Keywords

1 Introduction

The Southern Ocean is a very productive region of the global ocean with high stocks of commercial species and, in particular, stocks of Antarctic krill (Kerr et al. 2018). Due to the influence of climatic and anthropogenic factors, changes are observed in the ecosystem of the Southern Ocean, which negatively affects the stocks of commercial species (Atkinson et al. 2004; Clarke et al. 2007; Ducklow et al. 2007). Long-term monitoring is therefore critical for understanding and predicting possible changes in the ecosystem of the Southern Ocean (Kerr et al. 2018). Ocean color remote sensing is an important tool for operating monitoring of the state of this ecosystem in comparison with in situ observations in terms of spatial and temporal resolution. The food web in the ecosystem is based on the primary producer of organic matter, namely, phytoplankton. Therefore, the assessment of the ecosystem state requires the analysis of abundance of phytoplankton and its photosynthetic potential. Concentration of the photosynthetic-active pigment – chlorophyll a – is widely used as a marker of phytoplankton biomass, which can be assessed using satellite data. Previous studies showed that chlorophyll a was overestimated by up to two times calculated by the standard algorithm in the Southern Ocean (Dierssen and Smith 2000). Correct assessment of the chlorophyll a content from space-based scanners require a thorough understanding of in-water optical properties: absorbance and scattering by optically active components. Spectral light absorption coefficients of phytoplankton, non-algal particles (NAP), and colored dissolved organic matter (CDOM) are major parameters that determine the variability in the light field in the ocean. Hence, understanding their variability with respect to environmental factors is required for the development of the regional bio-optical models for assessment of water quality and productivity indicators based on ocean color remote sensing.

Total absorption a(λ) includes contribution of all optically active components (Kirk, 2011):

$$ a\left(\lambda \right)={a}_{ph}\left(\lambda \right)+{a}_{NAP}\left(\lambda \right)+{a}_{CDOM}\left(\lambda \right)+{a}_w\left(\lambda \right), $$
(16.1)
$$ {a}_p\left(\lambda \right)={a}_{ph}\left(\lambda \right)+{a}_{NAP}\left(\lambda \right), $$
(16.2)

where aph(λ) is the light absorption coefficient of phytoplankton at wavelength (λ), aNAP(λ) is the NAP light absorption coefficient, aCDOM(λ) is the CDOM light absorption coefficient, aw(λ) is the absorption coefficient of pure water, and ap(λ) is the particulate light absorption coefficient.

The overestimation of chlorophyll a concentration by standard algorithm is likely to be caused by regional peculiarities in chlorophyll a-specific phytoplankton absorption (Ferreira et al. 2017), and high and variable content of CDOM due to melting ice appeared under climate change (Clarke et al. 2007; Ducklow et al. 2007; Kerr et al. 2018). The spectral absorption by all optically active components has recently been studied only in the northwestern and northern parts of the Antarctic Peninsula (Ferreira et al. 2017, 2018). The lack of bio-optical data in the Powell Basin and the limited amount of spectral light absorption data in the Bransfield Strait require further detailed investigation of spectral bio-optical characteristics of waters in these regions of the Southern Ocean. The goal of this study is to assess variability in spectral light absorption coefficients by phytoplankton, NAP, and CDOM in the Bransfield Strait and Powell Basin in austral summer 2020 (January 11–February 4, 2020).

2 Methods

Bio-optical data were collected during the scientific cruise of R/V “Akademik Mstislav Keldysh” in the austral summer of 2020 (January 11–February 4). The study site covered the Bransfield Strait (BS), between the Southern Shetland Islands and the Antarctic Peninsula, and the Powell Basin (PB) (Fig. 16.1). Station locations that were sampled are shown in Fig. 16.1. The ocean bottom at stations in the Bransfield Strait varied as follows: 37 m (st 6585), 472 m (st 6587), 1260 m (st 6590), 1417 m (st 6591), 1436 m (st 6592), 748 m (st 6593), 196 m (st 6594), and 165 m (st 6595). In the Powell Basin, the stations were in the deepwater region with a bottom deeper than 1000 m, except for shelf stations 6596 (depth 360 m), 6597 (depth 828 m), 6599 (depth 579 m), and 6614 (depth 294 m).

