Journal of Marine Systems 78 (2009) 599–605

Contents lists available at ScienceDirect

Journal of Marine Systems

j ourna l homepage: www.e lsev ie r.com/ locate / jmarsys

Annual cycle of distribution of three ice-associated copepods along the coast nearDumont d'Urville, Terre Adélie (Antarctica)

Christophe Loots a, Kerrie M. Swadling b,⁎, Philippe Koubbi c

a IFREMER, Laboratoire de Ressources Halieutiques, 150, Quai Gambetta, BP 699, 62321 Boulogne-sur-mer Cedex, Franceb Tasmanian Aquaculture and Fisheries Institute and School of Zoology, University of Tasmania, Private Bag 49, Hobart, Tasmania 7001, Australiac Laboratoire d'Océanographie de Villefranche, CNRS UMR 7093, Université Paris VI, Station Zoologique, La Darse, BP 28, 06230 Villefranche-Sur-Mer, France

⁎ Corresponding author. Tel.: +61 6227 7218; fax: +6E-mail addresses: Christophe.Loots@ifremer.fr (C. Lo

Kerrie.Swadling@utas.edu.au (K.M. Swadling), koubbi@o

0924-7963/$ – see front matter © 2009 Elsevier B.V. Aldoi:10.1016/j.jmarsys.2009.01.003

a b s t r a c t

a r t i c l e i n f o

Article history:

In polar regions sea ice is Received 22 March 2007Received in revised form 23 October 2007Accepted 12 January 2009Available online 21 February 2009

Keywords:Paralabidocera antarcticaDrescheriella glacialisStephos longipesSouthern OceanSea iceLife cycles

a site of enhanced primary production during winter and provides importanthabitat for small grazers, such as copepods. We sampled zooplankton from the sea ice and water columnthroughout 2005, near Dumont d'Urville station (Terre Adélie, Antarctica). Three species of ice-associatedcopepods were found: two calanoid copepods Paralabidocera antarctica and Stephos longipes and theharpacticoid copepod Drescheriella glacialis. P. antarcticawas the most abundant of the three and was closelyassociated with the sea ice during most of the year. This species had a one year life cycle with a probableover-wintering period in the sea ice as nauplii and a short copepodite phase in spring. Reproduction andspawning occurred in early summer. A comparison with two other populations (near Syowa and Davisstations) along the east coast of Antarctica showed that there was a temporal shift in the life cycles of thethree populations, which was linked to variability in sea ice conditions. D. glacialis was the second mostabundant copepod and was more common during the winter than during summer, indicating its preferencefor the sea ice habitat. In autumn, the presence of D. glacialis in the deeper part of the water columnsuggested that this species colonised the sea ice from the benthos. S. longipes was found only in the watercolumn near Dumont d'Urville and was not very abundant. In Terre Adélie particular environmentalconditions, such as the absence of a permanent sea ice zone throughout the year, a longer time of openwater,strong katabatic winds and the presence of polynyas, have influenced both the abundance and distribution ofthe three common ice-associated copepods.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

Sea-ice is the dominant feature of the Antarctic marine ecosystem,covering up to 20 million km2 at its maximum extent (Fetterer andKnowles, 2002; Comiso, 2003) (Fig. 1). The ice plays a critical role inlarge-scale atmospheric and oceanographic processes and, at smallerscales, provides an important habitat for many plants and animals.The interstitial spaces that form the brine channel system in sea iceprovide both a refuge from predators and a foraging ground for smallgrazers. Ice algae can grow at very low levels of illumination andextend their growing seasonwell beyond that of phytoplankton in thewater column. For example, algae were present in measurablequantities in spring fast ice at least one month before phytoplanktonappeared in the water column near Davis Station (68°35′S, 77°58′E,Fig. 1). Similarly, in autumn the growth of ice algae extended wellbeyond that of phytoplankton (Swadling et al., 2004). Another

1 6227 8035.ots),bs-vlfr.fr (P. Koubbi).

l rights reserved.

possible function of sea ice for small grazers is as a refuge frompredation. Fish such as Pagothenia borchgrevinki and Trematomusnewnesi are associated with the undersurface of ice in coastal habitats(Williams, 1988; Dewitt et al., 1990; Vacchi and La Mesa, 1995). Thesespecies occupy crevices in the ice and prey on small copepods that livein or near the under ice surface (Hoshiai et al., 1989; La Mesa et al.,2000).

