Marine snow and fecal pellets are formed in the upper seawater, and as an important part of the biological pump (Longhurst and Harrison, 1988), their formation and changes are closely related to the life cycles of phytoplankton and zooplankton in the ocean. On the west coast of Sweden, during a diatom bloom, the number of diatoms increased significantly and the contact between them increased. The rapid aggregation and sedimentation of the bloom made it possible to observe that the in situ concentration of marine snow had reached a peak (Tiselius and Kuylenstierna, 1996). Moreover, during the diatom bloom, the fecal pellet carbon fluxes in the sediment also increased. Graf (1989) concluded that a pulse of fecal pellets from the copepods accounted for 92% of the total carbon settling to the deep North Atlantic at the end of a spring bloom in May. According to the results of Station TJ-A-1, we can roughly divide the obtained data into two periods: the higher flux from May to early August 2014 and the lower flux in the rest of the year. However, between May 2014 and May 2015 in the traps at Station TJ-A-1 in the northern SCS, marine snow and fecal pellets were found in only the bottom sediment trap (1 950 m), even other planktons including planktonic foraminifera and pteropods were commonly found in the upper sediment trap (500 m) (Huang, unpublished data). If the marine snow and fecal pellets in the bottom sediment trap are from the in situ vertical settlement as indicated by previous results (Alldredge and Silver, 1988; Simon et al., 2002; Turner, 2002), then marine snow and fecal pellets should also be found in the upper sediment trap. However, this hypothesis is not consistent with the actual observations from Station TJ-A-1. In addition, the area where the sediment trap was placed, the concentration of chlorophyll a (Chl a) in the surface seawater was not at a high level while marine snow and fecal pellets were not observed in the upper sediment trap at Station TJ-A-1. Therefore, the sources of the bottom marine sediments are mainly due to horizontal transport rather than in situ vertical sinking.
Higher total particle flux (TPF) at greater depths in the northern SCS imply that lateral inputs are strong. At site XS1 (17°24.5′N, 110°55.0′E, Fig.1), TPF averaged 111 mg/(m2·d) at 500 m water depth, while at 1 500 m water depth, TPF averaged 418 mg/(m2·d), which was about four times that in the upper layer trap (Liu et al., 2014). At site M1 (21°32′N, 119°28′E), mean TPF increased from 284 mg/(m2·d) at 248 m depth to 486 mg/(m2·d) at 948 m water depth, and continuously increased to 554 mg/(m2·d) at 2 848 m water depth (Chung et al., 2004). Similarly at site M2 (19°01′N, 117°32′E) average TPF increased from about 200 mg/(m2·d) to 260 mg/(m2·d), corresponding to depths from 1 240 m to 3 230 m (Chung et al., 2004). Higher TPF at greater water depths could be explained by lateral sediment transport by contour current which is defined as the along-slope component of the deep water current (Fig 1, Zheng and Yan, 2012). Station TJ-A-1 was also located on/around a major current route, whose sediment may be a mixture from vertical and horizontal transportation. In addition, the clay mineral compositions obtained from station TJ-A-1 confirmed that tremendous amounts of sediment derived from Taiwan are transported southwestward to the northwestern SCS (Liu et al., 2016).
Reports have shown that contour currents are important transport and sedimentary phenomena that control a majority of deep-sea sedimentation (Rebesco et al., 2014). Equipped with an Acoustic Doppler Current Profiler (ADCP), the sediment trap at the mooring system TJ-A-1 yielded a continuous time-series of current velocity (zonal velocity and meridional velocity). Then, the sub-inertial along-slope and cross-slope velocities (u and v) were obtained. According to the contour direction and velocity from September, 2011 to May 2013, the velocity of the contour current generally varied in the range of –5 to 3 cm/s, and the dominant direction of the deep current was ~250° to true north (Zhao et al., 2015). The suspended sediment concentrations were generally low during intervals of northeastward flows because the Taiwanese rivers were the major sources of fine sediment to the region (Liu et al., 2010, 2016; Zhang et al., 2014; Zhao et al., 2015). Reflection seismic profiles indicate that there were some active NE–SW direction bottom current channels on the northern continental slope of the SCS, which led to the accumulation of discontinuous drifts. Moreover, these drifts propagated southwestward following the direction of the bottom currents (Shao et al., 2007). The marine snow and fecal pellets in the samples of the bottom sediment trap at Station TJ-A-1 may have been brought by the northeast contour current.
