this study. Eucampia antarctica is the obvious exception as it is not used in transfer function studies due to its ubiquitous nature in modern sediments (Zielinski and Gersonde, 1997; Armand, 1997; Crosta et al., 1998a). Although the ratio of Eucampia terminal and intercalary valves (viz. Eucampia index; Fryxell, 1991; Kaczmarska et al., 1993) may provide a source of discrimination to their geographic distribution or winter growth, our data set does not currently contain this level of detail useful in attaining this goal. Additionally, no benthic diatom species are used in quantitative reconstructions, and therefore, they are not presented here. The diatom reference data set is composed of 32 species or species group categories (see also Crosta et al., this volume; Romero et al., this volume). We present in this paper the 14 species or groups of species that are linked to the sea ice environment around Antarctica, otherwise identified as the Sea Ice Zone. 2.2. Surface sample data set The Crosta et al. (1998a) data set, a derivate of the Pichon et al. (1992a) database, represents the main data source used in our analysis. Surface sample selection followed several criteria such as (1) the sample must be of Late Holocene age younger than 4000 years old, (2) more than 300 diatom valves should be counted, (3) at least 2% of extant diatom species should be present and (4) valve preservation should be good. The latter point is based on light microscopy observation and refers to the chemical and mechanical alteration of the valve structure (e.g. poroid size), the number of fragments derived from mechanical dissolution and on species number

95

through a comparison between nearby surface sediment samples. The Crosta et al. data set is complemented by unpublished counts from material in the south east Pacific region (Armand, unpubl.) and from the Tropical/Subtropical sector of the Atlantic Ocean (Romero, unpubl.). The reference database does not include the experimental dissolution data set of Pichon et al. (1992b) as this represents altered duplicates of natural surface samples already used here. The reference data set is hereafter referred to as the Diatom Database 228 (DD228). The reference surface sediment sample data set covers the Atlantic and Indian sectors of the Southern Ocean from 338S to the Antarctic Continent plus the Ross Sea, the Prydz Bay and the western part of the Antarctic Peninsula (Fig. 1). Very few core-tops have been retrieved from the abyssal plains and the Pacific sector of the Southern Ocean. 2.3. Modern parameters and oceanography Our treatment of DD228 explores the abundance relationship to two physical parameters, sea ice and sea-surface temperature (SST). SST data are derived from the World Ocean Atlas (1994) and cover the mean monthly maximum SST in February (Austral summer). Several sea ice parameters, derived from Schweitzer (1995), are used here in an effort to reveal potential relationships of distribution with the seasonal cycle of sea ice duration and degree of sea ice cover. For the latter parameter, the sea ice data represent February and September ice concentration as a percentage of sea ice to open ocean cover for the summer maximum and winter maximum average extent. Under this percentage scheme, we define open ocean conditions where concentration values are between 0 to 15%. Above 15% and up to 40% sea ice cover is defined as unconsolidated (uncompacted sea ice), and greater than 40% as consolidated sea ice (compact sea ice). These defining values are provided through calibration studies and weather filter studies on the NASA algorithm used to interpret data from satellite observations (Gloersen and Cavalieri, 1986; Gloersen et al., 1992; Cavalieri et al., 1995). Annual sea ice presence is represented as the number of months per year of sea ice cover over a location. Calculation of yearly sea ice

96

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

Z

SI

SSI WSI PF SAF

Z

O

PO

Z

PF

Z

SA

Z

ST

STF

Fig. 1. Location of the 228 surface samples represented in the DD228 of this study, Crosta et al. (this volume) and Romero et al. (this volume). Oceanographic fronts after Orsi et al. (1995) and oceanographic zones follow the nomenclature provided by Tre´guer and Jacques (1992). Summer and winter sea ice edges are from Comiso (pers. comm. 2002) using the bootstrap algorithm derived 15% ice edge over the last 20 years of data (1982–2002). Abbreviations: STF: Subtropical Front; SAF: Subantarctic Front; PF: Antarctic Polar Front; WSI: Maximum average winter sea ice edge; and SSI: Maximum average summer sea ice edge; SZI: Sea Ice Zone; POOZ: Permanently Open Ocean Zone; PFZ: Polar Frontal Zone; SAZ: Subantarctic Zone; STZ: Subtropical Zone. Boxed regions represent regions of denser surface sample coverage, which are detailed in Fig. 2A–C.

presence is based on abovementioned concentration limits where (1) unconsolidated sea ice is attributed a value of 0.5 and consolidated sea ice is attributed a value of 1 for each monthly period, and (2) yearly presence in number of months is the sum of the monthly presence. Oceanic Fronts follow the description of Orsi et al. (1995) and oceanographic zones follow the nomenclature provided by Tre´guer and Jacques (1992). Summer and winter sea ice edges are from Comiso (pers. comm., 2002) using the bootstrap algorithm derived 15% ice edge over 20 years of data (1982–2002).

3. Observations and discussion The biogeographical representation of the 14 sea ice species/groups around Antarctica is described with reference to their relative abundance against February sea-surface temperature (SST), sea ice concentration and annual duration in months per year. Taxonomic issues specific to individual species or their groupings are addressed. Due to the geographical detail required, but not possible in this work, the positioning of samples in the Ross Sea, Prydz Bay and Antarctic Peninsula region has

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

been summarized by regional boxes in all 14 maps, which are identified in Fig. 2A–C. Reports on the sedimentary distribution of Southern Ocean diatom species are numerous, yet the relative abundance data on which the distributions are based are more rarely published. We schematically illustrate such previous work in Fig. 3, which is also used as a basis for discussion in this section and in the two linked papers (Crosta et al., this volume; Romero et al., this volume). Studies that document sea ice habitats or sea ice associations for certain species are also noted in the respective species descriptions. Generalized reports on species distributions in the sediments were not included, nor was the phytoplankton distribution at the ocean surface covered with exception to some specific cases. It must also be kept in mind that comparison with diatom distributions and abundances reported in these published accounts incorporate differences between the preparation and counting methods reported here, but more importantly also to the taxonomic concept of each of the identified species (particularly members of the Chaetoceros, Fragilariopsis, Porosira and Stellarima genera). Thus, although the evolution of the current data set presented here (viz. Pichon et al., 1992a; Armand, 1997; Crosta et al., 1998a and DD228) addresses such confusion due to conflicting descriptions and illustrations, evolving descriptions, omissions of taxa or human error, it is not to say that the species and their abundances reported by others and used as a

A

B

97

basis for ecological comparison in this work have not suffered from similar problems addressed by DD228. Summary plots of a species or species group’s relative abundance against the physical parameters discussed through the remainder of the text are provided in Fig. 4A–D. 3.1. Actinocyclus actinochilus (Ehrenberg) Simonsen Figures in: Villareal and Fryxell (1983), Figs. 21– 24; Zielinski (1993), Pl. 1, Figs. 1 and 3. Actinocyclus actinochilus has a maximum relative abundance of 2.9% in the DD228, which is associated with February SST between 0 and 1 8C (Table 1, Fig. 4A). Where A. actinochilus is observed with abundances over 2%, sea ice duration is greater than 7 months per year, with an optimum of 8–9 months per year where A. actinochilus has its greatest abundances (Table 1, Fig. 4B). Increasing sedimentary abundances of this species falls in line with an ice-free region during summer (b 40% concentration) and a strongly compact sea ice covered region in winter (70–90% concentration) (Table 1, Fig. 4C,D). Geographically (Fig. 5), the species abundance distribution declines sharply from the coastal regions of Antarctica to the maximum winter sea ice edge. Actinocyclus actinochilus is commonly reported in the sediment record. Its distribution in the sediments has been described as confined to the Antarctic Divergence (Kozlova, 1966), or more commonly

C

Fig. 2. Regional boxed sample locations from Fig. 1. (A) 35 samples (24 + 11 respectively boxed) from the Antarctic Peninsula; (B) 27 samples from the Prydz Bay; (C) Ross Sea; 17 samples from the western Ross Sea.

98

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

0°

5 to 8 30°E

30°W

9

4

60°W

60°E

10

24 2 75°S

90°W

60°S

30°S

45°S

1

11

3 14

15 90°E

12,13 21 23

19, 20

22

18

16

17

120°W

120°E

KEY 150°W

150°E

1

Gersonde & Wefer (1987)

14

Fryxell (1991)

2

Prasad & Nienow (1986)

15

3

Gersonde (1984)

4

Zielinski & Gersonde (1997)

Kozlova (1966) Indian sector only. van Iperen et al. (1993) Indonesian Archipelago.

5

Pokras & Molfino (1986) SE Atlantic

17

Abbott (1973)

6

van Iperen et al. (1987) SE Atlantic

18

Leventer (1992)

7

Treppke et al. (1996) SE Atlantic

19

8

Schuette and Schrader (1981) SE Atlantic 20

9

DeFelice & Wise (1981)

10

Tanimura (1992)

11

Jouse et al. (1962 a, b)

12

Stockwell et al. (1991)

13

Taylor et al. (1997)

16

Truesdale & Kellogg (1979) Cunningham & Leventer (1998)

21

Kellogg & Kellogg (1987)

22

Donahue (1973)

23

Kozlova & Muchina (1967)

24

Schuette & Schrader (1979a,b) SE Pacific

Fig. 3. Schematic map representing study areas of previous studies dealing with diatoms in surface sediments from the Southern Ocean. Studies listed principally include those that document count or relative abundance data pertinent to the species/taxa discussed in this study, Crosta et al. (this volume) and Romero et al. (this volume). Numbers refer to the publications identified in the key (Jouse´ et al., 1962b; Schuette and Schrader, 1979a,b, 1981; Pokras and Molfino, 1986; van Iperen et al., 1993; Kozlova and Muchina, 1967; Treppke et al., 1996). Six publications refer to diatom distribution in surface sediments north of 308S (these are indicated out of the map).

0

0

8

4

0

0

0

gp. um id

tica

4

T. t

nta rc

0 3

99

a

s tria icro

ent . 8

S. m

seu dod

ialis

is

lac

ear

P. g 0 7

T. a

5

ubl in

epa r F. s

itsc

F. r 0

20

P. p

0

F. s

4

her i

uec

and

a

ost ata hom bica

s ylin dru 0 80

F. r

0

F. o bliq

100

F. c

urta

0 4

F. c

s chil u

Ch ae

ctin o A. a

A

February Sea Surface Temperatures (°C)

0

to. r.s.

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

0

3

40

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

D

February Sea Ice Concentration (% of covered water)

C

September Sea Ice Concentration (% of covered water)

B

Sea Ice Duration (Months per year)

0 1 2 3 4 5 6 7 8 9 10 11 12 0 10 20 30 40 50 60 70 80 90

100 0 10 20 30 40 50 60 70 80 90 100

Fig. 4. Relative abundances of the 14 major sea ice related species and/or species groups in DD228 against February sea-surface temperatures (A), sea ice duration in months per year (B), February sea ice concentration (C) and September sea ice concentration (D).

