Ecological Indicators 117 (2020) 106702

Contents lists available at ScienceDirect

Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind

Cyanobacteria dynamics and phytoplankton species richness as a measure of T

waterbody recovery: Response to phosphorus removal treatment in a
tropical eutrophic reservoir
⁎
Bruna Barçantea, Nilo O. Nascimentob, Talita F.G. Silvab, Leandro A. Reisa, Alessandra Giania,
a
Department of Botany, Universidade Federal de Minas Gerais, 6627 av. Antônio Carlos, CEP 31270-901 Belo Horizonte, Minas Gerais, Brazil
b
Department of Hydraulic Engineering and Water Resources, Universidade Federal de Minas Gerais, 6627 av. Antônio Carlos, CEP 31270-901 Belo Horizonte, Minas
Gerais, Brazil

A R T I C LE I N FO A B S T R A C T

Keywords: External nutrient inputs account for the main cause of eutrophication of aquatic systems and intensification of
Phoslock® toxic cyanobacteria blooms. But internal loads from sediments significantly contribute to environmental de-
Biodiversity gradation. In-lake treatments may therefore be combined to accelerate the restoration process. A tropical, eu-
Shallow urban lakes trophic reservoir with persistent cyanobacteria blooms was subjected to applications of a lanthanum-enriched
Pampulha reservoir
clay (Phoslock®) designed to control internal phosphorus loading. This study aimed to evaluate if the treatment
Water management
could lead to cyanobacteria reduction and improvement of the phytoplankton community and environmental
status. Data were obtained during the application and post-application months, and compared with pre-treat-
ment datasets. Significant decrease in phosphorus concentration occurred, but only in the second treatment year.
Cyanobacteria responded after five months of applications, reducing biomass with a 4-fold increase in water
transparency, but blooms were able to re-establish. However, by the end of the second year and until the in-
terruption of the treatment, cyanobacteria dominance was significantly reduced and replaced by a more diverse
phytoplankton community rarely seen in this environment. At this time, the number of species belonging to
groups other than cyanobacteria increased, characterizing more than 80% of the total of species. While pointing
to the system resilience and the treatment limitations, this study also indicated that under a long-term control of
phosphorus load, lakes could be restored and their aquatic communities recover. Decrease of cyanobacteria
biomass and increase in non-cyanobacteria biodiversity were key factors to evaluate the treatment success. The
study reinforces the importance of species and communities as sentinels of a healthier ecosystem.

1. Introduction bloom development and represents the main challenge in the restora-
tion of eutrophic systems (Carpenter, 2008; Schindler et al., 2008;
Nutrient input in aquatic ecosystems, mainly phosphorus (P) and Schindler et al., 2016), leading to a consensus that the control of ex-
nitrogen (N), is among the primary causes of eutrophication of urban ternal P inputs is essential to reverse eutrophication (Sas, 1989; Cooke
reservoirs worldwide, resulting in water quality deterioration and in- et al., 2005). In many cases, however, internal loads from P-rich sedi-
crease in phytoplankton biomass, favoring cyanobacteria blooms ments are very high and can delay the recovery of the aquatic system
(Smith, 2003; Smith and Schindler, 2009; Schindler et al., 2016). In (Søndergaard et al., 2003; Spears et al., 2012). Specific products to limit
addition to the problems associated with their high biomass (odor, P release from sediments have been tested and applied in restoration
anoxic deep layers), cyanobacteria can produce potent toxins that pose programs (Meis et al., 2013), in order to quickly reverse the ecological
a serious risk to aquatic life and the environment, as well to human damages caused by eutrophication (Copetti et al., 2016).
health (Chorus et al., 2000; Smith, 2003; Paerl and Otten, 2013). The lanthanum-modified bentonite clay, commercially known as
Therefore, eutrophication threatens ecosystem freshwater services, Phoslock®, is a geo-engineered material developed in the 1990s at the
such as fish production, drinking water supply and recreation Commonwealth Scientific and Industrial Research Organization
(Brönmark and Hansson, 2002; Carvalho et al., 2013). (CSIRO) in Australia as a tool to remove dissolved phosphorus from the
Phosphorus has been considered a critical factor in cyanobacteria water column of eutrophic lakes (Douglas et al., 1999; Douglas, 2002).

⁎
Corresponding author.
E-mail address: agiani@ufmg.br (A. Giani).

https://doi.org/10.1016/j.ecolind.2020.106702
Received 9 April 2020; Received in revised form 23 May 2020; Accepted 6 July 2020
1470-160X/ © 2020 Elsevier Ltd. All rights reserved.
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

