Getting good particles: Accurate sampling of particles by large volume in-situ filtration

• Published in: Limnology and oceanography, methods methods, 2012-09, Vol.10 (9), p.681-710
• Main Authors: Bishop, James K.B., Lam, Phoebe J., Wood, Todd J.
• Format: Article
• Language: English
• ISSN: 1541-5856; 1541-5856
• DOI: 10.4319/lom.2012.10.681

We deployed the Multiple Unit Large Volume in‐situ Filtration System (MULVFS) to simultaneously sample 12 depths between 10 and 900 m during the US GEOTRACES Atlantic and Pacific intercalibration experiments. Sampling was designed to simultaneously compare large (>51 µm) and small (micron or sub‐micron to 51 µm) size particulates collected by four classes of 142 mm filter holders against those collected using the main MULVFS filter holder. We evaluated Whatman “QMA,” Pall “Supor” (0.8, 0.45, 0.2 µm), and Millipore “MF” (0.45 µm) filters. Paired QMA filters had best particle loading and uniformity and sample particles to 0.8 µm. Paired 0.8 µm Supor filters sample particles to 0.45 µm and provide the best compromise in terms of sample loading and evenness of particle distribution for elements that require total sample digestion. We also found, under the most benign oceanographic conditions, that many 142 mm single and double baffle filter holders lost of up to 90% of the large particle size fraction, especially in the upper 150 m. We designed and validated a new 142 mm filter holder that solves this problem. We further studied the effect of filtration flow rate on large particle size distribution and chemistry; samples from 500‐800 m in the oligotrophic Pacific showed a 50% decrease of > 51 µm Mn, Ba, and Ca (but no effect on P) over a flow velocity range of 0.1 to 1.6 cm s−1. We recommend sampling below 1 cm s−1. The methodology of bottle and in‐situ filtration is also discussed.