The community structure and predation impact of carnivorous macrozooplankton (>2 cm; chaetognaths, medusae, ctenophores and mysids), with particular emphasis on the chaetognaths Eukrohnia hamata and Sagitta gazellae, were investigated during three surveys conducted in late austral summer (April/May) of 2001, 2004 and 2005 in the Polar Frontal Zone in the vicinity of the Prince Edward Islands (46º45’S, 37º50’E), Southern Ocean. The 2001 survey formed part of the Marion Offshore Variability Ecosystem Study (MOVES II), while the 2004 and 2005 surveys formed part of the Dynamics of Eddy Impacts on Marion’s Ecosystem study (DEIMEC III and IV respectively). Macrozooplankton samples were collected using WP-2, RMT-8 and Bongo nets. Results of the hydrographic survey indicated that the region of investigation, the Polar Frontal Zone (PFZ), is an area of high mesoscale variability. During the 2004 survey the Antarctic Polar Front (APF) and the Subantarctic Front (SAF) merged to form an intense frontal feature with subsurface temperature and salinity ranging from 8.5-7.5ºC and 34.15-33.88, respectively. A cyclonic cold core eddy, believed to have been spawned from the APF, was observed during the 2005 survey. Macrozooplankton abundance and biomass ranged from 0 to 43.731 ind. m⁻³, and from 0 to 41.55 mg wwt m⁻³ respectively, during the three surveys. Among the carnivorous macrozooplankton, chaetognaths (Eukrohnia hamata and Sagitta gazellae) were most prominent, contributing up to 85% of the total biomass during all three surveys. Elevated biomass values were found near and within the frontal feature during the 2004 survey, and also along the eddy edge during the 2005 survey. However, hierarchical cluster analysis did not reveal the presence of distinct zooplankton groupings associated with the various water masses encountered during the surveys and this is probably due to the high mesoscale variability in oceanographic conditions that are characteristic of the PFZ. The total average predation impact of the selected carnivorous macrozooplankton during the 2001, 2004 and 2005 surveys accounted for 4.93 ± 6.76%, 0.55 ± 0.51% and 4.88 ± 4.45 of the mesozooplankton standing stock, respectively. S. gazellae had the highest consumption rate in all three surveys, consuming up to 800 g Dwt 1000m⁻³d⁻¹ during the study. Of the two chaetognaths, E. hamata dominated the chaetognath standing stock. The combined abundance and biomass values of E. hamata and S. gazellae ranged from 0 to 43.73 ind. m⁻³ and from 0 to 41.551 mg wwt m⁻³ respectively, during the three surveys. Inter-annual variability in the chaetognath densities was apparent. Highest abundances and biomasses tended to be associated with specific water masses, confirming the existence of a relationship between zooplankton community structure and hydrographic conditions. Generally, about 90% of the chaetognaths contained no food in their guts. S. gazellae consumed a wider variety of prey. Oil droplets occurred in the guts of ≈ 51% of E. hamata. Cannibalism was low in both species, but greater in S. gazellae than E. hamata. During the three surveys, the feeding rate values of E. hamata and S. gazellae went up to 0.48 and 2.099 prey d⁻¹ respectively. S. gazellae also had a greater predation impact on the mesozooplankton standing stock than E. hamata. The mean predation impact of the chaetognaths combined was 0.31 ± 0.291%, 0.52 ± 0.28% and 0.53 ± 0.56% of the mesozooplankton standing stock during the 2001, 2004 and 2005 surveys, respectively. During all three surveys, the majority of individuals (≈ 76%) of the chaetognaths were at stage I maturity, suggesting that during the time of study the chaetognaths were not reproducing. In both species a significant difference (log-linear analysis, p < 0.05) in maturities between the years investigated was observed. In general, there were no differences in lengths and maturities between the different water masses encountered during the surveys. The lengths of E. hamata and S. gazellae ranged from 5 to 24 mm and from 9.4 to 63.6 mm, respectively.

