Pre-treatment
In the laboratory, sediment trap samples were sieved through 1 mm plastic mesh to eliminate “swimmer”. Then samples were divided into 10 or more fractions by a splitter (McLane splitter). Some fractions were rinsed with a small amount of distilled water on pre-weighed Nuclepore filters (0.4 mm pore size) and dried at 50°C for 24 hr. After dried samples were weighed, total mass fluxes (TMF: mg m-2 day-1) were computed with numbers of fraction, mouth-square (0.5 m2) and sampling period (day). Consequentially, samples on filters were scraped and pulverized with an agate mortar and stored.
Please note that some samples’ volume were too small to scrape from filters. In this case, Nuclepore filters with samples after weighing were used for ICP-AES measurement (see later). In other cases, a small volume samples were filtered with GF/F filters for measurements of carbon and nitrogen contents.
Organic carbon, inorganic carbon and nitrogen
For measurement of total carbon concentration (TC), samples of approximately 5 mg were precisely weighed in a tin capsule. Samples for measurement of organic carbon (OC) were precisely weighed on GF/F filters. GF/F filters with sample were placed in a container (plastic Tupperware or desiccator) filled with HCl mist to eliminate CaCO3 (decalcification) for approximately 24 hr. After decalcification, filter with sample was wrapped with a tin disk.
Samples in capsule and tin disk were placed in an auto-sampler of an elemental analyzer (Parkin-Elmer 2400) and TC and OC and concentration of nitrogen (N) were measured. Accurate values were determined using an official certified standard material (acetoanilide). Concentration of inorganic carbon (IC) was determined by difference between TC and OC. Mainly, measurement errors were less than 3%. However measurement errors were over 10% or IC becomes minus for small volume samples. In this case, IC was estimated with concentration of calcium (Ca) measured by ICP-AES.
Trace elements
Concentrations of trace elements were measured by using inductively coupled plasma-atomic emission spectrometry (ICP-AES). For this measurement, samples were decomposed in nitric acid solution by the fusion method (Huang et al., Limnol. Oceanogr.: Methods 5, 13-22, 2007 and references in this paper). Samples of approximately 20 mg were precisely weighed in a graphite crucible. Some samples were small and on Nuclepore filters. In this case, Nuclepore filter with sample was directly placed in a graphite crucible. Lithium metaborate (LiBO4) of approximately 140 mg (or approximate 7 times of sample volume) was then add and mixed each other well. Samples in graphite crucible were placed in the muffle kiln (furnace) and kept for 15 minutes under 950°C. Sequentially “lava-like” samples were placed in a teflon beaker with 6% nitric acid solution (HNO3) of approximately 20 ml and stirred for 1 hr. This solution was sequentially filtered with GF/F filter to eliminate graphite debris and, consequently, adjusted to its volume of 50 ml by adding 6% HNO3 solution and weighed precisely. Concentrations of trace elements (Si, Ca, Fe, Ti, Mn, Mg, Ba) were measured by ICP-AES (Perkin-Elmer Optima 3300DV). Concentrations of elements were determined by comparison with HNO3 based ICP-AES standard solutions (Kanto chemicals nos. 20,277-20,280). Rock standards of granite and basalt supplied by National Institute of Advanced Industrial Science and Technology (former Geological Survey of Japan) (Imai, 2000) were used. These standard solutions were prepared by using the same method as for sediment trap sample preparation. Measurement errors were less than 5% for every elements. However concentrations of Ti, Mn, Ba for small volume sample (usually shallower traps) were sometimes minus (under detection limit) and large uncertainty remains in these data.
Note
Detail of sediment trap mooring system is usually described in following our preliminary cruise reports.