Regional Studies in Marine Science 36 (2020) 101311

Contents lists available at ScienceDirect

Regional Studies in Marine Science

journal homepage: www.elsevier.com/locate/rsma

Spatio-temporal quantification and distribution of diatoms and

dinoflagellates associated with algal blooms and human activities in
Algiers Bay (Algeria) using Landsat-8 satellite imagery
∗
R.L. Boufeniza a , , M.M. Alsahli b , N.I. Bachari c , F. Houma Bachari a
a
Higher National School of marine Sciences and coastal Management, University Campus Dely Ibrahim, Bois des cars, PB 19, 16320, Algiers, Algeria
b
Kuwait University, Collage of Social Sciences, Department of Geography, P.O. Box: 68168, Kaifan 71962, Kuwait
c
University of Sciences and Technology Houari Boumediene, Department of Biology, BP 32 El Alia 16111 Bab Ezzouar Algiers, Algeria

article info a b s t r a c t

Article history: Quantifying phytoplankton and its changes in coastal waters is an important aspect of marine ecosys-
Received 25 March 2019 tems. Synoptic observations of phytoplankton communities became possible with remote sensing
Received in revised form 11 April 2020 technologies and spatial analysis methods. Using these advancements would assist the scientific
Accepted 8 May 2020
community to gain better understanding of marine ecosystem dynamics and consequently contribute
Available online 12 May 2020
in sustaining and maintaining these valuable ecosystems. Thus, this study aimed to evaluate the
Keywords: water quality and aquatic ecosystem status of Algiers Bay, Algeria, through the quantification of
Remote sensing diatoms and dinoflagellates species using in-situ measurements and Landsat images. We developed
GIS empirical models to assess phytoplankton abundance by comparing in-situ measurements of diatoms
Algae bloom and dinoflagellates with various spectral bands and band ratioing products derived from Landsat-8
Marine biogeography OLI images acquired in 4 April 2017 and 9 May 2018. The results showed a wide variation between
Mediterranean sea coastal and offshore stations. The number of diatoms was less than that of dinoflagellates. These two
Aquatic ecosystem
algae communities illustrated a different spectral behavior. Diatoms showed a significant relationship
with one of the ratioing products (i.e., log B7/log B2), whereas dinoflagellates exhibited an insignificant
relationship with the remotely sensed dataset. Also, we learned that the southeastern area of the bay
can be influenced by harmful algae blooms developed offshore because of hydrodynamic forces that
push these blooms toward this area of the bay.
© 2020 Elsevier B.V. All rights reserved.

1. Introduction fish death, respiratory disorders of organisms and decreased wa-

ter transparency (Al-Shehhi et al., 2012; Ghanea et al., 2016;
Understanding the spatial and temporal distributions of phy- Richlen et al., 2010). Eutrophication events can also have a neg-
toplankton communities provides a clear perspective of marine ative economic impact including losses of coastal aquaculture,
and coastal ecosystems (Torrecilla et al., 2009). Quantifying phy- increasing costs of water desalination and degrading of recre-
toplankton species, primary producers, is an efficient means to ational sector value (Harvey and Menden-Deuer, 2012; Lee et al.,
evaluate the seawater quality and ecosystem status (Johnson 2016; Tillmann et al., 2016). Thus, an adequate coastal and ma-
et al., 2013; Lee et al., 2017; Wei et al., 2019), as these species rine monitoring system is a fundamental means to protect these
are highly responsive to environmental changes and eutrophica- ecosystems (Ogashawara et al., 2014; Tundisi et al., 2008).
tion (Tepanosayn et al., 2017), especially in coastal ecosystems Diatoms and dinoflagellates are dominant plankton species
where anthropogenic activities are intensive (Zhai et al., 2013). that have an essential role in marine ecosystems and biogeo-
Eutrophication accelerates the growth of harmful algae that sig- chemical cycles (Barton et al., 2013; Cushing, 1989; Illoul et al.,
nificantly disturb coastal ecosystems. Developing algae blooms 2008). Fluctuations in diatoms and dinoflagellates may therefore
under eutrophication conditions (e.g., increasing inland nutrient have an impact on processes in the entire food chain (Leterme
inputs and favorable temperatures) can significantly degrade the et al., 2006). These species are efficiently transported by marine
seawater quality (Richardson, 2004), causing a massive impact on currents (Hallegraeff and Bolch, 1992) for long distances extend-
the ecosystem including decreased dissolved oxygen, subsequent ing their roles not only at local scales but also at regional and
global scales (Ahn and Shanmugam, 2006). Thus, an accurate
∗ Corresponding author. observation of these species would broaden our understanding
E-mail address: [email protected] (R.L. Boufeniza). of marine and coastal ecosystems.

https://doi.org/10.1016/j.rsma.2020.101311
2352-4855/© 2020 Elsevier B.V. All rights reserved.
2 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311