The biomass of metazoan grazers found associated with sea ice isdominated by crustaceans. Small copepods are abundantwithin the seaice interstices, while amphipods, copepods and euphausiids areimportant at the ice–water interface (Arndt and Swadling, 2006). Todate, the threemost frequently observed copepod species have been thecalanoids Stephos longipes and Paralabidocera antarctica and theharpacticoid Drescheriella glacialis (e.g. Hoshiai and Tanimura, 1986;Swadling et al., 1997, 2000; Schnack-Schiel et al., 2001), and the highestabundances have been recorded from Prydz Bay (Swadling et al., 1997,2000). Although all three species appear to have a circum-Antarcticdistribution there are differences in their dominance around Antarctica.S. longipes dominates in the Weddell, Amundsen and BellingshausenSeas (Fig. 1), reaching abundances up to 200,000 individuals m−2

(Schnack-Schiel et al., 1995, 1998), while along the coast of eastAntarctica it has not been observed to exceed 20,000 individuals m−2

Fig. 1. Maps of sea-ice extent in January (A) and July (B) 2005 in Antarctica (ftp://sidads.colorado.edu/DATASETS/NOAA/G02135/). Shows location of places, oceans and main seasdescribed in the text.

600 C. Loots et al. / Journal of Marine Systems 78 (2009) 599–605

(Hoshiai and Tanimura, 1986; Swadling et al., 2000). P. antarctica isabundant in the coastal fast ice of eastern Antarctica, where its naupliican reach up to 900,000 individuals m−2 (Hoshiai and Tanimura, 1986;Tanimura et al., 1996; Swadling et al., 1997, 2000). This species has beenobserved in much lower numbers in the Weddell Sea: up to26,000 individuals m−3 were recorded from platelet ice in DrescherInlet (Günther et al., 1999). Finally, the most widely distributedharpacticoid copepod in Antarctic sea ice is D. glacialis. It is commonin the Bellingshausen and Amundsen Seas, the Weddell Sea and alongthe coast of east Antarctica (e.g. Dahms et al.,1990; Schnack-Schiel et al.,1998; Swadling, 2001, Fig. 1).

Despite the knowledge that we now have about the broadscaledistribution of these key ice-associated species, there are regionswhere little, if any, information exists. The objective of the presentstudy was to examine the abundance and distribution of the threespecies over an annual cycle along the coast of Terre Adélie (Fig. 1).The aims were firstly to document the occurrences of the copepodsand secondly to compare their distribution and life cycles at this sitewith other regions along the Antarctic coast.

Fig. 2. Location of the sampling site in the Pointe Geologie Archipelago (Terre Adélie).

2. Materials and methods

2.1. Study site

A coastal sampling site facing Jacobsen rock on main Petrels Island(66°40 S, 140°E; Fig. 2) was chosen for the temporal survey. This sitewas located on the Eastern part of Pierre Lejay Bay in a small trench of40–60m depth running northward between the islands. The samplingsite had a water depth of 41 m.

In Terre Adélie the sea-ice begins to form in mid March, when airtemperatures decrease to approximately−15 °C. Sea ice growth alongTerre Adélie is unpredictable as a result of variable meteorological andoceanic conditions. The region is one of the windiest in Antarctica,with frequent katabatic winds, i.e. downslope gravitational strongwinds of cold air that come from the continent (Wendler et al., 1997).A maximum of 187 km h−1 was reached in August 2005. Theoccurrence of these katabatic winds means that the sea ice oftenbreaks out frequently during the period fromMarch to July or, in somecases, the ice just weakens, creating ice-free areas known as polynyas.Polynyas are particularly pronounced to the north of Terre Adélie.Strong winds concentrate snow into patches, which induces a rise intemperatures locally and causes the sea-ice to melt at its surface.Finally, in late spring, atmospheric depressions over Terre Adélie,combined with oceanic swell, destabilise the sea-ice even further andthe ice usually breaks out by late November. During summer thearchipelago is always surrounded by open water, along with icebergsthat have calved from the Astrolabe Glacier (Fig. 2).