Waniek et al. (2000) provided a calculation method for evaluating the effects of horizontal advection and particle sinking speed on particle fluxes as measured by moored sediment traps. The results showed that the distance and direction between a given sediment trap and the region where the particles were produced depends on the mean sinking velocity of the particles, the horizontal velocity field above the trap and the deployment depth of the trap (Waniek et al., 2000). In this study, since in situ data were not available at Station TJ-A-1, the velocities of horizontal advection at depths of 100, 400, 1 700, and 2 000 m (Zhao et al., 2015; Yang et al., 2008) were selected from the northern SCS (Table 1). The velocity of each water layer used was obtained by averaging the velocity of the upper and lower depths. We selected only the current velocity data in the dominant direction of the deep water current for calculation. Using the above method, the trajectories of particulate matter collected by the 1 950 m layer sediment trap at Station TJ-A-1 were calculated.
Table 1. Monthly average velocities used in our calculations
Sinking rates of marine snow and fecal pellets were related to their composition (Alldredge and Silver, 1988; Turner, 2002). And diatoms dominated in our marine snow and fecal pellet samples. As the peaks in fecal pellet carbon fluxes were accompanied by similar peaks in diatoms were found in sediment traps located in the eastern Fram Strait showing that diatoms sank at a similar speed to appendicularian fecal pellets (Lalande et al., 2016). In addition, the sinking velocity of the diatom-related particles at the site SCS-N (18.5°N, 116°E) from the sea surface to the deep SCS was estimated to be ~30–50 m/d by comparing the surface chlorophyll abundance and species of diatoms in sediment traps at depths of 1 003 m and 3 226 m (Ran et al., 2015). The position of the SCS-N site was close to that of Station TJ-A-1. Therefore, if the variation in sedimentation velocity due to the increase in depth during sedimentation was not considered, then the time required to reach the 1 950 m depth from the seawater surface was estimated to be approximately 40–66 days. Given that typical sinking speed of large particles is considered to be a constant 50 m/d, we calculated the sediment source distance in the northeast direction that could be collected by the 1 950 m trap. As the depth of the deployed trap increases, the range of source distance of particulate matter that the trap can collect increases. At the position of the sediment trap TJ-A-1, the maximum lateral displacement between the origin at the sea surface and the trap deployed at 2 000 m was greater than 400 km (Fig. 5).
Figure 5. Range of source distance of particulate matter that the trap can collect in the northeast direction.
The sea surface Chl a concentration data monthly were derived from the MODIS (Moderate Resolution Imaging Spectroradiometer) database in the areas between 15°N to 25°N and 107°E to 122°E, with a spatial resolution of 4 km×4 km. The Chl a concentration was used as an indicator of sea surface productivity. We compared the monthly Chl a concentration data from March 2014 to May 2015, and found that in April and May 2014, there was a significant area of high Chl a concentration off the southwestern shore of Taiwan, which was approximately 3 mg/m3 (Fig. 6). In addition, Chen et al. (2016) calculated the ten-year (2003-2012) median timing of the annual Chl a concentration climax in East Asian marginal seas. The results suggested that the peak time of diatom blooms in the southwestern part of Taiwan is during April of each year. This region is also located within the distance range where the particle matters was produced. Therefore, we speculated that marine snow and fecal pellets formed in large quantities during the diatom blooms around April and May, and after a period of time, they were packed and sank from the light transmission layer of the ocean to the deep ocean. These particles were then transported from the southwest of Taiwan by the northeast contour current to the northern SCS, and they were collected by the deep-layer sediment trap at Station TJ-A-1. The whole process of marine snow settlement, transportation and collection occurred over 2 months.
Observations of marine snow and fecal pellets in a sediment trap mooring in the northern South China Sea
- Received Date: 2019-05-23
- Available Online: 2020-04-21
- Publish Date: 2020-03-01
Abstract: Sediment traps are an important tool for studying the source, composition and sedimentation processes of sinking particulate matter in the ocean. An in situ observational mooring (TJ-A-1) is located in the northern South China Sea (20.05°N, 117.42°E) at a water depth of 2 100 m and equipped with two sediment traps deployed at 500 m and 1 950 m. Samples were collected at 18-day intervals, and 20 samples were obtained at both depths from May 2014 to May 2015. Large amounts of fecal matter and marine snow were collected in the lower trap. The fluxes of marine snow and fecal pellets exhibited a fluctuating decrease between May 2014 and early August 2014 and then stabilized at a relatively low level. Scanning electron microscopy observations revealed that the main components of the marine snow and fecal pellets were diatoms, coccolithophores, radiolarians, and other debris, all of which are planktons mostly produced in photic zone. Used in conjunction with the particle collection range estimates from the lower trap and data on ocean surface chlorophyll, these marine snow and fecal pellets were related to the lateral transport of deep water and not vertical migrations from overlying water column. Moreover, the source area might be southwest of Taiwan.
|Citation:||Meng Gao, Baoqi Huang, Zhifei Liu, Yulong Zhao, Yanwei Zhang. Observations of marine snow and fecal pellets in a sediment trap mooring in the northern South China Sea[J]. Acta Oceanologica Sinica, 2020, 39(3): 141-147. doi: 10.1007/s13131-020-1561-9|