Water 2021, 13, 2463. https://doi.org/10.3390/w13182463 www.mdpi.com/journal/water
Cyanobacteria and Cyanotoxins in a Changing Environment:
Concepts, Controversies, Challenges
*, Jutta Fastner
and Martin Welker
German Environment Agency, Schichauweg 58, 12307 Berlin, Germany; firstname.lastname@example.org
Independent Researcher, 10117 Berlin, Germany; email@example.com
* Correspondence: firstname.lastname@example.org
Abstract: Concern is widely being published that the occurrence of toxic cyanobacteria is increasing
in consequence of climate change and eutrophication, substantially threatening human health. Here,
we review evidence and pertinent publications to explore in which types of waterbodies climate
change is likely to exacerbate cyanobacterial blooms; whether controlling blooms and toxin concen-
trations requires a balanced approach of reducing not only the concentrations of phosphorus (P) but
also those of nitrogen (N); how trophic and climatic changes affect health risks caused by toxic cya-
nobacteria. We propose the following for further discussion: (i) Climate change is likely to promote
blooms in some waterbodies—not in those with low concentrations of P or N stringently limiting
biomass, and more so in shallow than in stratified waterbodies. Particularly in the latter, it can work
both ways—rendering conditions for cyanobacterial proliferation more favourable or less favoura-
ble. (ii) While N emissions to the environment need to be reduced for a number of reasons, control-
ling blooms can definitely be successful by reducing only P, provided concentrations of P can be
brought down to levels sufficiently low to stringently limit biomass. Not the N:P ratio, but the ab-
solute concentration of the limiting nutrient determines the maximum possible biomass of phyto-
plankton and thus of cyanobacteria. The absolute concentrations of N or P show which of the two
nutrients is currently limiting biomass. N can be the nutrient of choice to reduce if achieving suffi-
ciently low concentrations has chances of success. (iii) Where trophic and climate change cause
longer, stronger and more frequent blooms, they increase risks of exposure, and health risks depend
on the amount by which concentrations exceed those of current WHO cyanotoxin guideline values
for the respective exposure situation. Where trophic change reduces phytoplankton biomass in the
epilimnion, thus increasing transparency, cyanobacterial species composition may shift to those that
reside on benthic surfaces or in the metalimnion, changing risks of exposure. We conclude that
studying how environmental changes affect the genotype composition of cyanobacterial popula-
tions is a relatively new and exciting research field, holding promises for understanding the biolog-
ical function of the wide range of metabolites found in cyanobacteria, of which only a small fraction
is toxic to humans. Overall, management needs case-by-case assessments focusing on the impacts
of environmental change on the respective waterbody, rather than generalisations.
Keywords: cyanobacteria; cyanobacterial toxins; climate change; eutrophication; health risk
A large and increasing number of publications is voicing the concern that toxic cya-
nobacterial blooms will increase in consequence of climate change and eutrophication,
and that this will substantially threaten human health [1–3]. In response to accelerated
waterbody eutrophication in the second half of the 20th century, cyanobacterial blooms
became increasingly common, and their detrimental effects on water quality moved into
the focus of public and scientific attention. In parallel, research on phytoplankton moved
Citation: Chorus, I.; Fastner, J.;
Welker, M. Cyanobacteria and
Cyanotoxins in a Changing
Water 2021, 13, 2463.
Academic Editor: Maria
Received: 08 July 2021
Accepted: 01 September 2021
Published: 07 September 2021
Publisher’s Note: MDPI stays neu-
tral with regard to jurisdictional
claims in published maps and insti-
Copyright: © 2021 by the authors. Li-
censee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and con-
ditions of the Creative Commons At-
tribution (CC BY) license (http://crea-
Water 2021, 13, 2463 2 of 41
from a descriptive botanical focus towards concepts for understanding drivers of the oc-
currence and dominance of species. For this, laboratory experiments on resource limita-
tion differentiated between uptake rates limited by concentrations of dissolved nutrients
(i.e., Monod kinetics), rates of cell division limited by nutrient cell quotas (i.e., Droop ki-
netics) and total bioavailable nutrient fractions limiting the maximally attainable biomass
in situ, leading to the concept of carrying capacity in its limnological interpretation .
This fed into the other important focus on phytoplankton populations in waterbodies:
understanding the mechanisms of competition between species that determine occurrence
The understanding of cyanobacterial toxins was pioneered by a small number of re-
search groups in the 1960s and 1970s who elucidated the chemical structures of different
toxin groups. By 1992, a variety of toxins was identified, in particular, hepatotoxic pep-
tides (microcystins), neurotoxic alkaloids (anatoxins and saxitoxins), protein synthesis in-
hibiting alkaloids (cylindrospermopsins) and a neurotoxic organophosphate (anatoxin-
a(S), syn. guanitoxin ). Cyanotoxin research gained momentum in the 1990s as the in-
creasing availability of powerful methods of chemical analysis enabled wider surveys of
cyanotoxin occurrence. This led to the recognition that the majority of blooms include
globally occurring toxic species, and since the 1990s cyanobacterial blooms have been re-
ceiving massively increased research attention .
Cyanobacterial blooms continue to increase in response to eutrophication where ag-
riculture is intensified or urban settlements expand without sufficient wastewater treat-
ment. Elsewhere efforts to curb and reduce eutrophication began in the 1970s, focusing on
phosphorus (P). Measures to reduce P emissions have increasingly been implemented,
resulting in declining P concentrations in many waterbodies  and in some waterbodies
eventually to less phytoplankton biomass and less cyanobacterial blooms . Trophic
change, i.e., rapid (anthropogenic) change in nutrient concentrations, is ongoing in both
directions , eutrophication in many waterbodies and but also ‘oligotrophication’ in oth-
Regarding cyanotoxins, we now have a general, albeit incomplete, picture of their
occurrence in relation to cyanobacterial taxa, geography, and environmental conditions
(reviewed in ). We also know that the toxicity of field populations primarily depends
on their composition of toxic and non-toxic genotypes. Yet, our knowledge on conditions
driving the occurrence of toxic genotypes is still sparse, and hence our understanding of
the regulation of in situ toxin occurrence is quite limited.
Moreover, the currently known cyanotoxins are only a small fraction of (bio)chemi-
cally similar metabolites produced by cyanobacteria, and their function for the producing
cells is yet unclear [11,12]. Toxicity to vertebrates is an evolutionary coincidence rather
than an evolutionary adaptation. The implications of cyanotoxins for human health have
been a key motivation for the study of cyanotoxins, and for the target of protecting public
health this is undoubtedly justified. However, the focus on metabolites which happen to
be toxic to humans but represent only a small fraction of the cyanobacterial metabolites
hampers a better understanding of cyanobacterial biology and biochemistry.
More recently, a focus of research has been on understanding how trophic change
and climate change act together to drive the occurrence of potentially toxic cyanobacteria,
and the concern about health risks from their toxins has further promoted this line of re-
search. As a result, some generalisations are increasingly being proposed, in particular,
that climate change will exacerbate cyanobacterial blooms, that controlling blooms and
toxin concentrations requires a balanced approach of reducing not only the concentrations
of phosphorus (P) but also those of nitrogen (N), and also that the occurrence of cyano-
toxins is a ‘very serious’ threat to human health.
In the following review, we address these generalisations, focusing on those concepts
which are still subject to controversy and on newly arising concerns. We ask three ques-
tions and assess the respective evidence on the basis of both theoretical considerations and
Water 2021, 13, 2463 3 of 41
In which types of waterbodies will climate change promote blooms (Section 2)?
Can controlling blooms by focusing on the reduction of P loads be successful without
also reducing N, and where eutrophication does decline, can this cause other risks by
merely shifting cyanobacterial populations (Section 3)?
How do trophic and climatic changes affect health risks from cyanobacteria (Section
For each of these points, we discuss how more recent views relate to earlier para-
digms in limnology, proposing hypotheses for further discussion. Our aim is differentia-
tion between well-substantiated generalisations and a need for situation-specific answers.
2. In Which Type of Waterbody Will Climate Change Promote Blooms?
Publications of the last two decades have increasingly addressed the potential for
climate change, particularly warming, to influence cyanobacterial proliferation and
blooms (reviewed, e.g., in [13,14]). While warming is frequently proposed to exacerbate
cyanobacterial blooms, temperature itself is not a resource such as nutrients or light, and
hence some important effects of changing temperatures on cyanobacterial blooms are in-
direct. The competitive advantage of cyanobacteria at high temperatures is not the only—
and in many cases not the most relevant—response. Climate change impacts complex eco-
logical systems in many ways, with some of the demonstrated and expected mechanisms
causing conditions to be more favourable for cyanobacteria while others can have the op-
posite effect. In Section 2 we discuss both possibilities and attempt to narrow down the
type of waterbody in which climate change is more likely to increase duration and/or in-
tensity of cyanobacterial blooms.
2.1. Impact of Warming on Cyanobacteria and Microcystin Concentrations
Growth rates of cyanobacteria are generally lower than those of many eukaryotic
phytoplankton species, and this is one of the mechanisms explaining why they often attain
dominance only later in the growing season. The growth rates of some cyanobacterial
taxa, particularly of Microcystis, show particularly steep curves of temperature-depend-
ence: Coles and Jones  show this for three species of cyanobacteria relative to one dia-
tom species. The relationships of growth rates to temperature presented by Reynolds 
are still a popular reference for higher growth rates of cyanobacteria at higher tempera-
ture; however, they actually show a substantially steeper increase only for Microcystis, and
at all temperatures the growth rates for cyanobacteria are lower than those for the six algal
species (including three species of diatoms). The key message from this figure is that
growth rates of these cyanobacteria still increase up to almost 30 °C (and for one strain
even up to 40 °C) while for the eukaryotic algae they are generally higher, but increase
only up to 20 °C. This is in line with the results of Butterwick et al.  who show that
four out of six species of cyanobacteria still grow at 30 °C, while this was the case for only
6 out of 15 eukaryotic phytoplankton species. It is also in line with the results of Thomas
and Litchman  who show optimum growth rates of 12 strains of Microcystis, Raphidi-
opsis (syn. Cylindrospermopsis) and Dolichopsermum (syn. Anabaena) in the range of 27–37
°C. Paerl and Otten  compile data from nine studies showing maximum growth rates
of Cyanobacteria between 25 and 35 °C.