100

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

Table 1 Temperature ranges, sea ice parameter ranges and maximum relative abundances of the 14 sea ice diatom species and groups in DD228 Taxa/taxa group

Maximum relative abundance (MRA) (%)

February SST at MRA (8C)

Actinocyclus actinochilus Chaetoceros resting spores Fragilariopsis curta Fragilariopsis cylindrus Fragilariopsis obliquecostata Fragilariopsis rhombica Fragilariopsis ritscheri Fragilariopsis separanda Fragilariopsis sublinearis Porosira glacialis Porosira pseudodenticulata Stellarima microtrias Thalassiosira antarctica gp Thalassiosira tumida

2.9 91.8 64.6 2.9 10.2 3.8 2.9 6.9 6.1 6.4 2.2 3.2 31.8 2

0 to 1 0.5 to 1.5 0.5 to 1 0.5 to 1 1 to 0 0.5 to 1 0 to 3 0.5 to 0 0 to 1 0 to 0.5 0 to 0.5 0.5 to 0.5 0 to 0.5 0.5 to 0

a

February SST range (8C) 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3

to to to to to to to to to to to to to to

9 3.5 2.5 3 2 11.5 7.5 1.8 2.5 2 2 3.5 3.5 7

Maximum sea ice duration at MRA (m/yr)

Maximum Feb. SI conc. at MRA (%)

Maximum Sept. SI conc. at MRA (%)

9 7 10 8.5 8.5 9 0/9a 8.5 7.5 8 9.5 8.5 8.5 8.5

2 43 20 0 0 1 0/28a 0 24 0 28 0 3 0

82 62 83 87 90 93 0/72a 90 61 90 81 88 92 90

Alternative maximum value based on major abundance peak at 2.8%.

south of the Polar Front (Donahue, 1973; DeFelice and Wise, 1981; Zielinski and Gersonde, 1997; Semina, 2003). Along the Antarctic coast, A. actinochilus has been observed in commonly low to medium abundances (Table 2) with by far the greatest abundance (15.2%) being reported to date in the Amundsen Sea by Kellogg and Kellogg (1987). Northward of the Polar Front, occurrences of A. actinochilus are reported but lack abundance data (Abbott, 1973; Donahue, 1973; Jouse´ et al., 1962a; DeFelice and Wise, 1981). Zielinski and Gersonde (1997) describe A. actinochilus with a maximum abundance of 4.7% in the South Atlantic sector with a summer SST range of 2 to 2 8C. In relation to sea ice conditions, Actinocyclus actinochilus has been reported in newly formed sea ice (Gersonde, 1984; Tanimura et al., 1990; Garrison and Close, 1993) and from both fast and pack-ice samples (Horner, 1985; Krebs et al., 1987; Garrison and Buck, 1989; Garrison, 1991). The species was noted as having a higher presence in sea ice than in the adjacent water column (Garrison et al., 1983, 1987). The observations from previous studies and data in DD228 indicate that Actinocyclus actinochilus is a cool water Antarctic species limited to the north by the Polar Front and quite possibly the maximum winter sea ice edge. The species is most commonly linked with other sea ice taxa both in the sediments and the sea ice.

3.2. Chaetoceros resting spores Chaetoceros is one of the most abundant diatom genera in the modern ocean. It is present in most environments from coastal temperate to polar regions. The genus is composed of approximately 180 marine planktonic species with around 75 species forming resting spores as a resting or survival phase (Stockwell and Hargraves, 1984). These small silica forms pose problems to diatom taxonomy and statistical Table 2 Published abundances of Actinocyclus actinochilus in Antarctic sediments Location

Max. rel. abundance

Reference

Off Filschner-Ronne ice shelf Lqtzow-Holm Bay Outer Prydz Bay Prydz Bay

b 3%

Gersonde, 1984

b 1% 0.68% Increasing from 0.5% to 2.67% 0.6 to 4%

Tanimura, 1992 Stockwell et al., 1991 Taylor et al., 1997

Indian Ocean section George V Coast Ross Sea

2.1% 2.85%

Amundsen Sea

15.2%

Bransfield Strait and Drake Passage

b 1%

Kozlova, 1966 Leventer, 1992 Cunningham and Leventer, 1998 Kellogg and Kellogg, 1987 Gersonde and Wefer, 1987

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

101

25

13 7 6 9

13 3

A. actinochilus 0 - 1% 1 - 2% 2 - 5%

Fig. 5. Distribution of Actinocyclus actinochilus relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

approaches. Linking the resting spores to vegetative cells is still in its infancy even though many have identified relationships via SEM work or indicate such work is under way (Hargraves, 1979; Stockwell and Hargraves, 1984; Leventer et al., 1993; Riaux-Gobin, 1996). Chaetoceros resting spores were never noted in the original Pichon database (1992b) and ensuing statistical analyses, but were included after the reanalysis of Crosta et al. (1997), which clearly indicated the predominance of Chaetoceros resting spores surrounding the Antarctic continent. In DD228, we continued the grouped approach for all Chaetoceros resting spores. Chaetoceros resting spores have a maximum relative abundance of 91.8% occurring in waters with an SST of 0.5 to 1.5 8C. Their distribution

in the sediments is circumpolar with little regard for zonal boundaries and yet with greatest abundances (i.e. N 20%) occurring in the Antarctic Peninsula region and the Ross Sea. Such occurrences are associated with February SST between 1.3 and 3.5 8C (Table 1, Fig. 4A). Where Chaetoceros resting spores are observed with abundances over 20% sea ice duration is greater than 3 months per year, with an optimum coverage of 3–9 months per year (7 m/yr at the maximum abundance) (Table 1, Fig. 4B). Plotted against sea ice concentrations, the distribution of maximum abundances falls most notably in line with moderately consolidated sea ice conditions in the winter dropping off at concentrations greater than 80% (Fig. 4D). Summer sea ice concentration levels

102

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

Kerguelen islands where they represent up to 20% of the total diatom assemblage. Abundances of Chaetoceros resting spores north of the winter sea ice edge are observed between 0 and 20% and for the most part fall into the 2–10% abundance categories (Fig. 6). Discrete regions with no or less than 1% abundance are observed in the South Atlantic and south of Tasmania in the permanently open ocean zone (POOZ), however, north of the Polar Front abundances in the sediments are generally up to 5–10%. Other observations of Chaetoceros resting spores north of the Polar Front with less than 10% relative abundance have been reported (Zielinski, 1993 for the South Atlantic; Gersonde and Wefer, 1987 for Drake Passage and van Iperen et al., 1987 for the southeast Atlantic).

appear to have little impact on the distribution pattern of the resting spores, although the very highest abundances are observed in the ice affected regions (Figs. 4C, 6). The data of DD228 highlight several interesting points. An eastward plume of decreasing abundances is observed from the Antarctic Peninsula to reach 5% of the diatom assemblage at 108W (Fig. 6). A similar gradient is found in the Ross Sea, with minimum occurrences of 25% encountered in McMurdo Sound. Both patterns are seemingly due to transport by currents from regions of greatest production such as the Bransfield Strait and the western Ross Sea (Leventer, 1991; Crosta et al., 1997). Occurrences of the CRS group in the POOZ and SAZ sediments are generally below 5% except around the Crozet–

23 7 9 6

6 6

Chaetoceros RS. 0 - 1% 1 - 2% 2 - 5% 5 - 10% 10 - 20% 20 - 30% 30 - 50% 50 - 75% 75 - 100%

5 12 1

Fig. 6. Distribution of Chaetoceros resting spores relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

Possibilities for the presence of these resting spores northward of the winter sea ice zone include bottom water entrainment and the eastward displacement by the Antarctic circumpolar current from the neritic influence of islands or plateaus. A relationship between the grouped Chaetoceros taxa abundance distribution and sea-surface temperature was not detected in the work of Zielinski and Gersonde (1997). Alternatively, the influence of high productivity waters is thought to be of primary importance to Chaetoceros distribution (Donegan and Schrader, 1982; Leventer, 1991; Sancetta et al., 1992; Karpuz and Jansen, 1992; Zielinski and Gersonde, 1997). More recently, surface water stratification produced by sea ice melt water has been suggested as the main significant factor in Chaetoceros resting spore dominance in the sediment record (Leventer et al., 1993; Leventer et al., 1996). Several Q-mode factor analyses of surface sediment diatom samples from the Weddell Sea integrating Chaetoceros taxa (both vegetative and predominantly resting spores) illustrated the dominance of the raw grouped taxa and a heavy bias towards Antarctic Peninsula samples (Zielinski, 1993). Ranking of the species reduced this dominance but only to evenly distribute the effect of the Chaetoceros taxa, thus in effect provide statistical noise. Even the role of sea ice in their habitat is poorly understood, and this is supported by the negative correlation of Chaetoceros taxa to other sea ice diatom types in factor analysis (Zielinski, 1993). Several workers have suggested that Chaetoceros resting spores may, once their ecological relationship to sea ice is understood, prove important to estimating sea ice cover (Zielinski, 1993; Crosta et al., 1997; Leventer et al., 1996). Yet, such ecological information is unlikely to be forthcoming in the near future, deeming removal of the species in statistical analysis necessary. As such, the removal of Chaetoceros resting spores from future diatom btransfer functionsQ is in line with similar statistical treatments performed by others (Karpuz and Schrader, 1990; Schrader and Karpuz, 1990; Schrader and Sorknes, 1991; Zielinski, 1993). Furthermore, Schrader and Karpuz (1990) argue that grouping resting spores, to the loss of taxonomic and biogeographic information, serve little purpose for ecological and thus, statistical work. Thus, in line with the decisions and arguments presented by Schrader and Karpuz (1990), Zielinski

103

(1993) and Zielinski and Gersonde (1997), we also support the removal of Chaetoceros resting spore data from future statistical analysis involving physical parameter estimation. 3.3. Fragilariopsis curta (Van Heurck) Hustedt Figures in: Hustedt (1958), Pl. 11, Figs. 140–144; Hasle (1965), Pl. 12, Figs. 2–5, Pl. 13, Figs. 1–6. The maximum geographical range of Fragilariopsis curta in DD228 appears confined southward from the maximum winter sea ice extent (Fig. 7). Highest abundances are observed near Prydz Bay (maximum 64.6%, Table 1) and also in the Ross Sea region and along the George V Coast. Trace excursions from the winter sea ice edge are observed only in the South Atlantic Ocean and are potentially linked to iceberg pathways. The comparison between sea ice cover and sedimentary abundances is plotted in Fig. 4B–D. Here, the highest abundances are linked with locations that experience 9–11 m/yr sea ice cover, where highly consolidated ice conditions exist during winter (65– 90%) and where summer sea ice concentration does not appear to play a strong role in its distribution under 40% concentration. The temperature–abundance plot reveals a summer temperature preference range of 1.3 to 2.5 8C, with the maximum abundance falling within temperatures of 0.5 and 1 8C (Fig. 4A). Most reference to Fragilariopsis curta in sediment samples reveals highest abundances near the Antarctic coast (Kozlova, 1966; Truesdale and Kellogg, 1979; DeFelice and Wise, 1981; Gersonde, 1984; Gersonde and Wefer, 1987; Kellogg and Kellogg, 1987; Stockwell et al., 1991; Leventer, 1992; Tanimura, 1992; Taylor et al., 1997; Zielinski and Gersonde, 1997). Maximum abundances around the Antarctic continent are documented in Table 3. The northwardly scattered dispersion of F. curta in the South Atlantic, beyond the winter maximum sea ice edge, was previously noted by DeFelice and Wise (1981) and Zielinski and Gersonde (1997). Abbott (1973) also mentions the presence of F. curta (maximum 3.6%) in his sediment samples from the southeast Indian Ocean, but this was not observed in the samples of this study. Semina’s (2003) summary of plankton collections suggested that the distribution of F. curta is bounded by the northern most extent of sea ice.

