Laboratory of Ocean Ecosystem Dynamics

日本語

Tokyo University of Marine Science and Technology

header

Home >> Members >> Mamoru Tanaka

Profile

Mamoru Tanaka

Mamoru Tanaka (Doctoral Student)

M.S., Tokyo University of Marine Science and Technology, Graduate School of Marine Science and Technology, 2012

B.S., Tokyo University of Marine Science and Technology, 2010

Top of page

Phytoplankton Patch

The patch structure of phytoplankton is an aggregation of phytoplankton cells. The size of phytoplankton aggregates ranges from a few hundreds of or tens of μm to a few mm, and an aggregate is difficult to collect because it is easily broken up and destroyed by the physical shock from conventional water sampling methods. In addition, an aggregate is too small to detect using a conventional fluorescence sensor because the patch structure is smaller than the sampling volume.

Fig. 1. a) TurboMAP-L (Turbulence Ocean Microstructure Acquisition Profiler-Laser) and b) laser fluorescence sensor (reduced sampling volume is 32 μL) attached on TurboMAP-L.

To capture delicate structure such as this, TurboMAP-L (Turbulence Ocean Microstructure Acquisition Profiler-Laser, Fig. 1a) was developed (Doubell et al., 2009). TurboMAP-L is a free-falling microstructure profiler, which means that this profiler can observe the ocean accurately and without contamination from vibration of the instrument itself or the ship’s movement. TurboMAP-L has two turbulent velocity shear probes and a laser fluorescence sensor as well as a CTD (Conductivity, Temperature, Depth), other biological sensors and XYZ-direction accelerators.

With a sampling volume of 32 μL, the TurboMAP-L laser fluorescence sensor has the smallest sampling volume of any previous in situ fluorescence sensor (Fig. 1b). TurboMAP-L digitizes its data at 256 Hz and falls downward around 0.6 ms-1. Thus, the high-resolution fluorescence data are digitized every 2.3 mm depth. This data show us the fine-scale distribution of fluorescence.

Fig. 2. High-resolution fluorescence data acquired by laser fluorescence sensor. The range of the y-axis is a) 1 m and b) 5 cm. (b) is the figure zoomed in around 14.4 m depth on (a).

Fig. 2a is an example of the observation data acquired from the ocean. The y-axis shows spatial scale (1 m in this case), and the x-axis shows the intensity of fluorescence, which indicates the concentration of phytoplankton. You can see that the data have ‘spike’-like signals.

Fig. 2b shows the data zoomed-in on a spike near 14.4 m depth, and in this image, the range of the y-axis is 5 cm. You can see that there are the continuous high values of fluorescence around 14.44 m. This graph is made up of about 4 points of data, and its spatial range is about 10 mm. The fluorescence value reaches 500 μgL-1, although the background value is about 50 μgL-1. This means that phytoplankton exist in high concentration reative to the ambient environment in a small area. We call this a patch structure of phytoplankton.

We use the coefficient of variation (the standard deviation divided by the average) to assess the patch structure of phytoplankton.

Tanaka phytoplankton equation

You can calculate the average value of data in a specific area (if you use the data on Fig. 2a, this would be the range of the y-axis, e.g., 1 m) and, based on the average, you can also calculate the standard deviation. (If you use the data on Fig. 2a, the average is 58.15 μgL-1, and the standard deviation is 55.99 μgL-1. So, the coefficient of variation is 0.96.)

The average represents the amount of phytoplankton cells in a specific area, and the standard deviation represents the variance of the distribution of phytoplankton cells. If phytoplankton cells form patch structures, the standard deviation is high. (Considering an extreme example, if phytoplankton cells distribute completely uniformly, the standard deviation becomes zero.) So, because the coefficient of variation is the ratio of the standard deviation and the average, we think that it represents the tendency of phytoplankton cells to make patch structures.