Fig. 3. Annual development of snow and sea-ice thickness at the sampling site fromNovember 2004 to December 2005. (OW=open water).

Fig. 4. Paralabidocera antarctica. Total abundance in the sea-ice (A) and in the seawater(B) from November 2004 to December 2005.

601C. Loots et al. / Journal of Marine Systems 78 (2009) 599–605

2.2. Sampling

For most of the year access to the site was over the fast ice, while asmall boat was used during the open water period in summer. Thirty-five sampling trips were conducted between November 2004 andDecember 2005. Samples were collected weekly, subject to meteor-ological conditions, except at the beginnings of December 2004 and2005 and from March to mid-April 2005 when ice and sea conditionswere not suitable for field work.

A 2 m long umbrella net (100 µm mesh) was used to sample mostof thewater column (0–35m). A closing-WP2 net (200 µmmesh)was

Fig. 5. Paralabidocera antarctica. Proportions (%) of different developmental sta

used to sample two layers of the water column: the ice–waterinterface (0–2 m; hereafter designated as WP2 surface) and the depthinterval 2–35 m (designated as WP2 deep). From April to the end ofNovember 2005, sea-ice cores were collected with an ice-corer(diameter of 7.5 cm). Snow and sea-ice thickness were measured foreach core. The bottom 20 cm of each core was retained and melted in500 ml GF/F filtered seawater. The water was filtered through 200 µmmesh sieves to retrieve the zooplankton. It is possible that somesmaller naupliar stages were lost through this mesh size and so werenot sampled quantitatively. Samples of zooplankton collectedwith thenets and from the ice cores were preserved in 5% borax-bufferedformalin-seawater.

All counting and identification of the sampleswas performed usinga Leica MZ 6 binocular microscope. When abundances were high, thesample was split with a Motoda box to obtain at least 200 specimensper species for each sub-sample. P. antarctica were sorted intodevelopmental stages (including sex for stages CVI; Tanimura,1992). The other two species, S. longipes and D. glacialis, were notsorted into stages. Counts were transformed into abundances basedon the vertical distance covered by the mouths of the nets; thismethod assumed 100% filtration efficiency (Tranter and Smith, 1968).Calculated volumes filteredwere 997.5 L for the umbrella net, 566 L forthe WP2 surface and 9330 L for the WP2 deep. Abundances ofcopepods in the sea ice cores were directly calculated per litre fromthe filtered volume and permetre square from the diameter of the ice-corer.

3. Results

3.1. Snow and ice

The ice started to grow in mid-March 2005 and by mid-April hadgrown to a thickness of 40 cm (Fig. 3). Growth then slowed down andthe ice reached 70 cm by the end of June. A maximum thickness of1.5 m was reached by the end of September. At the end of Novemberthe sea-ice broke out. The ice cover consisted largely of congelation icecomposed of long, columnar crystals.

ges (NIV–CVI) recorded in the sea-ice on several sampling dates in 2005.

Fig. 6. Paralabidocera antarctica. Proportions (%) of different stages (NVI–CVI) recorded in the seawater (WP2 surface and umbrella net) on several sampling dates in 2004 and 2005.

Fig. 7. Drescheriella glacialis. Abundances in the sea-ice (A) and in the seawater (B, C, D),from November 2004 to December 2005.

602 C. Loots et al. / Journal of Marine Systems 78 (2009) 599–605

The thickness of snow cover on the ice was highly patchy (Fig. 3).When the sea ice first formed there was no snow cover, then frommid-April to mid-June therewas a period when snow cover was stableat about 30 cm. During the winter snow thickness varied from 0 to70 cm, as a result of the katabatic winds blowing it off the ice.Furthermore, during this period, there was high spatial variability insnow cover: areas without snow cover and solid sea ice wereinterspersed with areas of 70 cm snow cover and melting ice.