The mechanism of action of Phoslock® is based on the binding of the 2. Material and methods
lanthanum ion of the bentonite clay with phosphate ion from water,
forming the mineral rhabdophane (LaPO4). The rhabdophane is then 2.1. Study site
precipitated from the water column and immobilized on the lake se-
diment. The ability of Phoslock® to adsorb phosphate has been well Pampulha reservoir is a hypereutrophic reservoir located in the city
studied in laboratory experiments (Ross et al., 2008; Haghseresht et al., of Belo Horizonte (MG) (19°51′09″S, 43°58′42″W), Brazil. It has a
2009) and has already been applied in some lakes around the world, surface area of about 1.8 km2, volume of 9.9x106 m3 and mean and
most of them, however, located in temperate regions (Lürling and van maximum depth of about 5.1 m and 16.2 m, respectively (Figueredo
Oosterhout, 2013; Bishop et al., 2014; Nürnberg and LaZerte, 2016; et al., 2016, Silva et al., 2016). The reservoir has eight tributaries:
Spears et al., 2016; Epe et al., 2017; Nürnberg, 2017). In contrast, Ressaca and Sarandi streams are the two major contributors to external
monitoring of in situ applications during cyanobacteria blooms have nutrient input into reservoir. Together, they are responsible for 70% of
seldom been carried out (Bishop et al., 2014; Lang et al., 2016; Bishop the total water inflow of the reservoir (Torres et al., 2007). The climate
and Richardson, 2018) and the response of the entire phytoplankton of the region is characterized by two distinct seasons: a warm and rainy
community to Phoslock® are still poorly understood, especially in tro- season between October and March and a colder and dry season be-
pical environments where toxic cyanobacteria blooms can persist all tween April and September.
year round (McGreggor and Fabbro 2000; Figueredo and Giani, 2009;
Figueredo et al., 2016). 2.2. Phoslock® application
Cyanobacteria have numerous adaptations that allow them to tol-
erate diverse environmental conditions. Some groups are able to fix The Phoslock® treatment in Pampulha reservoir was performed
atmospheric nitrogen and can also take up and store P, when it is under request of the municipality of Belo Horizonte by a Brazilian li-
limiting in the epilimnion (Ganf and Oliver, 1982; Reynolds, 2006; censee (HydroScience Company) that represents the Australian CSIRO.
Paerl and Otten, 2013). Additionally, they have the ability to regulate According to the municipality and the company (pers. comm.), the
their buoyancy in the water column due to the presence of gas vacuoles, Phoslock® amount needed to inactivate phosphorus in Pampulha lake
thereby optimizing their access to both light and nutrients (Reynolds was calculated on a 1:1 M ratio of La/P, based on measurements of
et al., 1987; Paerl, 2018), While these characteristics give cyano- phosphorus concentration in the water column and in the sediments,
bacterial populations a competitive advantage, they pose a challenge to taking into account the potentially releasable organic and Fe/Mn bound
eutrophication control in lake management, as the systems become P. The clay was periodically added as slurry to the water surface after
more resilient. being mixed with lake water using a barge and distributed in the re-
It is therefore important to follow and understand the cyanobacteria servoir across four established regions (see map in Suppl. Mat. Fig. S1).
response to restoration treatments in order to develop more appropriate The four areas covered the entire lake and were established based on
strategies to control and mitigate their blooms. Results are not always location, water depth, sediment characteristics, and total amount of
as expected, since a number of environmental factors may affect the phosphorus targeted. Applications were made 5 days per week, starting
response to the in situ treatment (Lürling and Faassen, 2012; Lürling in April 2016 and until February 2018, for a total of 23 months.
and van Oosterhout, 2013; Moos et al., 2014; Lang et al., 2016). Tro- Average amounts of around 33 to 47 tons of Phoslock® were applied
pical lakes present a further challenge, as they have been less recorded every month in Pampulha reservoir, achieving a total of approximately
(Robb et al., 2003; Moos et al., 2014; Yamada-Ferraz et al., 2015) and 900 tons over nearly two years of treatment. The treatment design for
also due to their higher temperature condition that contributes to an this specific system was the company’s responsibility and approved by
almost interrupted growing season of the phytoplankton, allowing high the municipality.
biomass concentration.
Pampulha reservoir is a small, tropical, urban lake in southeastern 2.3. Sample collection and water quality measurements
Brazil. The reservoir was originally built for water supply to the city of
Belo Horizonte, however, from the 1970′s onward, it has deteriorated Sampling was carried out during the Phoslock® applications from
following the rapid urbanization of its watershed. As a consequence, a April/16 to February/18, with additional post-application follow-ups
strong eutrophication process started in the early eighties leading to the from March/18 to July/18. Throughout the entire period, samples were
appearance of periodic harmful cyanobacterial blooms and to the in- taken once per month at a fixed station located in the central region of
terruption of the lake use for water supply (Pinto-Coelho, 1998). Sev- the lake (maximal depth 9.5 m). The physical-chemical profiles of water
eral studies performed during the last decades indicated the high levels temperature (°C) (WT), dissolved oxygen (mg L−1) (DO), electric con-
of eutrophication and accentuated cyanobacterial dominance in the ductivity (µS cm−1) (EC) and pH were obtained in situ every 0.5 m from
reservoir (Pinto-Coelho 1998; Figueredo and Giani 2001; Figueredo the surface to the bottom layer using a multi-parameter probe Yellow
et al., 2016; Silva et al., 2016). However, the reservoir maintains a Springs Instruments-YSI, model 556. Water transparency was estimated
status as an important tourist spot and is listed as one of the UNESCO’s by Secchi disk. Water samples for analyses of dissolved nutrients,
World Heritage sites. Tourists and the local population are constantly chlorophyll-a and phytoplankton biomass were collected at 0.5 m depth
seen around its margins, occasionally fishing, despite the risk of con- with a Van Dorn sampler. Water samples were also collected at 8.0 m
tamination. depth for analysis of soluble reactive phosphorus and total phosphorus.
For these reasons, the local municipality has been investing in im- Samples for the analyses of the quantitative composition of phyto-
provement and recovery policies of the lake and in 2016 it implemented plankton were immediately fixed with Lugol's solution. Water samples
a treatment program based in periodic Phoslock® applications, per- for the analysis of nitrite (N-NO2−), nitrate (N-NO3−), ammonium (N-
formed by a private company (HidroScience, a CSIRO licensee in NH4+) and soluble reactive phosphorus (SRP) were filtered on glass
Brazil). In the present study, we aimed to take advantage of this on- fiber filters (GF-1, Macherey- Nagel, 0.7 μm pore diameter) and frozen
going treatment program and to follow the lake conditions through (-20 °C) until analysis. Concentrations of ammonium, nitrate and nitrite
monthly sampling. Our specific goals were to: (1) evaluate Phoslock®’s were summed up to obtain DIN (dissolved inorganic nitrogen). For the
potential role in improving water quality and reducing P-availability; analysis of total phosphorus (TP), unfiltered samples were used.
(2) investigate if Phoslock® application and subsequent declines in P- Nutrients measurements were performed by colorimetric methods ac-
availability would reduce cyanobacterial blooms; and (3) assess if re- cording to APHA (2005).
ductions in cyanobacterial blooms prompted a displacement towards Chlorophyll-a concentration was determined by hot ethanol ex-
more diverse phytoplankton communities. traction, according to Nusch (1980). The phytoplankton species were

