Sinking fluxes of living polycystine and phaeodarian radiolarians and their intact empty skeletons were measured from surface waters to 2000 m using free-floating particle interceptor traps at three sites ranging from highly productive coastal upwelling to oligotrophic central gyre waters in the northeast Pacific Ocean. Total radiolarian fluxes and living pheodarian fluxes were generally highest at the coastal site throughout the water column. There was no consistent site-specific pattern for fluxes of living polycystines, polycystine empty skeletons, and phaeodarian empty skeletons. Living phaeodarians were the only group that showed the same rank order of sites with respect to flux at both the base of the euphotic zone and at 2000 m. Thus different short-term processes occurring in the water column (e.g. destructive and non-destructive predation and midwater addition of living radiolarians) altered radiolarian fluxes. Neither radiolarian fluxes at the base of the euphotic zone nor fluxes at 2000 m showed a simple correspondence with primary production, indicating that short-term measurements did not reveal long-term patterns. At most depths at all sites, fluxes of living, skeleton-bearing phaeodarians outnumbered fluxes of empty phaeodarian skeletons. In contrast, at most depths at all sites, fluxes of empty polycystine skeletons outnumbered fluxes of living, skeleton-bearing polycystines. Relatively large, living skeleton-less phaeodarians were the numerically dominant radiolarian in the uppermost traps at the oceanic and coastal sites. These phaeodarians agglutinate siliceous skeletons of other plankton and contributed 5% of the silicoflagellate flux, 16% of the polycystine flux, and 2% of the centric diatom flux at these depths. The resemblance of skeleton-less phaeodarians to fecal pellets may cause them to be misidentified as fecal material when trap contents are dried for processing.
Read full abstract