104

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

1 19 3 1 13 11 3

4 5 1

F. curta 0 - 1% 1 - 2% 2 - 5% 5 - 10% 10 - 20% 20 - 30% 30 - 50% 50 - 75%

1 16

Fig. 7. Distribution of Fragilariopsis curta relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

Garrison (1991) ascribed Fragilariopsis curta as a species most common in both pack and fast-ice as in previous reports (e.g. Gersonde, 1984; Horner, 1985; Krebs et al., 1987; Garrison et al., 1983, 1987; Garrison and Buck, 1989; Tanimura et al., 1990). The species is also noted in very high abundance in the water column near the sea ice edge (Garrison et al., 1982; Fryxell, 1989; Tanimura et al., 1990; Kang and Fryxell, 1992, 1993; Andreoli et al., 1995; Leventer and Dunbar, 1996). Leventer and Dunbar (1996) suggest that in the western Ross Sea region the presence of congelation sea ice, the predominant ice type of fast sea ice, is associated with increased relative abundances of F. curta. The distributions of Fragilariopsis curta against temperature, sea ice and within the Southern Ocean

confirm previous work in regard to the confined and specified sea ice habitat in which F. curta is linked. Palaeo-sea ice interpretation will benefit from the retrieval of this species in core sediments, although the incidence of specimen distribution presumably affected by the distribution of icebergs in the South Atlantic, begs caution in its application specifically in this region, and potentially elsewhere, iceberg pathways are known. 3.4. Fragilariopsis cylindrus (Grunow) Krieger Figures in: Hustedt (1958), Pl. 11, Figs. 145–146; Hasle (1965), Pl. 12, Figs. 6–12; Pl. 14, Figs. 1–10. Fragilariopsis cylindrus has a maximum relative abundance of 2.9% in the DD228 (Table 1). This

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126 Table 3 Published abundances of Fragilariopsis curta in Antarctic sediments Location

Max. rel. abundance (%)

Reference

Off Filschner-Ronne ice shelf South Atlantic

42

Gersonde, 1984

57

Zielinski and Gersonde, 1997 Tanimura, 1992 Stockwell et al., 1991 Taylor et al., 1997 Kozlova, 1966

Lqtzow-Holm Bay Outer Prydz Bay Prydz Bay Enderby Land to Prydz Bay George V Coast Ross Sea

50 68 70.8 20–90

Amundsen Sea

60

38 13.4 to 86

Leventer, 1992 Cunningham and Leventer, 1998 Kellogg and Kellogg, 1987

abundance maximum is associated with February SST between 0.5 and 1 8C, although the range of temperature under which abundances fall in the DD228 is limited between 1.3 and 3 8C (Fig. 4A). Where F. cylindrus is observed with abundances over 2%, sea ice duration is greater than 7.5 months per year, with an optimum of 8.5 months per year where F. cylindrus has its maximum abundance (Fig. 4B). Summer sea ice concentrations appear irrelevant to the abundance distribution of the species (Fig. 4C). In contrast, increasing abundances are associated with heavily consolidated sea ice conditions in the winter (N 70%– 90%) (Fig. 4D). Geographically (Fig. 8), the species abundance distribution is limited between the coastal regions of Antarctica to the maximum summer sea ice edge. It has been well established that Fragilariopsis cylindrus occupies sea ice covered environments and the highest reported abundances have been recorded around near the Antarctic coast (Table 4). Fragilariopsis cylindrus has been intensely investigated and shown to be the dominant marginal sea ice edge species, with increased seasonal abundance during the summer (Kang and Fryxell, 1992; Kang et al., 1993; Kang and Fryxell, 1993). It is also well preserved in the sediments (Jouse´ et al., 1962a; Gersonde, 1984; Tanimura et al., 1990; Tanimura, 1992; Leventer, 1992; Zielinski, 1993; Zielinski and Gersonde, 1997). Zielinski and Gersonde (1997) report their highest abundance peak of F. cylindrus (29%) occurring at

105

SST’s between 0.5 and 1 8C, with all other occurrences found over a temperature range of 2 to 2 8C. Hasle (1976) and Semina (2003) consider Fragilariopsis cylindrus a bipolar ice-neritic form and generally in the literature, the species is considered a dominant sea ice species (e.g. Gersonde, 1984; Horner, 1985; Krebs et al., 1987; Garrison and Buck, 1989; Tanimura et al., 1990; Garrison and Close, 1993; Scott et al., 1994). The species is associated with both land-fast and pack sea ice (Garrison, 1991), and has been found in exceptionally high abundances in the water column particularly at the melting sea ice edge (Fryxell, 1989; Kang and Fryxell, 1992, 1993; Andreoli et al., 1995; Leventer and Dunbar, 1996). Fragilariopsis cylindrus has also been linked with the process of sea ice formation (Clarke and Ackley, 1983; Garrison and Buck, 1989; Garrison et al., 1989; Garrison, 1991). 3.5. Fragilariopsis obliquecostata (Van Heurck) Heiden in Heiden et Kolbe Figures in: Hasle (1965), Pl.7, Figs. 2–7; Zielinski (1993), Pl. 5, Figs. 15–17. The maximum geographical range of Fragilariopsis obliquecostata in DD228 appears confined on or just north of the maximum summer sea ice extent (Fig. 9). Highest abundances are observed in the Ross Sea region (maximum of 10.2%, Table 1) and

Table 4 Published abundances of Fragilariopsis cylindrus in Antarctic sediments Location

Max. rel. abundance (%)

Reference

Off Filschner-Ronne ice shelf South Atlantic Lqtzow-Holm Bay Outer Prydz Bay Prydz Bay George V Coast Ross Sea

N3

Gersonde, 1984

29 12 8.9 37.3 11.7 6.07

Zielinski and Gersonde, 1997 Tanimura, 1992 Stockwell et al., 1991 Taylor et al., 1997 Leventer, 1992 Cunningham and Leventer, 1998 Kellogg and Kellogg, 1987 Gersonde and Wefer, 1987

Amundsen Sea Bransfield Strait and Drake Passage

b 1.1 1.61

106

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

22 15 6 10 1

F. cylindrus 0 - 1% 1 - 3%

12 5

Fig. 8. Distribution of Fragilariopsis cylindrus relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

in Prydz Bay. Lower abundances are observed from the marine sediments of the Antarctic Peninsula and along the George V Coast. The comparison of sea ice cover to sedimentary abundance is plotted in Fig. 4B–D. Here the highest abundances are linked with locations that experience N 7 m/yr sea ice cover, where highly consolidated ice conditions exist during winter (65–90%) and where summer sea ice concentrations do not appear to play a strong role in the species distribution. The temperature–abundance plot reveals a summer temperature preference range of 1.3 to 2 8C, with the maximum abundance falling within temperatures between 1 and 0 8C (Fig. 4A). Fragilariopsis obliquecostata is reported as having a similar distribution as Fragilariopsis ritscheri and

Fragilariopsis sublinearis in the Antarctic Zone (Kozlova, 1966; Hasle, 1965; Hasle, 1976; Tanimura, 1992; Zielinski and Gersonde, 1997). Yet, more specific studies indicate that F. obliquecostata increases in abundance south of the Antarctic Divergence (Kozlova, 1966; Truesdale and Kellogg, 1979; Gersonde, 1984; Gersonde and Wefer, 1987; Kellogg and Kellogg, 1987; Tanimura, 1992; Zielinski and Gersonde, 1997). The highest abundances of the species have been reported from the Weddell Sea (16.9%, Zielinski and Gersonde, 1997) and the Ross Sea (13.59%, Cunningham and Leventer, 1998). Our data from the Ross Sea fit well with these observations and add the Prydz Bay embayment to regions of increased abundance observed in marine floor sediments.

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

Fragilariopsis obliquecostata has been associated with surface melt pools and with increased abundance in the water column under sea ice (McConville and Wetherbee, 1983; Gersonde, 1984; Garrison et al., 1983; 1987; Garrison, 1991), and has been observed in land-fast and pack-ice samples (McConville and Wetherbee, 1983; Gersonde, 1984; Horner, 1985; Garrison and Buck, 1989; Tanimura et al., 1990; Garrison, 1991). The species is clearly confined to the sea ice environment where winter, more so than summer, sea ice concentration may play a role in its distribution. 3.6. Fragilariopsis rhombica (O’Meara) Hustedt Figures in: Hustedt (1958), Pl. 10, Figs. 113–120; Hasle (1965), Pl. 10, Figs. 2–6; Pl. 9, Figs. 1–6; Pl. 4, Fig. 19; Pl. 1, Fig. 6. Fragilariopsis rhombica was often encountered in abundances between 1–3% in sediments close to the Antarctic coast (Fig. 10). The top two maximum abundances (3.8 and 2.9%) were located along the Ade´lie coastline (138–1408E). The northward limit to the species dispersion in the sediments appears linked to the maximum summer sea ice extent, although a few odd occurrences appear between this and the winter sea ice edge and additional rare observations up into the SAZ are observed (Fig. 10). The February temperature range for the species is 1.3 to 11.5 8C although the majority of elevated abundances occur between 1 and 1 8C (Fig. 4A, Table 1). In DD228, the highest abundances were also related to sea ice free conditions in February and generally highly consolidated sea ice conditions in September (between 65 and 90% concentration, Fig. 4C–D). In terms of annual sea ice cover, the major abundances of the species were observed in regions where sea ice cover persisted for 7–9 months of the year (Fig. 4A). Previous sedimentary records of Fragilariopsis rhombica are described as occurring near the Antarctic coast or ice shelves (Kozlova, 1966; Abbott, 1973; Truesdale and Kellogg, 1979; Gersonde, 1984; Prasad and Nienow, 1986; Gersonde and Wefer, 1987; Stockwell et al., 1991; Leventer, 1992; Tanimura, 1992; Taylor et al., 1997; Zielinski and Gersonde, 1997; Cunningham and Leventer, 1998). The highest abundance reported for this species in the sediments was 16% by Kozlova (1966), 15% by Leventer (1992)

107

and 7% by Taylor et al. (1997). All other studies report abundances less than 5%. Fragilariopsis rhombica has been noted in both fast and pack-ice samples (Garrison et al., 1982, 1983; Gersonde, 1984; Horner, 1985; Krebs et al., 1987; Garrison and Buck, 1989; Garrison, 1991), which links in with our observations of distribution affected by winter sea ice cover. The rare observations of the species northward of the sea ice zone in DD228 are most likely related to the expatriation of the species through bottom water currents from the Antarctic shelf region. DeFelice and Wise (1981) and Zielinski and Gersonde (1997) also report the species sporadically northward of the sea ice zone in their South Atlantic study area, which may also be related to such bottom water transport pathways (Semina, 2003). 3.7. Fragilariopsis ritscheri Hustedt Figures in: Hustedt (1958), Pl. 11 Figs. 133–136; Hasle (1965), Pl. 1, Fig. 20; Pl. 4, Figs. 1–10; Pl. 15, Figs. 12–13; Pl. 17, Fig. 8. Fragilariopsis ritscheri occurs in low abundances b 3% in the sea ice region, with presumably expatriated occurrences observed in the South Atlantic PFZ (Fig. 11). One of these expatriated cases has the highest recorded abundance of the species within DD228 (2.9%), although the true highest abundance is observed in the Prydz Bay region at 2.8% (Table 1). The most remarkable observation regarding F. ritscheri’s geographical dispersion in DD228 is large number of samples (63 sites, see Appendix A) with trace abundances less than 1% (i.e. equivalent to 1 or 2 specimens observed in a sample). These trace occurrences appear in all regions from 478S to 778S, however, the geographic presentation in Fig. 11 highlights only those abundances greater than 1%, which are predominately accumulated between Ade´lie Land and Prydz Bay. Observations in the literature suggest that the species is found with increasing abundances north of the Antarctic Divergence (Kozlova, 1966; DeFelice and Wise, 1981; Stockwell et al., 1991; Tanimura, 1992; Leventer, 1992; Zielinski and Gersonde, 1997). The highest reported abundance of F. ritscheri to date is 3.28% in the Ross Sea (Cunningham and Leventer, 1998), whereas Zielinski and Gersonde (1997) observed a maximum of 2.6% in the South Atlantic sector.