2
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

identified and counted at 400X magnification using a ZEISS Axio Ob- conditions with increasing depth (Fig. 1B). The lowest oxygen con-
server D1 inverted microscope following Utermöhl sedimentation centration at the surface was observed between September and De-
method (1958). The counts were performed until reaching a minimum cember 2016, with values as low as 2.2 mg L−1, corresponding to the
of 100 individuals of the dominant species (to increase precision, when period of the lowest phytoplankton biomass and chlorophyll recorded.
dominant species were extremely abundant, subdominants were used). Patterns in water temperature and DO profiles are a response to the
Biovolume was calculated for each species using geometric shape ap- seasonal climatic changes that characterize the region, which shows a
proximation (Rott, 1981; Hillebrand et al., 1999) measured on a drier and colder season from April to September, and a warmer and
minimum of 20 randomly selected individuals. Species richness was rainy season from October to March (Fig. S2).
evaluated by the number of species recorded in each phytoplankton Following the Phoslock® treatment, phosphorus (TP and SRP) con-
phylum. centration largely varied over the study period in Pampulha reservoir
Precipitation and air temperature data were obtained from the (Fig. 2). In the surface layer, TP minimum and maximum values ranged
INMET (National Meteorological Institute), referring to station 83,587 from 66 to 269 μg L−1, while at the bottom, TP concentration was al-
(Belo Horizonte city, MG). Pre-application data were obtained from ways higher and oscillated from 143 to 786 μg L−1 (Fig. 2A). SRP
sampling carried out in Pampulha reservoir on a monthly basis in 2014 varied from 8 to 96 μg L−1 at the surface, and SRP concentration ranged
and 2015. Data included surface TP, SRP, water transparency by Secchi from 8 to 572 μg L−1 at higher depth (Fig. 2B). Water transparency
disk, chlorophyll-a and cyanobacteria biovolume. Sampling metho- values in Pampulha reservoir during the treatment were often quite low
dology and laboratory analyses were the same during the entire study. and below than 0.5 m, except between September/16 and February/17,
when the water transparency increased nearly 4-fold, reaching ap-
2.4. Data analysis proximately 2.0 m in November/16 as the water became very clear
(Fig. 3A).
To analyze the effect of the treatment in the reservoir, a dataset of
two years previous to the start of the treatment (Jan/14 to Dec/15) was 3.2. Composition and biomass of the phytoplankton community
summarized in box-plots together with the period of applications (first
treatment year: April/16 to March/17; second treatment year: April/17 Fig. 3B shows changes in the quantitative and qualitative compo-
to February/18). Data of TP, SRP, cyanobacteria biomass, chlorophyll-a sition of the phytoplankton community and chlorophyll-a concentration
and water transparency were checked for normality and hetero- following the Phoslock® applications in Pampulha reservoir. At dif-
scedasticity on log-transformed data using the SigmaPlot 11.0. A one- ferent time periods, Phoslock® had a distinct effect on the composition
way Analysis of Variance (ANOVA) was used to test for significant and biomass of the phytoplankton, with the positive effect of reduction
changes among years in normally distributed variables (TP and chlor- of cyanobacteria biomass in two periods. The first response was ob-
ophyll-a), followed by a Tukey post-hoc test to identify significantly served between September/16 and February/17 and the second be-
different years. A Kruskal-Wallis ANOVA (KWA), followed by Dunn’s tween December/17 and June/18. Discrete effects or no effect were
test, was used to test for significance among years when variables were observed in the remaining months.
not normally distributed (SRP and water transparency; Shapiro-Wilk Cyanobacteria were dominant during most of the study period, ac-
test, P < 0.05) or violated criteria of equal variance (Cyanobacteria counting for more than 80% of the total phytoplankton biomass in
biovolume; Shapiro-Wilk test, P < 0.05). Species richness of cyano- several months. The most representative cyanobacteria species were
bacteria and of other phytoplankton organisms during the treatment Cylindrospermopsis raciborskii (Nostocales), Planktothrix agardhii, P. iso-
period was divided in three subsequent years, and difference among thrix (Oscillatoriales), and Microcystis aeruginosa (Chroococcales).
years was tested by ANOVA, followed by a Tukey post-hoc test to Chlorophyta, Bacillariophyta, Euglenophyta, and other taxonomic
identify significantly different years. Data were tested for normality, groups were less abundant and made small or no contribution to the
and analyses performed as above. total phytoplankton biomass in most months. Lower cyanobacteria
To extract the main trends of the treatment’s effect on environ- biomass was recorded from September/16 to February/17, when their
mental and biotic parameters, a redundancy analysis (RDA) on data populations declined dramatically and nearly disappeared from the
sampled from April/16 to July/2018 was performed following Borcard reservoir, along with most phytoplankton species. The only exception
et al. (2011). The dataset included five months of post-treatment was in December/16, when large colonies of M. aeruginosa developed in
follow-ups (see 2.3). All data, except pH, were log (x + 1) transformed. the reservoir dominating the phytoplankton community, even if at low
The statistical significance of the relationship was evaluated by Monte biomass.
Carlo permutation tests, under 999 permutations. The analysis was run Despite the disappearance of nearly all cyanobacteria between
using the R software and the vegan package (Oksanen et al., 2013; R September/16 and February/17, they became dominant again
Development Core Team, 2013). Environmental data included TP, SRP, throughout 2017, showing large variations in species biomass, com-
DIN, pH, DO, rainfall, WT and water transparency, while biotic data position and dominance. They peaked in November/17, when the
included chlorophyll-a and total biomass of cyanobacteria. Collinearity highest biomass of the whole period of study was recorded (202 mm3
was tested by vif (variance inflation factor) and collinear variables were L−1). However, in December/17, cyanobacteria dramatically decreased
not included in the analysis. and their biomass remained low until June/18, representing a period of
higher biodiversity. Growth of small green algae (Chlorophyta) im-
3. Results mediately followed the reduction of cyanobacteria in the lake. Then
diatoms, a group rarely registered in Pampulha, became dominant with
3.1. Physical and chemical parameters a large bloom of Synedra ulna in March/18. As a result, profound shifts
in the structure of the phytoplankton community were registered
Water temperature vertical profiles (Fig. 1A) exhibited a clear sea- starting at the end of the second year of treatment and were maintained
sonal pattern of thermal stratification of the water column in the during the first post-treatment months, culminating in the effective
warmest months (between October and March). The thermal amplitude decrease of cyanobacterial blooms, as well as changes in species com-
in the water column during summer stratification periods was around position, higher species richness and a progressive reconstruction of the
5 °C. During the coldest months (May to August), the water column phytoplankton community.
tended to be isothermal, normally well mixed or showing very small Chlorophyll-a concentration varied from 2 μg L−1 in September/16
thermal amplitude from the surface to the bottom. The dissolved to 232 μg L−1 in August/16, showing a pattern of increase at the end of
oxygen (DO) profiles showed a trend of oxygen depletion and anoxic the dry season, when chlorophyll peaks usually occurred (Fig. 3B).

3
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

Fig. 1. Temporal variation of the (a) thermal water profile (°C) and (b) vertical dissolved oxygen profile (mg L−1) in Pampulha reservoir during the study period.

Lower chlorophyll values were observed at the end of the first year of A significant increase in species richness of phytoplankton species
treatment (from September/16 to January/17) indicating a clear re- other than cyanobacteria was observed in 2018, i.e. after 21 months of
duction of phytoplankton biomass. Fluctuations in chlorophyll con- Phoslock treatment (Fig. 5A). Fig. 5B shows a negative relationship
centration were observed throughout 2017, and reached a peak in between biovolume of cyanobacteria and number of non-cyanobacteria
November/17 (223 μg L−1), which corresponded to the period of species and the model was able to explain 57% of the variation, in-
highest cyanobacteria biomass. A significant reduction of chlorophyll dicating that a lower cyanobacterial biomass was allowing the existence
concentration was observed again at the end of the second year of of richer species communities and higher biodiversity.
treatment (from December/17 to February/18) and remained low in The results of the redundancy analysis (RDA) are represented in
the post-application months (from March/18 to June/18), following the Fig. 6. The multivariate analysis was used to identify patterns of change
strong shifts recorded in the phytoplankton composition and biomass. that occurred in environmental and biotic parameters as a consequence
of the treatment. The results showed that the model was highly sig-
nificant and explained 78% of the variability (F ratio = 6.48,
3.3. Statistical results P < 0.001). The first axis (RDA1) explained the larger proportion of
the variance (73%) and axis 2 (RDA 2) only 5%. The main variables
Fig. 4 represents box-plots of five variables measured in years pre- positively associated with the first ordination axis were water trans-
vious to the treatment, as well as during the treatment. Total phos- parency, rainfall and SRP, and, negatively, TP and dissolved oxygen
phorus showed a decreasing trend, but a significant difference was (DO) (Table 1). The biotic variables, chlorophyll and cyanobacteria
observed only in the second year of treatment. Transparency (by Secchi biomass, were both negatively correlated to axis 1(Table 1). Thus,
disk) showed significant increase, but a large variation was detected in RDA1 showed a pattern of decreasing TP, chlorophyll and cyano-
the first year of treatment. Chlorophyll-a decreased and the difference bacteria, opposite to increasing transparency, SRP and rain. These
between years before and during the treatment was significant. Cya- variables represent either a direct response to the Phoslock® treatment
nobacteria biovolume also decreased significantly in the first year of (for example phosphorus or cyanobacteria), or a potential interference
treatment, but started to increase again in the second year. Interestingly factor (rain). By precipitating phosphorus, Phoslock® caused a decrease
SRP concentration increased, especially during the first year of treat- of cyanobacteria biomass and chlorophyll, consequently causing a
ment, which may indicate a phosphorus release from the dying cya- lower production of oxygen (DO) and higher transparency, and even an
nobacterial cells as their biomass decreased.