Fig. 3. T-S diagram from the mouth of Tokyo Bay, Japan in June 2011. The color shows the common logarithm of the coefficient of variation of fluorescence calculated for every 1-meter bin.

Fig. 3 is the T-S diagram (Temperature-Salinity diagram) created from data taken at the mouth of Tokyo Bay, Japan in June 2011. The x-axis shows salinity, the y-axis shows temperature, and the black lines represent equal density lines. We can see that there are two main water masses in the area where sea water density is higher than σθ=1025 kgm-3. One water mass is the coastal water mass (relatively low salinity and low temperature) and another is the pelagic water mass (relatively high salinity and high temperature). The color indicates the common logarithm of the coefficient of variation of fluorescence calculated for every 1-meter bin. Warmer (cooler) colors show higher (lower) values. Looking at this image, we can see a clear trend: in the pelagic (coastal) water mass, the coefficient of variation is higher (lower).

Fig. 4. T-S diagram from Otsuchi Bay, Iwate Prefecture, Japan, in July 2005. The color shows the common logarithm of the coefficient of variation of fluorescence calculated for every 1-meter bin.

Fig. 5. a) T-S diagram from the Kuroshio region (located in the northwestern part of the Pacific Ocean, 143-146 degrees east longitude, 35-37 degrees north latitude) in July 2005. The color shows the common logarithm of the coefficient of variation of fluorescence calculated for every 1-meter bin. b) The grid-averaged T-S diagram.

Fig. 4 and 5a are the T-S diagrams from Otsuchi Bay, Iwate Prefecture, Japan in May 2011 and the Kuroshio region (located in the northwestern part of the Pacific Ocean, 143-146 degrees east longitude, 35-37 degrees north latitude) in October 2009. As with Fig. 3, we can see that there are two water masses. In Fig. 4, one water mass is the coastal water mass and another is the pelagic water mass. In Fig. 5a, one water mass is the Oyashio water mass (relatively low salinity and low temperature, nutrient-rich and flows from the northern part of the Pacific Ocean) and another is the Kuroshio water mass (relatively high salinity and high temperature, nutrient-poor and flows from the southern part of the Pacific Ocean). In Fig. 5a, the trend is difficult to see. Fig. 5b is the grid-averaged T-S diagram. In Fig. 5b, we can see the clear trend of the distribution of the coefficient of variation between two kinds of water masses represented in this T-S diagram.

According to the T-S diagrams shown above, the coefficient of variation of fluorescence seems to be higher in the pelagic water mass than in the coastal water mass. Assuming that the coefficient of variation represents the tendency of phytoplankton cells to make the patch structures, the results indicate that phytoplankton cells distribute intermittently (perhaps constructing a patch structure) in the pelagic, generally oligotrophic water mass.

Fig. 6 is a photo taken simultaneously with fluorescence data by a mini-camera (usually used for bio-logging) attached on TurboMAP-L. The range of this photo is 2x2 cm, meaning the size of the particles which we can see in this photo are around 1 mm. Riebesell (Riebesell, 1991a, b) collected in situ organic microaggregates (range also around 1 mm) and, using a microscope, saw that phytoplankton were attached to organic microaggregates. The results of the mini-camera observation and Riebesell’s studies indicate that the patch structure we captured may be phytoplankton cells attached to organic microaggregates.

Fig. 6. A photo taken simultaneously with fluorescence data by a mini-camera (usually used for bio-logging) attached on TurboMAP-L. The range of this photo is 2x2 cm, and the color shows the intensity of light.

Goldman (Goldman, 1984) predicted that nutrient in the pelagic ocean is concentrated not so much in ambient water as in organic microaggregates, and, in or on such organic microaggregates, nutrient efficiently circulates to and is circulated by zooplankton, phytoplankton, bacteria and detritus. This hypothesis is consistent with our finding that phytoplankton cells distribute intermittently, constructing a patch structure, in pelagic, generally oligotrophic water masses.