Fig. 16.1
figure 1

Map of stations (●) of the bio-optical investigations carried out in the scientific cruise 79 of the R/V “Akademik Mstislav Keldysh” in the Bransfield Strait and in the Powell Basin from January 11 to February 4, 2020

Full size image

Salinity and temperature profiles were recorded with a Sea-Bird SBE 911 plus CTD unit. The chlorophyll a fluorescence profiles were recorded with fluorometer Minitracka II (Chelsea), which was installed on the transparency meter PUM-200. Photosynthetically active radiation (PAR) incident on the sea surface and its vertical profile were measured with a Li-COR set on the deck (LI-190SA) and with submersible device (LI-192). Water samples were collected at several depths (3–7 depths) using Niskin bottles attached to the CTD. The sampling depths were chosen based on the temperature, salinity, PAR, and fluorescence profiles.

The depth of the photosynthetic zone (Zeu) was taken as the depth with 1% of the PAR incident on the sea surface. The concentrations of chlorophyll a and phaeopigments were determined spectrophotometrically with 90% acetone (Jeffrey and Humphrey 1975; Lorenzen 1967) using a dual-beam spectrophotometer Lambda 35 (PerkinElmer). Water samples were filtered immediately after collection under a low vacuum on Whatman GF/F glass fiber filters and stored in liquid nitrogen until analysis for particulate absorption and chlorophyll a concentration.

Particulate absorption was measured following the standard quantitative filter technique (QFT) in accordance with the current NASA protocol (Boss et al. 2019). To separate the phytoplankton pigments within the particulate matter from NAP, methanol extraction was performed (Kishino et al. 1985). Subsequent extraction with hot water was applied to remove phycobilipigments (Roesler and Perry 1995). The phytoplankton pigment absorption coefficient (aph(λ)) was estimated by subtracting bleached spectral absorption (aNAP(λ)) from unbleached spectral absorption ap(λ). The chlorophyll a-specific light absorption coefficient of phytoplankton was obtained by dividing aph(λ) by chlorophyll a concentration together with phaeopigments (TChl-a). The optical densities were corrected for path length amplification using beta (β) correction algorithm (Mitchell 1990). Optical density of the samples was measured from 350 to 750 nm with dual-beam spectrophotometer Lambda 35 (PerkinElmer) equipped with an integrating sphere.

The aNAP(λ) spectra were described by an exponential function (Babin et al. 2003):

$$ {a}_{NAP}\left(\lambda \right)={a}_{NAP}\left({\lambda}_r\right)\cdotp {e}^{-{S}_{NAP}n\left(\lambda -{\lambda}_r\right)} $$
(16.3)

Wavelength λr is the reference wavelength, 440 nm; SNAP represents the spectral slope for aNAP(λ). SNAP was calculated using a nonlinear least square fitting from 400 to 700 nm.

The CDOM absorption was determined following the current NASA protocol (Boss et al. 2019). Discrete samples were filtered (under low vacuum) through 47 mm 0.2 mm nylon filters (Sartorius Nuclepore) (0.2 μm pore size, 47 mm diameter) using a prefiltration with GF/F (Whatman). Both filters were rinsed with deionized water before filtration. The CDOM absorption was measured with a dual-beam spectrophotometer Lambda 35 (PerkinElmer) using a 10 cm path length quartz cuvette and deionized water for reference. Parameterization of the aCDOM(λ) was done similarly to NAP (Eq. 16.3), but SCDOM, which is the spectral slope for aCDOM(λ), was retrieved for a wavelength of 350–500 nm (Bricaud et al. 1981).

3 Results and Discussion

In the study site, hydrographical characteristics in the investigated layer (30–120 m) varied significantly at the stations (Fig. 16.2). The surface temperature varied from −0.47 °C to 3.19 °C, and the average value was 1.29 ± 1.4 °C in the BS. In the PB the surface temperature was lower than in the BS: it varied from −0.77 °C to 2.01 °C and was 0.72 ± 0.89 °C on average. The largest change in temperature with depth (between 0 m and 80 m) reached ~3 °C in the BS. Vertical profiles of temperature at some stations showed the existence of the upper mixed layer (Fig. 16.2). Conversely, at several stations (at two stations in the BS and at seven stations in the PB), the studied layer (up to 120 m) was almost uniform in temperature.