4
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

Fig. 2. Temporal variation of total phosphorus- TP (A) and soluble reactive phosphorus – SRP (B) at the surface (0.5 m) and bottom layer (8.0 m) in Pampulha
reservoir during the Phoslock treatment period.

increase of dissolved phosphorus (SRP), released from the lysis of dying Lang et al., 2016). The present study aimed to follow changes in the
cyanobacterial cells. Axis 2 explained a smaller portion of the variance, water quality of a tropical, highly eutrophic reservoir subjected to an
but it was positively associated with dissolved inorganic nitrogen (as in-lake treatment designed to remove phosphorus from the water
DIN, the sum of ammonium -NH4, nitrite - NO2 and nitrate –NO3) and, column. The results showed that the frequent Phoslock® applications
negatively, with water temperature (Table 1). reduced phosphorus availability in some periods, leading to short-lived
The analysis revealed three well-defined groups of months, re- changes in the phytoplankton community.
presented in Fig. 6 by three separated ellipses. The green ellipse, on the The initial response to the Phoslock® treatment was seen after five
negative side of axis 1, encloses months with the highest cyanobacteria months of almost daily applications. The resilience observed in the
biomass, which are found when TP and DO concentrations were also cyanobacteria response was likely due to the time needed for the
high. The blue ellipse, on the positive side of axis1, is characterized by treatment to be effective in an extremely P-rich environment, as well as
months of the second part of 2016, when a sudden decrease of cyano- to their capacity to store intracellular phosphorus (Mur et al., 1999). At
bacteria biomass and increased transparency were observed. Interest- the beginning of the treatment, cyanobacteria biomass in Pampulha
ingly, the orange ellipse encloses the summer months of the beginning reservoir was very high and cyanobacteria cells had probably enough
of both 2017 and 2018, when algal groups other than cyanobacteria intracellular phosphorus to keep growing even after the start of Phos-
were dominant and the highest phytoplankton richness was observed. lock® applications. At this time, most P was in particulate form, thus
These months are found on the negative side of axis 2, characterized by incorporated into the cyanobacteria biomass as TP (Fig. 2A) and was
higher water temperature and rain (representing summer in this region) quickly recycled and consumed by microorganisms when released.
and lower concentration of nitrogenous nutrients. Phoslock® has no effect on particulate P, therefore it will not remove P
from inside the phytoplankton cells (Lürling and van Oosterhout,
2013). However, as the phosphorus supply inside the cells was de-
4. Discussion
pleted, cyanobacteria died progressively, which also resulted in de-
creasing in TP concentration and increasing SRP (Fig. 2B). A similar
Limiting nutrient inputs into aquatic systems is the main strategy to
pattern was observed by Epe et al. (2017), following Phoslock® appli-
be adopted to successfully reduce eutrophication. Even though several
cations in Bärensee lake (Germany), as the progressive death of the
management methods focus on controlling external loading, sediments
biomass occurred, TP was reduced by almost 50% and SRP was si-
are known to promote significant release of phosphorus by internal
multaneously released in the water.
load, which becomes easily available to phytoplankton growth and may
In Pampulha reservoir, there was a relatively rapid initial response
eventually support serious cyanobacteria blooms (Nürnberg, 2009;

5
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

Fig. 3. Temporal variation of (A) Secchi disk depth and (B) phytoplankton biomass and composition and chlorophyll-a concentration in Pampulha reservoir during
the study period.

to the treatment, when cyanobacteria and total phytoplankton biovo- However, there was a second period from December/17 to July/18,
lumes significantly decreased from September/16 to February/17 when changes in the phytoplankton community were observed again
(Fig. 3B), prompting a 4-fold increase in water transparency and the (Fig. 3B). Unlike in September/16, when a pronounced decrease of the
lowest chlorophyll concentration in the whole period of study, and entire phytoplankton biomass was registered, in this second period the
resulting in a separate cluster in the RDA (Fig. 6). However, during phytoplankton community presented a more diverse species composi-
2017 this positive treatment outcome was not further sustained and tion and the dominance moved from cyanobacteria towards other algal
cyanobacteria biomass increased and persisted until December/17. groups. A larger contribution of chlorophytes, diatoms, cryptophytes
Lang et al. (2016) reported that after a Phoslock® treatment in Loch and euglenophytes was observed. Diatoms, for example, have been
Flemington (Scotland, UK) cyanobacteria blooms continued to develop, found to be abundant at moderate eutrophication levels (Ptacnik et al.,
even though at lower intensities. The authors attributed the limited 2008). Such phytoplankton diversity and changes in the structure of the
cyanobacteria reduction to weather conditions, arguing that the occa- phytoplankton community were rarely reported in Pampulha reservoir
sional drier and warmer summer that was recorded would have favored in recent years (Batista et al., 2018) and indicated a remarkable
the persistence of high TP concentration in the lake. It is possible that achievement during the Phoslock® treatment. Only few studies have
climatic factors also affected the response to the treatment in Pampulha previously reported the recovering of the phytoplankton community
reservoir. As previously reported, external loads from the runoff are following Phoslock® applications, observing a gradual decrease in cy-
quite high at the beginning of the rainy season in the reservoir and may anobacteria populations, the establishment of other algal groups and
cause a rapid increase in the concentration of nutrients (Silva et al., higher community diversity (see for example: Bishop et al., 2014; Lang
2016; Batista et al., 2018; Silva et al., 2019). The sudden inputs of et al., 2016; Bishop and Richardson, 2018).
nutrients as a consequence of the initial rainfalls (Fig. S2) may, there- Even if moderate, the decrease in chlorophyll-a and cyanobacteria
fore, have restrained the effectiveness of the treatment in the first year. biomass observed in a few months during this study indicates some

6
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

Fig. 4. Box-plots representing variables before and during the Phoslock treatment: (A) Total phosphorous; (B) Soluble reactive phosphorus; (C) Cyanobacteria
biovolume; (D) Chlorophyll-a; (E) Transparency by Secchi depth. Medians are denoted by central lines and are bounded by the 25th and 75th percentiles. Whiskers
show 10th and 90th percentiles. Dots outside boxes and whiskers are outliers. Different letters represent significant difference at level p ≤ 0.05. See methods for
details on statistics.

ecological improvements in the lake and show the Phoslock® potential low efficiency of resource use at the next trophic level, because of the
in controlling blooms in this environment. Before the treatment, cya- cyanobacteria poor palatability, resulting in low system turnover rates
nobacteria were dominant throughout the year in Pampulha reservoir, (Filstrup et al., 2014; Heathcote et al., 2016; Lehtinen et al., 2017). In
while other phytoplankton groups were rarely found and never at high our study, the decrease of cyanobacteria biomass and dominance in the
densities. During this study, cyanobacteria populations suffered strong total phytoplankton at the end of the second treatment year was cor-
oscillations, abruptly decreasing their biomass in two episodes, as dis- related to an increase in non-cyanobacterial species richness (Fig. 5B),
cussed before, therefore allowing the diversification of the phyto- suggesting the beginning of the development of a community that
plankton community (Fig. 3B; Fig. 5A). Biodiversity loss has been would better support herbivores and increase the ecosystem func-
connected to environmental changes and serious effects on ecosystem tioning.
functioning (Hooper et al., 2012). Highly productive, cyanobacteria The ability of Phoslock® to capture and remove phosphorus has
dominated, aquatic environments are quite exceptional since even at already been reported in several studies (Robb et al., 2003; Ross et al.,
low diversity, they are characterized at the primary producer level by 2008; Haghseresht et al., 2009; Spears et al., 2016). Comparing phos-
efficient use of resources by the dominant species, however they show phorous concentration in Pampulha reservoir with previous years (see:

7
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

the end of this study (Fig. 3B). Lürling and van Oosterhout (2013),
following Phoslock® applications in Lake Het Groene Eiland (The
Netherlands), observed that the results did not meet the expectations
and no significant decrease in phosphorus concentration was observed.
In Pampulha, TP concentration was almost always above 100 µg L−1,
demonstrating the high degree of eutrophication of the reservoir
(OECD, 1982). Indeed, TP concentration higher than 100 µg L−1 was
suggested as a turning point that leads to cyanobacteria dominance in
aquatic ecosystems (Downing et al., 2001; Giani et al., 2005). Even
though in Pampulha a significant cyanobacteria decrease was observed
after the first year of treatment, in 81% of the samples cyanobacteria
biovolume was still higher than 5 mm3 L, characterizing a hypereu-
trophic reservoir (Smith, 2003). Interestingly, even though phos-
phorous was still high, the small changes that occurred in the en-
vironmental conditions at the end of the second year of treatment were
enough to allow potential phytoplankton competitors to gain space and
to induce shifts in the community structure, increasing richness at least
for a short period of time. This indicates how the community may be a
better ecological indicator than, for example, chemical variables,
thanks to the organisms’ fast response to environmental variations and
water quality change detection, before they become measurable by
nutrient concentration (Brettum and Andersen, 2005).
Finally, despite its clear relevance, phosphorus is not the only im-
portant factor for the occurrence of high cyanobacteria biomass in
Pampulha reservoir, where ammonium and nitrate concentrations are,
for example, also extremely high (Pinto-Coelho et al., 1997; Figueredo
et al., 2016). The RDA indicated that nitrogenous nutrients (DIN) have
a different trend than phosphorus, thus they should be taken into ac-
count in the reservoir management program, when aiming for the full
recovery of this waterbody. The importance of the dual control of
phosphorous and nitrogen in freshwater systems has frequently been
Fig. 5. Species richness of cyanobacteria and other phytoplankton species in emphasized (Leavitt et al., 2006; Conley et al., 2009; Smith et al., 2016)
Pampulha reservoir during the period of study (A) and relationship and re- and it was suggested that nitrogen may control cyanobacteria growth
spective regression equation between percent of cyanobacteria biovolume and and toxicity (Gobler et al., 2016).
percent of non-cyanobacteria number of species (B). Small bars above columns In Pampulha reservoir, Phoslock® was applied almost daily for ap-
in (A) represent standard error; different letters at the top indicate significant proximately two years, but even at high frequency, the results were not
difference. completely satisfactory. It must be taken into account that a large
amount of nutrients still enters in the reservoir by the tributaries (Silva
Fig. 4A; Batista et al., 2018; Silva et al., 2019), TP showed a trend of et al., 2019), demonstrating that in order to achieve adequate results,
decreasing concentration with significant change especially in the Phoslock® applications should be preceded by a substantial reduction of
second year of treatment, indicating a positive effect. However, SRP external nutrient loads (Cooke et al., 2005). In Lake Het Groene Eiland,
increased (Fig. 4B). During the treatment period, phosphorus con- constant Phoslock® reapplications did not seem to be an effective and
centration declined significantly between September/16 and January/ sustainable measure for the ecological recovery of the lake (Lürling and
17. But phosphorus concentration increased close to the bottom van Oosterhout, 2013). On the other hand, Meis et al. (2013) found that
(Fig. 2B), suggesting that the amount of product applied in Pampulha smaller multiple doses were likely to increase the effectiveness of the
reservoir may not be enough to permanently bind P in the sediment and treatment. But they emphasized that the applications should always be
to sustain long-term improvements in water quality. In a study in Loch managed in parallel with the continuous reduction of external nutrient
Flemington, Lang et al. (2016) also found that the Phoslock® dosage load. According to these authors, the calculation of an “effective dose”
was probably insufficient to control the P release from the sediment, is nearly impossible due to the complex mechanisms inherent to
supporting cyanobacterial blooms of smaller magnitude in post-appli- phosphorus cycling, pointing to the need of constant monitoring and
cation years. Laboratory studies performed by Meis et al. (2013) with periodical dose readjustments to ensure better results.
sediments from the same Loch Flemington demonstrated that, following
Phoslock® applications under aerobic conditions, no sediment P release 5. Conclusions
occurred, while significant amounts of P were released under anaerobic
conditions. According to Meis et al. (2013), additional Phoslock® ap- Our study presents the results of a high frequency and long-term
plications could be a possible option under prolonged periods of anoxia. Phoslock® treatment in an urban lake in a tropical region, Pampulha
In Pampulha reservoir, the hypolimnion is almost permanently anoxic reservoir. The treatment showed a statistically significant reduction of
(Fig. 1B), which enables continuous phosphorus release from the sedi- phosphorus concentration in the reservoir only in the second year of
ment to the water column. This may indicate that the Phoslock® dosage treatment, but values often above 100 µg L−1 were still registered that
applied in Pampulha should probably be increased in future recovery prompted the maintenance of cyanobacteria blooms. Nevertheless,
programs to avoid P release from the anoxic sediments. Phoslock® positive effect on the phytoplankton community was re-
Unfortunately, phosphorus levels in Pampulha reservoir seem to be markable in two periods: the first positive outcome occurred at the end
still too high to sustain the ecological improvements observed in the of the first year of treatment and, the second, at the end of the second
phytoplankton community composition between January/18 and July/ year. In these occasions, a reduction of cyanobacteria biomass and a
18, and a tendency of cyanobacteria return was already noticeable at shift towards dominance of other phytoplankton groups was observed,
promoting greater community biodiversity. Yet, despite the fact that

8
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

Fig. 6. Triplot redundancy analysis of the relationships found between abiotic parameters (blue arrows), biotic parameters (red arrows) and samples distribution
(circles: the numbers inside the circles indicate month and year) in Pampulha reservoir (April/2016 to July/2018). The arrows indicate the direction of increase. The
angles between variables reflect their correlations (angles close to 90° indicate no correlation, angles close to 0° indicate a high positive correlation and angles close
to 180° indicate a high negative correlation). WT = water temperature; Rainf = rainfall; Transp = water transparency; DO = dissolved oxygen;
pH = hydrogenation potential; TP = total phosphorus; DIN = dissolved inorganic nitrogen; SRP = soluble reactive phosphorus; Chla = chlorophyll-a;
CyanoTotal = total biomass of cyanobacteria. See text for explanation of the three colored ellipses.