108

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

23 1 5 22 1

9 1

F. obliquecostata

6 10 1

0 - 1% 1 - 2% 2 - 5% 5 - 10% 10 - 20%

Fig. 9. Distribution of Fragilariopsis obliquecostata relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

With respect to the physical parameters and excluding the assumed expatriated samples in the South Atlantic, the highest abundance of Fragilariopsis ritscheri is observed under SST’s of 08 to 3 8C and have an annual sea ice duration varying widely from 2 to 10.5 months, peaking at 9 m/yr (Fig. 4A,B). Such a wide range of sea ice conditions is also reflected in the February sea ice concentrations comparison plot (Fig. 4C), where abundances greater than 1% appear either in ice free regions or in regions with unconsolidated sea ice cover (30% concentration). The locations with the greatest abundances are covered with consolidated sea ice conditions N 70% concentration during winter (Fig. 4D). F. ritscheri has been observed in both surface melt pools, land-fast and pack-ice samples in the past (McConville and

Wetherbee, 1983; Horner, 1985; Garrison and Buck, 1989; Tanimura et al., 1990; Garrison, 1991), but F. ritscheri, not unlike Fragilariopsis obliquecostata, has been found in higher abundances in the adjacent water column than in sea ice samples (Gersonde, 1984; Garrison et al., 1982, 1987), suggesting that it is potentially a species that prefers melt water conditions. 3.8. Fragilariopsis separanda Hustedt Figures in: Hustedt (1958), Pl. 10, Figs. 108–112; Hasle (1965), Pl. 9, Figs. 7–10; Pl. 10 Fig. 1. Fragilariopsis separanda is observed with relative abundances up to 6.9% of the total diatom assemblage in DD228 (Table 1). The species is most dominant in

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

109

23 1 10 14 4

10

F. rhombica

15 1

0 - 1% 1 - 2% 2 - 5%

Fig. 10. Distribution of Fragilariopsis rhombica relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

the Ross Sea and Prydz Bay regions and generally around the Antarctic coast (Fig. 12). The distribution of this species throughout the Southern Ocean, north of the winter sea ice extent, may indicate this species ability to be more easily transported in bottom water currents, which is particularly evident in the Southwest Pacific sector to the south and east of New Zealand. The highest abundances are associated with the February SST signature of 0.5 to 0 8C, although of all the species reported in this paper, it appears to have the widest temperature dispersion with abundances N 2% covering a temperature range of 18 to 8 8C (Table 1, Fig. 4A). As a result of the wide temperature range, presumably related to transport, F. separanda has a superficially poor relationship to sea ice parameters. Nevertheless, where maximum abundan-

ces are observed sea ice duration lasts 4.5 to 9 m/yr, summer concentration is generally ice free or less than 30% and September sea ice concentration is representative of highly consolidated sea ice conditions (Fig. 4C–D). In the literature, the highest maximum abundances of Fragilariopsis separanda are observed as slightly increased at offshore locations rather than in-shore coastal environments (Kozlova, 1966; DeFelice and Wise, 1981; Gersonde and Wefer, 1987; Stockwell et al., 1991; Leventer, 1992; Taylor et al., 1997; Cunningham and Leventer, 1998). The exception being sediment samples from Lqtzow-Holm Bay, which has the highest recorded abundance in the sediments (10%, Tanimura, 1992). Both Zielinski and Gersonde (1997) and DeFelice and Wise (1981) found

110

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

24 20 5 1 10

F. ritscheri

17

0 - 1% 1 - 2% 2 - 5%

Fig. 11. Distribution of Fragilariopsis ritscheri relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

F. separanda confined to the south by the Polar Front in the South Atlantic sector. In our data set, we find trace occurrences north of the Polar Front in all sectors, which is not the first time that F. separanda has been noted north of this oceanographic feature in the Indian Ocean (Kozlova, 1966; Abbott, 1973). Semina (2003), although remarking on the direction of expatriated Antarctic species north and eastward of New Zealand, did not specifically list the Antarctic species that contribute to this signature. The most likely candidate, however, in this southwest Pacific sector appears to be F. separanda. Curiously, for a diatom considered associated with the sea ice region, Fragilariopsis separanda has never been reported from any studies of sea ice samples. The possibility of a lack of distinction

between the morphologically similar Fragilariopsis rhombica and F. separanda may mean previous reports embraced both occurrences under the F. rhombica banner, and so some caution in applying distribution histories for either species is required. Our temperature range and maximum abundance are nonetheless similar to that reported from the sediments by Zielinski and Gersonde (1997) and extends the range from 12 to 17.8 8C. 3.9. Fragilariopsis sublinearis (Van Heurck) Heiden Figures in: Hasle (1965), Pl. 7, Fig.1; Pl. 11, Figs. 1– 10; Pl. 12, Fig. 1; Zielinski (1993), Pl. 5, Figs. 25–26. Fragilariopsis sublinearis has a maximum relative abundance of 6.1% in DD228, where maximum

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

111

23 1 5 3 17 1 10

F. separanda

1 15 1

0 - 1% 1 - 2% 2 - 5% 5 - 10%

Fig. 12. Distribution of Fragilariopsis separanda relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

abundances are associated with February SST between 1.3 and 2.5 8C (Table 1, Fig. 4A). In instances where abundances of F. sublinearis are greater than 2%, sea ice duration is greater than 7.5 m/yr (Fig. 4B). There is little differentiation observed in the contrast of summer sea ice concentration against relative species abundance (Fig. 4C), yet September sea ice concentration is clearly associated with abundances greater than 2% (Fig. 4D). Geographically plotted, these major abundances of F. sublinearis are located almost exclusively in Prydz Bay, Ross Sea and along Wilkes Land within the maximum February sea ice extent (Fig. 13). Fragilariopsis sublinearis is reported as having a similar distribution in the Antarctic Zone as Fragilariopsis ritscheri and Fragilariopsis obliquecostata

(Kozlova, 1966; Hasle, 1965; Hasle, 1976; Tanimura, 1992; Zielinski and Gersonde, 1997), and like the latter species, it is considered to have an heightened abundance in the sediments south of the Antarctic Divergence (Kozlova, 1966; Gersonde, 1984; Gersonde and Wefer, 1987; Kellogg and Kellogg, 1987; Tanimura, 1992; Taylor et al., 1997; Zielinski and Gersonde, 1997). Leventer (1992) documented the only exception with equal maximum proportions of F. ritscheri and F. sublinearis along the George V Coast. Within the Ross Sea, F. sublinearis has been reported at an abundance maximum of 7.31% (Cunningham and Leventer, 1998), whereas in the South Atlantic the greatest abundance ever noted in the sediments stands at 13% in an area where SST is below 1 8C (Zielinski and Gersonde, 1997). It is surprising then to

112

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

24 1 1 18 1 10

F. sublinearis

1 10 6

0 - 1% 1 - 2% 2 - 5% 5 - 10%

Fig. 13. Distribution of Fragilariopsis sublinearis relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

find reports of F. sublinearis predominantly from land-fast and less so from pack-ice environments in the literature based on a tendency for low temperatures and our perceived relationship with sea ice concentrations greater than 70% (Garrison et al., 1982, 1983; Gersonde, 1984; Horner, 1985; Garrison and Buck, 1989; Tanimura et al., 1990; Garrison, 1991; Scott et al., 1994). 3.10. Porosira glacialis (Grunow) Jbrgensen Figures in: Hasle (1972), Figs. 62–63; Zielinski, 1993, Pl. 3 Figs. 1 and 2. Porosira glacialis and Porosira pseudodenticulata were provided emended descriptions by Hasle (1973) and classified as neritic Antarctic species. Porosira

glacialis, a bipolar species, is reportedly observed in waters adjacent to sea ice or the coast (Hasle, 1973). Previous studies by Pichon et al. (1992a) and Armand (1997) did not formally differentiate between P. glacialis and P. pseudodenticulata, although Armand (1997) tried to split the group based on observations described by the work of Zielinski and Gersonde (1997). In DD228, P. glacialis is observed in low abundances around the Antarctic continent, most predominately in the Prydz Bay region above 2%, with a maximum abundance in the sediments of 6.4% (Fig. 14, Table 1). Against February SST, the species’ distribution is sharply confined to a small range of 1.3 to 2 8C (Fig. 4A). Maximum abundances are observed at an SST of 0 to 0.5 8C. With respect to sea ice conditions,

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

113

23 1 12 6 7 1 10

P. glacialis

17

0 - 1% 1 - 2% 2 - 5% 5 - 10%

Fig. 14. Distribution of Porosira glacialis relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

Porosira glacialis is found in sediments that have at least 7.5 m/yr sea ice cover overhead, with the majority of locations experiencing marginal to ice free conditions in summer (b 30%) and relatively highly compacted sea ice cover in winter (65–85%, Fig. 4B–D). Abundances of P. glacialis notably decrease when winter sea ice concentration is N 85%. Porosira glacialis has been reported rarely as a sea ice species from earlier sea ice investigations (Horner, 1985; Krebs et al., 1987; Garrison, 1991). Zielinski and Gersonde (1997) reported the temperature signal of P. glacialis as 1 to 1.5 8C summer SST in the South Atlantic sector. Along the Antarctic coast, both P. glacialis and Porosira pseudodenticulata have been recorded in the sediments at less than 1% and at a maximum of ~ 3–4% along the George V Coast,

the Ross Sea and Prydz Bay shelf (Prasad and Nienow, 1986; Kellogg and Kellogg, 1987; Stockwell et al., 1991; Leventer, 1992; Taylor et al., 1997; Cunningham and Leventer, 1998). Transgressions north of the Porosira Group sedimentary record, particularly P. glacialis, have only been observed in the Bransfield Strait (Gersonde and Wefer, 1987) and are inferred from the abundance plot of the species by Zielinski and Gersonde (1997, Fig. 17.27). 3.11. Porosira pseudodenticulata (Hustedt) Jouse´ Figures in: Hasle (1972), Figs. 64–65; Hasle (1973), Pl.5, Fig. 30–31; Pl.6, Fig. 32; Pl. 7, Fig. 38. Porosira pseudodenticulata has a maximum relative abundance of 2.2% in DD228. We find it

114

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

associated with February SST between 1.3 and 2 8C (Table 1, Fig. 4A), which is similar to that observed with Porosira glacialis. Abundances greater than 1% are isolated in Prydz Bay and eastwardly offshore of Mawson Coast (Fig. 15). Although P. pseudodenticulata’s abundances are lower than those of P. glacialis, it portrays similar physical distribution characteristics with respect to sea ice cover with less than 30% sea ice cover occurring in summer, N 70% sea ice cover occurring in winter and 7.5 to 11 m/yr of sea ice duration covering their sites of deposition annually (Fig. 4B–D). Porosira pseudodenticulata has been attributed as an abundant constituent within sea ice (Hasle, 1973). This was supported with physical data from Zielinski and Gersonde (1997), which they reported as 2 8C

summer SST for this species. As noted in the preceding related species description, P. pseudodenticulata has been recorded at very low abundances around the Antarctic coastline. In addition to those studies, a rare occurrence (b 1%) of P. pseudodenticulata was described in Lqtzow-Holm Bay (Tanimura, 1992), whereas the largest reported abundance from the sediments to date (b 5%) was found along the Filschner-Ronne ice shelf (Gersonde, 1984). Most phytoplankton studies have located Porosira pseudodenticulata in both fast and pack-ice samples (e.g. Garrison et al., 1983; Gersonde, 1984; Krebs et al., 1987). However, more detailed studies have confounding results with respect to regions of preferred habitat. Some studies have indicated P. pseudodenticulata with an increased abundance in the

24 15 5 1 9

P. pseudodenticulata

17

0 - 1% 1 - 2% 2 - 5%

Fig. 15. Distribution of Porosira pseudodenticulata relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

115

Prydz Bay (Fig. 16). The distribution of the species south of the summer sea ice extent and in comparison with the sea ice data indicates that the species prefers extended ice conditions (N 7.5 m/yr cover, winter concentrations N 65%, Fig. 4B,D) but with a period of open water conditions in summer (b 30% concentration, Fig. 4C). The summer SST range covers 1.3 to 3.5 8C, with a distinct maximum temperature occurrence of 0.5 to 0.5 8C (Table 1, Fig. 4A). Previous records from the literature related to the distribution of the species must be treated cautiously to account for taxonomic changes and corrections (Syvertsen, 1985; Hasle et al., 1988). Previous identifications under the name Coscinodiscus stellaris var. symbolophorus are now considered the resting stage of the species, whereas the vegetative stage was

water column than in sea ice (Garrison et al., 1987; Tanimura et al., 1990) and others with P. pseudodenticulata as the dominant species in fast-ice (Watanabe, 1982). The latter feature may be a function of accumulation of this species by ice platelet layers under the consolidated fast-ice (Garrison, 1991). 3.12. Stellarima microtrias (Ehrenberg) Hasle et Sims Figures in: Syvertsen (1985), Figs. 1–9; Hasle et al. (1988), Figs. 1–12. Stellarima microtrias, the Antarctic cool water species of the genus Stellarima is observed with relative abundances up to 3.2% of the total diatom assemblage in DD228 (Table 1). The species is located in sediments taken from the Ross Sea and

24

20 3 4 10

6

S. microtrias

8

3

0 - 1% 1 - 2% 2 - 5%

Fig. 16. Distribution of Stellarima microtrias relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

