Diatom Key Species in the Southern Ocean

Although about 130 diatom species have been described from the Antarctic Circumpolar Current (ACC) and the seasonally ice-covered areas, only a few contribute the bulk of biomass under bloom conditions. Long chain-forming (Fragilariopsis kerguelensis, Pseudonitzschia spp. and Chaetoceros spp.) and large-celled (Corethron pennatum and Thalassiothrix antarctica) diatoms are the major species in the open, iron-limited ACC. Most have thick silica frustules or robust spines with silica to nitrogen (Si:N) ratios above 2. The frustules of F. kerguelensis and T. antarctica together contribute the bulk of the diatom ooze accumulating under the ACC. In contrast the dominant diatoms in near-shore, iron-rich environments are more similar to diatoms from the continental shelves of the world ocean: weakly silicified species of the genera Thalassiosira, Chaetoceros. In many regions, Phaeocystis colonies also dominate biomass. The plankton of the continental margins tends to have Si:N ratios around 1. As a result, silica burial under the productive regions is much less but carbon burial is significantly higher than in the land-remote ACC.

We hypothesize that the dominance of large, robust diatoms in the open ACC is a result of heavy grazing pressure relative to the low, iron-limited primary production which leads to selection and persistence of species with effective defences. Despite their heavy frustules, the life cycle strategy of these species is to persist in the surface layer where their carbon is recycled but from which their silica frustles sink out. In productive environments on the other hand, fast-growing species build-up biomass till iron depletion. Similar species from other regions tend to form aggregates of living cells, resting spores and phytodetritus that sink out en masse from the surface layer following nutrient depletion. This universal phenomenon has been interpreted as a seeding strategy whereby overwintering cells are dispersed in deeper layers where grazing pressure is lower. Winter mixing transports some of these cells back to the surface layer together with a fresh supply of the limiting nutrient.

Such mass sinking events have been observed or recorded by sediment traps in some regions such as the Bransfield Strait and the southern Weddell Sea.
It follows that three broad categories of diatoms can be differentiated in the Southern Ocean: the background species, the fast-growing, boom-and-bust, carbon sinkers of iron-replete regions and the slower growing, persistent silica sinkers of the iron-limited ACC. Hence the intrinsic properties of the latter, key diatom species ultimately shape the biogeochemical cycles of the Southern Ocean.

Support for this hypothesis has come from iron-fertilization experiments carried out in the ACC that were all dominated by diatoms. Detailed floristic analyses of the phytoplankton from the EisenEx experiment revealed changes in species dominance that shed light on the ecology of these key species. The persistent species dominated the phytoplankton initially and responded to fertilization by increasing their growth rates albeit for brief periods. In contrast, two less silicified species exhibited much higher growth rates which were sustained throughout the 3 weeks of the experiment. One of these species Pseudonitzschia lineola contributed more than 50% of the total biomass. The other fast-growing species Chaetoceros curvisetus - a known carbon sinker from shelf regions - built up less biomass because of its low initial population size. It appears that in the presence of adequate resources fast growth rates are attained and sustained by only some species or even strains and that all the other species maintain much lower growth rates or do not respond to radical improvement in growth conditions at all. These results cast doubt on evolutionary selection for fast growth rates as a universal principle and hence on the theory of competitive exclusion as a shaping force in the plankton. We intend testing this hypothesis in the next iron fertilization experiment scheduled for 2004 (EIFEX).