The most common approach to observe diatoms and dinoflag- wastewater is still a significant issue due to the lack of effective
ellates is mainly through field sampling at defined locations treatment strategies and management of wastewater discharge
and subsequent laboratory analysis (Giardino et al., 2001). The at different flows (Seridi et al., 2007). These point and nonpoint
quantification of these species at larger geographical scales is sources of water pollution are disturbing Algiers Bay ecosystems
logistically difficult (Gobler et al., 2017; Ogashawara et al., 2014; and are believed to have a potential role in fluctuating phyto-
Shin et al., 2018). An alternative approach to overcome this issue plankton biomass (Morán et al., 2001; Raimbault et al., 1993;
is to use remote sensing imagery that can significantly estimate Taupier-Letage et al., 2003).
variations in density, size and pigment composition of species’ Phytoplankton biomass and communities within Algerian
cells from a synoptic perspective. Phytoplankton communities coastal waters were not significantly investigated (Illoul et al.,
influence the apparent optical properties (AOP) of seawaters that 2008). In the Algiers Bay, the most common species as reported
can be modeled using remote sensing imagery (Hoepffner and by Illoul et al. (2008) were diatoms and dinoflagellates, with a
Sathyendranath, 1993; Roesler et al., 1989; Ulloa et al., 1994). total of 14 toxic species, and 18 bloom-forming species identified.
Most of phytoplankton models were developed using Advanced Some phytoplankton species discovered in Algerian coastal wa-
Very-High-Resolution Radiometer (AVHRR), Sea-viewing Wide
ters, such as Alexandrium minutum, Dinophysis sacculus (Dinoflag-
Field-of-View Sensor (SeaWiFS), Moderate Resolution Imaging
ellates) and Pseudo-nitschia calliantha (Diatom) are considered as
Spectroradiometer (MODIS) and Geostationary Ocean Color Im-
toxin producers, whereas Lepidodinium chlorophorum (Dinoflag-
ager (GOCI). These sensors have spatial resolutions that range
ellate), Cyclotella meneghiniana (Diatom) and Holococcolithophora
from 250 to 1000 m. This coarse spatial resolution limits the use
sphaeroidea are bloom-forming species.
of these sensors in small coastal areas (Shin et al., 2018; Vincent
et al., 2004). Landsat-8 with high spatial and spectral resolutions
can be an alternative for these sensors in such areas. Landsat-8 3. Methodology
images have been efficiently applied in various aquatic studies
including phytoplankton monitoring (Bresciani et al., 2018; Wei Investigating phytoplankton biomass within Algiers Bay went
et al., 2019; Yadav et al., 2019). through seven stages (Fig. 2): in-situ data collection, laboratory
A comprehensive monitoring of phytoplankton communities analysis, geospatial data acquisition, image pre-processing, im-
in Algiers Bay using in-situ and remotely sensed data have not age processing, in-situ-satellite data matching up and statistical
been established yet. Thus, Landsat-8 images and in-situ data modeling. Phytoplankton algae samples were collected in the
were used in this study to understand the spatial and temporal Algiers Bay within ±2 h of Landsat-8 overpass. These samples
distributions of diatoms and dinoflagellates in Algiers Bay. The were analyzed in laboratory to quantify diatoms and dinoflag-
study objectives were to evaluate the water quality and aquatic
ellates cells using an inversed microscope. Geospatial data (GIS
ecosystem status through: (1) the quantification and identifica-
layers and Landsat-8 images) were used to model and map
tion of diatoms and dinoflagellates; and (2) estimation of tempo-
diatoms and dinoflagellates spatial and temporal patterns within
ral and spatial distributions by developing empirical models that
Algiers Bay. A shapefile of Algiers Bay boundary was downloaded
integrate in-situ measurements of diatoms and dinoflagellates
from the national map of National Institute for Cartography and
algae with various spectral bands and band combinations derived
Remote Sensing (NICRS) (http://www.inct.mdn.dz), whereas the
from the Landsat-8 OLI images in spring of 2017 and 2018.
bay’s bathymetry dataset was obtained from the Houma et al.
2. Study area (2006) study. Cloud-free Landsat-8 images of the study area
were downloaded from the US Geological Survey (USGS) website
Algeria faces the Mediterranean Sea with a coast extending to (https://www.usgs.gov). These images were radiometrically and
about 1300 km. Algiers Bay is semi-circular shape with an average atmospherically corrected during the pre-processing stage. Cor-
depth of 38 m located in the middle of this coast at coordinates rected images with near-concurrent in-situ data were matched
of 36.82◦ North and 03.19◦ East. The bay has an area of 180 km2 to investigate the relationship between water surface reflectance
extending 18 km from East to West and 7 km from North to from different spectral regions and the phytoplankton cell density
South. The bay is surrounded by the Mitidja plain in the South, (i.e., diatoms and dinoflagellate algae). The following subsections
tip of the Pescade (Raïs Hamidou) in the West and Cape Matifou described the methodology in detail.
(El Marsa) in the East (Fig. 1).
Algiers Bay has a significant value to the local and national 3.1. In-situ measurement
socio-economic developments. It is highly populated coast that
hosts more than 3 million people (10% of the Algerian popula-
Phytoplankton algae samples were collected in the Algiers Bay
tion). The bay experiences various anthropogenic activities that
on 4 April 2017 and 9 May 2018 within ±2 h of the satellite
contribute in degrading the bay’s ecosystems through point and
overpass. Such a narrow time window between in-situ data col-
nonpoint sources of water pollution. The coastal zone of Algiers
Bay hosts about one fourth of national companies and Algiers lection and satellite overpass is essential to minimize the effect
Port, one of the most important ports in the country. The bay of environmental condition changes over the time (Butt and
receives industrial water from Wadi El-Hamiz, Wadi El-Harrach Nazeer, 2015; Mouw et al., 2015). The phytoplankton samples
and El Hamma seawater desalination plants (Illoul et al., 2008). were taken from the surface water (1 m) using Niskin bottle of 5 L
Also, the coastal zones experienced over the last decades many at 14 sites in the first voyage and 16 sites in the second voyage
coastal management projects that have contributed significantly (Fig. 1). The water samples were filled in 250 ml opaque and clean
in increasing the amount of nutrients and suspended sediments polypropylene bottles using a tube to protect the planktonic cells
(Woodcock et al., 2008). from deterioration and to take a representative sample. Water
Wastewater is another significant factor that contributes in samples were fixed with acidic Lugol’s solution that changed the
degrading the bay’s water quality by increasing concentrations water color to light brown. The samples were stored at 4 ◦ C in
of organic matter, suspended solids, and chemical and biologi- coolers, and then taken to the laboratory for analysis. The voyages
cal pollutants. Although the Algerian government spent consid- were conducted in favorable weather conditions with clear skies
erable efforts to protect the environments of this region, the and calm seas.
 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311 3

Fig. 1. Algiers Bay and sampling stations.

Fig. 2. Flowchart for the method proposed in this study.

3.1.1. Phytoplankton quantification Mcmanus and Katz, 2009; Palmer and Maloney, 1954). Among
Several techniques can be used to quantify and identify phy- these techniques, we followed the standard guide for phytoplank-
toplankton cell densities (Downing and Rigler, 1984; Hasle, 1978; ton counting by inverse microscopy (SGPCIM) - Standard NF EN
Havens et al., 1996; Hötzel and Croome, 1999; Lund et al., 1958; 15204, 2006 (AFNOR, 2006). The guide is based on the Utermöhl
4 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311