3.2. Copepod abundance

3.2.1. P. antarcticaFrom January to August 2005 no specimens of P. antarctica were

found in the sea-ice or in the seawater.In the sea ice, from August to the end of November, this species

showed three successive peaks in abundance (Fig. 4A): one in Augustwith a maximum of 10,548 individuals m−2 (51 individuals l−1), asecond at the end of September with 8740 individuals m−2

(28 individuals l−1), and a third in October–November with a max-imum of 25,617 individuals m−2 (65 individuals l−1). In August thepopulation of P. antarctica was composed of late naupliar stages(Fig. 5A): NIV (4%), NV (34%) and NVI (62%). During the time of peakabundance at the end of September naupliar stages were stilldominant (83%), although the first copepodite stages, CI and CII(17%), had also appeared (Fig. 5B). Development through successivecopepodite stages occurred from October to the end of November(Fig. 5C–G). In October, CIII–CVI developed rapidly (Fig. 5C–E),although adult stages were not yet dominant. By November (Fig. 5Fand G) CV and adult stages were dominant. CVI females were foundwith spermatophores attached and were more abundant than CVImales.

From 0 to 2 m in the water column, (WP2 surface, Fig. 4B) therewas a peak in abundance (563 individuals m−3) at the end ofNovember 2004, just as the sea ice started to break out. Onlycopepodite stages were found and the population of P. antarctica wasmainly composed of CVI males (N70%; Fig. 6A). There was still a stockof CII–CV present and, in December, CVI females comprised approxi-mately 25% of the population (Fig. 6B). In 2005, the first individuals ofP. antarctica appeared in the water in August (2 individuals m−3), andreached a maximum of 144 individuals m−3 at the end of October(Fig. 4B). During this period, the population consisted of all stagesbetween NVI and CV (Fig. 6C).

Results for the water column (umbrella net, Fig. 4B) showed therewas a peak in abundance in December (160 individuals m−3),

following a peak at the end of November in the WP2 surface.However, abundances were about four times lower than those in theWP2 surface. In December, CVI females were the only stage found in

603C. Loots et al. / Journal of Marine Systems 78 (2009) 599–605

the water column (Fig. 6D). As was found for the WP2 surface net,the population in October was mainly composed of stages CI to CV(Fig. 6E).

Throughout 2005, the abundances from WP2 deep were very lowand have not been included for this species. Eggs, and stages NI–NIIIwere never observed in the sea ice or the water column. Note that themesh size of the umbrella net (100 µm) was sufficient to sample theseearly naupliar stages quantitatively if they were present.

3.2.2. D. glacialisD. glacialis was found in both the sea-ice and the water column

during 2005 (Fig. 7). In the sea-ice (Fig. 7A), abundanceswere relativelylow fromMay to September, with a maximum of 2500 individuals m−2

(4 individuals l−1), while they increased during spring to a maximumof 6500 individuals m−2 (16 individuals l−1) in October. In thewater column, abundances were lower during summer (7–15 indivi-duals m−3) than during winter (16–50 individuals m−3). Duringwinter trends in abundance as assessed by the three different nets weresimilar. D. glacialis reached their maximum in June, with 51 individualsm−3 in the WP2 surface (Fig. 7B), 16 individuals m−3 in the WP2 deep(Fig. 7C) and 30 individuals m−3 in the umbrella net (Fig. 7D). AfterJune, abundanceswere still high, although therewas a general decrease.