Fig. 16.2
figure 2

Vertical profiles of the chlorophyll a fluorescence intensity (F) (a), temperature (T) (b), and salinity (S) (c) in the Bransfield Strait and in the Powell Basin from January 11 to February 4, 2020: green line shows the variability in euphotic depth (Zeu) between stations

Full size image

Salinity of the surface layer varied from 31.6 to 35.8 PSU, while the average value was 33.6 ± 1.0 PSU. In the salinity profile, there was a minimum in the layer 0 ~ 25 m. In this minimum region, the salinity was by 0.5–5.0 PSU lower in comparison with that in the surface layer. In a deeper water layer (> ~ 25 m), a salinity of ~34.5 PSU remained constant at all stations (Fig. 16.2).

The euphotic zone varied between stations in a narrow range from 50 m at station 6614 to 83 m at station 6619 (Fig. 16.2), with the exception of two stations (6609 and 6613), where the euphotic layer was shallower (Zeu = 28 and 35 m). The mean Zeu was equal to 64 ± 12 m.

The variability in the profiles of hydrographical characteristics resulted in different vertical distributions of chlorophyll a fluorescence (F) (Fig. 16.2). At several stations, the values of F were distributed homogenously within the layer. But at other stations, depth-dependent variability in F was evident: F profiles are characterized by a maximum, and the intensity and location of maximum varied between stations (Fig. 16.2). The variability in the F profiles is likely to be caused by the different stability in the water column. A similar variability in the profiles of hydrographical characteristics and chlorophyll a fluorescence was found in previous research studies in the austral summer in the northwestern Antarctic Peninsula (Figueiras et al. 1999; Ferreira et al. 2017).

The TChl-a in the surface layer varied by more than one order of magnitude, from 0.25 to 4.44 mg m−3 (Fig. 16.3). The maximum values of TChl-a were found in the Powell Basin: 4.44 mg m−3 at station 6609 and 2.18 mg m−3 at station 6613. The mean value of TChl-a in the surface layer at all stations with the exception of these two stations (6609 and 6613) was 0.72 ± 0.35 mg m−3. In the euphotic layer, TChl-a was in a range of 0.15–1.81 mg m−3 except for stations 6609 and 6613, where TChl-a reached 4.44 and 2.18 mg m−3, respectively. In the layer of chlorophyll a maximum, which was observed at some stations, the TChl-a exceeded 1.3–1.5 times that in the surface layer. On average, in the euphotic layer, the TChl-a was equal to 0.46 ± 0.26 mg m−3.

Fig. 16.3
figure 3

Vertical distribution of the chlorophyll a concentration together with phaeopigments (TChl-a) in the Bransfield Strait and in the Powell Basin from January 11 to February 4, 2020: stations 6609 (), 6613 (), and other stations (Δ)

Full size image

In the vertical distribution of TChl-a, three types of profiles could be distinguished (Fig. 16.4): (1) profiles with relatively uniform distribution of TChl-a (e.g., at st. 6602, Fig. 16.4), (2) profiles with a maximum near the bottom of euphotic zone (~ (1–0.1%) PAR) (e.g., at st. 6604, Fig. 16.4), and (3) profiles with TChl-a decreasing with depth (e.g., at st. 6614, Fig. 16.4). The third type of profile was most common: 16 stations of a total of 26 stations. The obtained TChl-a range of variability was in a good agreement with the results in austral summer in the northwestern Antarctic Peninsula (Ferreira et al. 2017, 2018).

