Ma. Brzezinski et Dm. Nelson, THE ANNUAL SILICA CYCLE IN THE SARGASSO SEA NEAR BERMUDA, Deep-sea research. Part 1. Oceanographic research papers, 42(7), 1995, pp. 1215-1237
The annual cycles of silicic acid and biogenic particulate silica (BSi
O2) concentrations were examined from October 1988 through December 19
92 at the JGOFS time-series site near Bermuda. Lithogenic particulate
silica (LSiO(2)) concentrations were measured from August 1991 through
December 1992 at the same site. Distinctly different seasonal pattern
s were observed in the concentration of biogenic and lithogenic phases
. Integrated BSiO2 concentrations in the upper 160 m were greatest (7.
6-56.3 mmol BSiO2 m(-2)) during an annual diatom bloom that occurred e
ach year between January and April. In contrast, integrated LSiO(2) co
ncentrations in the upper 160 m showed an annual maximum during July a
nd August (2.05-2.12 mmol LSiO(2) m(-2)) probably due to greater aeoli
an dust inputs during summer. The export of both BSiO2 and LSiO(2) was
examined with sediment traps deployed for 4 days each month from Augu
st 1991 through August 1992, The annual export of biogenic silica at 1
50 m (47.6 +/- 8-6 mmol BSiO2 m(-2) y(-1), s, d.) was dominated by a w
inter diatom bloom, which was responsible for 62% of the annual flux.
In contrast, the export of LSiO(2) through 150 m was highest during Ju
ly and August (0.039-0.064 mmol LSiO(2) m(-2) day(-1)), coincident wit
h the summer maxima in suspended LSiO(2) concentrations. The close tem
poral coupling between maxima in suspended concentrations and vertical
flux indicates that any delay between the appearance of BSiO2 and LSi
O(2) in the surface water and their subsequent export was less than ou
r 30 day sampling interval. The standing stock of BSiO2 in the upper 1
60 m during non-bloom periods was very small with little temporal vari
ability (mean = 2.7 +/- 0.9 mmol m(-2), s, d.). High dissolution rates
of BSiO2 within sediment traps at that time (mean = 0.072 +/- 0.036 d
ay(-1),s. d.) suggested that significant silica production was occurri
ng to maintain the observed standing stocks of BSiO2 against losses du
e to dissolution. Comparison of estimated BSiO2 production and export
during this period indicates that at least 64-82% of the BSiO2 produce
d during the non-bloom period was recycled within the euphotic zone. E
stimates of the annual new production by diatoms, with a mole ratio of
1:1 for Si:N within sinking siliceous particles assumed, suggest that
diatoms account for up to 26-48% of the new production occurring at t
his site.