Regarding the more complex mechanisms determining species composition in the
field, using microcosm experiments and modelling de Senerpont Domis et al.  show
that although growth rates of cyanobacteria responded more strongly to warming than
those of other species, this did not affect seasonal succession from diatoms over chloro-
phytes to cyanobacteria. Comparing growth rates of three strains of native cyanobacterial
species to three strains of invasive species from warmer climates, Mehnert et al.  found
that while the native ones can build populations earlier in the season, there is a potential
for the invasive Nostocales, particularly Chrysosporum (syn. Aphanizomenon) ovalisporum,
to outcompete them later in the season. Using the example of Raphidiopsis raciborskii, these
Water 2021, 13, 2463 4 of 41
authors add a further consideration: to unfold maximum growth rates at high tempera-
tures, even shade tolerant species need high light intensity. In highly eutrophic and thus
turbid waterbodies, this may counteract their competitive advantage given by high tem-
A further direct impact of warming postulated by some authors is that higher tem-
peratures favour microcystin-producing strains of Microcystis over non-producing ones.
Lei et al.  discuss respective publications (see also Section 4.1), while for the results of
their own co-culture experiments they do not consider microcystin production the key
mechanism leading to dominance. Ninio et al.  show that in laboratory experiments
the Microcystis strain with a low MC content outcompetes the one with a high MC content
particularly at high temperatures (20 and 25 °C), and this is in line with their field obser-
vation of the non-toxic strain increasingly dominating in Lake Kinneret in the wake of
warming. These authors propose slight changes in temperature to be decisive for the out-
come of competition between strains and suggest that there is no per se advantage of MC-
producing over non-producing genotypes at elevated temperature; rather, this may vary
between strains. Ninio et al.  conclude that “this widely accepted model for the expansion
and persistence of Microcystis blooms should be re-evaluated as higher temperatures may affect not
just the intensity of the bloom, but also the composition of the Microcystis population, species,
strains and genotypes”. As of now, a general claim of elevated temperatures favouring mi-
crocystin-producing cyanobacteria does not appear to be well evidenced by data.
2.2. Impact of Climate Change on Conditions in Waterbodies
Mechanisms through which overall climate change can shape the environmental con-
ditions in waterbodies are manyfold and begin with the catchment. Heavy precipitation
can cause nutrient input pulses from the catchment through soil erosion and run-off, but
it can also dilute nutrient concentration in a waterbody. Conditions acting within a water-
body include impacts on duration and stability of thermal stratification, frequency of mix-
ing events caused by storms, changes in external and internal nutrient loads and changes
in photon flux density (via many mechanisms affecting mixing and turbidity). Cotting-
ham et al.  emphasise the effect of changes in stratification patterns on the duration of
cyanobacterial life stages in the pelagic versus the benthic zones and call for research to
elucidate how these impact on population development. While elevated atmospheric CO
concentrations affect global temperatures, they can also influence the strain composition
of cyanobacterial populations and thus toxin concentrations through increasing concen-
trations of dissolved CO
, particularly in eutrophic and hypertrophic waterbodies
. Shuvo et al.  comprehensively discuss the wide range of interacting variables that
influence the extent of climate change effects. Besides temperature, this includes lake ba-
sin morphology and orientation, hydrology and trophic state but also solar irradiation
and cloud cover.
For the direct impact of hydrophysical conditions, Reichwaldt and Ghadouani 
give an overview of case reports in which rainfall events in lakes and reservoirs triggered
an increase or decrease of cyanobacterial biomass. These authors quote increased turbidity
and high turbulence as potential drivers of change—in some cases for a decrease and in
others for an increase of blooms. They also emphasise the rather small number of studies
published on the impact of rainfall patterns as well as the complexity of the mechanisms
involved and list “flushing, nutrient input, reduction of conductivity (especially after long dry
periods), and mixing of the water column due to water inflow and associated strong winds” as
mechanisms by which rainfall events shape the environment to favour or disfavour cya-
nobacteria in phytoplankton communities. Huisman et al.  point to the relevance of
time patterns of extreme events: where nutrient pulses from storms follow periods of
drought, they are most likely to feed cyanobacterial proliferation if these are nutrient-lim-
ited. Yang et al.  show this with an 18-year time series of data from Lake Taihu, China,
where high wind speeds and strong rainfall caused blooms to increase particularly since
2012, once nutrient concentrations had declined to levels potentially limiting biomass (i.e.,
Water 2021, 13, 2463 5 of 41
TP <80 µg/L, with TN still in the range of 1900–2400 µg/L). With a time series analysis
from 2000 to 2014 for the large, deep pre-alpine Lago Maggiore, Morabito et al.  show
that over longer time spans (months) pronounced rain events dilute rather than enrich
nutrients; nonetheless within short time spans (5–7 days) cyanobacterial blooms showed
some (albeit not significant) correlation to rainfall. Furthermore, soil characteristics in the
catchment, such as iron content, have an impact on P loads following storm events, and
Tang et al.  conclude that an increase in extreme precipitation events may substantially
mitigate eutrophication in some waterbodies while worsening the conditions in others.
Phytoplankton cells entrained in the turbulence of an epilimnion or shallow water-
body experience these hydrophysical changes in terms of exposure to light intensity and
light–dark cycles and, where not only wind but also rainfall is involved, also in terms of
nutrient pulses or dilution. The frequency at which waterbody conditions change can be
an important consequence of climate change, as disturbances at intermediate time scales
not only can disrupt cyanobacterial blooms, but generally increase phytoplankton diver-
sity. Litchman et al.  analyse how the traits of different functional groups of phyto-
plankton determine their responses to these environmental changes and outline how the
complex interplay of these mechanisms may be modelled. Anneville et al.  emphasise
the challenge of studying the impact of extreme events in the field, as these vary consid-
erably between years, possibly more so than the carrying capacity given by nutrient con-
Publications on the impact of climate change on phytoplankton follow two different
approaches: some evaluate data across a larger number of waterbodies for associations
between blooms and elevated temperatures, applying multivariate statistics or neural net-
works [27,35,36]. Associations thus found can provide valuable indicative information
particularly where different data sets and statistical approaches show similar results.
However, in face of the different mechanisms acting in different types of waterbodies the
results are sensitive to the choice of waterbodies and therefore may not well support gen-
eralisations. A smaller number of publications has studied time series of data from indi-
vidual waterbodies to assess the impact of extreme events such as heat waves [30,34,37],
extreme winters , rainfall patterns  or reduced mixing [39,40]. While many studies
conclude that climate warming will exacerbate cyanobacterial blooms, some (discussed
below) show the opposite. Key to understanding contradictory evidence is an understand-
ing of the possible mechanisms and hierarchy of conditions determining the access of phy-
toplankton cells to nutrients and light. In the following, we propose that predictions about
the impact of climate change are most reliable for nutrient-poor and for shallow water-
2.2.1. Climate Change Is Not Likely to Exacerbate Blooms in Oligotrophic and
Evaluating a large data set (2561 lakes across seven countries), Shuvo et al.  found
TP to be the main single predictor of phytoplankton biomass (measured as Chl-a concen-
tration), accounting for 42% of the variation, but spring and summer weather (tempera-
ture, precipitation and cloud cover) were almost equally relevant for the concentrations
of Chl-a, while physical characteristics (residence time, mean depth, surface area and ele-
vation) together accounted for 20% of the variation. However, this evaluation did not dif-
ferentiate lakes by trophic status, and these percentages are specific to this data set. Where
concentrations of the macronutrients N and P strongly limit total phytoplankton biomass,
dominance of cyanobacteria is not likely and if it does occur, the low concentrations of N
and P will not provide the capacity for the development of significant blooms (see Section
3.1.3). Rigosi et al.  analysed data of >1000 North American lakes classified by trophic
state and show that in oligotrophic lakes cyanobacterial biomass related more strongly to
nutrients, while in mesotrophic lakes temperature showed a stronger impact. In eutrophic
Water 2021, 13, 2463 6 of 41
and hyper-eutrophic lakes these authors observed a significant interaction between nutri-
ents and temperature. Further, they found that whether nutrients or temperature—or
both—promoted growth of cyanobacteria to be taxon-specific.
Elliott et al.  confirm this with a modelling study simulating a summer bloom of
Dolichospermum which showed that while temperature promoted the growth of this taxon,
the effect was much smaller than that of changes in the nutrient load. Using satellite data
reflecting lake surface temperature and concentrations of Chl-a from 188 lakes, Kraemer
et al.  found that in phytoplankton-rich lakes warmer conditions boosted Chl-a while
this was not the case in phytoplankton-poor lakes. Gallina et al.  studied the impact of
extreme weather events on 5 oligo- to moderately eutrophic, deep, stratified peri-alpine
lakes and found that warm summers did not lead to dominance of cyanobacteria: while
their biomass increased during hot periods, their contribution to total phytoplankton in-
creased only slightly, and the most striking response was a reduction in their species di-
versity. Anneville et al.  found that extremely warm summers and autumns did not
lead to cyanobacterial blooms in oligotrophic alpine Lake Annecy whereas in the meso-
trophic lakes Bourget and Geneva, they did increase, although not in summer but only in
autumn. These authors also conclude that milder autumn temperatures benefit not only
cyanobacteria, but other phytoplankton taxa as well, and this only if nutrient concentra-
tions are sufficiently high. In shallow subtropical Peri Lake (southern Brazil) with rather
low nutrient concentrations (5–20 µg/L TP), rainfall diluted TP rather than adding loading
to the lake because import from land in this conserved watershed was low .
Climate change therefore is not likely to have much impact on bloom formation in
oligotrophic waterbodies. Nonetheless, in very large oligotrophic waterbodies transient
surface films can recruit themselves from the large water volume and, under certain con-
ditions, accumulate to dense scums with potentially hazardous concentrations of cyano-
toxins . Further exceptions, discussed in Section 3.3, include cyanobacteria that grow
on surfaces (including macrophytes), as well as metalimnetic layers in deep, stratified
lakes, particularly Planktotrhix rubescens.
2.2.2. In Meso- and Eutrophic Waterbodies, Climate Change Can Act Both Ways: Exacer-
bating or Reducing Blooms
Where both P and N are ample or available beyond the amounts that phytoplankton
can utilise to form more biomass, the impacts of climate change will act primarily through
changing hydrophysical conditions. For example, in hypertrophic Nieuwe Meer with nu-
trient concentrations well above limiting phytoplankton biomass, the competitive ad-
vantage of Microcystis at high temperature and low mixing proved evident . Mecha-
nisms exacerbating blooms include the earlier onset of and prolongation of the growing
season, giving cyanobacterial populations more time to outcompete other phytoplankton.
Additionally, higher stability of stratification caused by earlier warming gives them a
competitive advantage by allowing more effective vertical migration.
