ABSTRACT-The role of silicon availability on cell-cycle progression in marine diatoms was examined using flow cytometric methods. Silicon deprivation halted the progression of cells through the cell cycle with cells arresting in G1. G2 and M in 6 of 7 species examined (5 centric and 1 pennate species). The exception, Phaeodactylurn tricornutum, did not require silicon for growth and did not have a silicondependent segment within its cell cycle. T h s species was also the only one lacking a light-dependent arrest point late in its cell cycle suggesting that the arrest of diatom cells in G2 and M in the dark is related to silicon metabolism. Chaetoceros spp. were unique in that they had 2 silicon-dependent segments In G1: one at the GI/S boundary apparently associated with a silicon requirement for DNA synthesis and a second earher In G1 associated with the d e p o s~t~o n of siliceous setae. Silicon limitation of Thalaseoslra weissflogii led to an increase in the duratlon of G2 with the duration of G1, S and M remaining as observed under nutrient-replete conditions. In severely llmited cells, G2 comprised 82 % of the cell cycle and lasted for over 2 d , hlore complicated responses were observed for Cylindrotheca fusiformis and Chaetocel-OS simplex. Modest silicon limitation of C. fuslf01-1111s led to increases in the duration of G2 and posslbly M. More severe silicon stress did not lengthen M further, but both G2 and G1 increased in duratlon. For C. simplex, modest silicon l~mitation led to the expansion of G1 alone, while more severe limitation lengthened G1, G2 and M . Changes in cell cycle durations in this species appeared related to a decline in the silicon content of siliceous setae deposited during G1 These results corroborate past observations that silicon metabolism is linked to specific segments of the cell cycle, but indicate that these regions can lengthen dramatically in response to silicon limitation.

We performed "Si tracer experiments to measure the uptake rates of silicic acid by natural diatom assemblaged in 2 Gulf Stream warm-core rings as a function of the extracellular silicic acid concentration. The experiments were performed a t and near the centers of rings 81-D and 82-B, and lncluded 5 sampling dates in late spring, summer and autumn, when silicic acid was undetectable to conventional analyses (< 0.2 yM) throughout the upper 20 to 30 m Uptake was measurably substratelimited only at those stations where silicic acid was undetectable. Concentrations of 2.5 to 3 1rM Si(OH)4, which typlfied ring 82-B before the development of the seasonal thermocline, were high enough to support maximum uptake rates. When substrate 1i.mitation was present the dependence of the uptake rate on sllicic acid concentration exhibited hyperbolic saturation kinetics, with half-saturation constants (K,) ranging from 0.53 to 0.90 \&M. This range is considerably lower than that reported previously for natural diatom asselnblages (1.5 to 3.7 pM). A numerical sllnulation indicates that a diatom species with a K, value typical of the ring assemblages would outcompete most diatom species whose silicon kinetics have been examined in culture, not only under conditions of continuously low sll~cic a c~d concentration but with the sporadic, pulsed nutrient injections that appear to characterne warm-core ocean eddies. Prevlous studies of silicic acid uptake kinetics in natural waters have all been conducted In relatively nutnent-rlch systems and the lower K, values we report here for 2 Gulf Stream warm-core rings Indicate that diatoms growing in severely nutrient-depleted surface waters have a higher afflnlty for slliclc acid than those growing in more eutrophic habitats. If this is true for other nutrient-depleted environments such as the tropics and mid-ocean gyres, the likelihood of significant slhcon limitation in such systems is diminished.

Profiles of silicic acid, chlorophyll a, biogenic silica, and lithogenic silica concentrations and the rate of sllica production were obtained from 7 depths in the upper 100 to 200 m at 8 stations in the western Sargasso Sea between May 7 and 18. 1989 Stations were distnbuted from the southern edge of the Gulf Stream to about 400 km south of Bermuda An additional set of profiles was obtained at Stn S near Bermuda on March 28 1989 Sillcic acid concentrations in the euphotic zone xvere generally between 0 6 and 0 9 pM Blogenic sillca concentrations ranged from 7 to 1400 nmol S1 1-' with concentratlons c 5 0 nmol S1 1-' being typical of stations south of 35" N well away from the Gulf Stream and its eddies Specific production rates of biogenic silica (V, were in the same range as those reported for the picophytoplankton In the Sargasso Sea s u g g e s t~n g that diatoms can gioxv as fast as the more numerous plcophytoplankton in oligotrophic oceans Subsurface maxima in biogenic silica concentration, chlorophyll biomass and silica production rates were observed neai the top of the nitracl~ne at 7 of the 9 stations Integrated silica production rates between the surface and the 0 1 % light depth were generally betxveen 0 2 and 0 7 inmol Si m -d-l with an aveiage of 47 % of that production occurring within the nitracline If diatoms taking up silicic acid xvithin the nitracline also ut~lize nitrate, they may account for between 0 56 and 0 84 m01 C m-' yr.' of nexv production lvhich is 16 to 24 % of the estimated annual total new production for the region)