method (Utermöhl, 1958) and adopted by many marine and noise. Pixels over shallow waters (<8 m in depth) were removed
coastal monitoring programs worldwide (Downing and Rigler, from the analysis to minimize the bottom reflectance (Lodhi,
1984; Havens et al., 1996). We counted and identified planktonic 2002). Excluding these invalid pixels reduced the uncertainty of
cells (diatoms and dinoflagellates) through systematic identifi- modeling phytoplankton cell density.
cation procedures using an inversed (optical) microscope. Cell
counting was performed with horizontal scanning in a prepared 3.3. Comparison of in-situ and satellite data
sedimentation tank (2.5 to 10 ml after a settling period of 48
h). The results were expressed in cells density per volume (Cell In-situ measurements taken within two hours or less from
l−1 ) (Barton et al., 2013; Druart and Rimet, 2008; Throndsen, the satellite overpass were matched with the processed Landsat
1995). The process of counting and identification took 20 days images by extracting the pixel values of 3-by-3 window centered
from the date of the sampling voyage, following the work of at each in-situ point. The mean of each extracted window was
Laplace-Treyture et al. (2009) work. computed using a filtering approach described by Bailey and
Werdell (2006). The filtered mean is mathematically expressed
3.2. Remote sensing data by the following equation:
.5 ∗ σ − X ) < Xi < (1.5 ∗ σ + X )
∑
3.2.1. Landsat-8 satellite imagery i (1
filtered Mean = (1)
Two Landsat-8 images downloaded on 4 April 2017 and 9 May n
2018 were used to model phytoplankton cell density within Al- where X and σ are the mean and the standard deviation of
giers Bay water as Landsat imagery was successfully used in mea- the nine-pixel values from each window. n is the number of
suring phytoplankton abundance (Bresciani et al., 2018; Nazeer pixel values within ± 1.5 ∗ σ from the mean. The filtered mean
et al., 2017; Wei et al., 2019). Landsat-8 operated by USGS was was disregarded whenever n was less than five. This matching
launched on 11 February 2013 (Bonansea et al., 2018; Roy et al., approach was applied to the first seven bands (i.e., from coastal
2014). The satellite carries two sensors: The Operational Land aerosol to SWIR-2 band) of processed Landsat OLI images.
Imager (OLI) and the Thermal Infrared Sensor (TIRS) with a spatial
resolution of 30 m and 100 m, respectively (Olmanson et al., 3.4. Statistical analysis and spatial analyses
2016; Roy et al., 2014). The temporal resolution of Landsat-8 is
16 days, and its images are delivered in 16-bit unsigned integer The in-situ datasets were statistically summarized to explore
format in level 1. The landsat-8 OLI sensor has four visible bands, the phytoplankton species temporal variability. Also, the spatial
four near infrared bands, and one panchromatic band, whereas variability of diatoms was compared to dinoflagellate spatial vari-
TIRS has two thermal infrared bands (Msi et al., 2019; Pahlevan ability at study sites using Chi-square test. The relationship of
and Schott, 2013; Vanhellemont and Ruddick, 2014). Landsat- plankton (i.e., diatoms and dinoflagellate algae) cell density to the
8 images are distributed by USGS through their website (https: seven bands of processed Landsat OLI images was investigated
//www.usgs.gov). using the correlation coefficient and regression analysis. The re-
lationship was investigated by comparing in-situ observation to
3.2.2. Sentinel-2A satellite imagery reflectance values of each band and the products of ratioing
A Sentinel-2A image obtained on 15 May 2018 (6 days after these bands. Data transformations were applied for non-normally
the second voyage) was used for visual interpretation purposes distributed variables to meet linear regression assumptions. The
to track algae blooms. Sentinel-2A operated by European Space accuracy of models was evaluated using root mean squared error
Agency was launched on 23 June 2015 carrying Multispectral (RMSE) and mean absolute percent difference (MAPD). Also, the
Imager (MSI) that records 12 spectral bands from blue to short- spatial interpolation technique was used to estimate the spatial
wave near infrared (SWIR) spectrum with a spatial resolution and temporal distribution of dinoflagellates in spring 2017 and
ranging from 10 m to 60 m (www.sentinel.esa.int). The sensor 2018. All statistical analyses were preformed using IBM SPSS
has a red-edge band with a spatial resolution of 20 m that Statistics 24 (IBM Corp, New York, USA), whereas the spatial
allows detecting chlorophyll-a absorptions in case II waters, such analysis was conducted using QGIS 3.6.
as Algiers Bay (Dörnhöfer et al., 2016). Sentinel-2A images are
distributed through the Copernicus Open Access Hub website 4. Result and discussion
(https://scihub.copernicus.eu/).
The mean diatom and dinoflagellate cell densities in 2017 and
3.2.3. Image preprocessing 2018 were 5084, 1502, 15 589 and 21 655 Cell l−1 , respectively
Landsat-8 level 1 images were radiometrically and atmospher- (Table 1). Dinoflagellate cell densities were higher than diatom
ically corrected using ENVI 5.3. Row values (i.e., digital numbers) cell densities in all stations in 2017 and 2018. In general, phy-
of the images were converted to surface reflectance values. Then, toplankton cell densities in the northwestern side of the bay is
the dark object subtraction (DOS) method was used to atmo- lower than densities in the eastern side of the bay. The max-
spherically corrected images. This method is largely used due to imum diatom cell density was observed at station 16 in 2017,
its simplicity. The DOS-based approach assumes that the mini- whereas the maximum dinoflagellate cell density was observed
mum value of each band is zero (Nazeer et al., 2014; Norjamaki at station 16 in 2018 (Fig. 3). The spatial distributions of diatoms
and Tokola, 2007), and minimum values above zero occur due and dinoflagellates at the study sites exhibited a distinct pattern
to scattering. The scattering effects are minimized by subtracting that divided the bay into two zones as revealed by the Chi-
the minimum value from each pixel value in a corresponding square test: One in the northwestern area of the bay (zone1)
band (Chavez, 1989). This approach is very practical with the characterized by spatial dependency of the two species; and the
absence of atmospheric measurements and can provide accurate other in the southeastern area of the bay (zone 2) where the
results comparable to those provided by complicated atmospheric observed dinoflagellates density was higher than what was ex-
algorithms (Jensen, 2015). pected (Fig. 4). This can be explained by the hydrodynamic factors
Three types of pixels were excluded from the analysis during in Algiers Bay, such as drift current and swell that disperses
the image preprocessing stage: mixed pixels (inland-water pix- the seawater masses to the southeastern zone. Also, effluents
els), shallow-water pixels and incorrect pixels induced by sensors of the wastewater discharging from Wadi El Harrach (near to
 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311 5

