Without small oceanic organisms atmospheric CO2 levels would be about 200 ppm higher than they are today; phytoplankton convert dissolved inorganic carbon (DIC) to particulate organic carbon (POC) during photosynthesis, influencing the air-sea exchange of CO2. Eventually some of this POC is exported out of the upper ocean, often as either phytodetrital aggregates or zooplankton faecal pellets. Because of the complexity of this biological carbon pump (BCP), the fate of the exported POC in the mesopelagic zone is difficult to predict. To make things more complex all of these processes vary temporally and spatially. Marine snow catchers (MSCs) were used to analyse fast and slow sinking particles separately, which is a unique approach as slow sinking POC fluxes are not often quantified. To investigate what controls the fate of particles in the upper mesopelagic zone (50 - 500 m) particles were collected from three contrasting oceanic regions: the Southern Ocean (SO), Equatorial Tropical North Pacific (ETNP) oxygen minimum zone (OMZ) and the temperate North Atlantic. In all sampling areas the slow sinking POC flux was as large if not larger than the fast sinking POC flux. This emphasises the importance of slow sinking particles in the upper mesopelagic zone. The main outcome from this thesis is the importance of the role of zooplankton in BCP processes. For instance the efficiency which particles were exported from the mixed layer varied inversely with primary production in the SO, and was likely due to the zooplankton grazing down the phytoplankton. When extending the data to include the ETNP and the North Atlantic this relationship still held, conflicting the long-standing theory that as primary production increases export efficiency increases. In the ETNP oxygen minimum zone a high proportion of exported POC sank through the mesopelagic zone. Microbial oxygen uptake incubations showed for the first time that fast sinking particles are turned over significantly slower than slow sinking particles (0.13 d?1 and 5 d?1 respectively). Microbial degradation of POC could explain most of the fast sinking POC attenuation with depth, with the remainder lost due to abiotic fragmentation. Therefore it is likely that zooplankton degradation of particles is reduced in OMZs as their abundance and metabolism are lowered. This reduces the overall remineralistion of POC, hence a higher fraction of POC is transferred to depth in OMZs. Phytoplankton lipid biomarkers dominated lipid particle composition throughout the upper mesopelagic zone in the ETNP, further emphasising the minor role of zooplankton in OMZs. Comparing the observations with an ecosystem model output at all three oceanic sites further emphasised the importance of zooplankton in the BCP. The model poorly parameterises zooplankton processing of particles and thus the observations and model matched best in the ETNP, where zooplankton processing of particles is naturally low. Changes in climate will effect the abundance and distribution of these small organisms. Further understanding of how zooplankton community structure and metabolism may change in the future will be important to predict how atmosphericCO2 levels may change.