116

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

more commonly identified as Coscinodiscus furcatus (Syvertsen, 1985). All reports of Stellarima microtrias (under C. furcatus, Coscinodiscus symbolophorus and S. microtrias) are noted in sediments along the Antarctic coast and various ice shelves (Jouse´ et al., 1962a; Kozlova, 1966; Truesdale and Kellogg, 1979; Gersonde, 1984; Prasad and Nienow, 1986; Gersonde and Wefer, 1987; Kellogg and Kellogg, 1987; Stockwell et al., 1991; Leventer, 1992; Taylor et al., 1997; Zielinski and Gersonde, 1997; Cunningham and Leventer, 1998). Kozlova (1966) recorded the highest abundances of the species along the Antarctic coast in the Indian sector (maximum 8.7%) but most recent observations record less than 1% with exception to samples along the Filschner-Ronne Ice Shelf, which are noted up to 3% (Gersonde, 1984; Zielinski and Gersonde, 1997) and off Cape Darnley in Prydz Bay with abundances up to 2.5% (Taylor et al., 1997). Kozlova (1966) stated that the northern limit of the species in the sediments was to 608S and only single observations of bC. furcatusQ have ever been noted this far south in the South Atlantic sector (DeFelice and Wise, 1981). Semina (2003) notes as having found it rarely south of New Zealand. Stellarima microtrias has been reported in both land-fast and pack-ice samples (Horner, 1985; Garrison and Buck, 1989; Garrison, 1991). Through the seasons, S. microtrias has been located in several sea ice associated situations. Over spring, the vegetative cell has been found in abundance at depth away from the sea ice edge (Fryxell, 1989), whereas in the sea ice, the species is noted in high abundance (Garrison et al., 1987). In summer, the species has been noted in very high abundance in fast sea ice samples (Watanabe, 1982; Krebs et al., 1987; Tanimura et al., 1990). During autumn, the resting spore is found in high abundance under sea ice and has not been present in the open-ocean (Fryxell, 1989), and the species was commonly found in newly forming sea ice (Gersonde, 1984; Garrison and Close, 1993). 3.13. Thalassiosira antarctica Group Figures in: Thalassiosira antarctica: Hasle and Heimdal (1968), Figs. 7–13; Johansen and Fryxell (1985), Figs. 37–39. Thalassiosira scotia: Johansen et al. (1985), Figs. 5–8; Johansen and Fryxell (1985), Figs. 40–42.

Due to the close affinity of Thalassiosira scotia vegetative cells with those of Thalassiosira antarctica resting cells (Johansen et al., 1985) and the earlier provision by other workers (Zielinski, 1993; Zielinski and Gersonde, 1997), we combine the two species under T. antarctica in our sample observations. In DD228, Thalassiosira antarctica has a maximum relative abundance of 31.8%, where maximum abundances are associated with February SST between 0 and 0.5 8C (Table 1, Fig. 4A). Where abundances of T. antarctica are greater than 10%, sea ice duration is greater than 6 m/yr (Fig. 4B). Summer sea ice concentrations indicate a trend of preference for conditions in the unconsolidated sea ice region (Fig. 4C, 15–40% concentration), whereas in winter, highest abundances are clearly associated with sea ice concentrations N 70% (Fig. 4D). Geographically placed, the species occur in almost all sites south of the maximum winter sea ice zone, with highest abundances located in Prydz Bay and along Wilkes Land (Fig. 17). Several sporadic cases of the species are observed in the South Atlantic (~ 20–258W) and the Indian Ocean (~ 75–808E) sectors. Thalassiosira antarctica abundances reported from sediments surrounding Antarctica are incongruently reported as: rare (Kozlova, 1966; Abbott, 1973; Tanimura, 1992), in medium maximum abundances between 5 and 21% (Gersonde, 1984; Gersonde and Wefer, 1987; Stockwell et al., 1991; Leventer, 1992; Taylor et al., 1997), with maximum abundances of 32% in the coastal sections of the Weddell Sea (Zielinski and Gersonde, 1997) or equally as high in the Ross Sea (31%, Cunningham and Leventer, 1998). Furthermore, observations north of the Subantarctic Front have been interpreted as surface water transport of the Antarctic Circumpolar Current (Zielinski and Gersonde, 1997). Thalassiosira scotia has only been reported from the South Atlantic. Zielinski and Gersonde (1997) attributed part of their T. antarctica/scotia taxon as referable to T. scotia in the Scotia Sea and the adjacent Argentine Basin, and Ligowski (1993) observed T. scotia in the South Atlantic but with highest abundances (12.4%) in littoral habitats. Garrison (1991) notes Thalassiosira antarctica as present in both fast and pack-ice. Fast-ice accumulation of the species (Krebs et al., 1987) has been

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

117

10 13 1 2 6 14 2 1

10 1

T. antarctica group

2 5 12

0 - 1% 1 - 2% 2 - 5% 5 - 10% 10 - 20% 20 - 30% 30 - 50%

Fig. 17. Distribution of Thalassiosira antarctica group relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

speculated to be resultant of ice platelet accumulation under the consolidated sea ice (Garrison, 1991). The species has also been noted under sea ice or in the adjacent water column (Cassie, 1963; Garrison et al., 1987). Semina (2003) notes T. antarctica as rare in the Pacific sector. There are no records of Thalassiosira scotia noted in sea ice studies. Although both species have been linked to neritic environments associated with ice conditions of the Antarctic (Hasle and Heimdal, 1968; Johansen et al., 1985). Future studies, specifically those in the region of previous T. scotia occurrences in the South Atlantic, would benefit from discriminating the two species in an attempt to advance clear distributional trends and environmental relationships to sea ice conditions.

3.14. Thalassiosira tumida (Janisch) Hasle Figures in: Hasle et al. (1971), Figs. 13–25; Fryxell et al. (1984), Figs. 1, 7, 8, 12–15; Johansen and Fryxell (1985), Figs. 29–32. Like many of the Antarctic species described in DD228, Thalassiosira tumida has its highest abundances geographically located in the Ross Sea and Prydz Bay regions (Fig. 18). A discreet range of SST is related to these highest abundances (2.0%): 0.5 to 0 8C February SST (Fig. 4A, Table 1). These abundance maxima are in turn associated with near year round sea ice cover (N 8.5m/yr), where there are values of 80–90% sea ice concentration during September and open-ocean to sea ice edge conditions during February (Fig. 4B–D).

118

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

24 25 6 10

T. tumida

8 9

0 - 1% 1 - 2%

Fig. 18. Distribution of Thalassiosira tumida relative abundances in DD228. Location of the oceanic fronts from Orsi et al. (1995), and location of February and September maximum sea ice extent from Comiso (pers. comm. 2002).

Previous sediment investigations revealed Thalassiosira tumida in low to rare abundances (b 3%) around the Antarctic coast (Truesdale and Kellogg, 1979; Gersonde, 1984; Prasad and Nienow, 1986; Stockwell et al., 1991; Leventer, 1992; Cunningham and Leventer, 1998). Further north, both DeFelice and Wise (1981) and Abbott (1973) noted the species as very rare in open-ocean sediments. Thalassiosira tumida has been categorized as a pack-ice species (Garrison and Buck, 1989; Garrison, 1991). However, it has been recovered from both fast and pack-ice samples of the Weddell Sea and Antarctic Peninsula region (Garrison et al., 1983; Gersonde, 1984; Krebs et al., 1987). In spring, the species is also found in abundance in the adjacent water column at the sea ice edge (Garrison et al., 1987). Semina (2003)

categorized the species as Notal-Antarctic and found it rarely in material from the Atlantic and Pacific sectors.

4. Discussion and conclusions 4.1. Environmental relationships In plotting the environmental data against diatom relative abundances in the Diatom database, we have illustrated that each species or taxa has strong affinities to SST and the degree of sea ice cover both annually and at seasonal extremes. Zielinski and Gersonde (1997) clearly defined the association between relative abundance of diatoms with Summer SST from the South Atlantic. Applied around Antarctica for 14 ice-

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

related species of DD228, many small extensions to the February SST signatures against maximum abundances were observed. SST ranges with respect to Zielinski and Gersonde (1997) were increased for all species with exception to Fragilariopsis sublinearis and the Thalassiosira antarctica group. A new SST range for the abundances of Thalassiosira tumida and Chaetoceros resting spores (Table 1) provides an addition to the list of diatom physical relationships now documented. In terms of increases to species relative abundances, only four species in DD228 (Fragilariopsis curta, Fragilariopsis ritscheri, Porosira glacialis, Stellarima microtrias) were observed at greater percentages than originally reported for these species in Zielinski and Gersonde (1997, Table 2). Yet, only P. glacialis, with 6.4% as its maximum abundance, has now the highest reported sedimentary abundance for the species in the literature to date. The remaining species abundances remain within the highest reported abundances observed elsewhere. The relationship between sea ice parameters and major Southern Ocean species distribution is illustrated for the first time in this suite of related papers (this work, Crosta et al., this volume; Romero et al., this volume). As expected, all species in this paper reveal a relationship to sea ice cover annually and with respect to the seasonal sea ice concentration. Several species indicate a very strong annual relationship to sea ice duration in particularly Fragilariopsis curta, Fragilariopsis obliquecostata, Fragilariopsis sublinearis, both Porosira species, Stellarima microtrias and Thalassiosira tumida. In these instances, the species abundances sharply increase from no or trace values once more than 7 m/yr of sea ice is present. Within this group, only Porosira glacialis appears to clearly decrease with respect to continued sea ice duration greater than 9 m/ yr, providing it alone with a clear abundance/sea ice duration range of 7.5–9 m/yr. Other species such as Actinocyclus actinochilus, Fragilariopsis cylindrus, Fragilariopsis rhombica and the Thalassiosira antarctica group also indicate a steady increasing abundance response to sea ice duration, which peaks around 8.5–9 m/yr. Some additional species within the 14 discussed may also be susceptible to decreased abundances with increased annual cover over 10 m/yr (i.e. A. actinochilus, Chaetoceros resting spores F. rhombica, Fragilariopsis ritscheri, Fragilariopsis separanda and S. microtrias), although such trends

119

may be an artifact of the sample locations available to us. Two species provided poor relationships to sea ice duration; F. ritscheri and F. separanda. We attribute this poor concurrence to the expatriated occurrences of the species north of the sea ice zone and/or due to taxonomic error between F. kergeulensis and F. ritscheri (Hasle, 1965; Armand, 1997). Chaeotoceros resting spore abundances provided an interesting contrast to the other species’ observations, whereby highest relative abundances were positioned between the values of 3 and 8 m/yr duration. An environmental signature may be present here since the zone of 3–8 m/ yr duration would cover much of the stormy ice melt back zone that would serve as a region of environmental stress conducive to spore formation. However, with no real indication of the species represented by this resting spore category and their true dispersion as individual species, and with the demise of this species category within climatic statistical applications, it is misleading to forward further conjecture without additional life cycle, taxonomic and ecological information on this group. February and September sea ice concentration comparisons against abundance dispersion revealed strongest associations with September sea ice concentration levels within the sea ice related species studied. The majority of species abundances were highest under ice free conditions in the Austral summer (February) with clearly decreasing abundance as sea ice concentration levels increased through unconsolidated (15–30% conc.) to lowest/nil abundances in compact ice conditions with N 30–40% concentration (for example Actinocyclus actinochilus, Fragilariopsis obliquecostata, Fragilariopsis rhombica, Fragilariopsis separanda, Porosira glacialis, Thalassiosira tumida). Some species appeared to equally tolerate uncompacted sea ice cover or icefree conditions (Chaetoceros resting spores, Fragilariopsis curta, Fragilariopsis cylindrus and Stellarima microtrias), where an increase over 30–40% concentration generally signaled a decrease to abundances, with exception to F. cylindrus and Chaetoceros resting spores. The latter of which appeared essentially unassociated with February sea ice concentrations. Fragilariopsis sublinearis, Porosira pseudodenticulata and the Thalassiosira antarctica group are all species that show a slight increase in abundance within uncompacted sea ice concentrations

120

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

(15–40%) during summer. All these species have been reported predominately in land-fast or pack-ice conditions and it may not then be surprising to find that their summer association as viewed by abundances here is similarly tolerant of unconsolidated sea ice conditions, presumably while they are in full reproductive bloom. The least clear association to summer sea ice conditions is observed against abundances of Fragilariopsis ritscheri. Even when disregarding presumed expatriated occurrences, outlying cases of heightened abundances in the unconsolidated sea ice zone make interpretation of this species abundance pattern difficult and we conclude that the species in general appears to prefer ice-free summer conditions. The relationships to winter sea ice conditions viewed through September concentrations can be grouped into two major and two minor responses from the 14 species. Both major responses show a pattern of increased abundances in direct association with consolidated sea ice concentrations particularly commencing around 60–70% concentration. The difference between the major groups is defined on whether the species abundances visibly decrease (e.g. Fragilariopsis curta, Fragilariopsis cylindrus, Fragilariopsis sublinearis, Porosira glacialis, Thalassiosira antarctica group) or remain constant (e.g. Fragilariopsis obliquecostata, Fragilariopsis rhombica, Fragilariopsis separanda, Stellarima microtrias, Thalassiosira tumida and potentially Fragilariopsis ritscheri) when September sea ice concentration peaks above 80%. Again, our division and interpretation may be biased by our current sample coverage near the Antarctic coast. The first of the minor responses is exemplified by Actinocyclus actinochilus’ and possibly F. cylindrus’ step-wise increases in abundance in hand with increases from ice-free to unconsolidated and then consolidated sea ice concentrations. The second form is represented by Chaetoceros resting spores whereby increased abundances range between unconsolidated and consolidated sea ice (30–85% conc.). These differences among the species, as viewed through the three sea ice parameters, provide clear evidence that the use of sea ice related diatom taxa and those that are equally delimited by sea ice cover (Crosta et al., this volume; Romero et al., this volume) is as equally useful for reconstruction or potentially forecasting sea ice conditions through statistical regressions.