In contrast, however, pronounced stability of thermal stratification reduces upward
diffusion of nutrients from the hypolimnion to the epilimnion and erosion of the metalim-
nion; it can trap nutrients in the hypolimnion as particles sink down from the epilimnion
and are mineralised. Wentzky et al.  show this for Rappbode Reservoir, Germany:
while the increase of the surface water temperature reflected that of air temperatures, the
temperature in the hypolimnion showed a slight decrease, attributed to the significantly
earlier onset and higher stability of thermal stratification. For Lake Garda, Salmaso et al.
 show that warming induced incomplete mixing, thus causing “climate warming-in-
duced oligotrophication” of the epilimnion, reducing the biomass of Planktothrix. Verburg et
al.  describe a similar observation in tropical Lake Tanganyika, quoting reduced ver-
tical mixing due to more pronounced water density gradients also for Lake Malawi.
Independently of trophic status, intermittent rainfall and storms, particularly if in-
tensive, can disrupt blooms , re-setting seasonal species succession to an earlier stage.
Water 2021, 13, 2463 7 of 41
Using a 40-year data series, Posch et al.  show the pronounced impact of weather-
induced variations in stratification on the biomass of Planktothrix rubescens in Lake Zürich.
2.2.3. Climate Change Is More Likely to Enhance Cyanobacterial Blooms in Shallow
Than in Thermally Stratified Waterbodies
The impact of climate-driven hydrophysical conditions on the sediment–water inter-
face tends to be more pronounced in shallow than in stably stratified waterbodies. One
mechanism is phosphorus release from sediments: in shallow water warming reaches the
sediment surface much more strongly, enhancing the mineralisation rate of organic matter
and thus directly releasing P . Hupfer et al.  show that more rapid mineralisation
may, in turn, speed up the development of anoxic zones on the sediment surface, thus
promoting release of redox-sensitively adsorbed P (provided this is a significant fraction
of P in the sediment).
A further mechanism is sediment resuspension by storms and heavy winds in shal-
low water, even more so if increased drought reduces their water level, and the minerali-
sation of detritus thus resuspended provides pulses of P which can feed blooms. Tam-
meorg et al.  demonstrate the relevance of stronger wind and wave action in Lake
Peipsi, resuspending the seston that formed during the previous summer weeks which is
readily biodegradable: it releases P that supplies late summer cyanobacterial population
growth. In a modelling study for Lake Tegel, in which stratification is not very stable,
Chorus and Schauser  show that storms occurring in late spring/early summer can
prevent thermal stratification from developing at all. Thus, for the same lake, climate
change can lead to more stable stratification in some years and destabilisation in others,
depending on time patterns of warming and storms. Analysing data from 143 shallow
lakes in Europe and South America, Kosten et al.  show that with increasing tempera-
ture the percentage that cyanobacteria contribute to total phytoplankton biomass in-
creases steeply. These authors conclude that in a warmer climate controlling cyanobacte-
rial dominance will require more pronounced nutrient reduction.
2.3. Conclusions for the Impact of Climate Change on Waterbody Conditions Favouring Blooms
A hierarchy of conditions determines the success of cyanobacteria: while nutrients
“set the overall stage”, weather events shape the more immediate reactions. They can en-
hance the frequency, duration and intensity of cyanobacterial blooms only within the
frame given by the carrying capacity of resources for phytoplankton biomass, and in oli-
gotrophic waterbodies this is low. This also applies to shallow waterbodies, provided they
are oligotrophic, but due to the generally higher relevance of sediment–water contact, this
is less likely to be the case. Thus, waterbody morphology can be seen as a secondary var-
iable determining how climate change can affect the dominance and biomass of cyano-
Of course, there are exceptions to such a general pattern. Here we did not address
the possible impacts of changes on food webs, in particular on grazing (discussed in
Winder and Sommer  and Moustaka-Gouni and Sommer ) because of their rela-
tively low relevance to cyanobacteria. However, Selmeczy et al.  describe the fascinat-
ing case of formerly oligotrophic, deep and stratified Lake Stechlin which is developing
towards eutrophication without any recognisable external nutrient input, possibly be-
cause a change in hydrophysical conditions was induced by a climatic anomaly, leading
to a chain reaction of shifts in the biota which in turn triggered release of P stored in the
The occasionally voiced perception of a “scientific consensus” that climate change
will exacerbate cyanobacterial blooms is challenged by the diversity of waterbody condi-
tions driving responses to change. A key conclusion that Richardson et al.  draw from
their analysis of data from 494 lakes in central and northern Europe is the pronounced
variability of cyanobacterial biomass relative to gradients in temperature, even though
these authors differentiated their data by lake types (alkalinity, colour, mixing).
Water 2021, 13, 2463 8 of 41
Reichwaldt and Ghadouani  expect a strong impact of climate change but emphasise
that it will be “very complex and will strongly depend on site-specific dynamics, cyanobacterial
species composition, and cyanobacterial strain succession”. Furthermore, Mooij et al.  warn
that the “reuse of eutrophication models for studying climate change is a logical step but should
be done with great care, because the validity of the outcomes has generally not yet been properly
tested against empirical data, and field studies show clear synergistic effects that are not well cov-
ered by existing models.“ From the research reviewed above we share the view that while
the evidence base for generalisations on the impact of climate change is still limited, for
the target of avoiding cyanobacterial blooms it is clearly prudent to strengthen the efforts
to mitigate eutrophication.
3. Concepts for Mitigating Eutrophication—Do They Need to Include N, and Can Re-
ducing Trophic State Risk Shifting Cyanobacteria to Metalimnetic or Benthic Habi-
Based on Liebig’s law of the limiting nutrient and the recognition of the relevance of
P in this role by Vollenweider  and others, many of the efforts to mitigate eutrophica-
tion of freshwaters have strongly focused on reducing P loads. This is also due to P being
easier to control: banning it from detergents almost halved loads from some domestic
wastewaters within a few months [58,59]. However, in most situations this alone is insuf-
ficient for controlling eutrophication, and implementing P removal in sewage treatment,
at minimum through simultaneous precipitation, is also necessary . While efficient P
removal is widely implemented , cost-effective and efficient procedures for N removal
are not yet equally established, largely for technical and economic reasons [62,63]. For
coastal waters the prevalence of N-limitation was recognised, and some regulations also
address N loads (e.g., those of the European Union and some intergovernmental river-
basin management treaties); however, generally, these efforts have been less successful:
where load reduction has occurred, it is substantially more pronounced for P than for N,
particularly for loads from agriculture (reviewed by Glibert et al. ). These authors also
show that in regions with rapidly developing and growing economies, N loads are in-
creasing dramatically, and that urea is increasingly replacing the ammonium-nitrate fer-
tilisers which are more challenging to handle because of being explosive. As cyanobacteria
more readily use reduced forms of N, this can promote their proliferation .
The more recent awareness of the relevance of forms of nitrogen for promoting cya-
nobacterial blooms  has led to a strong focus on the ratio of N:P in numerous publica-
tions, with high N:P-ratios being interpreted as indicating a need to also reduce the con-
centrations of N in the respective waterbody and often without explicit attention to the
absolute concentrations of either nutrient. This also includes experiments adding nutri-
ents to field samples, from which the same conclusion is frequently drawn if adding not
only P but also N enhances growth more than adding only N or only P. These approaches
appear to be in conflict with the state of knowledge on the mechanisms of nutrient limita-
tion established in the last decades as well as with the increasing amount of data showing
substantial decline of phytoplankton biomass, particularly that of cyanobacteria, once P
concentrations undercut threshold concentrations. Clarity on the most effective ap-
proaches to bloom control is, however, important for implementing measures to curb the
biomass they can attain in the respective waterbody. This includes the need to explicitly
differentiate between the limitation of nutrient uptake rates (which can limit growth rates)
by dissolved N and P, and the limitation of biomass (sometimes referred to as ‘yield’ or
‘standing stock’) by total N and P. The widely used term ‘growth limitation’ is unclear
and can imply either or both. It is also important to understand whether or not measures
in an upstream waterbody are likely to ‘displace eutrophication downstream’  and where
a ‘dual strategy’, addressing both nutrients [14,19], is the more effective way forward.
In this section we discuss the conclusions drawn from N:P ratios and nutrient addi-
tion experiments relative to the state of knowledge established in the last century. From
Water 2021, 13, 2463 9 of 41
this we draw conclusions for aspects that a ‘dual strategy’ needs to consider in order to
develop a differentiated approach, based on case-by-case assessments.
3.1. N:P Ratios and Growth Responses to Adding Both Nutrients Relative to the Absolute Con-
centrations of N and P
Particularly during the 1980s and 1990s, research in phytoplankton ecology focused
on elucidating the mechanisms of nutrient limitation. A key outcome was the paradigm
that not the N:P ratio, but the absolute concentration of the limiting nutrient determines
the maximum possible biomass of phytoplankton and thus of cyanobacteria. It follows
that whether N or P are currently limiting biomass can be inferred from their absolute
concentrations. We postulate that this paradigm still holds true, for the reasons discussed
in Sections 3.1.1–3.1.6. While shifting N:P ratios certainly can have downstream effects,
we also postulate that these require case-by-case assessments rather than a generic ‘dual
strategy’ (Section 3.1.6).
3.1.1. Evidence Regarding Dissolved Nutrient Fractions
Phytoplankton cells deplete of P, N or both will quickly incorporate dissolved nutri-
ents whenever they become available and use them for building new biomass (reviewed
in Salmaso and Tolotti ). If dissolved P is available in excess of the current need for
cell division, it can be stored as polyphosphate for up to four further divisions . Cya-
nobacteria are also capable of storing excess N in phycobiliproteins and, in particular, in
cyanophycin granules . The latter can constitute up to 18% of cellular dry weight .
Their formation is triggered by depletion of other resources (P, light or sulfate), and they
are catabolised to enter the nitrogen metabolism when other resources no longer limit
growth . Further, a number of bloom-forming cyanobacterial taxa can access the at-
mospheric N-pool through diazotrophy . However, in turbid, eutrophic waterbodies
N-fixation rarely contributes significantly to compensating N-limitation because this pro-
cess is energy-intensive [74,75].
To some extent cyanobacterial cells can up- or downregulate their uptake rate of dis-
solved P and do this in particular in response to pulses of dissolved nutrients (see Cáceres
et al. , on P transporters). Nonetheless, uptake rates are limited by the cells’ nutrient
uptake mechanisms, and above a saturation concentration they cannot further increase
uptake rates. However, species differ in their nutrient uptake kinetics, generally charac-
terised by maximum uptake rates and half-saturation constants, and these differences are
important for the outcome of competition between species under nutrient limiting condi-
tions or fluctuating nutrient concentrations . Concentrations above which uptake rates
reach saturation can be assumed for all of the phytoplankton species and are well estab-
lished for a number of taxa. Generally, there appear to be no species for which uptake
rates are not saturated if DIN is >100–130 µg/L and DIP is >3–10 µg/L [78–81]. In conse-
quence, nutrient limitation is given only in situations in which the concentrations of dis-
solved nutrients are below these saturating levels.