Table 1 of nutrients; weather related factors (sudden runoff inputs at the be-
Redundancy analysis scores to axis-1 (RDA1) and axis-2 (RDA2) for biotic and ginning of the rain season); anoxic sediments, almost permanently
environmental variables. present in Pampulha reservoir, which facilitated the continuous release
Biotic variable scores of phosphorus from deeper layers.
The results presented in this paper show that in-lake treatment
RDA1 RDA2 based on lanthanum-enriched clay has a great potential to improve the
ecological status of eutrophic lakes in the Tropics and elsewhere.
Chla −1.758 −0.386
CyanoTotal −1.516 0.448 However, this study also pointed out the difficulties to treat an eco-
system in real scale, while dealing with a large number of uncontrolled
Abiotic variables scores
RDA1 RDA2
environment variables. Despite the treatment limitations here high-
lighted, it was encouraging to observe that even in a hypertrophic
TP −0.586 0.277 system such as Pampulha reservoir an efficient long-term control of
DIN −0.101 0.576 phosphorus loads could revert the loss of phytoplankton biodiversity,
DO −0.653 −0.343
therefore recovering its ecosystem functions and services. Decrease of
pH −0.043 −0.146
WT 0.490 −0.588 cyanobacteria biomass and increase in non-cyanobacteria biodiversity
Transp 0.886 0.199 were key factors to evaluate the treatment success, reinforcing the
Rainf 0.435 −0.317 importance of species and communities to evaluate the ecosystem
SRP 0.736 0.336
health.
Chla = Chlorophyll-a; CyanoTotal = Total biomass of cyanobacteria;
TP = Total phosphorus; DIN = dissolved inorganic nitrogen; DO = Dissolved
oxygen; pH = Hydrogenation potential; WT = Water temperature; CRediT authorship contribution statement
Transp = Water transparency; Rainf = rainfall; SRP = soluble reactive phos-
phorus. Bruna Barçante: Investigation, Visualization, Writing - original
draft, Writing - review & editing. Nilo O. Nascimento: Funding ac-
the Phoslock® treatment was efficient for the temporary mitigation of quisition, Project administration, Writing - review & editing. Talita
cyanobacteria blooms, our findings indicate that this technique was not F.G. Silva: Data acquisition, Writing - review & editing. Leandro A.
able to support effective and long-lasting ecological restoration of this Reis: Data acquisition. Alessandra Giani: Conceptualization, Funding
highly eutrophic system. Several factors may have been responsible for acquisition, Project administration, Supervision, Methodology, Writing
the lack of better results such as: no sufficient removal of external loads - review & editing.

9
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

Declaration of Competing Interest 1263693.