4.2. Diatom spatial pattern in the Southern Ocean and ecological conditions Fryxell (1989) reiterated the comment of Semina (1979) that the geography of diatoms as a function of environmental conditions was an urgent problem of that time. There have been small and significant advances in the Southern Ocean region, which includes this current work, driven in part by the use diatoms play in explaining hydrological conditions prior to the introduction to satellite and historical ships records. However, the study of ecological responses, life cycle stages and relationships to physio-chemical cycles at the species level remains one of the more pressing needs for modern phytoplankton studies generally. The 14 species investigated in this work represent those species that are preferentially preserved in the sediments of the Southern Ocean and thus not the true complement of all species present in the sea ice zone and not necessarily representative of the phytoplankton abundances in the ocean surface in which they exist. The biogeographic representations of these sea ice-related species illustrate a suite of variation in their geographical locations, that range from Chaetoceros resting spore presence in most locations increasing southward to localized enclaves of abundance in Antarctic embayments as observed on the extreme with Porosira pseudodenticulata, Stellarima microtrias and Thalassiosira tumida. Although our range of samples within the current winter sea ice edge number 107 sites, our locations are limited around Antarctica to the Antarctic Peninsula, the Ross and Prydz Bay embayments and the East Antarctic coast, which restrict our interpretation of biogeographic range. We have opted to compare our data with previous accounts rather than to include such data to ensure our taxonomic treatment and inclusion of taxa remain constant through the database. Nevertheless, we acknowledge that distributions reported elsewhere compliment the findings reported here and serve as a basis for ongoing studies that will increase the knowledge of qualitative diatom distribution within the Sea Ice Zone. The majority of sea ice related species (i.e. Actinocyclus actinochilus, Fragilariopsis curta, Fragilariopsis cylindrus, Fragilariopsis obliquecostata, Porosira glacialis, Thalassiosira antarctica group) reveal a distribution limited and decreasing in abun-

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

dance northward by the maximum averaged winter sea ice extent. These biogeographic realms fit largely with all reported occurrences detailed. Some species appear limited to the maximum averaged summer sea ice extent and/or embayment regions (i.e. P. glacialis, Porosira pseudodenticulata, Stellarima microtrias and Thalassiosira tumida). Surface water or sea ice studies have indicated that each of these species is most likely to be observed in regions of pack and land fast-ice occurrence (Horner, 1985; Garrison, 1991). It is also important to note that these preserved specimens have been noted as being, or bearing, cold adaptation stages (Fryxell et al., 1984; Fryxell, 1994) in their frustule morphology that inevitably will be linked to such an extreme geo-physical environment. The biogeographic distributions of four Fragilariopsis species (Fragilariopsis curta, Fragilariopsis separanda, Fragilariopsis rhombica and Fragilariopsis ritscheri) and the Thalassiosira antarctica group remain the exceptions to a simplistic sedimentary distribution as a result of occurrences north of the average winter sea ice extent. Fragilariopsis rhombica and F. ritscheri have sporadic occurrences in the South Atlantic and Southwest Pacific regions, respectively, and although such observations have been made for the northward expatriation of F. rhombica elsewhere, it has not been observed in reports of F. ritscheri distribution or with links to Antarctic Bottom Water (AABW) transport northwards (e.g. Burckle and Stanton, 1975; Booth and Burckle, 1976; Stickley et al., 2001; Semina, 2003). In spite of this, F. ritscheri remains a species that exists in regions of bottom water formation along the Antarctic coast and its potential entrainment and transport northward by AABW are possible. Thalassiosira antarctica group specimens are also observed sporadically in the Indian and Atlantic sectors of the Southern Ocean. These northward dispersions have been attributed to surface water or iceberg entrainment (Zielinski and Gersonde, 1997), however, due to their documented environmental relationship with regions of pack and fast-ice, there is room for entrainment by AABW as another mode of dispersion. Similarly, northward dispersion in the South Atlantic of F. curta, although noted, has yet to have it true source conclusively identified. The distribution of F. separanda is far less sporadic and involves abundances up to 5% of the total abundance of samples throughout the Permanently Open Ocean

121

Zone (POOZ). It is difficult to determine without any hard facts on the tolerance limit or reproductive capability of the species in bwarmer zonesQ of the Southern Ocean alongside the contrasting evidence of our Antarctic relative abundance maximums in the sediments and a lack of identification in previous sea ice studies as to whether the species is strictly a sea icerelated species. However, employing our current distribution and the weight of reference to Antarctic off-shore abundance maximums, we believe F. separanda is entrained during bottom-water formation. Fragilariopsis separanda is very heavily silicified and could survive displacement and burial in more northerly regions, which is quite likely the case in the Indian and southeast Indian sectors. The distribution in the South Atlantic sector could be attributed to this factor and iceberg transport. Thus, the species may in fact prove to be an allochthonous species that is unreliable for transfer function work in northern positioned cores.

5. Conclusions This work provides a modern Southern Ocean diatom, sediment-based, biogeography that refers to many separate and local reports on sediment distributions of diatom taxa documented since the work of Kozlova (1966). In addition to the biogeographic distribution, the abundance of diatom species or taxa from our DD228 has been compared to both sea ice and sea-surface temperature data, thus providing indications of environmental pressures that affect diatom distributions at the sea-surface, which are later preserved in the sea-floor sediments. Relationships previously inferred or documented by earlier regional studies concurred largely with the geographic and relative abundances of species observed in sediments of the Southern Ocean, and we are at a point where the evidence exists to show relationships between geography, environment, diatom abundance and diversity are interlocked. We have shown through our database of 228 Southern Ocean samples that diatom distributions in the sediments can provide the information necessary for multivariate statistical application for the benefit of deciphering the environmental conditions of Summer SST and sea ice parameters, such as annual duration and seasonal sea ice concentrations. The 14 sea ice-

122

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

related species in this work reveal individual species level signatures encompassing the environmental conditions in which they are successful in proliferating and being preserved in. The role that sea ice and Antarctic Bottom Water formation play in the expatriation of Antarctic diatoms northward of the Sea Ice Zone, has still to be conclusively proven, yet nevertheless can be tracked through a few species and in particularly through Fragilariopsis separanda. Such expatriation also comes with a cautionary note when utilized in statistical applications to derive past environmental scenarios. We consider most of the distributions comprehensive of current understanding, but expect due to the evolving nature of diatom research, advancing diatom taxonomy and integration of data sets that these distributions and environmental preferences will be refined in the future to the benefit of applied environmental assessment in the past. An important underlying goal of the biogeographic papers presented (this work; Crosta et al., this volume; Romero et al., this volume) was to bring about better awareness of the importance of careful taxonomic identification so that applied environmental analysis can be used more successfully. Our work provides a significant background to future ecological studies on the major species of the Southern Ocean, even though molecular studies are already starting to unravel diatom physiological responses within this extreme ice environment (Mock and Valentin, 2004).

Acknowledgments Discussions with U. Zielinski, L. Burckle, R. Gersonde, A. Leventer, P. De Deckker and D. Roberts during the first author’s research aided the development of this manuscript. We gratefully acknowledge L. Burckle (Lamont-Doherty Earth Observatory), D. Cassidy (FSU Antarctic Research Facility), and R. Gersonde (Alfred-Wegener-Institute), A. Leventer (Colgate University) and F. Taylor (IASOS, University of Tasmania) for their generous contribution of samples and/or percentage data. We thank Brian Harrold (Australian National University, Canberra) Rick Smith, Lisette Robertson and Nerida Bleakley (Antarctic/ACE CRC, Hobart) for help in various drafting stages of the maps presented. The reviewers

of this and the subsequent manuscripts provided helpful reviews and are thanked for their time in evaluating them. Leanne Armand’s research and preparation of the manuscript were funded by an Australian Postgraduate Research Award (ANU, Canberra, Australia) and an Australian Research Council Postdoctoral Fellowship (F39800347) at the Institute of Antarctic and Southern Ocean Studies (IASOS) at the University of Tasmania. Xavier Crosta and Jean-Jacques Pichon’s research was funded by CNRS (Centre National de la Recherche Scientifique), PNEDC (Programme National d’Etude de la Dynamique du Climat) and Missions Scientifiques des Terres Australes et Antarctiques Franc¸aises (IFRTPTAAF). Oscar Romero’s research was funded by SFB261 at the University Bremen.

Appendix A Samples in DD228 with less than 1% relative abundance of the species Fragilariopsis ritscheri Sample code

Latitude (8S)

4759 5008 5156 5201 5248 5348 5402 5406 5413 5535 5634 5638 5719 5817 5847 5853 6044 6217 6221 6222 6238 6317 6318 6343 6352 6410 6419 6435

47.98 50.14 51.93 52.02 52.8 53.81 54.01 54.09 54.22 55.58 56.57 56.63 57.32 58.3 58.78 58.88 60.74 62.3 62.35 62.36 62.64 63.28 63.3 63.72 63.93 64.17 64.31 64.58

Longitude 21.58 6.79 42.88 20.47 54.08 8.22 19.79 0.34 3.52 0.72 34.18 25.72 7.98 16.03 15.43 15.2 86.39 57.62 57.96 57.37 59.54 59.34 117.26 60.05 56.61 56.81 61.88 62.65

Rel. abundance (%) 0.3 0.6 0.3 0.3 0.6 0.6 0.6 0.6 0.6 0.6 0.7 0.3 0.6 0.3 0.3 0.3 0.7 0.3 0.3 0.6 0.6 0.3 0.6 0.6 0.3 0.3 0.6 0.6

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126 Appendix A (continued) Sample code

Latitude (8S)

6445 6449 6450 6501 6545 6652 6711 6731 6749 6758 6801 6804 6806 6808 6811 6828 6837 6839 6840 6842 6843 6858 6914 7601 7605 7610 7621 7630 7631 7637 7641 7657 7711 7721 7730

64.76 64.82 64.83 65.02 65.75 66.87 67.68 67.51 67.62 67.97 68.01 68.08 68.1 68.14 68.18 68.41 68.61 68.65 68.67 68.7 68.71 68.96 69.23 76.02 76.09 76.17 76.35 76.5 76.52 76.62 76.69 76.95 77.17 77.35 77.5

Longitude 62.76 126.72 62.63 63.26 138.2 63.16 68.5 68.2 67.98 67.62 76.55 68.28 72.25 67.7 75.87 72.01 74.52 76.72 77.27 77.51 76.74 75.19 76.1 167.2 166.7 168.96 167.2 166 170.09 164.35 167.82 166.33 165.8 165.88 165.8

Rel. abundance (%) 0.5 0.7 0.3 0.6 0.9 0.3 0.3 0.6 0.3 0.6 0.3 0.3 0.3 0.3 0.6 0.7 0.9 0.3 0.6 0.6 0.9 0.3 0.3 0.3 0.3 0.6 0.6 0.6 0.3 0.6 0.3 0.6 0.6 0.6 0.3

References Abbott Jr., W.H., 1973. Vertical and lateral patterns of diatomaceous ooze found between Australia and Antarctica. PhD Thesis, Department of Geology, University of South Carolina, USA. Abelmann, A., Gersonde, R., 1991. Biosiliceous particle flux in the Southern Ocean. Marine Chemistry 35, 503 – 536. Andreoli, C., Tolomio, C., Moror, I., Radice, M., Moschin, E., Bellato, S., 1995. Diatoms and dinoflagellates in Terra Nova Bay (Ross Sea–Antarctica) during austral summer 1990. Polar Biology 15, 465 – 475. Armand, L.K., 1997. The use of diatom transfer functions in estimating sea-surface temperature and sea-ice in cores from the southeast Indian Ocean. PhD Thesis, Australian National University, Canberra, Australia.