3.2.3. S. longipesS. longipeswas found only in the water column and never in the sea

ice throughout the sampling period (Fig. 8). In all three nets, therecorded abundances were high during the summer period, with amaximum of 22 individuals m−3 in the WP2 surface (Fig. 8A),44 individuals m−3 in the WP2 deep (Fig. 8B) and 132 individualsm−3 in the umbrella net (Fig. 8C). From April to December,abundances were lower in the WP2 deep (5 individuals m−3) andin the umbrella net (maximum of 30 individuals m−3), whereas they

Fig. 8. Stephos longipes. Abundances in the seawater from the WP2 surface (A), WP2depth (B), and umbrella net (C), from November 2004 to December 2005.

had a constant trend of higher values in the upper part of the watercolumn (WP2 surface, Fig. 8A).

4. Discussion

At Dumont d'Urville, three copepods generally described as ice-associated species were collected. The calanoid copepod P. antarcticaand the harpacticoid D. glacialiswere found in the seawater and in thesea ice, whereas the calanoid S. longipes was only found in theseawater. P. antarctica was the dominant of the three species andclosely associated with the sea ice. In the Weddell Sea, Günther et al.(1999) recorded 16 species of harpacticoids in the platelet layers ofDrescher Inlet and Dahms et al. (1990) found seven harpacticoidspecies to be associatedwith the sea ice. On the east coast of AntarcticaHoshiai and Tanimura (1986) observed three species, while Swadlinget al. (2000) found D. glacialis near Davis Station and noted severalother species, though in very low numbers. The importance of S.longipes in sea ice tends to decrease from theWeddell Sea (Kurbjeweitet al., 1993; Menshenina and Melnikov, 1995; Schnack-Schiel et al.,1995, 2001) around to the east coast of Antarctica (Hoshiai andTanimura, 1986; Tanimura et al., 1996; Swadling et al., 2000; Swadling,2001; present study), whereas P. antarctica becomes more and moreabundant. In the sections below we will consider the distributions ofthe three species in coastal waters along Terre Adélie in more detail.

4.1. P. antarctica

The year long results from Dumont d'Urville allow us to (1)describe the life cycle of P. antarctica in Terre Adélie in the sea-ice andin seawater and (2) compare it with other locations along the eastcoast of Antarctica where this species is also dominant. While theexact location of its eggs remains uncertain, it is probable thatP. antarctica underwent a long overwintering period, from March toSeptember, in the sea-ice as naupliar stages. While the earlier stages(NI–NIII) were not collected, the presence of considerable numbers oflate naupliar stages in the sea ice from August onwards was consistentwith other observations on overwintering nauplii of P. antarctica(Tanimura et al., 1996; Swadling et al., 2004). The first copepoditestages appeared in spring (end of September), when daylight andatmospheric temperatures were increasing, and light was sufficient toactivate primary production in the sea ice. Development through thecopepodite stages was very fast (approximately 1 month) andoccurred predominantly in the sea ice. This strategy, consisting of aprolonged naupliar phase and short development through thecopepodite stages, is quite unusual in marine copepods. The springalgal bloom facilitates very fast growth from CI to adult (Hoshiai andTanimura, 1986), while the proliferation of ice algae in autumn doesnot appear to induce any decrease in naupliar developmental time.Food availability can explain how overwintering as nauplii andsubsequent fast copepodite development are sustained, but it doesnot explain why P. antarctica adopted this life cycle. Tanimura et al.(1996) hypothesised that, compared with late copepodite stages,small naupliar stages alleviate the potential for overcrowding of thepopulation in autumn. Such overcrowding could result in a highaccumulation of excretory products, which could lead to thedeterioration of an unstable sea-ice environment. Moreover, it islikely that copepodite stages are restricted from achieving highdensities during the winter because of their greater size.

P. antarctica reached adulthood rapidly in spring and reproductionoccurred in or near the sea ice. At the end of November adults werereleased into thewater just before the break out of sea ice and this waspossibly triggered by the general deterioration of the ice. At that timeadult males were dominant over adult females in surfacewaters. Later,in December, adult females were almost the only stage present,though not in the surface waters, which suggests that females spawnin deeper water, probably near the substrate.