Figueredo, C.C., Giani, A., 2001. Seasonal variation in the diversity and species richness
of the phytoplankton in a tropical eutrophic reservoir. Hydrobiologia 445, 165–174.
The authors declare that they have no known competing financial https://doi.org/10.1023/A:1017513731393.
interests or personal relationships that could have appeared to influ- Figueredo, C.C., Giani, A., 2009. Phytoplankton community in the tropical lake of Lagoa
Santa (Brazil): conditions favoring a persistent bloom of Cylindrospermopsis raci-
ence the work reported in this paper. borskii. Limnologica 39, 264–272. https://doi.org/10.1016/j.limno.2009.06.009.
Figueredo, C.C., Pinto-Coelho, R.M., Lopes, A.M.M.B., Lima, P.H.O., Gücker, B., Giani, A.,
Acknowledgements 2016. From intermittent to persistent cyanobacterial blooms: identifying the main
drivers in an urban tropical reservoir. J. Limnol. 75, 445–454. https://doi.org/10.
4081/jlimnol.2016.1330.
This study was supported by a scholarship from CAPES Filstrup, C.T., Hillebrand, H., Heathcote, A.J., Harpole, W.S., Downing, J.A., 2014.
(Coordenação de Aperfeiçoamento de Pessoal de Nível Superior) to B.B. Cyanobacteria dominance influences resource use efficiency and community turnover
in phytoplankton and zooplankton communities. Ecol. Lett. 17, 464–474. https://doi.
and by funds provided by FAPEMIG (Fundação de Apoio à Pesquisa de
org/10.1111/ele.12246.
Minas Gerais) to A.G. and by CNPq (Conselho Nacional de Ganf, G.G., Oliver, R.L., 1982. Vertical separation of the light and available nutrients as a
Desenvolvimento Científico e Tecnólogico) to N.O.N. The second and factor of causing replacement of Green algae by Blue-green algae in the plankton of a
the last authors are CNPq-PQ fellows. We thank the Municipality of stratified lake. J. Ecol. 70: 829-844.http://doi.org/10.2307/2260107.
Giani, A., Bird, D., Prairie, Y., Lawrence, J., 2005. Empirical study of cyanobacterial
Belo Horizonte city (PBH, Prefeitura de Belo Horizonte) for sampling toxicity along a trophic gradient of lakes. Can. J. Fish. Aquat. Sci. 62, 2100–2109.
support and especially for providing information about the Phoslock® https://doi.org/10.1139/f05-124.
application in the reservoir. We also thank the members of the Gobler, C.J., Burkholder, J.A.M., Davis, T.W., Harke, M.J., Johengen, T., Stow, C.A., Van
de Waal, D.B., 2016. The dual role of nitrogen supply in controlling the growth and
Phycology laboratory who helped with the sampling and laboratory toxicity of cyanobacterial blooms. Harmful Algae 54, 87–97. https://doi.org/10.
analyses. 1016/j.hal.2016.01.010.
Haghseresht, F., Wang, S., Do, D.D., 2009. A novel lanthanum modified bentonite,
Phoslock, for phosphate removal from wastewaters. Appl. Clay Sci. 46, 369–375.
Appendix A. Supplementary data https://doi.org/10.1016/j.clay.2009.09.009.
Heathcote, A.J., Filstrup, C.T., Kendall, D., Downing, J.A., 2016. Biomass pyramids in
Supplementary data to this article can be found online at https:// lake plankton: influence of Cyanobacteria size and abundance. Inland Waters 6,
250–257. https://doi.org/10.5268/IW-6.2.941.
doi.org/10.1016/j.ecolind.2020.106702.
Hillebrand, H., Durselen, C.D.D., Kirschtel, U., Pollingher, T., Zohary, T., 1999.
Biovolume calculation for pelagic and benthic microalgae. J. Phycol. 35, 403–424.
References https://doi.org/10.1046/j.1529-8817.1999.3520403.x.
Hooper, D.U., Adair, E.C., Cardinale, B.J., Byrnes, J.E.K., Hungate, B.A., Matulich, K.L.,
Gonzalez, A., Duffy, E., Gamfeldt, L., O’Connor, M.I., 2012. A global synthesis reveals
APHA – American Public Health Association. 2005. Standard methods for the examina- biodiversity loss as a major driver of ecosystem change. Nature 486, 105–108.
tion of water and wastewater. 20th ed. Washington, EUA: American Public Health https://doi.org/10.1038/nature11118.
Association. Lang, P., Meis, S., Procházková, L., Carvalho, L., Mackay, E.B., Woods, H.J., Pottie, J.,
Batista, A.M.M., Figueredo, C.C., Giani, A., 2018. Variability in a permanent cyano- Milne, I., Taylor, C., Maberly, S.C., Spears, B.M., 2016. Phytoplankton community
bacterial bloom: species-specific responses to environmental drivers. FEMS responses in a shallow lake following lanthanum-bentonite application. Water Res.
Microbiol. Ecol. 94, fiy197. https://doi.org/10.1093/femsec/fiy197. 97, 55–68. https://doi.org/10.1016/j.watres.2016.03.018.
Bishop, W.M., Hyde, S., McNabb, T., Cormican, I., Willis, B.E., 2014. Operational eva- Leavitt, P.R., Brock, C.S., Ebel, C., Patoine, A., 2006. Landscape- scale effects of urban
luation of Phoslock phosphorus locking technology in Laguna Niguel Lake, CA. nitrogen on a chain of freshwater lakes in central North America. Limnol. Oceanogr.
Water, Air, Soil Pollut. 225, 1–11. https://doi.org/10.1007/s11270-014-2018-6. 51, 2262–2277. https://doi.org/10.4319/lo.2006.51.5.2262.
Bishop, W.M., Richardson, R.J., 2018. Influence of Phoslock® on legacy phosphorus, Lehtinen, S., Tamminen, T., Ptacnik, R., Andersen, T., 2017. Phytoplankton species
nutrient ratios, and algal assemblage composition in hypereutrophic water resources. richness, evenness, and production in relation to nutrient availability and imbalance.
Environ. Sci. Pollut. Res. 25, 4544–4557. https://doi.org/10.1007/s11356-017- Limnol. Oceanogr. 62, 1393–1408. https://doi.org/10.1002/lno.10506.
0832-2. Lürling, M., Faassen, E.J., 2012. Controlling toxic cyanobacteria: Effects of dredging and
Borcard, D., Gillet, F., Legendre, P., 2011. Numerical ecology with R. Use R! series. phosphorus-binding clay on cyanobacteria and microcystins. Water Res. 46: 1447-
Springer Science, New York, NY, USA. https://doi.org/10.1007/978-1-4419-7976-6. 1459. http://doi.org/doi:%2010.1016/j.watres.2011.11.008.
Brettum, P., Andersen, T., 2005. The use of phytoplankton as indicators of water quality. Lürling, M., van Oosterhout, F., 2013. Case study on the efficacy of a lanthanum-enriched
NIVA Rep. SNO 4818-2004. clay (Phoslock®) in controlling eutrophication in Lake Het Groene Eiland (The
Brönmark, C., Hansson, L., 2002. Environmental issues in lakes and ponds: current state Netherlands). Hydrobiologia 710, 253–263. https://doi.org/10.1007/s10750-012-
and perspectives. Environ. Conserv. 29, 290–306. https://doi.org/10.1017/ 1141-x.
S0376892902000218. McGreggor, G.B., Fabbro, L.D., 2000. Dominance of Cylindrospermopsis raciborskii
Carpenter, S.R., 2008. Phosphorus control is critical to mitigating eutrophication. PNAS (Nostocales, Cyanoptokaryota) in Queensland tropical and subtropical reservoirs:
105, 11039–11040. https://doi.org/10.1073/pnas.0806112105. implications for monitoring and management. Lakes Reserv. Res. Manage. 5,
Carvalho, L., McDonald, C., Hoyos, C., Mischke, U., Phillips, G., Borics, G., Poikane, S., 195–205. https://doi.org/10.1046/j.1440-1770.2000.00115.x.
Skjelbred, B., Lyche Solheim, A., Van Wichelen, J., Cardoso, A.C., 2013. Sustaining Meis, S., Spears, B.M., Maberly, S.C., Perkins, R.G., 2013. Assessing the mode of action of
recreational quality of European lakes: minimizing the health risks from algal blooms Phoslock® in the control of phosphorus release from the bed sediments in a shallow
through phosphorus control. J. Appl. Ecol. 50, 315–323. https://doi.org/10.1111/ lake (Loch Flemington, UK). Water Res. 47, 4460–4473. https://doi.org/10.1016/j.
1365-2664.12059. watres.2013.05.017.
Chorus, I., Falconer, I.R., Salas, H.J., Bartram, J., 2000. Health risks caused by freshwater Moos, M.T., Taffs, K.H., Longstaff, B.J., Ginn, B.K., 2014. Establishing ecological reference
cyanobacteria in recreational waters. J. Toxicol. Environ. Health Part B 3, 323–347. conditions and tracking post-application effectiveness of lanthanum-saturated ben-
https://doi.org/10.1080/109374000436364. tonite clay (Phoslock®) for reducing phosphorus in aquatic systems: an applied pa-
Cooke, G.D., Welch, E.B., Peterson, S.A., Nichols, S.A., 2005. Restoration and leolimnological approach. J. Environ. Manage. 141, 77–85. https://doi.org/10.1016/
Management of Lakes and Reservoirs, 3rd ed. CRC Press, Taylor & Francis Group, j.jenvman.2014.02.038.
Boca Raton. Mur, L.R., Skulberg, O.M., Utkilen, H., 1999. Cyanobacteria in the environment. In:
Conley, D.J., Paerl, H.W., Howarth, R.W., Boesch, D.F., Seitzinger, S.P., Havens, K.E., Chorus, I., Bartram, J. (Eds.), Toxic Cyanobacteria in Water: A Guide to Their Public
Lancelot, C., Likens, G.E., 2009. Controlling eutrophication: nitrogen and phos- Health, Consequences, Monitoring and Management. St. ent. St. Edmundsbury Press,
phorus. Science 323, 1014–1015. https://doi.org/10.1126/science.1167755. Suffolk, pp. 15-40.
Copetti, D., Finsterle, K., Marziali, L., Stefani, F., Tartari, G., Douglas, G., Reitzel, K., Nürnberg, G.K., 2009. Assessing internal phosphorus load— problems to be solved. Lake
Spears, B.M., Winfield, I.J., Crosa, G., D'haese, P., Yasseri, S., Lürling, M., 2016. Reserv. Manage. 25, 419–432. https://doi.org/10.1080/00357520903458848.
Eutrophication management in surface waters using lanthanum modified bentonite: a Nürnberg, G.K., LaZerte, B.D., 2016. Trophic state decrease after lanthanum-modified
review. Water Res. 97, 162–174. https://doi.org/10.1016/j.watres.2015.11.056. bentonite (Phoslock) application to a hyper-eutrophic polymictic urban lake fre-
Douglas, G.B., Adeney, J.A., Robb, M., 1999. A novel technique for reducing bioavailable quented by Canada geese (Branta canadensis). Lake Reserv. Manage. 32, 74–88.
phosphorus in water and sediments. International Association Water Quality https://doi.org/10.1080/10402381.2015.1133739.
Conference on Diffuse Pollution 517–523. Nürnberg, G.K., 2017. Attempted management of cyanobacteria by Phoslock (lanthanum-
Douglas, G.B., 2002. US Patent 6350383: remediation material and remediation process modified clay) in Canadian lakes: water quality results and predictions. Lakes Reserv.
for sediments. Alexandria, VA: U.S. Patent and Trademark Office. Manage. 33, 163–170. https://doi.org/10.1080/10402381.2016.1265618.
Downing, J.A., Watson, S.B., McCauley, E., 2001. Predicting Cyanobacteria dominance in Nusch, E.A., 1980. Comparison of different methods for chlorophyll and pheopigment
lakes. Can. J. Fish. Aquat. Sci. 58, 1905–1908. https://doi.org/10.1139/f01-143. determination. Arch. Hydrobiol. Beih. Ergeb. Limnol. 14, 14–36.
Epe, T.S., Finsterle, K., Yasseri, S., 2017. Nine years of phosphorus management with OECD- Organization for Economic Co‐Operation and Development, 1982. Eutrophication
lanthanum modified bentonite (Phoslock) in a eutrophic, shallow swimming lake in of waters; monitoring, assessment and control. OECD Paris, 154 pp.
Germany. Lake Reserv. Manage. 33, 1–11. https://doi.org/10.1080/10402381.2016. Oksanen, J., Kindt, R., Legendre, P., O’Hara, B., Simpsons, G.L., Solymos, P., Stevens, M.

