Temporal variability of suspended mass and composition in the Northeast Pacific water column: relationships to sinking flux and lateral advection

Temporal variations in full water column profiles of 1–53 μm suspended mass and bulk composition (calcium carbonate, opal, aluminosilicate and organic matter fractions) were determined by large-volume in situ filtration at station M, 220 km west of Point Conception, California (4100 m depth) in the NE Pacific. Occupations during June 1991, October 1991, February 1992, June 1992 and November 1993 demonstrated that suspended mass varied seasonally by a factor of two throughout the mid- and deep-water column, suggesting a dynamic relationship with concurrently measured variations in deep water sinking flux. Suspended mass in the deep water column was highest in the two June occupations and November 1993, and lower during October 1991 and February 1992. Subsurface maxima in suspended Al and opal occurred throughout the upper 1500 m, and are interpreted as lateral advection of shelf and slope derived sediments, occurring mostly in summer and fall. A factor-of-three discrepancy between water column losses of suspended mass between June and October 1991 and integrated near-bottom vertical flux over this period supports the conclusion that sediment traps do not capture important large aggregate flux events which efficiently remove deep suspended matter. Lateral dispersion of advected lithogenous material must occur, but calculations based on the composition and estimated flux of large aggregates suggest that much of the laterally injected mass can be removed vertically on time scales of weeks at this site. As a result, deep suspended particles have a residence time of <1 year with respect to vertical removal at this site, roughly an order of magnitude shorter than in the oligotrophic deep ocean. The short residence times and seasonally varying inventory have important implications for the role of particle dynamics in “boundary scavenging” of reactive elements in ocean margins. Use of the suspended Al data with a simple mixing model between new surface-derived suspended organic matter and old carbon advected at a constant C/Al ratio can account for much of the seasonal and vertical variation in suspended Δ14C (Druffel et al., 1998) for the February and June 1992 profiles, lending support to the hypothesis that sedimentary carbon or dissolved organic matter adsorbed to lithogenous mineral particles determines the apparent age of suspended carbon throughout this water column.