Table 1 in order to evaluate the performance, whereas the spatial and

Descriptive statistics of the in-situ phytoplankton. temporal distribution of dinoflagellates was mapped using a spa-
Variables (Cells .l-1 ) Min Max Median Mean tial interpolation technique due to the dinoflagellate insignificant
2017
Diatoms 1748 11 816 5008 5084 relationship with OLI bands. The temporal distribution of both
Dinoflagellates 156 45 772 10 010 15 589 species showed that cell density in spring 2018 was higher than
Diatoms 125 4300 1374 1502 what it was in spring 2017. Also, spatial distributions of diatoms
2018
Dinoflagellates 90 126 010 3461 21 655
and dinoflagellates seemed to be influenced by various factors.
Diatoms 125 11 816 2521 3173 Mapping diatom cell density in spring 2017 and 2018 revealed
Total
Dinoflagellates 90 126 010 6651 18 824
a general spatial pattern for diatoms; nearshore waters from the
northeastern area to the south of bay were characterized with
Table 2 high cell density of diatoms (Fig. 6). Such elevated density in
Linear regression models for Operational Land Imager reflectance values where these waters can be attributed to high inland nutrient inputs
log-transformed diatom density is the dependent variable for all stations. B in
and hydrodynamic systems that provide favorable conditions for
the equation column refers to Operational Land Imager reflectance bands.
phytoplankton species to grow. Also, waters near Algiers Port
Phytoplankton Equation RMSE MAPD r2 n
were characterized with a relatively high cell density possibly in-
1.1953 (logB7/LogB2) + 1.4434 0.28 2.31 0.62
duced by anthropogenic activities. The diatom abundance clearly
1.0965 (logB6/LogB2) + 1.6524 0.29 2.31 0.59
0.6986∗log (B2/B6) + 2.7891 0.30 2.31 0.57 declined in offshore waters, especially at the middle of bay. This
Log (diatom) 0.8107∗log (B3/B7) + 2.6906 0.30 2.31 0.57 30 can be explained by two factors: longshore current forces that
0.7537∗log (B3/B6) + 2.7855 0.31 2.31 0.55 disperse nutrients and suspended solids, which enhances phyto-
1.0227∗log (B4/B6) + 2.8796 0.31 2.31 0,53 plankton growth, along the coast rather than offshore waters; and
1.0834∗log (B4/B7) + 2.7806 0.31 2.31 0.53
the waves induced by the northwestern wind that oriented the
flow of phytoplankton, nutrients and suspended sediments to the
southeastern part of the bay.
station 16) and Wadi El Hamiz as well as pollutants from El The spatial distribution of dinoflagellates seemed to be con-
Marsa harbor and management projects (Chabane et al., 2018), trolled by hydrodynamic systems and anthropogenic activities;
contributed in increasing nutrient concentrations and suspended in general, these species were abundant in nearshore waters. An-
solids providing a favorable environment for harmful algae to other noteworthy spatial pattern was the inverse relationship be-
grow in the southeastern zone (Kouhanestani et al., 2019). tween cell density of dinoflagellates and diatoms at S13 and S16.
This might indicate the presence of heterotrophic dinoflagellates
4.1. Modeling phytoplankton densities at these sites that graze on diatoms.
During the second voyage (on 5 May 2018), a massive algae
Bands and band ratioing products of atmospherically corrected bloom was observed with red-orange color close to S4 and S5
OLI images were correlated to in-situ cell densities of diatoms, stations coming from the northwest. Six days later, the algae
dinoflagellates and total phytoplankton (i.e., diatom and dinoflag- bloom expanded and reached nearshore waters as illustrated by
ellate cell densities) (n = 30). The correlation between OLI bands, the Sentinel-2A image obtained on 15 May 2018 (Fig. 7). The
when analyzed independently, and in situ datasets was insignif- samples taken from this bloom contained Noctiluca scintillans, one
icant (r2 < 0.24) and thus disregarded for the sake of brevity. of dinoflagellate species. This strongly indicates that the species
Several OLI band ratioing products showed a significant rela- was subject to a considerable proliferation in Algiers Bay. The
tionship (Fig. 5) with diatom cell density (Table 2), whereas proliferation of Noctiluca scintillans can be either autotrophically
dinoflagellate and the total phytoplankton densities had insignif- or heterotrophically by N. scintillans grazing on phytoplankton
icant relationship with OLI ratioing products (Tables 3 and 4). such as diatoms (Sweeney, 1976; Wang et al., 2016). Thus, the
These band ratioing products were used to model diatom cell
emergence of heterotrophic planktons is often accompanied by a
density. The band ratioing of logB7/ logB2 exhibited the most
decline in diatoms (Wang et al., 2016), as observed in S13 and
significant relationship with diatom cell density (r2 = 0.62, RMSE
S16.
= 0.28, MAPD = 2.31%).
The algae blooms coming from open sea revealed that the
The cell structure of diatoms and dinoflagellates might explain
water quality and ecosystems of Algiers Bay are significantly
the relationships of OLI bands with the cell density of these
interconnected with regional factors that cannot be comprehen-
species. Diatoms are autotrophs, and thus their density is highly
sively detected without using remotely sensed imagery. Cities and
associated with their pigments, especially chlorophyll-a (Godhe
ports along Mediterranean Sea coast can be a source of various
et al., 2008) that significantly contributes in changing AOP of
pollutants that freely move toward many places including Algiers
water (Gordon et al., 1988). This might explain the significant
relationship between OLI bands and diatom cell density. Dinoflag- Bay. Thus, a remote sensing based-alert system is mandatory to
ellates species act differently. They can be either autotrophs or sustain and maintain Algiers Bay ecosystems.
heterotrophs that lack chloroplasts (Leakey et al., 1996) meaning
that the presence of dinoflagellates, does not necessarily correlate 5. Conclusion
with chlorophyll-a concentrations. Heterotrophic dinoflagellates,
about the half of dinoflagellate species, change AOP of seawater Dinoflagellates were more abundant than diatoms in Algiers
in a different matter than what autotrophic dinoflagellates do. An Bay. The two species exhibited spatial and temporal variabilities.
increasing number of heterotrophic dinoflagellates changes the Nearshore waters were enriched with diatoms and dinoflagel-
color of seawater to brown causing red tide events (Jeong et al., lates due to hydrodynamic systems and anthropogenic activities.
1999; Park et al., 2013; Phlips et al., 2015). Also, the species were more abundant in spring 2018 than what
they were in 2017. During May 2018, the Algiers Bay experi-
4.2. Spatial and temporal distributions of diatoms and dinoflagel- enced an algae bloom coming from the open sea that influenced
lates nearshore waters: This event provided a clear case of how the
bay is interconnected with Mediterranean Sea ecosystems. Such
The spatial and temporal distribution of diatoms was mapped interconnection between the bay and Mediterranean Sea requires
using the most accurate Landsat-8 model (i.e., logB7/ logB2) and a synoptic monitoring using remote sensing imagery.
6 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311

Fig. 3. In-situ surface distributions of diatom and dinoflagellate cell densities.

Fig. 4. The distribution of chi-square for (log10 ) diatoms and dinoflagellates cell densities 2017 and 2018.