10
B. Barçante, et al. Ecological Indicators 117 (2020) 106702

H.H., Wagner, H., 2013. The vegan package. http://cran.r-project.org/, http://vegan. Schindler, D.W., Carpenter, S.R., Chapra, S.C., Hecky, R.E., Orihel, D.M., 2016. Reducing
r-forge.r-project.org/. phosphorus to Curb Lake eutrophication is a success. Environ. Sci. Technol. 50,
Paerl, H.W., Otten, T.G., 2013. Harmful cyanobacterial blooms: causes, consequences, 8923–8929. https://doi.org/10.1021/acs.est.6b02204.
and controls. Microb. Ecol. 65: 995-1010. https://doi.org/10.1007/s00248-012- Silva, T.F.G., Giani, A., Figueredo, C.C., Viana, P.S., Khac, V.T., Lemaire, B.J., Tassin, B.,
0159-y. Nascimento, N.O., Vinçon-Leite, B., 2016. Comparison of cyanobacteria monitoring
Paerl, H.W., 2018. Mitigating toxic planktonic cyano.bacterial blooms in aquatic eco- methods in a tropical reservoir by in vivo and in situ spectrofluorometry. Ecol. Eng.
systems facing increasing anthropogenic and climatic pressures. Toxins 10, 1–16. 94, 79–87. https://doi.org/10.1016/j.ecoleng.2016.06.037.
https://doi.org/10.3390/toxins10020076. Silva, T., Vinçon-Leite, B., Lemaire, B.J., Petrucci, G., Giani, A., Figueredo, C.C.,
Pinto-Coelho, R.M., Moura, R.T., Moreira, A., 1997. Zooplankton and bacteria contribu- Nascimento, N.O., 2019. Urban stormwater runoff impacts on cyanobacteria dy-
tion to phosphorus and nitrogen internal cycling in a tropical and eutrophic reservoir: namics in a tropical urban lake. Water 11, 1–28. https://doi.org/10.3390/
Pampulha Lake. Brazil. Int. Revue ges. Hydrobiol. 82, 185–200. https://doi.org/10. w11050946.
1002/iroh.19970820206. Smith, V.H., 2003. Eutrophication of freshwater and coastal marine ecosystems: a global
Pinto-Coelho, R.M., 1998. Effects of eutrophication on seasonal patterns of mesozoo- problem. Environ. Sci. Pollut. Res. 10, 126–139. https://doi.org/10.1065/espr2002.
plankton in a tropical reservoir: a 4-year study in Pampulha Lake. Brazil. Freshwater 12.142.
Biol. 40, 159–173. https://doi.org/10.1046/j.1365-2427.1998.00327.x. Smith, V.H., Schindler, D.W., 2009. Eutrophication science: where do we go from here?
Ptacnik, R., Lepistö, L., Willén, E., Brettum, P., Andersen, T., Rekolainen, S., Lyche Trends Ecol. Evol. 24, 201–207. https://doi.org/10.1016/j.tree.2008.11.009.
Solheim, A., Carvalho, A., 2008. Quantitative responses of lake phytoplankton to Smith, V.H., Wood, S.A., McBride, C.G., Atalah, J., Hamilton, D.P., Abell, J., 2016.
eutrophication in Northern Europe. Aquat. Ecol. 42, 227–236. https://doi.org/10. Phosphorus and nitrogen loading restraints are essential for successful eutrophication
1007/s10452-008-9181-z. control of Lake Rotorua, New Zealand. Inland Waters 6 (2), 273–283. https://doi.
R Development Core Team. 2013. R Development Core Team R: a language and en- org/10.5268/IW-6.2.998.
vironment for statistical computing. R Foundation for Statistical Computing, Vienna, Søndergaard, M., Jensen, P.J., Jeppesen, E., 2003. Role of sediment and internal loading
Austria (2013). http://www.R-project.org. of phosphorus in shallow lakes. Hydrobiologia 506, 135–145. https://doi.org/10.
Reynolds, C.S., Oliver, R.L., Walsby, A.E., 1987. Cyanobacterial dominance: the role of 1023/B:HYDR.0000008611.12704.dd.
buoyancy regulation in dynamic lake environments. New Zeal. J. Mar. Fresh. 21, Spears, B.M., Carvalho, L., Perkins, R., Kirika, A., Paterson, D.M., 2012. Long-term var-
379–390. https://doi.org/10.1080/00288330.1987.9516234. iation and regulation of internal phosphorus loading in Loch Leven. Hydrobiologia
Reynolds, C.S., 2006. Ecology of Phytoplankton. Cambridge University Press, Cambridge. 681, 23–33. https://doi.org/10.1007/s10750-011-0921-z.
https://doi.org/10.1017/CBO9780511542145. Spears, B.M., Mackay, E.B., Yasseri, S., Gunn, I.D.M., Waters, K.E., Andrews, C., Cole, S.,
Robb, M.S., Greenop, B., Goss, Z., Douglas, G., Adeney, J., 2003. Application of Phoslock, De Ville, M., Kelly, A., Meis, S., Moore, A.L., Nürnberg, G.K., van Oosterhout, F., Pitt,
an innovative phosphorus binding clay, to two Western Australian waterways: pre- J.A., Madgwick, G., Woods, H.J., Lürling, M., 2016. A meta-analysis of water quality
liminary findings. Hydrobiologia 494, 237–243. https://doi.org/10.1023/ and aquatic macrophyte responses in 18 lakes treated with lanthanum modified
A:1025478618611. bentonite (Phoslock®). Water Res. 97, 111–121. https://doi.org/10.1016/j.watres.
Ross, G., Haghseresht, F., Cloete, T.E., 2008. The effect of pH and anoxia on the perfor- 2015.08.020.
mance of Phoslock, a phosphorus binding clay. Harmful Algae 7, 545–550. https:// Torres, I.C., Resck, R., Pinto-Coelho, R.M., 2007. Mass balance estimation of nitrogen,
doi.org/10.1016/j.hal.2007.12.007. carbon, phosphorus and total suspended solids in the urban eutrophic Pampulha
Rott, E., 1981. Some results from phytoplankton counting intercalibrations. Schweiz. Z. reservoir, Brazil. Acta Limnol. Bras. 19, 79–91.
Hydrol. 43, 34–62. https://doi.org/10.1007/BF02502471. Utermöhl, H., 1958. Zur vervollkommung der quantitativen phytoplankton-methodik.
Sas, H., 1989. Lake restoration by reduction of nutrient loading: expectations, experi- Mitt. Int. Ver. Theor. Angew. Limnol. 9, 1–38. https://doi.org/10.1080/05384680.
ences, extrapolations. St. Augustin: Academia Verlag Richarz 497, pp. https://doi. 1958.11904091.
org/10.1080/03680770.1989.11898731. Yamada-Ferraz, T.M., Sueitt, A.P.E., Oliveira, A.F., Botta, C.M.R., Fadini, P.S.,
Schindler, D.W., Hecky, R.E., Findlay, D.L., Stainton, M.P., Parker, B.R., Paterson, M., Nascimento, M.R.L., Faria, B.M., Mozeto, A.A., 2015. Assessment of Phoslock® ap-
Beaty, K.G., Lyng, M., Kasian, S.E.M., 2008. Eutrophication of lakes cannot be con- plication in a tropical eutrophic reservoir: an integrated evaluation from laboratory
trolled by reducing nitrogen input: results of a 37 year whole ecosystem experiment. to field experiments. Environ. Technol. Innov. 4, 194–205. https://doi.org/10.1016/
PNAS 105, 11254–11258. https://doi.org/10.1073/pnas.0805108105. j.eti.2015.07.002.

11