3.1.2. Evidence Regarding Total Nutrient Fractions
There is an upper limit to the maximally attainable biomass that can be sustained by
the ambient resource at a given point in time, established in phytoplankton ecology since
the 1990s as “carrying capacity”. Using the example of Rostherne Mere, Reynolds and
Bellinger  explains how resources alternate over the course of a year to set this upper
limit: in temperate climates this is typically light during winter, and in oligotrophic to
slightly eutrophic waterbodies limitation may shift to total P and/or total N during sum-
mer, once all dissolved nutrient is incorporated into biomass and detritus. At high nutri-
ent concentrations cell density may reach a point at which it causes such pronounced tur-
bidity that light availability limits any further increase of phytoplankton biomass .
Accordingly, a range of studies has shown threshold concentrations of TP above which
Water 2021, 13, 2463 10 of 41
higher concentrations do not lead to more biomass. Some of these studies compare annual
or seasonal means between many lakes [68,78,84–86]. Others follow the trophic develop-
ment of individual lakes (see case studies in Fastner et al.  and Salonen et al. ).
Taken together, these data show that in the plots of biomass versus TP concentrations,
biomass typically levels off at 50–100 µg/L TP (see also Quinlan et al. ). Of course, in
multi-lake studies the scatter is considerable, due to further specific conditions constrain-
ing growth such as background turbidity or losses through grazing and sedimentation. In
particular, the stability and depth of thermal stratification influences light limitation: the
TP-threshold at which biomass levels off can be higher in very shallow waterbodies and
lower in those with a deeply mixed epilimnion.
A TN-threshold concentration below which N is likely to limit further biomass in-
crease can be tentatively inferred from the threshold concentration for TP by applying the
Redfield Ratio of 16:1 as atomic ratio of N:P or as mass ratio of 7:1, as first introduced by
Redfield [90,91] from observations on marine phytoplankton. This ratio is still widely
used, although it is only an average, and more recent work, including on freshwater phy-
toplankton, has shown that atomic ratios of N:P range up to 45:1 (mass ratio 20:1), in ex-
treme cases even up to 133:1 (mass ratio 60:1) for the biomass of some (marine) species
. Thus, based on a TP-threshold of 50 µg/L, a TN threshold would amount to 350 µg/L
when using the Redfield Ratio and to 2000 µg/L using a TP-threshold of 100 µg/L and a
mass ratio of N:P = 20:1. This leads to TN concentrations above which N-limitation of
phytoplankton biomass becomes unlikely in the range of 350–2000 µg/L. In their evalua-
tion of 102 lakes in northern Germany, Dolman et al.  found biovolumes plotted
against TN did not level off up to a TN concentration of 2000 µg/L, while for more than
800 Danish lakes Søndergaard et al.  found that above 500 µg/L TN, biomass was
strongly dependent on TP and scarcely influenced by TN.
3.1.3. Evidence Regarding Cyanobacterial Blooms
TP-concentrations indirectly influence species composition because they limit the
maximum phytoplankton biomass and thus turbidity: where water stays clear because
low TP-concentrations cannot support a high amount of biomass, species that achieve a
high growth rate at high light intensity tend to outcompete cyanobacteria. This was first
demonstrated experimentally by Mur et al.  with the example of Scenedesmus quadri-
cauda and Planktothrix agardhii and has been widely confirmed with field data showing
that cyanobacteria tend to reach high levels of biomass and to dominate the phytoplank-
ton where nutrient concentrations and thus biomass are high. Relative to nutrient concen-
trations, field data also show similar patterns for cyanobacterial biomass or dominance:
Downing et al.  conducted a meta-analysis covering 99 temperate lakes worldwide
and show that the summer mean percentage contribution of cyanobacteria to total phyto-
plankton biomass increases with the concentration of TP “from a minimal fraction in nutri-
ent-poor oligotrophic lakes to an asymptotic average of 60% above total P of 80–90 µg/L”. Other
field studies on large data sets followed, albeit chiefly on waterbodies in temperate cli-
mates: Jeppesen et al.  evaluated 35 studies of trophic recovery encompassing a range
of lake types, differentiating between shallow and stratified lakes; these authors found a
substantial contribution of cyanobacteria to total phytoplankton biovolume above a
threshold of 50 µg/L TP for shallow lakes and occasionally as low as 10–15 µg/L in strati-
fied lakes. Carvalho et al.  analysed summer mean phytoplankton data from >1500
lakes in 16 countries reported under the EU Water Framework Directive and found that
in the TP concentration range of 10–100 µg/L cyanobacterial biomass increased, but the
relationship flattened at higher TP concentrations. Vuorio et al.  analysed data from
5678 phytoplankton samples from 2029 boreal lakes, differentiating between genera and
water colour to determine TP thresholds at which regressions of biomass versus TP con-
centrations level off; these ranged from 10 µg/L for Planktothrix in oligohumic lakes to 50
µg/L for Microcystis in polyhumic lakes. Salmaso and Tolotti  review a number of stud-
ies, largely from temperate climates, and conclude an “increasing dominance of cyanobacteria
Water 2021, 13, 2463 11 of 41
with TP increase”, with cyanobacteria typically dominating at higher total biovolume lev-
els: they contributed >50% if total biovolume was >3 mm³/L and >75% if total biovolume
was >40 mm³/L.
Field data on cyanobacterial dominance for other climates are scarce. For 14 tropical
reservoirs in Venezuela, González and Quirós  confirm the pattern observed in tem-
perate climates: their data show dominance of cyanobacteria in 8 of the 10 thermally strat-
ified reservoirs in which TP concentrations exceeded 18–20 µg/L, and the only two excep-
tions were lakes with residence times of only 12 days. Such hydraulic conditions that keep
the system in a constant early stage of plankton succession where populations of cyano-
bacteria simply do not have the time to develop . Further, although the TN concentra-
tions in this study correlated closely with those of TP, the atomic N:P ratios above 12 sug-
gests that phytoplankton biomass in all of these lakes was not limited by nitrogen.
3.1.4. Misinterpretation of the Role of N and of N:P Ratios
Smith  published the first evaluation of cyanobacterial dominance in relation to
N:P ratios, and this is often quoted to argue a need to assess limitation using N:P ratios.
Downing et al.  discussed the contemporary controversy about the relevance of N:P
ratios for cyanobacterial dominance, quoting Trimbee and Prepas  who added fur-
ther data to Smith’s assessment and concluded that the absolute concentrations of TP and
TN were better predictors of cyanobacterial dominance than the ratio of TN:TP. Indeed,
plots of biomass data (for total phytoplankton or for cyanobacteria) against TN or the ratio
of TN:TP often show patterns similar to those of plots of biomass against TP, with similar
or even higher correlation coefficients [95,98] or close associations in multivariate statisti-
cal analyses such as PCA. Such statistical results are sometimes interpreted as indicating
co-limitation. However, a key mechanism causing such statistical outcomes is that the
loads of both nutrients typically increase in parallel [95,102], reflecting their shared origin
from sewage, mineral fertilisers or manure. In consequence, biomass can correlate with N
or N:P even if the levels of N are far too high to be limiting, and actually P is the limiting
nutrient. Hence, as basic rule, in nutrient-saturated situations N:P ratios are meaningless.
If N:P ratios are applied to assess which of the two nutrients may currently be limiting it
is therefore important to report whether data above a threshold concentration were ex-
3.1.5. Co-Limitation and the Misinterpretation of Enhanced Growth in Experiments
Adding Both N and P
Results of experiments at different scales (from micro- over mesocosms to whole
lake) adding N and P separately as well as in combination have been widely published,
and often they show the highest biomass increase when adding both N and P as compared
to that when adding only one nutrient . For two reasons such a growth response does
not, however, imply that both nutrients are limiting phytoplankton biomass in the field.
A straightforward reason is that light as further growth-limiting resource has rarely been
adequately controlled in such experiments while in highly eutrophic waterbodies it is of-
ten the limiting resource. The other reason is that even if adding both nutrients leads to
more biomass, the inverse does not apply: if one of the two nutrients is minimised, the
amount of biomass that can form cannot exceed the amount supported by that nutrient
(Figure 1). For field situations, this not only requires the load to be reduced, but that water
exchange also leads to an export of the respective nutrient (including of the plankton and
detritus containing it) from the waterbody.
Water 2021, 13, 2463 12 of 41
Figure 1. Schematic illustration of the impact of increasing or reducing nitrogen (N, blue squares)
and phosphorus (P, pink triangles) concentrations, either individually or both, in a moderately eu-
trophic situation in which cellular uptake has depleted dissolved N and P and both nutrients are
incorporated in cells (green circles). Nutrients that are not incorporated into biomass remain in the
dissolved fraction outside of cells.
Where the concentrations of both dissolved, bioavailable N and P are low enough to
limit uptake rates (as discussed above: <3–10 µg/L for DIP and <100–130 µg/L for DIN),
limitation may oscillate between N and P as pulses of nutrients reach the waterbody ex-
ternally or from recycling processes such as biomass degradation and as phytoplankton
uptake draws down their concentration. This is, however, not a co-limitation by N and P
in a strict sense but rather a sequential shift in limiting factors. This mechanism has been
termed sequential co-limitation by Saito et al.  or serial co-limitation by Harpole et al.
 and is in agreement with Liebig’s Law of one nutrient limiting at a time.
Liebig’s law, developed for fertilising agriculture to increase the yield of crops in the
mid 19th century, has a long history of debate on its applicability in aquatic ecology. De
Baar  traced this with a pronounced emphasis on the limitation of growth processes.
Dolman and Wiedner  show that when applying Liebig’s law not to processes but to
biomass—or to use agricultural terminology, to yield or standing stock—it holds: from
analysing three large data sets from Germany and the USA, these authors “find that indeed
Water 2021, 13, 2463 13 of 41
phytoplankton biomass, measured across a set of lakes, is better predicted separately from either TN
or TP according to whichever is relatively least available … than from either TP alone … or jointly
using both TP and TN for each lake”. They emphasise the decisive role of cell-internal nutri-
ent concentrations (cell quota) for growth and propose a role of nutrient ratios, averaged
over time, to indicate “which nutrient is most likely to have the lower cell quota”.
Why then conduct nutrient addition experiments? For management decisions, iden-
tifying sequential or serial co-limitation is not relevant because the timescale of fluctuating
limitation is much smaller (days to weeks) than the timescale for planning and imple-
menting policies and technical measures (years to decades). Restoration measures are
most likely to be effective if they address the nutrient that has the highest probability of
effectively reducing total phytoplankton biomass and thus cyanobacterial blooms, regard-
less of which nutrient is currently limiting .
