This study investigates oceanic biogeochemical cycles of nutrients and silicate.An adjoint model is used to calculate the 3D large scale ocean circulation and biogeochemicalfluxes of nutrients and silicate simultaneously. Advection and diffusion of dissolved nutrients, productionof particulate matter, and vertical particle fluxes are parameterizedto achieve a 3D flow field and biogeneous particlefluxes consistent with hydrographic and nutrient data.Vertical fluxes are parameterized for particulate organic carbon, calcite, and opal separately.It is shown that simulated distributions of temperature, salinity, nutrients and silicate can indeedbe brought to good agreement with data.The resulting flow field is consistent with geostrophic dynamics and contains major current.Resulting biogeneous particle fluxes are reasonable in their spatial distribution and magnitude.A major goal of the model calculations is to examine whether particle fluxes determinedwith the adjoint method conflict with direct flux measurements.Sediment trap data fromthe German Joint Research Project SFB261 ``The South Atlantic in the LateQuaternary'' are assimilated into the model. The comparison of model results and sediment trap datareveals that model fluxes are generally higher than direct measurements. Even if the model is forcedto reproduce sediment trap data, systematic deviations remain. A solution which gives particle fluxesin agreement with sediment trap data and data of dissolved nutrients cannot be obtained.The results from adjoint modeling indicate that sediment traps do not catch sinking particles quantitativelybut trapping efficiency seems to be low, especially at shallow water depths.An extension of the model is the calculation of sediment accumulation rates.Budgets of dissolved nutrients in the bottom layer are used for indirect determination of accumulationrates for organic carbon, calcite, and opal. The accumulation rates derived with the adjoint modelare partly agree with recent independent estimates.Model values for calcite accumulation are lower than literature values.The resulting model fields of physical circulation and biogeochemical fluxes bear very specialconditions in the Weddell Sea. The Southern Ocean is generally identified as a HNLC (High Nutrients Low Chlorophyll)region, i.e. organic carbon production is low inspite of high nutrient concentrations in surface waters.Further it is known that the Southern Ocean plays a major role in silica cycling which is reflected in high modelopal productivity. On the other hand, opal fluxes in the deep Weddell Sea were found to be rather low.The 'high production low flux anomaly' is also reproduced in the model results.The results presented in this study give confidence thatadjoint modeling can be used to calculate vertical particle fluxesfrom water column nutrient distributions.It is proposed that further refinement of the model grid and the inclusion of more, independenttracers in the model calculations can be used for a better understanding of biogeochemical cycles.