Table 3
Linear regression models between Operational Land Imager reflectance values and for all
phytoplankton cell densities in all stations.
Variable Equation RMSE MAPD r2 n
0.2592(LogB6/LogB3) + 0.4355 3.25 3.81 0.23
0.1345∗(LogB6/LogB4) + 0.7075 3.19 3.80 0.11
Log (phytoplankton) 0.2185∗(LogB7/LogB2) + 0.6903 3.04 3.76 0.17 30
0.3194∗Log(B2/B6) − 0.5264 4.34 4.09 0.13
0.2186∗Log(B4/B7) − 0.3825 4.33 4.09 0.17

In this study, we demonstrated the efficiency of using Landsat- their pigment concentrations, especially chlorophyll-a that signif-
8 images to model phytoplankton cell density. Landsat-8 OLI icantly contributes in changing AOP of water.
bands successfully captured the diatom cell density variations, The methodology used in this study to model phytoplankton
whereas dinoflagellate cell density exhibited insignificant rela-
tionship with OLI bands. The reason for this insignificant re- density can be applied in future investigations to include more
lationship was probably because almost half of dinoflagellate in-situ data. The proposed model can be improved by using more
species are heterotrophs whose abundance does not covary with in-situ observations well distributed over the year. Including a
 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311 7

Fig. 5. The relationship between diatoms and the band ratio (REF logB7/REF logB2).

Fig. 6. Predicted distributions of surface diatoms using the most accurate Landsat model: (A in Apr 04, 2017) and (B in May 09, 2018).

Table 4
Linear regression models between Operational Land Imager reflectance values and for dinoflagellates cell densities
in all stations.
Variable Equation RMSE MAPD r2 n
−1,1845(R5/R3) + 4,2606 0,726 2,700 0,189
Log (dinoflagellates) −2,4036(B3/B4) + 5,8196 0,728 2,701 0,185
0,19*log(B4/B6) + 3,6666 0,804 2,689 0,005
101,09*(B1/B3) + 32,22 80,821 105,339 0,100
Root (dinoflagellates) −82,494*(R5/R3) + 143,15 81,764 105,391 0,080 30
16,985Log(R4/R6) + 100,17 85,132 105,242 0,004
−977,43*(LogB5/LogB7) + 952,2 53,209 72,993 0,169
Root (dinoflagellates) 31,425*Log (B2/B4) + 62,55 57,466 72,827 0,032
4,3577*(B2/B4) + 60,541 57,478 72,826 0,031
8 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311

Fig. 7. (A and B): phytoplankton bloom in the surface waters of Algiers Bay (09/05/2018), (C): Noctiluca scintillans algal blooms (small picture) and its color (large
picture) in the Algiers bay coastal water (May 2018) and (D): hydrodynamic of plankton bloom and suspended matter to the southeastern part of the Bay observed
by Sentinel-A2 in 15/5/2018 (R = Band4, G = Band3, B = Band2).

large number of in-situ observations well distributed over dif- Ahn, Y.-H., Shanmugam, P., 2006. Detecting the red tide algal blooms from
ferent months would allow for the development of more reliable satellite ocean color observations in optically complex Northeast-Asia Coastal
waters. Remote Sens. Environ. 103, 419–437. http://dx.doi.org/10.1016/j.rse.
models to estimate phytoplankton cell density during any time of
2006.04.007.
the year. Such models would be a key factor for successful marine Al-Shehhi, M.R., Gherboudj, I., Ghedira, H., 2012. A study on the effect of
and coastal monitoring programs. dust and wind on phytoplankton activities in the Arabian Gulf. In: 2012
IEEE International Geoscience and Remote Sensing Symposium. IEEE, pp.
CRediT authorship contribution statement 2571–2574. http://dx.doi.org/10.1109/IGARSS.2012.6350954.
Bailey, S.W., Werdell, P.J., 2006. A multi-sensor approach for the on-orbit
validation of ocean color satellite data products. Remote Sens. Environ. 102,
R.L. Boufeniza: Conceptualization, Software, Validation, For- 12–23. http://dx.doi.org/10.1016/j.rse.2006.01.015.
mal analysis, Writing - original draft. M.M. Alsahli: conceptual- Barton, A.D., Finkel, Z.V., Ward, B.A., Johns, D.G., Follows, M.J., 2013. On the roles
ization, Methodology, Writing - review & editing. N.I. Bachari: In- of cell size and trophic strategy in North Atlantic diatom and dinoflagellate
vestigation, Resources, Data curation. F. Houma Bachari: Project communities. Limnol. Oceanogr. 58, 254–266. http://dx.doi.org/10.4319/lo.
2013.58.1.0254.
administration,Supervision, Formal analysis. Bonansea, M., Ledesma, M., Rodriguez, C., Pinotti, L., 2018. Using new remote
sensing satellites for assessing water quality in a reservoir. Hydrol. Sci. J.
Declaration of competing interest http://dx.doi.org/10.1080/02626667.2018.1552001.
Bresciani, M., Cazzaniga, I., Austoni, M., Sforzi, T., Buzzi, F., Morabito, G.,
Giardino, C., 2018. Mapping phytoplankton blooms in deep subalpine lakes
The authors declare that they have no known competing finan-
from Sentinel-2A and Landsat-8. Hydrobiologia 824, 197–214. http://dx.doi.
cial interests or personal relationships that could have appeared org/10.1007/s10750-017-3462-2.
to influence the work reported in this paper. Butt, M.J., Nazeer, M., 2015. Landsat ETM+ Secchi Disc Transparency (SDT)
retrievals for Rawal Lake. Pak. Adv. Sp. Res. 56, 1428–1440. http://dx.doi.
Acknowledgments org/10.1016/j.asr.2015.06.041.
Chabane, K., Bahbeh, L., Seridi, H., 2018. Ecological Quality Status of the Algiers
coastal waters by using macroalgae assemblages as bioindicators (Algeria,
The authors would like to express their gratitude to Mediterranean Sea). Mediterr. Mar. Sci. 305. http://dx.doi.org/10.12681/mms.
Dr. Amarouche Khalid for his assistance in obtaining in-situ data 15951.
sampling. Also, the authors would like to thank the entire team at Chavez, Jr., P.S., 1989. Radiometric calibration of Landsat Thematic Mapper
the Space Oceanography Laboratory for the advice and support. multispectral images. Photogramm. Eng. Remote Sens. 55, 1285–1294.
Cushing, D.H., 1989. A difference in structure between ecosystems in strongly
stratified waters and in those that are only weakly stratified. J. Plankton Res.
References 11, 1–13. http://dx.doi.org/10.1093/plankt/11.1.1.
Dörnhöfer, K., Göritz, A., Gege, P., Pflug, B., Oppelt, N., 2016. Water con-
AFNOR, 2006. Qualité de l’eau-Norme guide pour le dénombrement du stituents and water depth retrieval from Sentinel-2A—A first evaluation in an
phytoplancton par microscopie inversée (méthode Utermöhl). oligotrophic lake. Remote Sens. 8, 941. http://dx.doi.org/10.3390/rs8110941.
 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311 9