Investigating nutrient limitation in field populations is relevant for understanding
species dominance, as differences between phytoplankton taxa in nutrient uptake rates
are decisive for the outcome of competition between them. A worthwhile scientific target
for experiments with adding nutrients to field populations in micro- or mesocosm there-
fore is to elucidate which taxa are the superior competitors for the limiting resource(s) and
end up dominating, both under stringent nutrient limitation and in response to additions
of N, P or both, or to investigate how pulsed nutrient addition can promote species’ di-
versity (pioneered, e.g., by Tilman et al.  and Sommer ). Such experiments are
particularly valuable if their design includes assessing light limitation, as done, e.g., by
Maberly et al.  and Kolzau et al. . This target requires differentiating the growth
rates observed in the experiments by species or genera, but most published nutrient ad-
dition experiments are limited to a bulk biomass parameter, usually chlorophyll-a, and
differentiation by taxa, as done by Müller and Mitrovic  is rare.
In consequence, nutrient addition experiments can provide insights on the drivers of
species composition—including cyanobacterial dominance or even the outcome of com-
petition between cyanotoxin-producing genotypes and non-producers (see below). For
management decisions on the nutrient to reduce for the target of controlling cyanobacte-
rial blooms, neither N:P ratios nor nutrient enrichment experiments add insights. To con-
trol trophic state and curb blooms it is scarcely relevant to determine which nutrient is
currently limiting phytoplankton biomass: sufficiently reducing one nutrient will render
that one limiting, regardless of which one was limiting before load reduction . The
key criterion is that sufficient reduction is feasible.
3.1.6. Criteria for the Choice of N or P to Control Blooms
There is no doubt that for overall environmental protection it is desirable to reduce
the loads of both excess N and P to the environment (see below). Nevertheless, for the
target of mitigating cyanobacterial blooms of health-relevant dimensions in a specific wa-
terbody, setting a priority on N or P may be necessary. A key criterion for the specific
waterbody is a practical one, i.e., the feasibility of achieving a sufficiently strong reduction
to limit biomass. For P this has been achieved in an increasing number of cases, particu-
larly through controlling point sources from wastewater (see above). While adding the
necessary steps to sewage treatment has become feasible for controlling relevant point
sources of P, reducing N loads from sewage may be more costly , although tech-
niques are increasingly being developed and established .
Controlling nutrient loads from non-point sources, in particular from agriculture, is
generally more challenging , as this requires collaboration with and compliance of a
larger number of stakeholders, often with opposing interests, in the catchment.
3.2. Criteria for Addressing P or N and for a ‘Dual Strategy’ Addressing Both
One argument for the need of a dual strategy of reducing P and N is failure to suffi-
ciently reduce external P loads. However, where restoration by P load reduction has not
been successful, the key question (discussed in Section 3.2.1) is whether there are chances
Water 2021, 13, 2463 14 of 41
for a sufficiently pronounced reduction of the N load. A further argument is that internal
P loads prevent TP from reaching target concentrations that effectively curb biomass, i.e.,
P cycled internally between water and anaerobic sediments. The question is, however,
whether this internal load is indeed chiefly from redox-sensitive release of legacy P, or
from mineralisation of freshly sedimented organic matter. Hupfer and Lewandowski 
pick up on the paradigm that the former precludes restoration success: “The apparent cor-
relation between hypolimnetic oxygen depletion and hypolimnetic P accumulation as two typical
symptoms of increased eutrophication in thermally stratified lakes was misinterpreted by reducing
the complex processes at the SWI (sediment-water interface) to the statement ‘oxygen depletion
leads to release of phosphorus’, which subsequently has become the accepted paradigm of the day”.
These authors show that in many cases mineralisation is actually the major source of in-
ternal loading. This has important cause–effect implications because once external loading
decreases to levels effectively reducing primary production, there is less organic matter to
sediment and decompose, releasing P. This is why internal measures to intercept P release,
such as sediment capping or treatment, need to be periodically repeated if the external
load is not sufficiently reduced. Before concluding that P cannot be sufficiently reduced
because of high internal loading it is therefore important to understand the mechanisms
causing P release and the relevance of redox conditions for this process (Section 3.2.1).
A further aspect discussed by Conley et al.  is that where P control is successful
and biomass substantially declines, this can lead to more N being ‘left over’ and dis-
charged from the waterbody. In consequence, reducing only P may result in more N
reaching downstream aquatic ecosystems. However, whether or not this is the case and if
so, whether or not it triggers adverse effects critically depends on conditions in the respec-
tive downstream waterbodies (Section 3.2.2). Both aspects require careful case-by-case as-
sessments (Section 3.2.3).
3.2.1. Impacts of N within the Waterbody
In contrast to P load reduction, success stories are scarcely known from load reduc-
tion measures focusing on N . One is the case of Müggelsee in Berlin, Germany,
which is shallow and subject to substantial internal P loads from the sediment during
summer as a legacy of decades of intensive agriculture upstream. As a result of restruc-
tured agriculture following the German reunification, fertiliser input declined, leading to
substantial N-load reduction and, through in-lake denitrification, to phases of about 100
days of N-limitation during late summer. In consequence, cyanobacterial biomass de-
clined by 89% compared to the average during the 1980s, interestingly without a shift to
diazotrophic taxa such as Aphanizomenon sp. . This demonstrates that substantial N
load reduction can lead to N concentrations in the lake which are sufficiently low to effec-
tively limit cyanobacterial blooms. In contrast, the simultaneous reduction of the P load
to this lake was compensated by high internal loads which are likely to remain high for
some years to come, i.e., until water exchange has exported sufficient amounts of phos-
phorus out of the system.
Candidate waterbodies for a focus on N include those in which N is already limiting
or concentrations are not far above a potentially limiting level, thus rendering further re-
duction is a realistic option, as is typical for many shallow waterbodies during late sum-
mer. As one possible option of a fairly low-cost measure with the potential for a strong
impact in such situations, Kolzau et al.  propose agreements with farmers in the catch-
ment to time their application of fertilisers or manure so that run-off and leaching from
agricultural soils is minimised during the cyanobacterial growth season. Chorus and
Spijkerman  discuss the perspectives of controlling eutrophication by reducing N
loads in some detail.
In some situations, in which internal P loading due to release of redox-sensitively
bound P from sediments is a relevant source of P, maintaining some nitrate in the water
above the sediment may even be beneficial. Steinman and Spears  emphasise the cou-
Water 2021, 13, 2463 15 of 41
pling of internal processes of N and P cycling through the role of nitrate for redox-sensi-
tive P release mentioned above, and hypolimnetic concentrations of nitrate N > 0.5 mg/L
are well known to effectively suppress the release of P from sediments by maintaining a
sufficiently high redox potential at the sediment surface. Andersen  showed this with
data from nine Danish lakes in which, even in anoxic hypolimnia, no P release was ob-
served if nitrate N concentrations were in the range of 1 mg/L. For lakes without thermal
stratification he reports a nitrate-N threshold of 0.5 mg/L below which P release may occur
even if the water was ‘well oxygenated’, possibly as nitrate transports oxygen into the in-
terstitial water very effectively . Salonen et al.  show pronounced hyperbolic
curves for the concentrations of both TP and TN in the sediment-near water of Lake
Vesijärvi relative to the concentration of nitrate-N: the concentrations of both TP and TN
increased massively once nitrate was depleted. Schauser et al.  also discuss this for
Lake Tegel. The case studies for both these lakes, however, confirm the above-mentioned
assessment by Hupfer and Lewandowski  that redox-sensitive release may be a minor
source of P relative to mineralisation of P bound in the large amounts of biomass formed
in the lake on the basis of external P sources.
The overall ecological assessment of purposely adding nitrate products to suppress
internal P loading  may be questionable (perhaps unless its consumption through
denitrification can be demonstrated). The available experience does, however, show that
making use of nitrate for sediment oxidation can be an option, provided there is a good
understanding of the respective sediment–water system and its nutrient budgets. In par-
ticular, before counting on nitrate to suppress internal P loading it is important to assess
the relevance of redox-sensitive P release relative to that through mineralisation.
3.2.2. Downstream Impacts of N
In waterbodies previously N-limited during summer, reducing biomass by reducing
TP concentrations can lead to dissolved N then being ‘left over’ by the phytoplankton;
also, less biomass production leads to less organic substance available to fuel denitrifica-
tion. In consequence, more N is exported to downstream ecosystems. This concern is ad-
dressed by Paerl et al.  with the example of the Neuse River (and further studies
quoted therein) where upstream P reduction led the increase of downstream phytoplank-
ton biomass in previously N-limited estuarine and coastal waters. An important criterion
for a ‘dual strategy’ therefore is possible downstream impacts of an imbalance in the re-
duction of N and P.
Ample evidence shows that with some exceptions most coastal and estuarine waters
are currently N-limited (see reviews by Howarth and Marino  or Malone and Newton
). However, as discussed above for limnetic systems, reducing P loads to sufficiently
low levels can switch limitation to P also for coastal waters, particularly where water ex-
change rates are high and thus dilute or export P, including that adsorbed to sediments
which is resuspended during strong wind events and storms. In some coastal settings the
success chances for management measures to achieve stringent nutrient limitation may be
higher for P than for N merely because of the dramatic increase of N loads relative to those
of P. Reviewing N loads to coastal areas, Malone and Newton  quote projections for
2050 that propose a doubling of global N input to the oceans as compared to 1990 due to
increased synthetic fertiliser use (particularly in Asia). Statham  shows atmospheric
deposition to be an important source, accounting for 15–50% of the N loads to the estuaries
studied. The regulations and management measures established in many countries during
the past decades that were effective mostly for reducing P loads are showing success also
for some coastal waters, as discussed by Howarth and Marino  for the North Sea
estuaries and Laholm Bay in Sweden.
A strong argument for eutrophication control through limiting N loads is that deni-
trification can permanently remove N from aquatic systems to the atmosphere as N
applies to freshwater as well as to marine systems: the share of N that is transformed to
is estimated to be 40–50% for the northern Gulf of Mexico, 25% for the Great Barrier
Water 2021, 13, 2463 16 of 41
Reef, 42–96% for the Baltic Sea, 40–42% for the Northern Adriatic Sea and 40% for the
Chesapeake Bay . Phosphorus can also be effectively lost through burial in the sedi-
ment: in low salinity sediments a relevant fraction of P may be permanently bound in
insoluble vivianite . In shallow waters subject to sediment being resuspended, P ad-
sorbed to sediment particles can be remobilised through wind-driven turbulence. None-
theless, in their review of nutrient retention in the coastal areas of the Baltic Sea Carstensen
et al.  show that in some estuarine regions removal of P by burial in sediments is
more pronounced than losses of N, both through burial of organic matter and through
denitrification. Losses of both N and P from the euphotic coastal zones also strongly de-
pend on water retention times, this is, transport to the open ocean.
