Limnol. Oceanogr., 44(5), 1999, 1282–1292

The contribution of Rhizosolenia mats to silica cycling in the central North Pacific and the coupling of mat silicon metabolism and their vertical migration was examined in areas to the west of the Hawaiian Islands (23–288N and 159–1758W) in 1995 and to the east of Hawaii along 318N (160–1278W) in 1996. The biogenic silica content of Rhizosolenia mats sampled in 1995 averaged 1.82 6 1.87 (SD) mmol Si mat21. Larger mats that averaged 4.56 6 3.54 (SD) mmol Si mat21 were observed in 1996. Kinetic experiments indicated that substrate limitation of mat silica production was widespread across the study region, with ambient [Si(OH)4] restricting silica production to 33% of maximum potential rates. Three lines of evidence indicate that silicon metabolism is not tightly coupled to the migration of mats to and from the nutricline. In 1996, mats in surface waters could double their Si content in 0.55 d on average without migrating to the nutricline to obtain Si. However, average doubling times (9.8 d) in 1996 were of the same order as a migration cycle, necessitating significant Si uptake at depth. Si uptake rates did not differ significantly between ascending and descending mats, suggesting that mats ascending from the nutricline had not fulfilled their Si uptake requirements. Finally, small internal pools of Si in ascending mats indicated that if significant amounts of Si were taken up at depth, they were not stored for use in the surface waters. The biomass and silica production rates of mats collected using SCUBA in the upper 20 m were extrapolated to 150 m by using abundances determined using a video plankton recorder (VPR). The results suggest that mats account for about 3% of the standing stock of biogenic silica and 26% of silica production in the upper 150 m. The daily silica production by Rhizosolenia mats (317 mmol Si m22 d21) is 50–76% of the total silica production in the Sargasso Sea. This high rate of silica production combined with the wide geographic distribution of mats throughout several mid-ocean gyres suggests that mats may contribute significantly to global silica production. Macroscopic aggregates of diatoms of the genus Rhizosolenia were first described in detail by Carpenter et al. (1977) in the Sargasso Sea. These ‘‘mats’’ range in size from 1–30 cm and have been reported in several oligotrophic warm seas, including the Indian Ocean (Wallich 1960), the California Current (Alldredge 1982), and in greatest abundance in the North Pacific Gyre (Villareal and Carpenter 1989). Regionally, these diatom mats contribute significantly to oceanic nutrient budgets. Episodic blooms of large diatoms such as Rhizosolenia have the potential to contribute significantly to oceanic new production (Goldman 1993), and several observations have supported this concept (Sancetta et al. 1991; Blain et al. 1997). Observation of an extremely intense and extensive bloom of Rhizosolenia was documented in the equatorial Pacific in 1992 (Yoder et al. 1994). Large diatoms, including several Rhizosolenia sp., are