Net biogenic silica production and nitrate regeneration determine the strength of the silica pump in the Eastern Equatorial Pacific

Abstract

The high-nitrate, low-silicic acid character of the eastern equatorial Pacific (EEP) has been attributed to the preferential export of diatom biogenic silica (bSiO 2) over particulate organic nitrogen due to less efficient recycling of Si in surface waters. To gain insight into the strength of this silica pump, we examined [Si(OH) 4] and [bSiO 2] distributions and net bSiO 2 production rates in two regions of the EEP, one spanning the equator from 4°N to 3.25°S at 140°W and the other along a tropical instability wave (TIW) at 0.5°N between 132.5 and 123.4°W, ending within a cold vortex at 1.75°N by 125°W. Large uncertainty in the net bSiO 2 production rate measurement precluded a detailed examination of trends at high spatial resolution, but averaged data revealed clear differences in Si cycling between these two sampling areas. Surface [Si(OH) 4] generally remained at levels <4 μM across both, but [bSiO 2] nearly doubled to values as high as 226 nmol Si L −1 along the TIW. The mean integrated net rate of bSiO 2 production along the meridional transect was no more than 0.29 mmol Si m −2 d −1 to the 0.1% light level depth and −0.31 mmol Si m −2 d −1 to a depth of 300 m, implying net loss of bSiO 2 to dissolution in the upper 300 m in this area. In contrast, integrated net bSiO 2 production rates were five times higher on average in the zonal sampling area, exhibiting a mean of 1.45 mmol Si m −2 d −1 within the euphotic zone that declined by only ∼15% to a depth of 300 m, suggesting a significant potential for silica export to deeper waters along the TIW. In total, the fraction of bSiO 2 produced in the euphotic zone that was supported by new inputs of Si(OH) 4 was at least 3.7 times greater on average than the fraction of inorganic nitrogen taken up as NO 3 − , consistent with expectations for a silica pump. However, the mean integrated rate of NO 3 − uptake exceeded that of new Si(OH) 4 uptake by at least five times, implying preferential nitrate depletion in contradiction to nutrient distributions that indicated preferential Si(OH) 4 drawdown. This discrepancy can be resolved if, on average, ∼70% of the nitrate taken up was regenerated in the euphotic zone rather than being supplied by ‘new’ upwelled sources. These results suggest that the silica pump in the EEP arises only in part from the less efficient recycling of biogenic silica compared to organic matter, but also requires efficient recycling of nitrate in surface waters to reduce the rate of new nitrogen production well below the rate of gross nitrate uptake. Additional work will be necessary to confirm this apparent dependency and its relationship to variations in the physical and biological processes driving new production in this system. A relationship between bacterial protease activity and rates of silica recycling was also identified, but, unlike in coastal systems, the contribution of particle-associated bacteria was relatively unimportant in effecting bSiO 2 dissolution in the EEP.