Thus, increased N concentrations in effluents of a waterbody undergoing oligotroph-
ication through P-load reduction may or may not lead to increased downstream eutroph-
ication, strongly depending on conditions in the downstream waterbodies: if P is already
limiting in these, some additional N should not have an effect. If N concentrations are far
above any chance of limiting biomass, slightly augmenting them through upstream oligo-
trophication should also have no effect. Furthermore, it should have no effect if the efflu-
ent of the upstream waterbody flows through wetlands with massive denitrification. It is
therefore important that site-specific assessments are conducted at a larger scale, taking
into account the downstream impact of measures intended to control blooms in a specific
3.2.3. Case-by-Case Assessments
While this may take some years, the increasing experience with oligotrophication
shows that both for lakes and coastal areas, later on in the recovery process sediments can
again become a sink rather than a source for P [8,125], provided that the external load is
sufficiently reduced and that water residence times are not too high. In conclusion of com-
piling data from a large number of lakes, Steinman and Spears  show that sediment
TP concentrations of many lakes are quite similar across a wide trophic spectrum: they
are elevated only in some eutrophic lakes. These authors strongly emphasise the need for
“lake-specific analyses to determine the best management strategy; a one-size-fits-all approach is
doomed to failure”. Lewis and Wurtsbaugh  comprehensively discuss situations in
which N-limitation of total phytoplankton biomass is likely, they emphasise that in most
cases restriction of P loads is the “most promising management tool” while also acknowledg-
ing the role of N-limitation in phosphorus-saturated waterbodies. Already more than 10
years ago, these authors pointed out the need to reduce atmospheric N deposition to pre-
vent currently N-limited lakes from becoming more eutrophic.
Taken together, the current state of knowledge shows that limiting only one nutrient
is highly effective if it is brought down to sufficiently low concentrations, both for fresh-
water and marine systems. However, whether or not such target concentrations can be
achieved within the desired timeframe and whether or not negative downstream impacts
are likely depends on the specific conditions. Generally requiring the reduction of the
loads of both nutrients can dilute the available funding and by this increase the risk that
neither measure for N nor P reduction is sufficiently efficient to bring concentrations
down low enough to sustainably limit phytoplankton biomass.
We therefore propose to supplement the call for a ‘dual strategy’ with a call for basing
this on case-by-case assessments that include not only an assessment of the technical, po-
litical and societal feasibility of attaining sufficiently low target concentrations, but also of
impacts on downstream ecosystems.
3.3. Shifts in Cyanobacterial Populations in Consequence of Oligotrophication
A specific aspect of trophic change is the reduction of trophic state leading to more
transparent water. This (re-)opens habitats for metalimnetic, tychoplanktonic and benthic
cyanobacteria. In Section 4.3 we discuss how the challenges these present for human
health risk assessment differ from those of epilimnetic cyanobacteria.
Water 2021, 13, 2463 17 of 41
3.3.1. Disappearance and Re-Appearance of Planktothrix rubescens in the Metalimnion
Planktotrhix rubescens is the most prominent and best studied example. Most strains
of this species produce microcystins [126,127] and dense populations typically occur in
meso-eutrophic deep, stratified lakes primarily in the metalimnion (mostly between 10
and 20 m) . This low-light adapted species can effectively use the very low intensity
of green light in these depths as well as adapt to variable light conditions by means of
buoyancy regulation [128–130].
The occurrence of P. rubescens is closely linked to trophic state. It appeared in many
formerly oligotrophic deep lakes such as Lake Mondsee, Lake Zürich and other peri- and
subalpine lakes as they shifted from oligotrophic to mesotrophic [131–133]. With further
eutrophication, however, P. rubescens disappeared due lack of light in the metalimnion
caused by high epilimnetic biomass . In times of highest trophic state of these lakes,
other cyanobacteria such as Microcystis, Aphanizomenon and Dolichospermum, but also
green algae or diatoms were reported as dominant taxa in the epilimnion (e.g., Chiemsee
, Lake Zürich , Lake Baldegg , Lac du Bourget ). As successful man-
agement measures to reduce trophic state led to a return of higher transparency, epilim-
netic cyanobacteria disappeared and P. rubescens (re-)appeared in many of these lakes, a
process termed “paradoxical outcome” of lake restoration by Jacquet et al. .
P. rubescens can thrive at temporarily lower P-concentrations because, like other cya-
nobacteria, it can effectively compensate the lower phosphorus concentrations in the epi-
limnion by storing P well in excess of the current need and/or by excreting alkaline phos-
phatase . However, primarily it benefits from the usually higher nutrient concentra-
tions in the metalimnion compared to the epilimnion. In the phase of the decline of TP
concentrations P. rubescens attained biovolumes up to ~ 10 mm
/L in the metalimnion and
dominated the total phytoplankton biomass. This changed once total phosphorus (TP)
declined below 10 µg/L and in consequence P. rubescens declined to biovolumes <1 mm
or even was absent [128,132,136,138].
During winter mixing of waterbodies, P. rubescens can occur dispersed over the entire
water column or, once mixing subsides, it can even form dense surface blooms rendering
the water deeply reddish-brown  and in which microcystin concentrations can range
up to >30 mg/L [140,141]. In the metalimnion microcystin concentrations are generally
much lower with reported values rarely exceeding 10 µg/L .
Where TP concentrations are around or slightly above 10 µg/L the abundance of P.
rubescens can vary substantially between years, including periodic absence as, e.g., in Lac
du Bourget and Ammersee [136,143]. Such annual variability has been explained by fluc-
tuations of water transparency and thus of ratios of euphotic depths to mixing depths
. In contrast, constantly elevated biomass of P. rubescens is attributed to changes in
the patterns of stratification and mixing due to climate change. During the last decades,
years with incomplete mixing have become more frequent, for example, in Lake Zürich,
allowing P. rubescens to increase its population as more cells survive as inoculum for the
next season’s population while holomixis would diminish cell density more strongly due
to higher hydrostatic pressure in deep water layers [39,144]. On the other hand, reduced
mixing can also result in nutrient depletion of the epilimnion through sedimentation and
thus lead to lower abundance of P. rubescens, as observed in Lake Garda . Higher tem-
peratures early in the year and thus an earlier onset of stratification have been found to
favour P. rubescens in Mondsee  and Lake Geneva .
The current state of knowledge of conditions conducive to P. rubescens shows that in
many deep, thermally stratified lakes and reservoirs the combined effects of changing nu-
trient concentrations and climate-driven changes in stratification patterns can work both
ways, this is, either favour or disfavour the proliferation of this species. For water body
management and use—especially for the abstraction of raw water for drinking water pro-
duction—this means that the metalimnetic phytoplankton populations need to be ob-
served closely in order to adjust offtake depths and/or water treatment. An increasing
understanding of the combined impacts of stratification stability, water transparency and
Water 2021, 13, 2463 18 of 41
nutrient concentrations may allow the development of models that can predict the wax or
wane of this species in such waterbodies .
3.3.2. Benthic and Tychoplanktic Cyanobacteria
Benthic (toxic) cyanobacteria are often reported to have increased both in frequency
of occurrence as well as distribution, though in part this perception might be due to in-
creased awareness, triggered by dog poisoning that typically receives broad public atten-
tion . In contrast to planktic cyanobacteria, for which mass developments are linked
to higher trophic levels, benthic cyanobacteria are frequently reported from clear-water
rivers and lakes  in which light can penetrate to the bottom, allowing mats of, e.g.,
Phormidium (Microcoleus) to grow or allowing macrophytes to develop which provide sur-
faces for epiphytic cyanobacteria. Nutrient mobilisation from co-occurring bacteria and
fungi can enable benthic cyanobacteria to cover large areas of riverbeds .
The shift from planktic to benthic production in the course of water quality improve-
ment and higher transparency has been shown for large European rivers such as the River
Loire and the Danube [149,150]. Recent data from the River Loire indicate an increase of
Phormidium at some sites . Phormidium (including synonymous taxa ) is a glob-
ally widespread largely benthic and often toxic cyanobacterium, reported from many riv-
ers in, e.g., New Zealand , France  and California .
Potentially toxic benthic cyanobacteria such as Phormidium, Microcoleus, Geitlerinema,
Oscillatoria or Microseira wollei are also reported from lakes and reservoirs in which water
quality has improved from eutrophic to meso- or oligotrophic [154–156]. Lowe  iden-
tified the littoral of oligotrophic lakes with ample light as the preferred habitat of benthic
cyanobacteria. As with benthic cyanobacteria in rivers, information on the abundance of
benthic cyanobacteria half a century ago, prior to eutrophication, is scarce. Examples in-
clude two German waterbodies with a massively reduced P load since the 1980s which
now have reached an oligo- to mesotrophic state, with planktonic cyanobacteria scarcely
occurring: both lakes now show recurrent abundance of benthic cyanobacteria. In the
Wahnbach reservoir (used for drinking water production) eutrophication control led to a
drastic reduction of P. rubescens, but thereafter sporadic episodes with geosmin producing
cyanobacteria inhabiting the littoral between 1 and 11 m occurred . In Lake Tegel
(Berlin), macrophytes are re-colonising the littoral as a result of increased water transpar-
ency, but in 2017 tychoplanktic Tychonema producing anatoxins was found in tufts of wa-
ter moss (Fontinalis sp.) following an episode of dog poisoning  and since then is
leading to recurrent dog death incidents.
Benthic cyanobacteria appear to have received attention particularly after such inci-
dences, and thus the timepoint of their initial occurrence and of conditions favouring their
abundance cannot be elucidated retrospectively. While optimal light conditions certainly
are important, hydrophysical conditions such as storm-driven detachment of submerged
macrophytes or water level fluctuations in reservoirs may further influence their abun-
dance and in particular, whether or not they appear in parts of a water body where people
or pets might be exposed.
4. How Do Trophic and Climatic Changes Affect Health Risks from Cyanotoxins?
Health risks from cyanobacterial blooms depend on the levels of biomass that they
attain and on the contents of various toxin types in that biomass. Toxin contents in turn
depend on the taxonomic composition of a bloom, as the potential to synthesise specific
types of toxins is highly variable between cyanobacterial taxa (species and genera). Some
toxin types such as microcystins are produced by a high variety of taxonomically distant
taxa, while the number and diversity of taxa potentially producing cylindrospermopsins
is much lower. While the different specific combinations of environmental conditions fa-
vouring the proliferation and dominance of particular taxa is well described [160,161], for
inferring the likely toxin types and concentrations, these at best allow worst-case estimates
based on the upper range of known toxin/biomass ratios. This is because the taxonomic
Water 2021, 13, 2463 19 of 41
category of a ‘species’ encompasses a range of different genotypes which can vary consid-
erably in their content of secondary metabolites, including cyanotoxins [142,162–164].