123

Armand, L.K., 2000. An ocean of ice-advances in the estimation of past sea-ice in the Southern Ocean. GSA Today 10 (3), 1 – 7. Armand, L.K., Leventer, A., 2003. Palaeo sea ice distribution– reconstruction and palaeoclimatic significance. In: Thomas, D., Dieckmann, G. (Eds.), Sea Ice: Physics, Chemistry and Biology. Blackwell Science Ltd, Oxford, pp. 333 – 372. Booth, J.D., Burckle, L.H., 1976. Displaced Antarctic diatoms in the southwestern and central Pacific. Pacific Geology 11, 99 – 108. Burckle, L.H., Stanton, D., 1975. Distribution of displaced Antarctic diatoms in the Argentine Basin. Third Symposium on Recent and Fossil Marine Diatoms, Kiel, F.R. Germany, 283 – 292. Cassie, V., 1963. Distribution of surface phytoplankton between New Zealand and Antarctica December 1957. Trans-Antarctic Expedition 1955–1958 Scientific Reports 7. Cavalieri, D.J., St. Germain, K.M., Swift, C.T., 1995. Reduction of weather effects in the calculation of sea ice concentration with the DMSP SSM/I. Journal of Glaciology 41 (139), 455 – 464. Clarke, D.B., Ackley, S.F., 1983. Relative abundance of diatoms in Weddell Sea pack ice. Antarctic Journal of the United States 18, 181 – 182. Crosta, X., Pichon, J.-J., Labracherie, M., 1997. Distribution of Chaetoceros resting spores in modern peri-Antarctic sediments. Marine Micropaleontology 29, 283 – 299. Crosta, X., Pichon, J.-J., Burckle, L.H., 1998a. Application of modern analog technique to marine Antarctic diatoms: reconstruction of maximum sea-ice extent at the Last Glacial Maximum. Paleoceanography 13 (3), 286 – 297. Crosta, X., Pichon, J.-J., Burckle, L.H., 1998b. Reappraisal of Antarctic seasonal sea-ice at the last glacial maximum. Geophysical Research Letters 25 (14), 2703 – 2706. Crosta, X., Sturm, A., Armand, L., Pichon, J.-J., 2004. Late quaternary sea ice history in the Indian sector of the Southern Ocean as recorded by diatom assemblages. Marine Micropalaeontology 50, 209 – 226. Crosta, X., Romero, O.E., Armand, L.K., Pichon, J.-J., this volume. The biogeography of major diatom taxa in Southern Ocean surface sediments: 2. Open-ocean related species. Palaeogeography, Palaeoclimatology, Palaeoecology. Cunningham, W., Leventer, A., 1998. Diatom assemblages in surface sediments of the Ross Sea: relationship to present oceanographic conditions. Antarctic Science 10 (2), 134 – 146. DeFelice, D.R., Wise, S.W.J., 1981. Surface lithofacies, biofacies and diatom diverity patterns as models for delineation of climatic change in the southeast Atlantic Ocean. Marine Micropaleontolgy 6, 29 – 70. Donahue, J.G., 1973. Distribution of planktonic diatoms in surface sediments of the Southern South Pacific. In: Goodell, H.G. (Ed.), Marine Sediments of the Southern Ocean, American Geophysical Society, Antarctic Map Folio Series, vol. 17, p. 18. pl. 9. Donegan, D., Schrader, H., 1982. Biogenic and abiogenic components of laminated hemipelagic sediments in the central Gulf of California. Marine Geology 48, 215 – 237. Fryxell, G.A., 1989. Marine phytoplankton at the Weddell Sea ice edge: seasonal changes at the specific level. Polar Biology 10, 1 – 18.

124

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

Fryxell, G.A., 1991. Comparison of winter and summer growth stages of the diatom Eucampia antarctica from the Kerguelen Plateau and south of the Antarctic Convergence Zone. In: Barron, J., Larsen, B., et al. (Eds.), Proc. Ocean Drilling Program, Scientific Results, vol. 119. Ocean Drilling Program, College Station, pp. 675 – 685. Fryxell, G.A., 1994. Planktonic marine diatom winter stages: Antarctic alternatives to resting spores. Memoirs of the California Academy of Sciences 17, 437 – 448. Fryxell, G.A., Hasle, G.R., Carty, S.V., 1984. Thalassiosira tumida (Janisch) Hasle: observations from field and clonal cultures. In: Ricard, M. (Ed.), 8th International Diatom Symposium, pp. 11 – 21. Garrison, D.L., 1991. Antarctic sea ice biota. American Zoologist 31, 17 – 33. Garrison, D.L., Buck, K.R., 1989. The biota of Antarctic pack ice in the Weddell Sea and Antarctic Peninsula regions. Polar Biology 10, 211 – 219. Garrison, D.L., Close, A.R., 1993. Winter ecology of the sea ice biota in Weddell Sea pack ice. Marine Ecology Progress Series 96, 17 – 31. Garrison, D.L., Buck, K.R., Silver, M.W., 1982. Ice algal communities in the Weddell Sea. Antarctic Journal of the United States 17, 157 – 159. Garrison, D.L., Buck, K.R., Silver, M.W., 1983. Studies of ice-algal communities in the Weddell Sea. Antarctic Journal of the United States 18, 179 – 181. Garrison, D.L., Buck, K.R., Fryxell, G.A., 1987. Algal assemblages in Antarctic pack ice and in ice-edge plankton. Journal of Phycology 25, 564 – 572. Garrison, D.L., Close, A.R., Reimnitz, E., 1989. Algae concentrated by frazil ice: evidence from laboratory experiments and field measurements. Antarctic Science 1, 313 – 316. Gersonde, R., 1984. Siliceous microorganisms in sea ice and their record in sediments in the southern Weddell Sea (Antarctica). Proceedings 8th Diatom Symposium Paris, Koeltz Koenigstein, 549 – 566. Gersonde, R., Crosta, X., Abelmann, A., Armand, L., 2005. Seasurface temperature and sea ice distribution of the Southern Ocean at the EPILOG Last Glacial Maximum—a circumAntarctic view based on siliceous microfossil records. Quaternary Science Reviews 24, 869 – 896. Gersonde, R., Wefer, G., 1987. Sedimentation of biogenic siliceous particles in Antarctic waters from the Atlantic sector. Marine Micropaleontology 11, 311 – 322. Gloersen, P., Cavalieri, D.J., 1986. Reduction of weather effects in the calculation of sea ice concentration from microwave radiances. Journal of Geophysical Research 91 (C3), 3913 – 3919. Gloersen, P., Campbell, W.J., Cavalieri, D.J., Comiso, J.C., Parkinson, C.L., Zwally, H.J., 1992. Arctic and Antarctic Sea Ice, 1978–1987: satellite passive-microwave observations and analysis. NASA Scientific and Technical Information Program, NASA SP-511. Washington. Hasle, G.R., 1965. Nitzschia and Fragilariopsis species studied in the light and electron microscopes: III. The genus Fragilariopsis. Skrifter utgitt av det Norske Videnskaps-Akademi. I. Matematisk-Naturvidenskaplig Klasse 21, 1 – 49 (pl. 1–17).

Hasle, G.R., 1972. Thalassiosira subtilis (Bacillariophyceae) and two allied species. Norwegian Journal of Botany 19, 111 – 137. Hasle, G.R., 1973. Some marine plankton genera of the diatom family Thalassiosiraceae. Second symposium on recent and fossil marine diatoms London, Sept. 4–9, 1972. Beihefte zur Nova Hedwigia 45, 1 – 49 (pl. 1–18.) Hasle, G.R., 1976. The biogeography of some marine planktonic diatoms. Deep-Sea Research 23, 319 – 338. Hasle, G.R., Heimdal, B.R., 1968. Morphology and distribution of the marine centric diatom Thalassiosira antarctica Comber. Journal of the Royal Microscopical Society 88 (3), 357 – 369. Hasle, G.R., Heimdal, B.R., Fryxell, G.A., 1971. Morphologic variability in fasciculated diatoms as exemplified by Thalassiosira tumida (Janisch) Hasle, comb. nov. In: Llano, G.A., Wallen, I.E. (Eds.), Biology of the Antarctic Seas IV, Antarctic Research Series, vol. 17. American Geophysical Union, Washington, pp. 313 – 333. Hasle, G.R., Sims, P.A., Syvertsen, E.E., 1988. Two recent Stellarima species: S. microtrias, S. stellaris (Bacillariophyceae). Botanica Marina 31, 195 – 206. Hargraves, P.E., 1979. Studies on marine plankton diatoms: IV. Morphology of Chaetoceros resting spores. Beiheft zur Nova Hedwigia 64, 99 – 120. Horner, R., 1985. Sea Ice Biota. CRC Press, Boca Ranton. Hustedt, F., 1958. Diatomeen aus der Antarktis und dem sqdatlantik. Reprinted from bDeutsche Antarktishe Expedition 1938/1939Q Band II, Geographisch-kartographische Anstalt bMundusQ. Hamburg. 191 pp, pl. 3–13 (in German). Johansen, J.R., Fryxell, G.A., 1985. The genus Thalassiosira (Bacillariophyceae): studies on species occurring south of the Antarctic convergence zone. Phycologia 24 (2), 155 – 179. Johansen, J.R., Doucette, G.J., Fryxell, G.A., 1985. The genus Thalassiosira (Bacillariophyceae): morphology of heterovalvate resting spores of T. scotia. American Journal of Botany 72 (12), 1861 – 1870. Jouse´, A.P., Koroleva, G.S., Nagaeva, G.A., 1962a. Diatoms in the surface layer of sediment in the Indian sector of the Antarctic. Trudy Instituta Okeanologii Akademiya Nauk SSSR 61, 20 – 91. Jouse´, A.P., Koroleva, G., Nagaeva, G.A., 1962b. Stratigraphic and paleographic investigations in the Indian sector of the Southern Ocean. Akademiia NAUK SSSR, Moscow 8, 137 – 161. Kaczmarska, I., Barbrick, N.E., Ehrman, J.M., Cant, G.P., 1993. Eucampia Index as an indicator of the Late Pleistocene oscillations of the winter sea-ice extent at the ODP Leg 119 Site 745B at the Kerguelen Plateau. Hydrobiologia 269/270, 103 – 112. Kang, S.-H., Fryxell, G.A., 1992. Fragilariopsis cylindrus (Grunow) Krieger: the most abundant diatom in water column assemblages of Antarctic marginal ice-edge zones. Polar Biology 12, 609 – 627. Kang, S.-H., Fryxell, G.A., 1993. Phytoplankton in the Weddell Sea, Antarctica: composition, abundance and distribution in water-column assemblages of the marginal ice-edge zone during austral autumn. Marine Biology 116, 335 – 348. Kang, S.-H., Fryxell, G.A., Roelke, D.L., 1993. Fragilariopsis cylindrus compared with other species of the diatom family

