Seawater samples were collected in the water column of the Canda Basin and the Bering Sea from aboard the R/B Xue Long during August 1999. Activity concentrations of dissolved and particulate 234Th were measured using beta countion techniques to quantify the scavenging and residence time of 234Th and organic carbon export fluxes. Prinary production (PP)and bacterial production were also determined in the study areas through in situ incubation experiments. Singificant 234Th scavenging was observed in the upper 100m of the water column in both study ares. with up to 40% of 234Th defictit found at Bering Sea stations and -15% of 234Th defictit at the Canada Basin station. Measured PP decreased from -12.5μmol C/m3/hin in surface water to near zero at -100m depth. with an integrated PP of 3.83mmol C/m2/d in the Canada Basin. Bacterial production. on the other hand. was on the order of 2.0mmol C/m2/d. which is up to 52% fo the integrated PP. Particulate organic carbon (POC) export fluxes derived from 234Th/238U disequilibrium were -1mmol C/m2/d in the Canda Basin and -10mmol C/m2/d in the Bering Sea. with fluxes in the latter areaheing 5 to 10 times higher than those found in the Caada Basin. These expon fluxes correspond to a ThE ratio (the ratio of 234Th-derived POC export to primary production) of 0.23 for the Canada Basin and 0.7 for the Bering Sea. The highe ThE rations in the study areas suggest a decoupling of production and particulate export in the high-latiude ocean. Rations fo POC to particlate 234Th (μmol C/dpm) decreased consistertly with incerasing depth, suggestion that organic carbon is preferentially remineralized relative to 234Th. Interestigly, The profile of particulate 234Th in the Canada Basin showed a unique characteristic: particulate 234Th activeites increased with increasing depth. suggesting a contiuous scavenging of 234Th and a rapid settling rate of the particles.

Natural colloids are abundant in seawater and are an intermediary in the fate, transport and bioavailability of many trace elements. Knowledge of the pathways and mechanisms of the biological uptake of colloidal Fe and other Fe species is of paramount importance in understanding Fe limitation on marine phytoplankton and thus carbon sequestration in the ocean. Whether the natural colloids serve as a source for the biological Fe requirements of marine phytoplankton, or just as a sink for particle-reactive metals in the oceans remains largely unknown. This study examined the bioavailability of Fe bound with colloids from different regions to a coastal diatom (Thalassiosira pseudonana). Natural colloids were isolated by cross-flow ultrafiltration and radiolabeled with 59Fe before being exposed to phytoplankton. Control experiments were conducted to ensure that 59Fe radiolabeled onto the colloids remained mostly in the colloidal phase. Both the natural oceanic and coastal colloidal organic matter complexed Fe (1nm-0.2Am) can be biologically available to the marine diatom even though its uptake was lower than the low molecular weight counterparts. By comparing the measured Fe internalization fluxes and the alculated maximum diffusive uptake fluxes, it is evident that ligand exchange kinetics on the cell surface may control the internalization of macromolecular Fe. The calculated concentration factors under dark and light conditions were generally comparable. Colloidal Fe, as an important intermediary phase, can be actively involved in the planktonic food web transfer through biological uptake and regeneration processes. The bioavailable fraction of Fe may be substantially underestimated by only considering the truly dissolved Fe or overestimated when using the external fluxes, such as aerosol Fe, as the bioavailable fraction.

Iron (Fe) is mostly complexed with the organic ligands or colloids that are very abundant in natural seawater. In this study, natural colloids were isolated by ultrafiltration and radiolabeled with 59Fe. The biological uptake of radiolabeled colloid-bound Fe of different sizes (1-10kDa and 10kDa-0.2mm) and ages (1 and 15d) by diatoms (Thalassiosira pseudonana) and copepods (Acartia spinicauda) was then determined. The uptake of radiolabeled colloid-bound Fe was compared with the uptake of low molecular weight (LMW) complexed Fe (1 kDa) or that of EDTA-Fe (at a ratio of 1:2 for Fe: EDTA). Our study demonstrates that the colloid-bound Fe of different sizes and ages was bioavailable to the diatoms. The uptake of colloidal bound Fe was, however, 6-31 times lower than the uptake of LMW Fe, suggesting that colloidal binding reduced Fe bioavailability to diatoms. Fe bound with small colloids (1-10kDa) was taken up at a higher rate than Fe bound with large colloids (10kDa-0.2mm) at typical colloidal organic carbon concentrations. The uptake of colloid-bound Fe was also much higher when the Fe had been bound with the colloids for 1d rather than for 15d. Differences in the colloidal organic carbon concentration did not appreciably affect the uptake of colloid-bound Fe by the diatoms. Similarly, copepods accumulated colloid-bound Fe at a much higher rate when the Fe was associated with small colloids rather than with large colloids. Direct ingestion of colloidal particles by the copepods appeared to be negligible. In both diatoms and copepods, Fe uptake may involve its dissociation from the colloids before being accumulated by the organisms. The study therefore demonstrates that colloid-bound Fe is less available for marine phytoplankton and that colloidal size and thus the colloidal speciation may be critical in controlling colloid-bound Fe uptake by aquatic organisms.