Mesozooplankton community structure and grazing impact in the Polar Frontal Zone (PFZ) of the Southern Ocean were investigated during two cruises of the South African National Antarctic Programme (SANAP), the Marion Offshore Ecosystem Variability Study I & II (MOEVS). During the first cruise (MOEVS I), a meso-scale oceanographic grid survey was conducted in the upstream region of the Prince Edward Islands (PEI) in austral autumn (April) 2001. Mesozooplankton samples, collected using a Bongo net (fitted with 200 and 300µm mesh nets) at depths between 200 and 300 m, were separated into three size fractions: 200-500 µm; 500-1000 µm; 1000-2000 µm by reverse filtration. Total surface (depth <5 m) chlorophyll-a (chl-a) concentration (measured fluorometrically) during the study ranged between 0.11 and 0.34 µg 1^(-1) and was always dominated by picophytoplankton (<2.0 µm). Total mesozooplankton abundance and biomass during the survey ranged between 49 and 1512 ind. m^(-3) and between 0.7 and 25 mg Dwt. m^(-3), respectively. Throughout the survey, the 200-500 µm class numerically dominated the mesozooplankton community, comprising an average of ~ 69% (SD = ± 12.3%). The dominant species in the 200-500 µm size fraction were the copepods Oithona similis, Calanus simillimus and Metridia lucens and the pteropod, Limacina retroversa. However, in terms of biomass, the 1000-2000 µm group was predominant, with dry weight values constituting an average of ~ 66% (SD = ± 10.2%). Biomass was dominated by carnivorous zooplankton, particularly the euphausiids, Euphausia vallentini and Thysanoessa vicina and the chaetognaths, Sagitta gazellae and Eukrohnia hamata. Three distinct groupings of stations were identified by multivariate analysis. The different station groupings identified reflect changes in the relative contributions of the rather than different species assemblages. During the second cruise (MOEVS II), conducted in April 2002 (austral autumn), mesozooplankton community structure and grazing impact were investigated at 13 stations in the west Indian sector of the PFZ. Total integrated chl-a biomass ranged between 11.17 and 28.34 mg chl-a m^(-2) and was always dominated by nano- and picophytoplankton (<20 µm). Throughout the study, small copepods, mainly Oithona similis and Ctenocalanus vanus, numerically dominated the mesozooplankton community comprising up to 85% (range 30 to 85%) of the total abundance. Grazing activity of the four most abundant copepods (O. similis, C. vanus, Calanus simillimus and Clausocalanus spp.), which comprised up to 93% of total mesozooplankton abundance, was investigated using the gut fluorescent technique. Results of gut fluorescence analyses indicated that C. simillimus, Clausocalanus spp. and Ctenocalanus vanus exhibited diel variability in gut pigments, with maximum values at various stages of the night. In contrast, O. similis did not demonstrate diel variation in gut pigment contents. Ingestion rates of the four copepods ranged from 23.23 to 1462.02 ng (pigm.) ind^(-1) day^(-1), depending on the species. The combined grazing impact of the four copepods, ranged between 1 and 36% of the phytoplankton standing stock per day, with the highest daily impact (~ 35.86%) occurring at stations in the vicinity of the Antarctic Polar Front. Among the copepods, O. similis and C. vanus were generally the most important consumers of phytoplankton biomass; together they were responsible for up to 89% (range 15 to 89%) of the total daily grazing impact. Carbon specific ingestion rates of the copepods varied between 42 and 320% body carbon per day, depending on the species. The study highlights the importance of small copepods in terms of both their significant contribution to total mesozooplankton numbers and their grazing impact on the phytoplankton standing stocks in the PFZ during austral autumn.

The aim of the present study was to assess the ecological role of the euthecosome pteropod, Limacina retroversa, in particular, and the mesozooplankton community, in general, in the pelagic ecosystem of the Polar Frontal Zone (PFZ), Southern Ocean. Data were collected from four oceanographic surveys to the Indian sector of the PFZ during austral autumn 2000, 2002, 2004 and 2005. Copepods, mainly Calanus simillimus, Oithona similis, Clausocalanus spp. and Ctenocalanus spp., typically dominated total mesozooplankton counts, accounting for, on average, between 75.5 % and 88.1 % (Mean = 77.4 %; SD = 13.4 %) of the total, during the present investigation. Results of the study indicate that L. retroversa may, at times, contribute substantially to total mesozooplankton abundances. During the study, L. retroversa contributed between 0.0 and 30.0 % (Mean = 5.3 %; SD = 7.1 %) to total mesozooplankton numbers. Significant small-scale variability in abundance and size structure of