Further investigations on why POC concentrations differ in samples collected by Niskin bottle and in situ pump

Particulate organic carbon (POC) concentrations measured in bottles are often higher than those measured by in situ pumps when samples are taken concurrently. In previous work, we suggested that differential collection of zooplankton might explain this systematic discrepancy in POC between these small volume (bottle) and large volume (in situ pump) techniques. We have now further quantified the carbon contributed by zooplankton collected in the >70-μm particulate fractions from both bottles and pumps at sites in the Mediterranean Sea and Long Island Sound. Our results show that zooplankton abundance and lipid concentrations from zooplankton are ∼one order of magnitude higher in the bottles than in the pumps, supporting the idea that part of the pump–bottle difference is due to collection of more zooplankton by the bottles. Particle washout off the 70-μm mesh used in the in situ pump may cause loss of some particles as well. However, zooplankton in the >70-μm fraction from the bottles contributed only about 1–2 μM POC, which cannot explain the up to 20 μM POC differences observed in this study. Thus, the mechanisms leading to such a large POC difference are still unclear and need to be further investigated. POC concentrations measured using microquartz filters were similar to those using glass fiber filters, suggesting that filter types cannot explain the higher POC observed in bottles, where glass fiber filters are normally used. Furthermore, we investigated several different pump inlet designs to determine how these might affect the ability of pumps to collect and retain large (>70 μm) particles, including zooplankton. The comparison among different pump inlets suggests that inlet design affects the efficiency and retention of large particles and that a sealed filter holder with a narrow right-angle tubular opening is the most efficient at catching/retaining zooplankton.