Downing, J.A., Rigler, F.H., 1984. A manual on methods for the assessment of Lodhi, M.A., 2002. Influence of bright bottom depth on reflectance from clear
secondary productivity. Fresh Waters 2. and turbid water bodies. Geocarto Int. 17, 7–16. http://dx.doi.org/10.1080/
Druart, J.-C., Rimet, F., 2008. Protocols for analysis of phytoplankton INRA: 10106040208542239.
sampling, counting and biovolumes. INRA-Thon. Rapp. SHL 283, 96. Lund, J.W.G., Kipling, C., Le Cren, E.D., 1958. The inverted microscope method of
Ghanea, M., Moradi, M., Kabiri, K., 2016. A novel method for characterizing estimating algal numbers and the statistical basis of estimations by counting.
harmful algal blooms in the Persian Gulf using MODIS measurements. Adv. Hydrobiologia 11, 143–170. http://dx.doi.org/10.1007/BF00007865.
Sp. Res. 58, 1348–1361. http://dx.doi.org/10.1016/j.asr.2016.06.005. Mcmanus, G.B., Katz, L.A., 2009. Molecular and morphological methods for
Giardino, C., Pepe, M., Brivio, P.A., Ghezzi, P., Zilioli, E., 2001. Detecting chloro- identifying plankton, : what makes a successful marriage? J. Plankton Res.
phyll, Secchi disk depth and surface temperature in a sub-alpine lake using 31, 1119–1129. http://dx.doi.org/10.1093/plankt/fbp061.
Landsat imagery. Sci. Total Environ. 268, 19–29. http://dx.doi.org/10.1016/ Morán, X.A.G., Taupier-Letage, I., Vázquez-Domínguez, E., Ruiz, S., Arin, L., Raim-
S0048-9697(00)00692-6. bault, P., Estrada, M., 2001. Physical-biological coupling in the Algerian basin
Gobler, C.J., Doherty, O.M., Hattenrath-Lehmann, T.K., Griffith, A.W., Kang, Y., (SW Mediterranean): Influence of mesoscale instabilities on the biomass and
Litaker, R.W., 2017. Ocean warming since 1982 has expanded the niche of production of phytoplankton and bacterioplankton. Deep Res. I Oceanogr.
toxic algal blooms in the North Atlantic and North Pacific oceans. Proc. Natl. Res. Pap. 48, 405–437. http://dx.doi.org/10.1016/S0967-0637(00)00042-X.
Acad. Sci. 114, 4975–4980. http://dx.doi.org/10.1073/pnas.1619575114. Mouw, C.B., Greb, S., Aurin, D., DiGiacomo, P.M., Lee, Z., Twardowski, M.,
Godhe, A., Asplund, M.E., Harnstrom, K., Saravanan, V., Tyagi, A., Karunasagar, I., Binding, C., Hu, C., Ma, R., Moore, T., Moses, W., Craig, S.E., 2015. Aquatic
2008. Quantification of diatom and dinoflagellate biomasses in coastal color radiometry remote sensing of coastal and inland waters: Challenges
marine seawater samples by real-time PCR. Appl. Environ. Microbiol. 74, and recommendations for future satellite missions. Remote Sens. Environ.
7174–7182. http://dx.doi.org/10.1128/AEM.01298-08. 160, 15–30. http://dx.doi.org/10.1016/j.rse.2015.02.001.
Gordon, H., Brown, O.B., Evans, R.H., Brown, J.W., Smith, R.C., Baker, K.S., Msi, S., Duforêt-gaurier, L., Doan, C., Mangin, A., 2019. Remote Sensing of
Clark, D.K., 1988. A semianalytic radiance model of ocean color. J. Geophys. Environment Coastal and inland water pixels extraction algorithm (WiPE)
Res. 93, 10909. http://dx.doi.org/10.1029/JD093iD09p10909. from spectral shape analysis and HSV transformation applied to Landsat 8
Hallegraeff, G.M., Bolch, C.J., 1992. Transport of diatom and dinoflagellate resting OLI and. Remote Sens. Environ. 223, 208–228. http://dx.doi.org/10.1016/j.rse.
spores in ships ’ Ballast Water, : Implications for plankton biogeography.. J. 2019.01.024.
Plankton Res. 14, 1067–1084. http://dx.doi.org/10.1093/plankt/14.8.1067. Nazeer, M., Nichol, J.E., Yung, Y.-K., 2014. Evaluation of atmospheric correc-
Harvey, E.L., Menden-Deuer, S., 2012. Predator-induced fleeing behaviors in tion models and Landsat surface reflectance product in an urban coastal
phytoplankton: A new mechanism for harmful algal bloom formation? PLoS environment. Int. J. Remote Sens. 35, 6271–6291. http://dx.doi.org/10.1080/
One 7, e46438. http://dx.doi.org/10.1371/journal.pone.0046438. 01431161.2014.951742.
Nazeer, M., Wong, M.S., Nichol, J.E., 2017. Science of the Total Environment A
Hasle, G.R., 1978. The inverted microscope method. Phytoplankt. Man 19, 1–196.
new approach for the estimation of phytoplankton cell counts associated
Havens, K., East, T., Beaver, J., 1996. Experimental studies of zooplankton–
with algal blooms. Sci. Total Environ. http://dx.doi.org/10.1016/j.scitotenv.
phytoplankton–nutrient interactions in a large subtropical lake (Lake
2017.02.182.
Okeechobee, Florida, U.S.A.). Freshw. Biol. 36, 579–597. http://dx.doi.org/10.
Norjamaki, I., Tokola, T., 2007. Comparison of atmospheric correction methods
1046/j.1365-2427.1996.00122.x.
in mapping timber volume with multitemporal Landsat images in Kainuu,
Hoepffner, N., Sathyendranath, S., 1993. Determination of the major groups of
Finland. Photogramm. Eng. Remote Sens. 73, 155–163. http://dx.doi.org/10.
phytoplankton pigments from the absorption spectra of total particulate
14358/PERS.73.2.155.
matter. J. Geophys. Res. 98, 22789. http://dx.doi.org/10.1029/93JC01273.
Ogashawara, I., De Alcântara, E.H., Stech, J.L., Tundisi, J.G., 2014. Cyanobacteria
Hötzel, G., Croome, R., 1999. A Phytoplankton Methods Manual for Aus-
detection in Guarapiranga Reservoir (São Paulo State, Brazil) using Landsat
tralian Freshwaters. Land and Water Resources Research and Development
TM and ETM+ images. Ambient. e Agua - An Interdiscip. J. Appl. Sci. 9,
Corporation.
224–238. http://dx.doi.org/10.4136/ambi-agua.1327.