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126 Bacillariaceae in Antarctic marginal ice-edge zones. Nova Hedwigia Heft 106, 335 – 352. Karpuz, N.K., Jansen, E., 1992. A high-resolution diatom record of the last deglaciation from the SE Norwegian Sea: documentation of rapid climatic changes. Paleoceanography 7 (4), 499 – 520. Karpuz, N.K., Schrader, H., 1990. Surface sediment diatom distribution and Holocene paleotemperature variations in the Greenland, Iceland and Norwegian Sea. Paleoceanography 5 (4), 557 – 580. Kellogg, D.A., Kellogg, T.B., 1987. Microfossil distributions in modern Amundsen Sea sediments. Marine Micropaleontology 12, 203 – 222. Kozlova, O.G., 1966. Diatoms of the Indian and Pacific sectors of the Antarctic. National Science Foundation, Israel Program for Scientific Translations, Washington, DC. 185 pp, pl. I–VI. Kozlova, O.G., Muchina, V.V., 1967. Diatoms and silicoflagellates in suspension and floor sediments of the Pacific Ocean. International Geology Review (translated from Geokhimiya Kremnezema, N.K. 65–12(51), 192–218, 1967) 9, 1322–1342. Krebs, W.N., Lipps, J.H., Burckle, L.H., 1987. Ice diatom floras, Arthur Harbour, Antarctica. Polar Biology 7, 163 – 171. Labeyrie, L.D., Labracherie, M., Gorfti, N., Pichon, J.-J., Vautravers, M., Arnold, M., Duplessy, J.-C., Paterne, M., Michel, E., Duprat, J., Caralp, M., Turon, J.-J., 1996. Hydrographic changes of the Southern Ocean (southeast Indian sector) over the last 230 kyr. Paleoceanography 11 (1), 57 – 76. Laws, R.A., 1983. Preparing strewn slides for quantitative microscopical analysis: a test using calibrated microspheres. Micropaleontology 29, 60 – 65. Leventer, A., 1991. Sediment trap diatom assemblages from the northern Antarctic Peninsula region. Deep-Sea Research 38 (8/9), 1127 – 1143. Leventer, A., 1992. Modern distribution of diatoms in sediments from the George V Coast, Antarctica. Marine Micropaleontology 19, 315 – 332. Leventer, A., 2003. Particulate flux from sea ice in polar waters. In: Thomas, D.N., Dieckmann, G.S. (Eds.), Sea Ice. An Introduction to its Physics, Chemistry, Biology and Geology. Blackwell Science Ltd, Oxford, pp. 303 – 332. Leventer, A., Dunbar, R.B., 1996. Factors influencing the distribution of diatoms and other algae in the Ross Sea. Journal of Geophysical Research 101 (C8), 18489 – 18500. Leventer, A., Dunbar, R., DeMaster, D.J., 1993. Diatom evidence for late Holocene climatic events in Granite Harbor, Antarctica. Paleoceanography 8 (3), 373 – 386. Leventer, A., Domack, E.W., Ishman, S.E., Brachfeld, S., McClennen, C.E., Manley, P., 1996. Productivity cycles of 200–300 years in the Antarctic Peninsula region: understanding linkages among the sun, atmosphere, oceans, sea ice, and biota. Geological Society of America, Bulletin 108 (12), 1626 – 1644. Ligowski, R., 1993. Marine diatoms (Bacillariophyceae) in the Antarctic ecosystem and their importance as an indicator of food soure of krill (Euphausia superba Dana). Wydawnictwo Uniwersytetu Lo´dzkiego, Lo´dz´, Poland, pp. 242 (In Polish with English summary).

125

McConville, M., Wetherbee, R., 1983. The bottom-ice microalgal community from annual ice in the inshore waters of East Antarctica. Journal of Phycology 19, 431 – 439. Mock, T., Valentin, K., 2004. Photosynthesis and cold acclimation: molecular evidence from a polar diatom. Journal of Phycology 40, 732 – 741. Orsi, A.H., Whitworth III, T., Nowlin, W.D., 1995. On the meridonal extent and fronts of the Antarctic circumpolar current. Deep-Sea Research 42 (5), 641 – 673. Pichon, J.-J., Labeyrie, L.D., Bareille, G., Labracherie, M., Duprat, J., Jouzel, J., 1992a. Surface water temperature changes in the high latitudes of the Southern Hemisphere over the last glacial–interglacial cycle. Paleoceanography 7 (3), 289 – 318. Pichon, J.-J., Bareille, G., Labracherie, M., Labeyrie, L.D., Baudrimont, A., Turon, J.-L., 1992b. Quantification of the biogenic silica dissolution in Southern Ocean sediments. Quaternary Research 37, 361 – 378. Pokras, E.M., Molfino, B., 1986. Oceanographic control of diatom abundances and species distributions in surface sediments of the tropical and southeast Atlantic. Marine Micropaleontology 10, 165 – 188. Prasad, A.K.S.K., Nienow, J.A., 1986. Marine diatoms of sediments from Croft Bay, Antarctica. Antarctic Journal of the United States 21, 157 – 159. Rathburn, A.E., Pichon, J.J., Ayress, M.A., DeDeckker, P., 1997. Microfossil and stable-isotope evidence for changes in Late Holocene paleoproductivity and paleoceanographic conditions in the Prydz Bay region of Antarctica. Paleogeography, Paleoclimatology, Paleoecology 131, 485 – 510. Riaux-Gobin, C., 1996. Deep subantarctic sediments: revival tests and comparison of diatom spores from cultures and sediment samples. 14th International Diatom Symposium, Tokyo 1996, Tokyo Diatom Institute, p. 99. Romero, O., Hensen, C., 2002. Oceanographic control of biogenic opal and diatoms in surface sediments of the Southwestern Atlantic. Marine Geology 186, 263 – 280. Romero, O., Armand, L.K., Crosta, X., Pichon, J.J., this volume. The biogeography of major diatom taxa in Southern Ocean sediments: 3. Tropical/Subtropical Species. Palaeogeography, Palaeoclimatology, Palaeoecology. Sancetta, C., Heusser, L., Hall, M.A., 1992. Late Pliocene climate in the Southeast Atlantic: preliminary results from a multidisciplinary study of DSDP Site 532. Marine Micropaleontology 20, 59 – 75. Schrader, H.J., Gersonde, R., 1978. Diatoms and Silicoflagellates. Micropaleontological counting methods and techniques—an exercise on an eight metres section of the Lower Pliocene of Capo Rossello, Sicily. In: Zachariasse, W., Riedel, J., Sanfilippo, R., et al. (Eds.), Utrecht Micropaleontological Bulletin, vol. 17, pp. 129 – 176. Schrader, H., Karpuz, N.K., 1990. Norwegian–Iceland seas: transfer functions between marine planktic diatoms and surface water temperature. In: Bleil, U., Thiede, J. (Eds.), Geological history of the Polar Oceans: Arctic versus Antarctic, vol. 308. Kluwer Academic Publishers, Dordrecht, pp. 337 – 361.

126

L.K. Armand et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 223 (2005) 93–126

Schrader, H., Sorknes, R., 1991. Peruvian coastal upwelling: late quaternary productivity changes revealed by diatoms. Marine Geology 97, 233 – 249. Schuette, G., Schrader, H., 1979a. Diatom taphocoenoses in the coastal upwelling area off western South America. Nova Hedwigia 64, 359 – 378. Schuette, G., Schrader, H., 1979b. Diatom taphocoenoses in the coastal upwelling areas off western South America. Data Report 73, Reference 79-8. Oregon State University. Schuette, G., Schrader, H., 1981. Diatom taphocoenoses in the coastal upwelling area off south west Africa. Marine Micropalaeontology 6, 131 – 155. Schweitzer, P.N., 1995. Monthly averaged polar sea-ice concentration. U.S. Geological Survey Digital Data Series: Virginia. CD rom, Ed 1. DDS-27. Scott, P., McMinn, A., Hosie, G., 1994. Physical parameters influencing diatom community structure in eastern Antarctic sea ice. Polar Biology 14, 507 – 517. Semina, H.J., 1979. The geography of plankton diatoms of the Southern Ocean. Beiheft zur Nova Hedwigia 64, 341 – 358. Semina, H.J., 2003. SEM-studied diatoms of different regions of the World Ocean. Iconographia Diatomologica 10, 1 – 362. Stickley, C.E., Carter, L., McCave, N., Weaver, P.E., 2001. Lower circumpolar deep water flow through the SW Pacific gateway for the last 190 ky: evidence from Antarctic diatoms. The oceans and rapid climate change: past, present, and future. Geophysical Monograph 126, 101 – 116. Stockwell, D.A., Hargraves, P.E., 1984. Morphological variability within resting spores of the marine diatom genus Chaetoceros Ehrenberg. Proceedings of the Eighth International Diatom Symposium, Paris, Koenigstein, 81 – 95. Stockwell, D.A., Kang, S.-H., Fryxell, G.A., 1991. Comparisons of diatom biocoenoses with Holocene sediment assemblages in Prydz Bay, Antarctica. In: Barron, J., Larsen, B., et al. (Eds.), Proceedings of the Ocean Drilling Program. Scientific Results, vol. 119. Ocean Drilling Program, College Station, TX, pp. 667 – 673. Syvertsen, E.E., 1985. Resting spore formation in the Antarctic diatoms Coscinodiscus furcatus Karsten and Thalassiosira australis Pergallo. Polar Biology 4, 113 – 119. Tanimura, Y., 1992. Distribution of diatom species in the surface sediments of Lqtzow-Holm Bay, Antarctica. In: Ishizaki, K., Saito, T. (Eds.), Centenary of Japanese Micropaleontology. Terra Scientific Publishing Company, Tokyo, pp. 399 – 411.

Tanimura, Y., Fukuchi, M., Watanabe, K., Moriwaki, K., 1990. Diatoms in water column and sea-ice in Lqtzow-Holm Bay, Antarctica, and their preservation in preservation in the underlying sediments. Bulletin of the National Science Museum, Tokyo, Series C 18 (1), 15 – 39. Taylor, F., McMinn, A., Franklin, D., 1997. Distribution of diatoms in surface sediments of Prydz Bay, Antarctica. Marine Micropaleontology 32, 209 – 229. Tre´ guer, P., Jacques, G., 1992. Dynamics of nutrients and phytoplankton, and fluxes of carbon, nitrogen and silicon in the Antarctic Ocean. Polar Biology 12, 149 – 162. Treppke, U.F., Lange, C.B., Wefer, G., 1996. Vertical fluxes of diatoms and silicoflagellates in the eastern equatorial Atlantic, and their contribution to the sedimentary record. Marine Micropalaontology 28, 73 – 96. Truesdale, R.S., Kellogg, T., 1979. Ross Sea diatoms: modern assemblage distributions and their relationship to ecologic, oceanographic, and sedimentary conditions. Marine Micropaleontology 4, 13 – 31. van Iperen, J.M., van Weering, T.C.E., Jansen, J.H.F., van Bennekom, A.J., 1987. Diatoms in surface sediments of the Zaire deep-sea fan (SE Atlantic Ocean) and their relation to overlying water masses. Netherlands Journal of Sea Research 21 (3), 203 – 217. van Iperen, J.M., van Bennekom, A.J., van Weering, T.C.E., 1993. Diatoms in surface sediments of the Indonesian Archipelago and their relation to hydrography. Hydrobiologia 269/270, 113 – 128. Villareal, T.A., Fryxell, G.A., 1983. The Genus Actinocyclus (Bacillariophyceae): frustule morphology of A. sagittulus sp. nov. and two related species. Journal of Phycology 19, 266 – 452. Watanabe, K., 1982. Centric diatom communities found in the Antarctic sea ice, Antarctic Record. National Institute of Polar Research 74, 119 – 126. World Ocean Atlas, 1994. NOAA(NODC) CD-rom no. 1. Zielinski, U., 1993. Quantitative estimation of palaeoenvironmental parameters of the Antarctic surface water in the Late Quaternary using transfer functions with diatoms. Berichte zur Polarforschung 126, 1 – 148. Zielinski, U., Gersonde, R., 1997. Diatom distribution in Southern Ocean surface sediments (Atlantic sector): implications for paleoenvironmental reconstructions. Palaeogeography, Palaeoclimatology, Palaeoecology 129, 213 – 250.