Factors determining toxin contents—and hence toxin concentrations—discussed in
this section act on different levels of regulation, schematically illustrated in Figure 2. An
impact of nutrient availability on cellular toxin content is under discussion particularly
for N and the peptide toxins, the microcystins, and in Section 4.1.1 we review the current
state evidence for this, as well as approaches for clarifying this question. However, on the
level of distinction between genotypes it is still poorly understood which environmental
conditions favour which genotypes producing particular cyanotoxins, and in consequence
impacts of changes in environmental conditions on dominance of genotypes with and
without the production of certain toxins are still largely unclear. While we address this in
Section 4.1.2, understanding why certain genotypes with specific metabolites dominate
under certain conditions requires a broader approach, including the wide array of second-
ary metabolites in cyanobacteria that are synthesised by similar pathways and possibly
have similar functions in producing strains. It is still enigmatic which competitive ad-
vantage some cyanobacterial taxa gain from the metabolically costly production of a huge
variety of secondary metabolites, and toxicity to vertebrates is rather a coincidence than a
result of evolutionary selection—cyanobacteria produced microcystins long before verte-
brates have evolved . In Section 4.2 we discuss whether the key groups of cyanotoxins
are now known with the four major groups for which the World Health Organisation
provided further guideline values in 2020 . Section 4.3 explains the considerations
leading to their derivation, as these are important for assessing the implications that the
in situ concentrations observed may have for human health. Section 4.4 discusses how the
species shifts to metalimnetic and benthic habitats caused by oligotrophication change
risks of human exposure to cyanotoxins.
Water 2021, 13, 2463 20 of 41
Figure 2. Levels at which environmental conditions drive cyanotoxin occurrence and concentra-
4.1. Impact of Changes in the Availability of a Resource on Cyanotoxin Concentrations
For the first three levels of the environmental conditions driving cyanotoxin occur-
rence and concentrations illustrated in Figure 2, a good understanding is attained, provid-
ing a sound basis for developing management measures to control bloom occurrence. For
the following two levels, the drivers of genotype occurrence, scientific knowledge is
scarce, while for the lowest level, the intracellular processes of toxin biosynthesis are mod-
erately well understood.
Differences in toxin production between individual genotypes adds further levels of
complexity to the understanding of toxin occurrence in waterbodies. This limits the pre-
diction of in situ toxin concentrations from biomass levels of the producing species to
worst-case estimates. Many studies refrain from breaking taxonomic identification down
below the genus level because of the difficulties of species distinction and identification
(as well as ongoing changes in taxonomy ). However, for studying toxin occurrence
yet a further differentiation below the species level is important because within most po-
tentially toxigenic species genotypes with and without genes for the production of specific
metabolites coexist and co-occur in natural populations. For an understanding of the dy-
namics of toxin concentrations, a distinction only between toxigenic and non-toxigenic
genotypes is presumably an oversimplification.
Toxigenic genotypes, this is, those that possess functional genes for the biosynthesis
of a particular toxin, can be further sub-divided into clones or clonal complexes with a
specific metabolome (i.e., producing individual structural variants of toxins and other
peptide metabolites), and most likely also by specific ranges of cell quota (i.e., the amount
of individual structural variants per cell).
A number of studies have assessed the diversity of genotypes within natural popu-
lations, showing that generally multiple toxigenic as well as non-toxigenic genotypes co-
exist that are often indistinguishable by conventional analysis such as light microscopy
[23,166–168]. However, most of these studies reflect only snapshots of individual or a few
populations while the dynamics of genotype occurrence are rarely described [169–173].
Still fewer studies have addressed the competition between genotypes and if so, in most
cases with a focus on only one toxic and one non-toxic strain (Table 1). To our knowledge,
no studies have been published on the competition of multiple toxic and non-toxic geno-
types in an experimental system of a complexity that approaches that of natural systems.
Any extrapolation of results obtained from an experimental setup of low complexity, for
example, co-culture of one toxic and one non-toxic strain, risks being premature.
4.1.1. Cellular Level of Peptide Content and Biosynthesis
For the lowest level in Figure 2, the genetic background  and the cellular regu-
lation of peptide biosynthesis, a wealth of studies is available ranging from conventional
culture studies to more recent molecular approaches focusing on the transcriptional re-
sponse, for example, to light intensity , iron availability  or nutrient stress .
The growth conditions that have been found to influence toxin biosynthesis vary between
studies, and they do not show coherent outcomes: results vary between species, strains
and experimental conditions, showing that stress through resource limitation or temper-
ature can lead to more or to less microcystin produced per cell. Moreover, few experi-
ments have been done with continuous or semi-continuous cultures, and only in these is
limitation constant and clearly focused on one resource while in batch cultures, conditions
(in particular, photon flux density) change as cell density increases.
The results of these many culture studies do, however, allow two overarching in-
sights: one is that for microcystins and cylindrospermopsins the variability of toxin con-
tent in response to experimental conditions rarely exceeds a factor of 2–4 [142,162]. The
other is that the majority of studies agree on the finding that no factor or growth condition
Water 2021, 13, 2463 21 of 41
has been found that completely suppresses toxin production (with few exceptions ),
in line with the observation that microcystin production is coupled to growth rate .
This is remarkably different from regulation of comparable secondary metabolite produc-
tion in filamentous fungi, where a large share of biosynthetical gene clusters is silent un-
der certain conditions  and secondary metabolite production is subject to complex
regulation, including epigenetics . Therefore, toxin production by a cyanobacterial
strain generally results in a toxin content within a comparatively narrow range that seems
to be strain-specific. This means, for example, that one strain produces microcystins at cell
quota of 5–20 fg/cell while in another strain of the same species the range is 50–200 fg/cell.
Presence of biosynthesis gene clusters such as the mcy-cluster in a strain that has not been
found to produce microcystins is, thus far, only confirmed for strains with a mutation in
the gene cluster . Future studies on various factors affecting toxin production by cy-
anobacteria most likely will not produce results that fundamentally overturn what has
been found in hundreds of preceding studies.
The more recent field of ecological stoichiometry has addressed cellular microcystin
content in the wider context of analysing the impact of nutrient availability on the relative
atomic composition of biomass, emphasising that nitrogen amounts to 14% of the molec-
ular mass of microcystins. It thus seems plausible that N-limitation should reduce the cel-
lular content of microcystins, and indeed a number of laboratory culture studies report
this finding (e.g., [22,181]; see also the discussion in Gobler et al. ). However, others
find the opposite, e.g., Peng et al.  report an inverse relationship of average micro-
cystin content per cell to N concentration from experiments at 20–30 °C but not at lower
Brandenburg et al.  conducted a meta-analysis for N-rich toxins in a range of
freshwater and marine genera from 37 published laboratory culture studies of cellular
toxin contents in response to N and P limitation. The authors conclude that “responses of
N-rich toxins to nutrient limitation were less consistent than previously reported”: cellular toxin
contents decreased by only about 60% under N limitation, and this was statistically sig-
nificant only when pooling the data for all 37 studies across all toxins and all freshwater
and marine phyla. For the 24 studies on microcystins alone, extremes ranged from almost
3-fold decrease to almost 3-fold increase, and while some decrease was visible in the data
for about 11 of these studies, this was not statistically significant. Additionally, three of
these studies showed an increase and about 10 showed no response of cellular microcystin
contents to changes in N-availability. Brandenburg et al.  also evaluated 16 studies
addressing the impact of P-limitation on cellular microcystin content, and here they found
more pronounced responses, with an overall increase by 88% and only two studies show-
ing a decrease. A further important result of this meta-analysis is that cellular content of
N-rich toxins increased or decreased in parallel with the overall N-content of biomass, this
is, with C:N ratios and N:P ratios. This finding is interpreted as reflection of N being pri-
marily allocated to population growth and not specifically to the production of N-rich
This finding highlights that a focus on the metabolites that happen to be toxic to ver-
tebrates is too narrow an approach for the target of elucidating the environmental condi-
tions driving cellular cyanotoxin production and content. This pertains in particular to
oligopeptides: several hundreds of other oligopeptides from cyanobacteria have been de-
scribed , all with a relative N-content similar to that of microcystins, many with some
sort of bioactivity at least demonstrated in vitro . Broadening the perspective beyond
microcystins to cover the entity of N-rich secondary metabolites suggests that the availa-
bility of dissolved inorganic N (nitrate, urea or ammonia) is unlikely to generally favour
microcystin production over that of other cyanopeptides. Rather, this broader perspective
requires analysing changes in the total content of the cells’ oligopeptides and possible
shifts in their relative proportions. While the high in situ diversity of cyanobacterial pep-
tides has been acknowledged for nearly two decades [186–189], very few studies have
quantified oligopeptides other than microcystins  because this broader approach is
Water 2021, 13, 2463 22 of 41
hampered by the lack of standards for quantifying other cyanopeptides. Natumi and
Janssen  addressed this gap for multiple peptide classes, including microcystins, in
Microcystis PCC 7806 and Dolichospermum NIVA-CYA 269/6 by recording relative changes
and thus showed that the cellular peptide content responded synchronously to changes
in nutrient concentrations.
4.1.2. Genotype Level of Peptide Diversity and Dominance
As reviewed in Suominen et al. , the differences in microcystin content between
genotypes by far outweighs differences induced by resource limitation or temperature.
Co-occurrence of genotypes with high and low or no cellular contents of specific metabo-
lites in field populations explains why the microcystin cell quota found in field samples
are typically several-fold lower than the maximum cell quota observed in some strains in
the laboratory [142,192]. As indicated in Figure 2, at the level of individual genotypes
within cyanobacterial populations, knowledge about drivers of dominance is scarce. A
number of publications report the sub-specific in situ diversity of bloom forming taxa such
as Planktothrix, Microcystis and Aphanizomenon either in individual or multiple waterbod-
ies [23,127,193–196] by various typing schemes. Further, genotype diversity has been stud-
ied on larger geographic scales [127,197]. Based on the available studies it seems fairly safe
to assume a multitude of co-existing genotypes in any cyanobacterial bloom—or to con-
sider a bloom formed by a single or only a few genotypes an exception. Only a few studies
have assessed the seasonal dynamics of multiple genotypes within cyanobacterial popu-
lations [169,170,173,198] or shifts over the course of years [23,172]. Although respective
studies in this field would be exciting, methodological problems need to be solved to al-
low a cost-effective, high-resolution monitoring.