Houma, F., Belkessa, R., Bachari, N., 2006. Contribution of multispectral satellite
Olmanson, Leif G., Brezonik, Patrick L., Finlay, Jacques C., Bauer Marvin, E., et
imagery to the bathymetric analysis of coastal sea bottom. Rev. Desénergies
al., 2016. Comparison of landsat 8 and landsat 7 for regional measurements
Renouv. 9, 165–172.
of cdom and water clarity in lakes. Remote Sens. Environ. 185, 119–128.
Illoul, H., Maso, M., Fortuño Alós, J.M., Cros, L., Morales-Blake, A., Seridji, R., http://dx.doi.org/10.1016/j.rse.2016.01.007.
2008. Potentially harmful microalgae in coastal waters of the Algiers area Pahlevan, N., Schott, J.R., 2013. Leveraging EO-1 to evaluate capability of new
(Southern Mediterranean Sea). Algologie 29, 261–278. http://dx.doi.org/10. generation of Landsat sensors for coastal/inland water studies. IEEE J. Sel.
10261/21115. Top. Appl. Earth Obs. Remote Sens. 6, 360–374.
Jensen, J.R., 2015. Introductory Digital Image Processing: A Remote Sensing Palmer, C.M., Maloney, T.E., 1954. A new counting chamber slide for
Perspective. Prentice Hall Press. nannoplankton. Am. Soc. Limnol. Ocean. Sp. Publ. 2.
Jeong, H.J., Shim, J.H., Lee, C.W., Kim, J.S., Koh, S.M., 1999. Growth and grazing Park, J., Jeong, H.J., Kwon, J.E., Kang, H.C., Kim, J.H., Jang, S.H., Park, J.Y.,
rates of the marine planktonic ciliate strombidinopsis sp. on red-tide and Yoon, E.Y., Kim, J.S., 2013. Mixotrophic dinoflagellate red tides in Korean
toxic dinoflagellates. J. Eukaryot. Microbiol. 46, 69–76. http://dx.doi.org/10. waters: Distribution and ecophysiology. Harmful Algae 30, S28–S40. http:
1111/j.1550-7408.1999.tb04586.x. //dx.doi.org/10.1016/j.hal.2013.10.004.
Johnson, M.D., Stoecker, D.K., Marshall, H.G., 2013. Seasonal dynamics of meso- Phlips, E.J., Badylak, S., Lasi, M.A., Chamberlain, R., Green, W.C., Hall, L.M.,
dinium rubrum in Chesapeake Bay. J. Plankton Res. 35, 877–893. http: Hart, J.A., Lockwood, J.C., Miller, J.D., Morris, L.J., Steward, J.S., 2015. From red
//dx.doi.org/10.1093/plankt/fbt028. tides to green and brown tides: Bloom dynamics in a restricted subtropical
Kouhanestani, Z.M., Roelke, D.L., Ghorbani, R., Fujiwara, M., 2019. Assessment lagoon under shifting climatic conditions. Estuar. Coast. 38, 886–904. http:
of spatiotemporal phytoplankton composition in relation to environmental //dx.doi.org/10.1007/s12237-014-9874-6.
conditions of gorgan bay, Iran. Estuar. Coast. 42, 173–189. http://dx.doi.org/ Raimbault, P., Coste, B., Boulhadid, M., Boudjellal, B., 1993. Origin of high
10.1007/s12237-018-0451-2. phytoplankton concentration in deep chlorophyll maximum (DCM) in a
Laplace-Treyture, C., Barbe, J., Dutartre, A., Druart, J.C., Rimet, F., An- frontal region of the Southwestern Mediterranean sea (algerian current).
neville, O., 2009. Protocole Standardisé D’échantillonnage, de Conservation, Deep Sea Res. I Oceanogr. Res. Pap. 40, 791–804. http://dx.doi.org/10.1016/
D’observation Et de Dénombrement Du Phytoplancton En Plan D’eau Pour 0967-0637(93)90072-B.
la Mise Enøeuvre de la DCE. CEMAGREF, Groupement de Bordeaux, Unité de Richardson, A.J., 2004. Climate impact on plankton ecosystems in the Northeast
Recherche Réseaux, Epuration et Qualité des Eaux. Rapport Version 3.3.1. Atlantic. Science (80-.) 305, 1609–1612. http://dx.doi.org/10.1126/science.
Leakey, R., Archer, S., Grey, J., 1996. Microbial dynamics in coastal waters of 1100958.
east antarctica:bacterial production and nanoflagellate bacterivory. Mar. Ecol. Richlen, M.L., Morton, S.L., Jamali, E.A., Rajan, A., Anderson, D.M., 2010. The
Prog. Ser. 142, 3–17. http://dx.doi.org/10.3354/meps142003. catastrophic 2008–2009 red tide in the Arabian gulf region, with observa-
Lee, K.H., Jeong, H.J., Kim, H.J., Lim, A.S., 2017. Nitrate uptake of the red tide tions on the identification and phylogeny of the fish-killing dinoflagellate
dinoflagellate prorocentrum micans measured using a nutrient repletion Cochlodinium polykrikoides. Harmful Algae 9, 163–172. http://dx.doi.org/10.
method: effect of light intensity. ALGAE 32, 139–153. http://dx.doi.org/10. 1016/j.hal.2009.08.013.
4490/algae.2017.32.5.20. Roesler, C.S., Perry, M.J., Carder, K.L., 1989. Modeling in situ phytoplankton
Lee, K.H., Jeong, H.J., Kwon, J.E., Kang, H.C., Kim, J.H., Jang, S.H., Park, J.Y., absorption from total absorption spectra in productive inland marine wa-
Yoon, E.Y., Kim, J.S., 2016. Mixotrophic ability of the phototrophic dinoflag- ters. Limnol. Oceanogr. 34, 1510–1523. http://dx.doi.org/10.4319/lo.1989.34.
ellates Alexandrium andersonii, A. affine, and A. fraterculus. Harmful Algae 8.1510.
59, 67–81. http://dx.doi.org/10.1016/j.hal.2016.09.008. Roy, D.P., Wulder, M.A., Loveland, T.R.C.E.W., Allen, R.G., Anderson, M.C.,
Leterme, S.C., Seuront, L., Edwards, M., 2006. Differential contribution of diatoms Helder, D., Irons, J.R., Johnson, D.M., Kennedy, R., Scambos, T.A., Schaaf, C.B.,
and dinoflagellates to phytoplankton biomass in the NE Atlantic Ocean and Schott, J.R., Sheng, Y., Vermote, E.F., Belward, A.S., Bindschadler, R., Co-
the North Sea. Mar. Ecol. Prog. Ser. 312, 57–65. http://dx.doi.org/10.3354/ hen, W.B., Gao, F., Hipple, J.D., Hostert, P., Huntington, J., Justice, C.O., Kilic, A.,
meps312057. Kovalskyy, V., Lee, Z.P., Lymburner, L., Masek, J.G., McCorkel, J., Shuai, Y.,
10 R.L. Boufeniza, M.M. Alsahli, N.I. Bachari et al. / Regional Studies in Marine Science 36 (2020) 101311