604 C. Loots et al. / Journal of Marine Systems 78 (2009) 599–605

Swadling et al. (2004) found many eggs of P. antarctica in thesediments. It seems that in an annual sea-ice zone (e.g. near Davisstation) eggs are spawned near the sediments. Naupliar stage I wasthen present in large numbers in the water column during the periodof open water before they colonised the reforming sea ice as stages NIor NII in early autumn (Swadling et al., 2004). Near Dumont d'Urville,the time of openwaterwas longer than near Davis, so it is inferred thatnauplii must consequently stay longer in the seawater, increasing theirrisk of predation. In a semi-permanent sea-ice zone (e.g. near Syowastation), eggs are spawned at the ice–seawater interface duringsummer and immediately attached to the sea ice (Tanimura et al.,1996).

The results from the present study were compared to those fromother sites around Antarctica, specifically near Davis station (68°35′S,77°58′E; Swadling et al., 2004) and Syowa station (69°00′S, 39°35′E;Tanimura et al., 1996). This comparison showed that, in spring, alongthe east coast of Antarctica there is a temporal shift of approximately1 month (Fig. 9) in the length of the life cycle of P. antarctica at thethree sites, which is linked to earlier sea ice break-out in the easternsector. The life cycle was completed earlier at Dumont d'Urville than atSyowa, with the Davis population intermediate between the two(Fig. 9). The timing of growth and reproduction by the adults thusappears to be flexible and related to the time of sea ice break-out andthe duration of open water. The capacity for shortening thedevelopment times of the copepodite stages has important implica-tions for life cycle flexibility when sea-ice conditions are uncertain.Such plasticity in their life history could be a key component to thecontinual success of P. antarctica under changing sea ice conditions.

4.2. D. glacialis

Although D. glacialis has been described as living in the sea icethroughout its life cycle (e.g. Dahms et al., 1990), at Dumont d'Urvilleit was also found in thewater column throughout 2005. Its occurrencein the sampling nets indicated that it was more abundant during thewinter than during the summer period, probably because it favourssea ice habitats. In autumn, D. glacialiswas present in higher numbersin deeper water, suggesting there is a benthic component to the lifecycle prior to sea-ice formation. This could explain the peak innumbers that was observed in autumn in the three nets, indicating therising of the population from the sediments and subsequent invasionof the sea water and then the sea ice. However, the method oftransport from the benthos to newly formed sea ice is not well known.Dahms et al. (1990) observed that, while the copepodites were goodswimmers, nauplii were not seen to swim, suggesting that it is thecopepodite stages that undertake active vertical migration throughthe water from the sediments to the sea ice. During the summer, thesurvival capacities of D. glacialis without sea ice (demonstrated bysuccessful laboratory cultivation by Dahms et al., 1990) indicate that itcan survive during the openwater period in summer and into autumn,

Fig. 9. Paralabidocera antarctica. Comparison of the life cycles at three sites (Syo

before establishing itself in the sea ice in winter. After June,abundances in the seawater near Dumont d'Urville tended to decreasewhile they increased in the sea ice, highlighting a shift in their mainhabitat.

4.3. S. longipes

In contrast to D. glacialis, S. longipes was as abundant in thesummer as in the winter. However, their partitioning of the watercolumn was not the same. In summer, S. longipes was found in theentire water column, whereas in winter it was more abundant in theupper part, suggesting a link with the ice–water interface even thoughit was not found in the sea ice. The dominance of P. antarctica in TerreAdélie and its potentially high inter-specific competition with S.longipes (Günther et al., 1999) may explain the absence of S. longipesin the sea-ice at Dumont d'Urville in 2005. The life cycle of S. longipesis also more complex (Kurbjeweit et al., 1993; Schnack-Schiel et al.,1995), as it needs sea ice at specific times in its development. Thiscould make it more sensitive to unpredictable sea ice conditions. InTerre Adélie, where environmental conditions vary considerably, thedevelopmental plasticity shown by P. antarctica may give it anadvantage over S. longipes. It is also probable that S. longipes prefersdeeper waters, such as those in areas of the South Atlantic part ofAntarctica (Kurbjeweit et al., 1993) and Terra Nova Bay (Costanzoet al., 2002), and is replaced by P. antarctica in the Indian sector inshallow coastal regions (Tanimura et al., 1996; Swadling et al., 2004,present study).