L. retroversa and abundance of copepods was minimal. Inter-annual variability, on the other hand, was significant between some years. Total pteropod numbers were greatest during April 2002 and 2004, while copepods exhibited greatest abundances during April 2004 only. Pearson’s Correlation analysis suggested that L. retroversa abundances were positively correlated to total surface chlorophyll-a (chl-a) concentrations. The significantly lower chl-a concentrations recorded during April 2005 may explain the reduced pteropod numbers observed during that survey. The size class structure of L. retroversa comprised mainly small and mediumsized individuals during all four surveys. This corresponds well with records from the northern hemisphere (sub-Arctic and Arctic waters) where Limacina spp. are reported to exhibit maximum spawning during mid to late-summer. Higher abundances of large individuals only occurred during April 2005, when chl-a concentrations were very low; possibly the result of delayed spawning, due to reduced food availability. Ingestion rates of the four most abundant copepods, determined using the gut fluorescence technique, ranged between 159.32 ng (pigm) ind⁻¹ day⁻¹ and 728.36 ng (pigm) ind⁻¹ day⁻¹ (Mean = 321.01 ng (pigm) ind⁻¹ day⁻¹; SD = 173.91 ng (pigm) ind⁻¹ day). Ingestion rates of L. retroversa were much higher, ranging from an average of 4 28.68 ng (pigm) ind⁻¹ day⁻¹ in April 2002 to 4 196.88 ng (pigm) ind⁻¹day⁻¹in April 2005 (Mean = 4157.36 ng (pigm) ind⁻¹ day⁻¹; SD = 35.37 ng (pigm) ind⁻¹day⁻¹). Average daily grazing rates for the pteropod varied between 0.39 mg (pigm) m⁻² day⁻¹ in April 2005 and 17.69 mg (pigm) m-2 day-1 in April 2004 (Mean = 6.13 mg (pigm) m⁻² day⁻¹; SD = 11.04 mg (pigm) m⁻² day⁻¹); corresponding average daily grazing impacts ranged between 8.4 % and 139.8 % of the phytoplankton standing stock in April 2005 and 2004, respectively (Mean = 48.5 %; SD = 84.5 %). Average daily grazing rates of the four copepods ranged from 4.58 mg (pigm) m⁻² day⁻¹ to 8.77 mg (pigm) m⁻² day⁻¹ -1, during April 2002 and 2004, respectively (Mean = 6.28 mg (pigm) m⁻² day⁻¹; SD = 5.94 mg (pigm) m⁻² day⁻¹). Collectively, the copepods removed an average of between 31.6 % and 89.8 % of the phytoplankton standing stock per day, during April 2002 and 2004, respectively (Mean = 70.8 %; SD = 86.7 %). The daily grazing impact of the copepods accounted for an average of between 40.4 % and 87.8 % of the total zooplankton grazing impact, during April 2004 and 2005, respectively (Mean = 75.0 %; SD = 65.5 %). L. retroversa was responsible for an average of 52.4 % and 59.5 % of the total zooplankton grazing impact, during April 2002 and 2004, respectively. However, during April 2005, when L. retroversa numbers were significantly lower than previous years, the pteropod contributed an average of only 7.5 % to the total zooplankton grazing impact. Thus, during the present investigation,the pteropod was responsible for removing a mean of 48.9 % of the available phytoplankton (SD = 74.9 %). The predation impact of the dominant carnivorous macrozooplankton and micronekton in the PFZ was determined during April 2004 and 2005 using daily ration estimates obtained from the literature. Additionally, gut content analysis was used to determine the contribution of L. retroversa to the diet of the dominant predators. Average predation impact ranged from 1.1 % and 5.7 % of the total mesozooplankton standing stock during April 2004 and 2005, respectively (Mean = 3.8 %; SD = 12.3 %). Chaetognaths and euphausiids dominated total carnivore numbers and made the greatest contributions to total predation impact during both years. Copepods appeared to be the main prey item of the dominant carnivorous macrozooplankton-micronekton in the region. L. retroversa was only detected in the gut contents of the amphipod, Themisto gaudichaudi, but not in either of the chaetognath species (Eukrohnia hamata and Sagitta gazellae) or the myctophid fish (Electrona spp.). The pteropod was found in 19 % of amphipod guts dissected. Pearson’s Correlation analyses showed that the four major predatory zooplankton groups found in the PFZ (chaetognaths, euphausiids, amphipods and myctophid fish) were positively correlated to abundances of L. retroversa, suggesting that the pteropod might be an important prey item for many of the carnivorous macrozooplankton/micronekton in the PFZ. To conclude, L. retroversa may play an important role in the pelagic ecosystem of the PFZ, in austral autumn. However, ocean acidification and calcium carbonate undersaturation (as a result of increased anthropogenic carbon dioxide emissions), that is predicted to occur within the next 50 – 100 years, will most likely have significant implications for the Sub-Antarctic pelagic ecosystem if L. retroversa cannot adapt quickly enough to the changes.

