Compositional and fluorescence characteristics of the giant diatom Ethmodiscus along a 3000 km transect (28°N) in the central North Pacific gyre

摘要

The giant diatom Ethmodiscus was examined along an east-west transect at 28-30°N during 2002 and 2003 to determine if abundance, chemical composition or physiological status of this largest of diatoms varied on the scale of 100's-1000's of km in North Pacific gyre. Abundance ranged from < 0.1 - > 2.0 cells m~(-3) and supported the notion of an abundance mosaic reported previously. However, there was only minimal support for the relationship between abundance and nutrient concentration at 125m reported previously. Cellular chlorophyll varied little along the transect (7.3-10.9 ng chlcell~(-1)) except at the westernmost station. Cellular N and P quotas co-varied 3 - 4.5 fold (mean = 50.8 ±3.7 and 3.7±0.8nmol N and Pcell~(-1) and yielded N:P ratios that closely clustered around the Redfield ratio (average = 14.6±1.1). Only low levels of chlorophyll-normalized alkaline phosphatase (APase) activity were observed (0.4-2.5 nmol P μg chl~(-1) h~(-1)) with APase activity lower than that in either the bulk water, or co-occurring Trichodesmium spp. and Pyrocystis noctiluca. The active fluorescence parameter F_v:F_m, a property sensitive to Fe stress, was uniformly high at all stations (average = 0.73 ± 0.04 for 2003, and 0.69 ± 0.05 for 2002), indicating sufficient Fe for optimum photosynthetic competence. These results contrasted sharply with results from Rhizosolenia mats reported along the same transect where there was a significant decline westward in F_v:F_m. Both ferredoxin (Fd) and flavodoxin accumulated in cells of Ethmodiscus, resulting in Fd Index values of < 0.6. Iron cell quotas ranged from 0.7-5.1 pmol Fecel~(-1). When normalized to cytoplasmic volume, the Fe μm~(-3) was comparable to that of Escherichia coli. We note that the disproportionate contribution of the vacuole (with its high organic content) to total volume typical of large diatoms is a potentially significant source of error in Fe:C ratios and suggest that Fe should be normalized to cytoplasmic volume whenever possible to permit valid intercomparisons between studies. The composition, F_v:F_m data and Fe:C ratio suggest a relatively uniform population experiencing little N, P or Fe stress. The uncoupling of the Fd Index from these measures is consistent with previous findings showing that the expression of flavodoxin can be characterized as an early stress response and that its accumulation is not necessarily correlated with physiological deficit. Ethmodiscus appears to be well adapted to some of the most oligotrophic waters in the ocean. Because it is an important sedimentary marker, the biology of living Ethmodiscus provides insights into the source of extensive Ethmodiscus oozes. Mass sedimentation after frontal accumulation has been suggested as a source for these oozes. Our data contain no evidence that the flux is linked directly to Fe, N or P stress.