Fig. 16.4
figure 4

Examples of typical vertical profiles of temperature (T, °C, red line), salinity (S, PSU, blue line), photosynthetically active radiation (PAR, μmol·m−2·s−1, black line), and chlorophyll a concentration (TChl-a, mg m−3 ) at 6602 (1, first type), 6604 (2, second type), and 6614 (3, third type) stations in the Bransfield Strait and in the Powell Basin (January 11–February 4, 2020)

Full size image

High variability of TChl-a was accompanied by significant change (more than an order of magnitude) in the spectral light absorption coefficients of phytoplankton. The spectrum aph(λ) has two main maxima: in the blue (at ~438 nm) and in the red (at ~678 nm) spectral ranges.

Light absorption coefficients at blue aph(438) and red maximum aph(678) varied within the euphotic layer from 0.005 to 0.29 m−1 and from 0.003 to 0.12 m−1, respectively. The maximum values were recorded at stations with a high TChl-a (st. 6609 and 6613), the minimum — in the lower part of the euphotic layer.

Vertical profiles of aph(λ) corresponded to different types of vertical distribution of TChl-a (Fig. 16.5). In the case of a uniform distribution of TChl-a (e.g., at station 6602, Fig. 16.5), the values of aph(λ) were almost univariable with depth. The ratio between the coefficients in the blue and red peaks (R) did not change with depth (~ 2.0). Moreover, the chlorophyll a-specific absorption coefficient \( {a}_{ph}^{\ast}\left(\lambda \right) \) did not vary within the euphotic layer. Morel and Bricaud (1981) reported that phytoplankton species composition, cell size, and cellular pigment complex did not change with depth. Such homogeneity in vertical distribution of pigments and functional characteristics of phytoplankton (in particular, specific light absorption) is likely the result of the mixing of waters and phytoplankton within the euphotic layer at the rate exceeding the rate of photoadaptive processes in phytoplankton (Cullen and Lewis 1988).

Fig. 16.5
figure 5

Light absorption coefficient of phytoplankton (aph(λ)) and chlorophyll a-specific light absorption coefficient (aph∗(λ)) at stations in the Bransfield Strait and in the Powell Basin (January 11–February 4, 2020)

Full size image

In the case of a depth-dependent vertical distribution of TChl-a (the second and third types of chlorophyll a fluorescence profiles), the aph(λ) changed markedly with depth (Fig. 16.5). The R decreased from 2.8–2.4 in the surface layer to 1.9–1.5 in the lower part of the euphotic zone. The red maximum of the spectrum is formed due to the absorption of light by chlorophyll a and phaeopigments. In the blue part of the spectrum, not only chlorophyll a but also accessory pigments absorb light (Jeffrey et al. 1997). Consequently, the obtained decreasing of R (smoothing of spectra) is associated with an adaptive decrease in the proportion of accessory pigments relative to chlorophyll a and with a package effect (MacIntyre et al. 2002; Morel and Bricaud 1981). The values of \( {a}_{ph}^{\ast }(678) \) decreased with depth from 0.025–0.023 m2 mg−1 at the surface to 0.019–0.017 m2 mg−1 near the bottom of the euphotic zone. The observed variations with depth (smoothing of the spectra and a decrease in chlorophyll a-specific absorption coefficients of phytoplankton) reflected the phytoplankton adjustment to the environmental conditions that varied with depth (Bricaud et al. 1995). Water stratification (formation of the upper layer with higher temperature/lower density) probably leads to an increase in the stability of the water column, which provides the conditions for the phytoplankton to adjust to the vertical environment gradient within the euphotic zone.

The aph(λ) linked with TChl-a. A relationship between aph(438)), aph(678), and TChl-a was revealed and described with the power function:

$$ {a}_{ph}\ (438)=0.046\times {C_a}^{1.05},{r}^2=0.84 $$
$$ {a}_{ph}\ (678)=0.021\times {C_a}^{1.04},{r}^2=0.89. $$

In the equations, the power exponent is slightly higher than unity, which indicates a slight increase in \( {a}_{ph}^{\ast}\left(\lambda \right) \) with an increase in the TChl-a in water. This reflects a decrease in the degree of pigments packing in phytoplankton cells (Morel and Bricaud 1981). This is probably due to the fact that in more trophic waters (where TChl-a values are higher), phytoplankton is represented by smaller-celled species. Comparison of the obtained equations with those revealed from a large dataset of ocean data (Bricaud et al. 1995) exhibited the difference mainly in terms of the power coefficient.