Trezza, R., Vogelmann, J., Wynne, R.H., Zhu, Z., 2014. Landsat-8: Science and Ulloa, O., Sathyendranath, S., Platt, T., 1994. Effect of the particle-size distribution
product vision for terrestrial global change research. Remote Sens. Environ. on the backscattering ratio in seawater. Appl. Opt. 33, 7070. http://dx.doi.
145, 154–172. http://dx.doi.org/10.1016/j.rse.2014.02.001. org/10.1364/AO.33.007070.
Seridi, H., Ruitton, S., Boudouresque, C.F., 2007. Is it possible to calibrate the Utermöhl, H., 1958. Zur Vervollkommnung der quantitativen phytoplankton-
pollution level of the region of Algiers (Mediterranean Sea) by exploiting methodik. SIL Commun. 1953-1996 9, 1–38. http://dx.doi.org/10.1080/
marine macrophytes? C. R. Biol. 330, 606–614. http://dx.doi.org/10.1016/j. 05384680.1958.11904091.
crvi.2007.05.001. Vanhellemont, Q., Ruddick, K., 2014. Turbid wakes associated with offshore wind
Shin, J., Kim, K., Son, Y.B., Ryu, J., 2018. Synergistic Effect of Multi-Sensor data turbines observed with Landsat 8. Remote Sens. Environ. 145, 105–115.
on the detection of Margalefidinium polykrikoides in the South Sea of Korea. http://dx.doi.org/10.1016/j.rse.2014.01.009.
http://dx.doi.org/10.3390/rs11010036. Vincent, R.K., Qin, X., McKay, R.M.L., Miner, J., Czajkowski, K., Savino, J., Bridge-
Sweeney, M., 1976. Pedinomonas Noctilucae (Prasinophyceae), the flagellate man, T., 2004. Phycocyanin detection from LANDSAT TM data for mapping
symbiotic in Noctiluca (dinophyceae) in southeast Asia 1. J. Phycol. 460–464. cyanobacterial blooms in Lake Erie. Remote Sens. Environ. 89, 381–392.
http://dx.doi.org/10.1111/j.1529-8817.1976.tb02874.x. http://dx.doi.org/10.1016/j.rse.2003.10.014.
Taupier-Letage, I., Puillat, I., Millot, C., Raimbault, P., 2003. Biological response Wang, L., Lin, X., Goes, J.I., Lin, S., 2016. Phylogenetic Analyses of Three
to mesoscale eddies in the Algerian Basin. J. Geophys. Res. 108, 3245. Genes of Pedinomonas noctilucae, the Green Endosymbiont of the Ma-
http://dx.doi.org/10.1029/1999JC000117. rine Dinoflagellate Noctiluca scintillans, Marsupiomonadales (Chlorophyta,
Tepanosayn, G., Muradyan, V., Hovsepyan, A., Minasyan, L., Asmaryan, S., 2017. A Pedinophyceae) under the Reinstated Name Protoeuglena noctilucae. Ann.
Landsat 8 OLI satellite data-based assessment of spatio-temporal variations Anat. 167, 205–216. http://dx.doi.org/10.1016/j.protis.2016.02.005.
of lake sevan phytoplankton biomass. Ann. Valahia Univ. Targ. Geogr. Ser. Wei, J., Lee, Z., Shang, S., Yu, X., 2019. Semi-analytical derivation of phyto-
17, 83–89. http://dx.doi.org/10.1515/avutgs-2017-0008. plankton, CDOM and detritus absorption coefficients from the Landsat 8/OLI
Throndsen, J., 1995. Estimating cell numbers. Man. Harmful Mar. Microalgae 6, reflectance in coastal waters. J. Geophys. Res. Ocean. http://dx.doi.org/10.
3–80. 1029/2019JC015125.
Tillmann, U., Krock, B., Alpermann, T.J., Cembella, A., 2016. Bioactive compounds Woodcock, C.E., Allen, R., Anderson, M., Belward, A., Bindschadler, R., Cohen, W.,
of marine dinoflagellate isolates from western Greenland and their phyloge- Gao, F., Goward, S.N., Helder, D., Helmer, E., Nemani, R., Oreopoulos, L.,
netic association within the genus Alexandrium. Harmful Algae 51, 67–80. Schott, J., Thenkabail, P.S., Vermote, E.F., Vogelmann, J., Wulder, M.A.,
http://dx.doi.org/10.1016/j.hal.2015.11.004. Wynne, R., 2008. Free access to Landsat imagery. Science (80-.) 320, 1011a.
Torrecilla, R.E., Piera, F.J., Stramski, D., Reynolds, R., Millán-Nuñez, E., 2009. http://dx.doi.org/10.1126/science.320.5879.1011a.
Identification of phytoplankton pigment assemblages using derivative spec- Yadav, S., Yamashiki, Y., Susaki, J., Yamashita, Y., Ishikawa, K., 2019. Chlorophyll
troscopy of hyperspectral remote-sensing reflectances. Program 3, 80, estimation of lake water and coastal water using Landsat-8 and Sentinel-2a
https://doi.org/2099/8925. satellite. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci. XLII 1, 2–14.
Tundisi, J.G., Matsumura-Tundisi, T., Tundisi, J.E.M., 2008. Reservoirs and human http://dx.doi.org/10.5194/isprs-archives-XLII-3-W7-77-2019.
well being: new challenges for evaluating impacts and benefits in the Zhai, L., Platt, T., Tang, C., Sathyendranath, S., Walne, A., 2013. The response
neotropics. Braz. J. Biol. 68, 1133–1135. http://dx.doi.org/10.1590/S1519- of phytoplankton to climate variability associated with the North Atlantic
69842008000500020. Oscillation. Deep Sea Res. II Top. Stud. Oceanogr. 93, 159–168. http://dx.doi.
org/10.1016/j.dsr2.2013.04.009.