The trophodynamics of the numerically dominant mesozooplankton (200-2000 m) in the vicinity of the Subtropical Convergence (STC) in the Indian sector of the Southern Ocean during austral autumn (April / May) 2007 were investigated as part of the Southern Ocean Ecosystem Variability Study. The survey consisted of six north-south transects each bisecting the STC between 38º to 43ºS and 38º to 41º45’E. In total, 48 stations situated at 30 nautical mile intervals were occupied over a period of ten days. Hydrographic data revealed a well defined surface and sub-surface expression of the STC, which appeared to meander considerably between 41ºS and 41º15’S. Surface chlorophyll-a (chla) concentrations were low, ranging between 0.08 and 0.68 mg chl-a.m-3 and were generally dominated by the picophytoplankton (<2 m) which made up 66.6% (SD±17.6) of the total pigment. Chl-a concentrations integrated over the top 150m of the water column ranged between 11.97 and 40.07 mg chl-a.m-2 and showed no significant spatial patterns (p>0.05). Total integrated mesozooplankton abundance and biomass during the study ranged between 3934.9 and 308521.4 ind.m-2 (mean = 47198.19; SD±62411.4 ind.m-2) and between 239.8 and 4614.3 mg Dwt.m-2 (mean = 1338.58; SD ±1060.5), respectively. Again, there were no significant spatial patterns in the total mesozooplankton abundance or biomass within the region of study (p>0.05). No significant correlations were found between biological (chlorophyll-a concentrations and zooplankton abundance) and physico-chemical variables (temperature and salinity) (p>0.05). The total mesozooplankton community was numerically dominated by copepods of the genera Pleuromamma, Calanus, Oncaea and Oithona. Other important representatives of the mesozooplankton community included the tunicate, Salpa thompsoni, and the pteropod, Limacina retroversa. At the 40% similarity level, numerical analysis identified five distinct mesozooplankton groupings within the survey area. Differences between the groupings were associated with changes in the relative contribution of numerically dominant species rather than the presence or absence of individual species. No groupings were associated with any specific feature of the front within the survey area. The feeding rates of the six most numerically abundant mesozooplankton species (Calanus simillimus, Limacina retroversa, Pleuromamma abdominalis, Clausocalanus breviceps, Oncaea conifera, Salpa thompsoni) accounting for on average 39% of the total mesozooplankton counts, were investigated using the gut fluorescence technique. For all species, the total gut pigment contents during the night time were significantly higher than the daytime values (p<0.05 for all species). The gut evacuation rates (k) for selected mesozooplankton ranged between 0.14 and 0.81 h-1. The ingestion rates ranged between 147.8 and 5495.4 ng(pigm)ind-1.day-1 which corresponded to a daily ration of between 2.4 and 10.9% body carbon. The combined grazing impact of the selected species on the daily phytoplankton standing stock was highly variable and ranged between 1.2 and 174.1% with an average of 27.3% (SD±38.78%) within the survey area. The highest grazing impact (>60%) was typically associated with those stations where the pteropod, L. retroversa, and the tunicate, S. thompsoni, contributed more than 5% of the total mesozooplankton counts. No significant differences were found in the grazing impact of any or all selected species situated either north, south or in the immediate vicinity of the front (p>0.05 in all cases). The lack of defined spatial patterns in the mesozooplankton abundance and community structure suggests that the STC did not act as a significant biogeographic barrier to the distribution of mesozooplankton during the study. It is presumed that the large scale mixing event caused by a storm prior to this study was responsible for the observed lack of elevated biological activity within the region of the STC.