At station 6609, where TChl-a was relatively higher, at depths of 0 and 15 m, a local maximum was observed at a wavelength of ~545 nm, corresponding to the absorption band of phycoerythrin (Fig. 16.5). The appearance of this local maximum on the spectra is probably due to the relatively higher abundance of phycoerythrin-containing species in the community, in particular: cryptophyte and cyanobacteria (Moore et al. 1995; Novarino 2003; Heidenreich and Richardson 2020). At other depths, no local maximum at ~545 nm was observed.

The NAP absorption is determined by the absorption of light by suspended particles not associated with phytoplankton pigments such as bacteria and detritus (nonliving organic particles) and mineral particles (Morel and Ahn 1991; Stramski and Kiefer 1998). The aNAP(440) varied from 0.002 to 0.027 m−1. The highest aNAP(440) values (0.018–0.027 m−1) were recorded at the shallow-water (37 m) station 6585 near the Bransfield Strait and at stations 6609 and 6613, where the high TChl-a values were found. It was found that aNAP(440) increased with TChl-a. In the euphotic layer, the ratio between aNAP(440) and ap(440) varied from 0.05 to 0.82 and was equal to 0.25 ± 0.12 on average. The spectral slope coefficient (SNAP) ranged from 0.006 to 0.016 nm−1 with mean equal to 0.010 ± 0.002 nm−1. The mean SNAP value is in good agreement with data published for multiple different regions of the global ocean (Bricaud et al. 1998; Babin et al. 2003; Churilova et al. 2017; Efimova et al. 2018).

The CDOM light absorption varied significantly. The aCDOM(440) was in a range of 0.016–0.19 m−1. No correlation was found between aCDOM(440) and TChl-a. The spectral slope coefficient (SCDOM) varied from 0.009 to 0.022 m−1 with inverse trend with aCDOM(440) as observed elsewhere (Nelson and Siegel 2013; Churilova et al. 2017). The mean of the SCDOM was 0.013 ± 0.003 nm−1. The SCDOM values obtained in our studies in the BS and PB regions agree with the data in the northwestern Antarctic Peninsula (Ferreira et al. 2018). The SCDOM values in the Antarctic waters were in the lower range of data for the open ocean (0.015–0.030 nm−1) (Nelson and Siegel 2013) and in the European coastal waters 0.014–0.020 nm−1 (Babin et al. 2003).

The relative contribution of all optical components to the total (without water) light absorption at 440 nm was assessed in the upper layer whose thickness was assumed equal to the first optical depth (Fig. 16.6). The contribution of CDOM to the total light absorption varied from 3 to 92% (on average 45 ± 26%), and phytoplankton varied from 6 to 87% (on average 42 ± 21%). The NAP contribution was the smallest: from 2 to 44%, (on average 13 ± 8%). The absorption budget at 440 nm showed that the contribution of optically active components to light absorption in the first optical depth varied significantly and CDOM and phytoplankton were the main optically active components.

Fig. 16.6
figure 6

Contribution of phytoplankton aph(λ), non-algal particles aNAP(λ), and colored dissolved organic matter aCDOM(λ) to the total light absorption (without water) atot(λ) at a wavelength of 438 nm in the first optical depth in the Bransfield Strait and in the Powell Basin from January 11 to February 4, 2020

Full size image

4 Conclusions

In this research, we examined variability and parameterized the spectral absorption coefficients of optically significant components in Antarctic waters (Bransfield Strait and Powell Basin) based on the in situ dataset in austral summer 2020. Analysis showed variability of more than an order of magnitude of the chlorophyll a concentration, spectral light absorption coefficients of phytoplankton, NAP, and CDOM. The results of parameterization of light absorption by all optically significant components in waters will provide the required scientific basis for the adjustment of the bio-optical algorithm for assessment of chlorophyll a and colored detrital matter absorption (Suslin and Churilova 2016) to the peculiarities of the bio-optical properties of the Antarctic waters (Bransfield Strait and Powell Basin).