Considering Goldman’s hypothesis and our results, the patch structure of phytoplankton cells clearly has a key role in the very foundation of the marine ecosystem.

References

Top of page

Biogenic Mixing

Recently, a report by E. Kunze (E. Kunze et al., 2006) and a counterargument from W. Visser (W. Visser, 2007), both featured in the journal Science, stirred up intense discussion about biogenic mixing. Kunze reported that when krill migrate upward, they disturb the sea water. The intensity of this disturbance was highest level of ocean turbulence. In addition, krill have a large population, and daily vertical migration is a common behavior for ocean organisms. Therefore this report attracted the interest of ocean researchers.

Counter to Kunze, Visser argued that krill are unable to generate big eddies given their small size. In other words, krill are too small to mix the ocean efficiently. The key is mixing efficiency Γ:

Tanaka biomixing equation

where ΔP is variation of potential energy, and ΔK is variation of kinematic energy. Even if ΔK is large (e.g. krill impart a lot of power to the water mass), if ΔP is small enough (e.g. krill make small eddies that are the same as their body length, about 1 cm. Small eddies cannot raise potential energy efficiently), Γ will be small. According to W. Visser (2007), Γ will be less than 0.01 when considering krill, meaning only 1% of krill’s kinematic energy contributes to ocean mixing.

If so, what about other marine organisms? If an organism has a large population and can generate large turbulent eddies, what will its impact on the ocean be?

To answer this question, we focused on sardines. Sardines have a large population and are bigger than krill. (So we can expect that Γ will be larger than in the case of krill.) We conducted an observational experiment using the living bodies of sardines in an aquarium located in an amusement park, Yokohama Hakkeijima Sea Paradise, Kanagawa Prefecture, Japan. ‘Seeing is believing’. This experiment was recorded by three video cameras (Mov. 1).

Mov. 1. A movie taken by one of three video cameras in an aquarium located in an amusement park, Yokohama Hakkeijima Sea Paradise, Kanagawa Prefecture, Japan. In the later part of this movie, TurboMAP acquired turbulent shear data in a school of sardines.

We deployed TurboMAP (Turbulence Ocean Microstructure Acquisition Profiler) to capture the turbulent condition generated by a school of sardines. TurboMAP is the instrument which appears in the aquarium in the video shown above (Mov. 1), and its length is about 1.2 m. At about 7m, the aquarium was deep enough to acquire turbulence data, and TurboMAP’s rising speed stabilized after rising up about 1 m.

Biomixing figure

Fig. 1. A time series of turbulent shear data acquired in an aquarium located in an amusement park, Yokohama Hakkeijima Sea Paradise, Kanagawa Prefecture, Japan. The strong signal after about two seconds corresponds with the period when TurboMAP was in a school of sardines.

Fig. 1 is a time series of the turbulent shear data acquired in the video shown above (Mov. 1). We can see the strong signal after about two seconds, which is the period when TurboMAP was in a school of sardines. The signal corresponds to a turbulent energy dissipation rate of more than 1x10-4 Wkg-1 (the highest level of ocean mixing).

We succeeded in capturing the turbulent condition generated by a school of sardines three times. Each of the three times, the turbulent energy dissipation rate was more than 1x10-4Wkg-1.

This was the first directly viewable experiment (observation) to capture the turbulent condition generated by a school of fish. We are publishing the results of this experiment now.

If sardines generate strong turbulent mixing and have an important role in ocean mixing, we may have to change our current assumptions about ocean mixing. Although the mechanism of ocean mixing seems to be driven by physical triggers (wind and tide) and biological triggers (daily vertical migration of organisms and migration of fish schools, Dewar et al., 2006), research concerning the latter trigger has long been neglected. The results of our observational experiment will contribute to the understanding of the ocean ecosystem.

References

Top of page

Home | Research | About Us | Members | Gallery | Publications | Contact