The population structure and feeding dynamics of the hyperiid amphipod, Themisto gaudichaudii, was investigated during two cruises of the South African National Antarctic Programme conducted in the Indian sector of the Polar Frontal Zone during austral autumn (April) 2004 and 2005. During the 2004 cruise the frontal features that delimit the PFZ converged to form a single distinctive feature. In 2005, the research cruise was conducted in the vicinity of a cold-core eddy which was spawned from the Antarctic Polar Front. Total mesozooplankton abundance and biomass during the 2004 study ranged from 55.19 to 860.57 ind. m⁻³, and from 2.60 to 38.42 mg dwt m⁻³, respectively. In 2005 the abundance and biomass ranged from 23.1 to 2160.64 ind. m⁻³, and from 0.76 to 35.16 mg dwt m⁻³, respectively. The mesozooplankton community was numerically dominated by copepods, pteropods, and ostracods during both surveys. The abundance and biomass of Themisto gaudichaudii in the region of investigation was < 0.2 ind. m⁻³ (range 0.01 to 0.15 ind. m⁻³) and < 0.06 mg dwt m⁻³ (range 0.02 to 0.06 mg dwt m⁻³) during 2004, while in 2005 the abundance and biomass of the amphipod ranged from < 0.01 to 0.2 ind. m⁻³ and < 0.01 to 0.04 mg dwt m⁻³, respectively. These values correspond to < 1% of the total mesozooplankton abundance and biomass during both surveys. T. gaudichaudii exhibited no significant spatial patterns in abundance, biomass and total length during both 2004 and 2005 (p > 0.05 in all cases). A key feature of the two investigations was the virtual absence of juveniles (total length < 15 mm) among the amphipod population, supporting the suggestion that they exhibit strong seasonal patterns in reproduction. Gut content analysis during both years indicated that for both the male and female amphipods’, copepods were the most prevalent prey species found in stomachs, followed by chaetognaths and pteropods. Results of electivity studies indicate that T. gaudichaudii is an opportunistic predator, generally feeding on the most abundant mesozooplankton prey. Results of in vitro incubations indicated that the total daily feeding rate of T. gaudichaudii during 2004 ranged from 11.45 to 20.90 ind. m⁻³ d⁻¹, which corresponds to between 0.12 and 1.64% of the total mesozooplankton standing stock. In 2005, the feeding rate ranged between 0.1 and 1.73% of the total mesozooplankton standing stock. The low predation impact of T. gaudichaudii during this study can be related to their low abundances and high interannual variability throughout the region of investigation.

Trophodynamics of carnivorous zooplankton in the region of the Subtropical Convergence (STC) in the Indian sector of the Southern Ocean was investigated during austral autumn (April 2007) as part of the first cruise of the Southern Ocean Ecosystem Variability Study. Within the region of the study, the STC was well defined by the 14°C surface isotherm which separated the Agulhas Return Current and Subtropical water in the north from Sub-Antarctic waters to the south. Total average abundance (3.89 ± 5.46ind 100m-3) and biomass (0.14 ± 0.27mg Dwt 100m-3) of carnivorous zooplankton south of the front were significantly higher than the total average abundance (1.33 ± 1.81ind 100m-3) and biomass (0.03 ± 0.05mg Dwt 100m-3) north of the front (p<0.001). There were no significant correlations between the selected physico-chemical (temperature and salinity) and the biological (mesozooplankton abundance and biomass) variables and the total abundance and biomass of the carnivorous zooplankton during the investigation (p>0.05 in all cases). There was no evidence of enhanced biomass and abundance values at stations occupied in the immediate vicinity of the front. Total average carnivorous zooplankton abundance was dominated by chaetognaths (Eukrohnia hamata Möbius 1875, Sagitta gazellae Ritler-Záhony 1909 and S. zetesios Fowler 1905) and euphausiids (Nematoscelis megalops Sars 1883, Euphausia longirostris Hansen 1908 and E. spinifera Sars 1883), which contributed up to 86.58 ± 32.91% of the total counts. The total average biomass was dominated by euphausiids and amphipods (Themisto gaudichaudii Guérin-Méneville 1825, Phronima sedentaria Forsskål 1775 and Vibilia armata Bovallius 1887) which contributed up to 71.45 ± 34.85% of the total counts. In general the populations of both the euphausiids and amphipods were dominated by females while the chaetognaths were dominated by juveniles. Numerical analysis identified two major zooplankton groupings within the survey area which did not coincide with the water masses within the survey area. The SIMPER procedure of the PRIMER package indicated differences between the groups were mainly attributed to changes in the abundance of the numerically dominant species rather than the presence or absence of individual species. The absence of any significant spatial patterns in the distribution of the carnivorous zooplankton suggests that the STC did not act as a biogeographical barrier during the present study. The mean feeding rates of the chaetognaths E. hamata, S. gazellae and S. zetesios were 1.82 ± 0.85prey d-1, 3.63 ± 2.08prey d-1 and 2.18 ± 0.59prey d-1, respectively. These rates correspond to a combined predation impact equivalent to <5% of the mesozooplankton standing stock or <10% of the mesozooplankton secondary production. Mesozooplankton, comprising mainly copepods was the dominant prey in the guts of the three chaetognath species. Total predation impact of the euphausiids, chaetognaths and amphipods, estimated using published daily ration data, on the mesozooplankton standing stock and secondary production ranged from 0.01% to 1.53% and from 0.03% to 30.54%, respectively. Among the carnivorous zooplankton, chaetognaths were generally identified as the dominant predators of mesozooplankton. Low predation impact of selected carnivorous zooplankton suggested that these organisms contributed little to the vertical carbon flux within the region of investigation during the study.