4.4. The influence of katabatic winds on copepod distributions

Meteorological conditions in Terre Adélie potentially play animportant role in establishing the distribution and abundance of ice-associated copepods. Winds in the region show high variability on aweekly basis, which impacts on both the sea ice condition and amountand duration of snow cover. As a result of the strong katabatic winds,Terre Adélie experiences longer periods of openwater in summer thanother sites along the east coast and the ice is more subject tounpredictable break-out at other times of the year. This means thatany copepods trying to establish populations must be able to copewith the ephemeral nature of the sea ice. Further, unstable sea ice inautumn, coupled with the presence of polynyas during winter,probably has substantial impacts on the mortality rates of the earlynaupliar stages and on the presence of individuals in the watercolumn. This might explain, in part, why the abundances of ice-associated copepods in this region are low compared to other sitesaround Antarctica.

Another explanation for low abundances might be their under-estimation due to the high horizontal patchiness of sea ice biota asdescribed by Swadling et al. (1997). The heterogeneity of the snowcover means that light penetration throughout sea ice varies and this

wa, Davis and Dumont d'Urville stations) along the east coast of Antarctica.

605C. Loots et al. / Journal of Marine Systems 78 (2009) 599–605

influences growth of the under-ice algae. Patches of algae becomeestablished in areas where snow cover is thinnest and, hence,illumination is strongest, and these patches attract grazers. Therefore,sea ice copepods are probably more abundant where snow is thinner.Near Dumont d'Urville, where katabatic winds cause snow cover tovary substantially at small temporal and spatial scales, there is likely tobemore pronounced patchiness in the distribution of the ice algae andthe copepods.

Acknowledgements

We would like to thank the French Polar Institute (IPEV) forfunding the program 281 ICOTA and also the Zone Atelier Antarctiqueof CNRS. We are grateful to all the people of the 55th mission in TerreAdélie for their technical help every week.

References

Arndt, C.E., Swadling, K.M., 2006. Crustacea in Arctic and Antarctic sea ice: distribution,diet and life history strategies. Adv. Mar. Biol. 51, 197–315.

Comiso, J.C., 2003. Large-scale characteristics and variability of the global sea ice cover.In: Thomas, D.N., Dieckmann, G.S. (Eds.), Sea Ice: An Introduction to its Physics,Chemistry, Biology and Ecology. Blackwell Publishing, Oxford, pp. 112–142.

Costanzo, G., Zagami, G., Crescenti, N., Granata, A.I., 2002. Naupliar development ofStephos longipes (Copepoda: Calanoida) from the annual sea ice of Terra NovaBay, Antarctica. J. Crust. Biol. 22, 855–860.

Dahms, H.-U., Bergmans, M., Schminke, H.K., 1990. Distribution and adaptations of seaice inhabiting Harpacticoida (Crustacea, Copepoda) of the Weddell Sea (Antarc-tica). Mar. Ecol. 11, 207–226.

Dewitt, H.H., Heemstra, P.C., Gon, O., 1990. Nototheniidae. In: Gon, O., Heemstra, P.C.(Eds.), Fishes of the Southern Ocean. J.L.B. Smith Institute of Ichthyology,Grahamstown, South Africa, pp. 279–331.

Fetterer, F., Knowles, K., 2002. 2006. Sea Ice Index. National Snow and Ice Data Center,Boulder, CO. Digital media.

Günther, S., George, K.H., Gleitz, M., 1999. High sympagic metazoan abundance inplatelet layers at Drescher Inlet, Weddell Sea, Antarctica. Polar Biol. 22, 82–89.

Hoshiai, T., Tanimura, A., 1986. Sea ice meiofauna at Syowa Station, Antarctica. Mem.Natl. Inst. Polar Res., Spec. Iss. 44, 118–124.

Hoshiai, T., Tanimura, A., Fukuchi, M., Watanabe, K., 1989. Feeding by the nototheniidfish, Pagothenia borchgrevinki on the ice-associated copepod, Paralabidoceraantarctica. Proc. NIPR Symp. Polar Biol. 2, 61–64.

Kurbjeweit, F., Gradinger, R., Weissenberger, J., 1993. The life cycle of Stephos longipes—an example for cryopelagic coupling in the Weddell Sea (Antarctica). Mar. Ecol.Prog. Ser. 98, 255–262.

La Mesa, M., Vacchi, M., Zunini Sertorio, T., 2000. Feeding plasticity of Trematomusnewnesi (Pisces, Nototheniidae) in Terra Nova Bay, Ross Sea, in relation toenvironmental conditions. Polar Biol. 23, 38–45.

Menshenina, L.L., Melnikov, I.A., 1995. Under-ice zooplankton of the western WeddellSea. Proc. NIPR Symp. Polar Biol. 3, 126–138.

Schnack-Schiel, S.B., Thomas, D.N., Dieckmann, G.S., Eicken, H., Gradinger, R., Spindler,M., Weissenberger, J., Mizdalski, E., Beyer, K., 1995. Life cycle strategy of theAntarctic calanoid copepod Stephos longipes. Prog. Oceanogr. 36, 45–75.

Schnack-Schiel, S.B., Thomas, D.N., Dahms, H.-U., Haas, C., Mizdalski, E., 1998. Copepodsin Antarctic sea ice. Ant. Res. Ser. 73, 173–182.

Schnack-Schiel, S.B., Thomas, D.N., Haas, C., Dieckmann, G., Alheit, R., 2001. Theoccurrence of copepods Stephos longipes (Calanoidea) and Drescheriella glacialis(Harpacticoida) in summer sea ice in the Weddell Sea, Antarctica. Ant. Sci. 13,150–157.

Swadling, K.M., 2001. Population structure of two Antarctic ice-associated copepods,Drescheriella glacialis and Paralabidocera antarctica, in winter sea ice. Mar. Biol.139, 597–603.

Swadling, K.M., Gibson, J.A.E., Ritz, D.A., Nichols, P.D., 1997. Horizontal patchiness insympagic organisms of the Antarctic fast ice. Ant. Sci. 9, 399–406.

Swadling, K.M., McPhee, A.D., McMinn, A., 2000. Spatial distribution of copepods in fastice of eastern Antarctica. Polar Biosci. 13, 55–65.

Swadling, K.M., McKinnon, A.D., De'ath, G., Gibson, J.A.E., 2004. Life cycle plasticity anddifferential growth and development in marine and lacustrine populations of anAntarctic copepod. Limnol. Oceanogr. 49, 644–655.

Tanimura, A., 1992. Postembryonic development of Paralabidocera antarctica (I.C.Thompson) (Copepoda, Calanoida) from the fast ice near Syowa Station, Antarctica.Hydrobiology 245, 109–128.

Tanimura, A., Hoshiai, T., Fukuchi, M., 1996. The life cycle strategy of the ice-associatedcopepod, Paralabidocera antarctica (Calanoida, Copepoda), at Syowa Station,Antarctica. Ant. Sci. 8, 257–266.

Tranter, D.J., Smith, P.E., 1968. Filtration performance. In: Tranter, D.J., Fraser, J.H. (Eds.),Zooplankton Sampling. Monographs on Oceanographic Methodology, vol. 2.UNESCO Press, Paris, pp. 27–56.

Vacchi, M., La Mesa, M., 1995. The diet of the Antarctic fish Trematomus newnesiBoulenger, 1902 (Nototheniidae) from Terra Nova Bay, Ross Sea. Ant. Sci. 7, 37–38.

Wendler, G., Stearns, C., Weidner, G., Dargaud, G., Parish, T., 1997. On the extraordinarykatabatic winds of Adélie Land. J. Geophys. Res. 102, 4463–4474.

Williams, R., 1988. The inshore marine fishes of the Vestfold Hills region, Antarctica.Hydrobiology 165, 161–167.