To elucidate drivers of genotype composition and dominance, approaches that are
well established at the species level in phytoplankton ecology can be applied on the gen-
otype level, using molecular tools to distinguish a few genotypes within a species. For the
species level Litchman and Klausmeier  review how traits affect the outcome of com-
petition, depending on conditions in the waterbody; i.e., resistance to grazing and para-
sites, modes of reproduction and survival of inocula, temperature-dependence of growth
rates and growth responses to resource limitation. Among these traits, studies of differ-
ences between cyanobacterial genotypes have chiefly addressed growth rate responses to
temperature and resources (for resistance of cyanobacteria to grazing, see Moustaka-
Gouni and Sommer ).
Data characterising uptake affinity and cellular storage have become available from
a large number of experiments, compiled, for example by Litchman et al.  and Ed-
wards et al.  who used such data to model the outcome of competition between spe-
cies. At the level of different cyanobacterial genotypes such data are rare. Differentiation
at this level may be more challenging because cell size and shape are regarded to be the
“master variable” that governs resource uptake , but the size and shape of strains
within a species of cyanobacteria usually show no difference. Indeed, Hesse and Kohl
 found no difference in maximum growth rates relative to light of a wildtype and a
microcystin-deficient mutant of Microcystis PCC 7806 grown in semi-continuous cultures,
and Briand et al.  found no difference between the maximum growth rates (µ
five strains of Planktothrix agardhii (three of which produce microcystins). However, some
studies addressing the species level show that not all resource affinity parameters neces-
sarily correlate with cell size and shape [77,199]; niche separation of the morphologically
very similar Planktothrix rubescens (red, metalimnetic) and P. agardhii (green, epilimnetic)
may serve as an example. Additionally, monoculture experiments with different strains
of cyanobacterial species have found distinct differences in their growth traits: Suominen
et al.  report differences in growth characteristics between the Microcystis PPC 7806
wildtype and mcy-knockout mutant, and these differences were sufficiently distinct to
model the outcome of competition between both strains with these parameters.
Water 2021, 13, 2463 23 of 41
Differences in nutrient or light acquisition characteristics between genotypes may be
too subtle to show as statistically significant in any of the growth parameters listed above,
but nonetheless determine the outcome of competition. In such cases, co-culture experi-
ments can be a more effective approach to elucidate their relative competitive power un-
der specific combinations of growth conditions, either by taking field populations into
micro- or mesocosm cultures to test which genotypes end up dominating, depending on
how conditions are manipulated, or by combining two culture strains to see who ‘wins’.
Results of respective studies are compiled in Table 1.
Table 1. Competition between strains: examples of outcomes of competition experiments between cyanobacterial strains
with (MC+) and without (MC−) microcystin production or pairs of wildtype (mcy+) and knockout mutants (mcy−). The
studies are organised by the limiting resource or growth condition, respectively (some studies occur twice).
Dominance Attained in
25 µE/m² × s
MC− for both strains
within <2 weeks for the 2
strains with pronounced dif-
ference in critical light inten-
sity; within >160 days for the
with similar critical
Strains differentiated semiquantita-
tively using DGGE bands, different
pigments as well as
MC+: UTCC 300;
MC−: UTCC 632;
20 versus 80 µE/m²×s
MC+ at low light
MC− at high light in co-cul-
ture with UTCC 632, but
±balance at high light in co-
In monoculture at low light no dif-
ference in growth rate between
strains, but at high light higher
growth rate for MC+
M. aeruginosa PCC 806:
MC−: mcy− mutant;
39 and 5 µE/m² × s
mcy− at optimal light;
± balance at low light
In monoculture growth characteris-
tics of both strains were not signifi-
allelopathy of mcy+
strain tested and
35–120 µE/m² × s
MC+ under all light and tem-
perature conditions tested
In monoculture MC− grew faster
than MC+; independently of growth
condition an un-identified allelo-
chemical of FACHB905 (not MCs!)
seemed to suppress FACHB469
MC+ at lower temperatures;
MC− at higher temperatures
In line with field observations from
Lake Kinneret: in recent years with
higher temperatures, brown (phyco-
erythrin-containing) non MC− pro-
ducing Microcystis dominate
mg/L ± balance up to 11 days
MC+: 3 strains
MC−: 2 strains;
Light and tempera-
MC+ at low light and low
either MC− or balance of
both strains at high light and
Interpretation as trade-off between
better fitness of MC+ under limiting
conditions and metabolic costs of
MC− production under non-
Result in line with field observations
of genotype composition during
N MC+ under N limitation;
MC− after N pulse
M. aeruginosa PCC 7806
MC−: mcy− mutant;
mcy− at optimal light (39
µE/m² × s) and low N;
± balance of both strains at
low light, low N
In monoculture growth characteris-
tics of both strains were not signifi-
Allelopathy of mcy+
strain tested and
Water 2021, 13, 2463 24 of 41
M. aeruginosa PCC 7806
MC−: mcy− mutant;
chemostat and semi-con-
mcy+ always wins: if limita-
tion is continuous, N is
applied as single strong
or at 3-day intervals
Adding P pulsewise leads to co-ex-
istence of wildtype and mutant 
N, P MC+ or
balance at lower N or P
In monocultures MC
Unidentified allelochemical of
FACHB905 seems to suppress
FACHB469 under all growth condi-
The only striking observation gleaned from the compilation in Table 1 is the contra-
dictory nature of “who wins”, including counter-intuitive outcomes such as the domi-
nance of the non-microcystin-producing strain under N-imitation in the study by Suomi-
nen et al. . The results of LeBlanc Renaud et al.  show that under the same co-
culture conditions the microcystin producer can outcompete one non-producing strain
but not the other.
Basic data on resource acquisition characteristics and competitive power are best ob-
tained under steady-state conditions, preferably in continuous or semi-continuous cul-
tures. Beyond this basic characterisation it is also important to test pulse-wise or fluctuat-
ing changes in resource availability because this is likely to enable co-existence, as has
been shown on the level of competition between species [109,205]. In nature, pulse-wise
changes in resource availability are the rule and the most likely explanation for coexist-
ence. Briand et al.  tested the response of co-cultures of strains of P. agardhii first
grown in N-deplete medium to which a pulse of nitrate was then added, and this led to a
switch in dominance from the microcystin-producing strain to the non-producer. In a sim-
ilar experiment with a wildtype of Microcystis and a mcy-knockout mutant, Briand et al.
 found a single pulse of nitrate to strengthen the pre-existing dominance of the non-
producing mutant. With the same wildtype and mutant strains, but using semi-continu-
ous cultures instead of batch cultures with nutrient pulses at 3-day intervals, Suominen et
al.  found the opposite response to pulses of nitrate, i.e., dominance of the producing
wildtype, but coexistence for up to 30 days in response to pulses of phosphate.
The outcome of competition between strains with and without a specific metabolite
may or may not be related to a biological function of that metabolite—it may well be due
to other differences between the strains. For example, Lei et al.  propose that the re-
duced competitive strength of a microcytin-producing strain (at high light intensity, low
concentrations of N and P and low temperature) may have been caused by reduced pro-
duction of an (unidentified) allelochemical sequestered by the non-producing strain (Ta-
ble 1); they discuss this relative to other publications, some of which found evidence for
such effects and some of which did not.
For the wide array of cyanopeptides in cyanobacteria it is indeed conceivable that the
various peptides either have a similar or a complementary function. Briand et al. 
discuss this as one possible interpretation of their results with a Microcystis wildtype pro-
ducing microcystins and its knockout mutant: the latter contained significantly higher
concentrations—up to 12-fold—of a range of other oligopeptides, although the only dif-
ference between both strains was the lack of a functioning gene for microcystin produc-
tion. These authors also found significantly higher cellular concentrations of oligopep-
tides when co-culturing strains than in monocultures. Interestingly, pure microcystin-LR
was not the substance triggering these responses. Suominen et al. , referring to the
experiments by Briand et al. , with P. agardhii suggest that growth conditions may
not determine the outcome of competition directly, but possibly indirectly, through affect-
ing the production of (as of yet unidentified) allelopathic substances.
Water 2021, 13, 2463 25 of 41
We summarise the current state of knowledge on drivers of genotype composition as
The numerous monoculture experiments subjecting strains to different resource lim-
itations or temperatures have not revealed any specific conditions that unambigu-
ously enhance or reduce cyanotoxin production across the range of stSrains and spe-
cies thus tested. Even nitrogen limitation seems to have little impact on cyanopeptide
production beyond its general impact on the nitrogen content of cellular biomass.
In field populations strains producing microcystins generally seem to co-exist with
those that do not, and this observation is supported by the outcome of co-culture
experiments run with pulsed nutrient application: pulses promote co-existence ra-
ther than dominance of one strain. A hypothesis worth testing is that this applies to
other cyanobacterial oligopeptides as well.
While all cyanotoxin groups include several congeners and the microcystins include
many, bloom-forming cyanobacteria often also contain numerous further secondary
metabolites, particularly peptides, of unknown function. Research focusing merely
on the production of microcystins is likely to fall short of recognising potential trade-
offs, complementary or substituting functions for the producing genotypes.
The “chemical cross-talk” of peptides indicated by the results of some co-culture ex-
periments opens exciting perspectives for understanding the biological function of
these metabolites. This applies equally to the possibility that peptide metabolites con-
fer resistance to fungal and possibly other parasites [207,208].
In conclusion, co-culture experiments seem to be the currently most promising ap-
proach for developing a better understanding of the function of cyanobacterial metabo-
lites. However, when limited to two strains, a producer and a non-producer of a single
type of metabolite, the outcome may well be coincidental, and with another combination
of strains results may well be different.
After decades of research that so far has not found an unambiguous answer to the
question of the biological function of cyanotoxins, cyanobacterial oligopeptides and other
secondary metabolites, contributions to understanding the drivers of genotype composi-
tion are likely to be of substantial scientific value in their own right. Whether or not such
insights are useful for planning management strategies then remains to be seen.
4.2. Can We Expect New Cyanobacterial Toxins in the Future?
The number of studies and publications about toxic cyanobacteria has been steadily
increasing over the past few decades. Similarly, although to a lesser extent, the search for
natural products with pharmacological potential has led to the characterisation of a large
number of cyanobacterial metabolites, some 2000 of which are listed in CyanoMetDB .
Despite these two developments, the number of cyanobacterial toxins has been stable over
the years, in particular regarding major toxin groups rather than individual toxin variants.
These groups, microcystins, cylindrospermopsins, anatoxins and saxitoxins, make up the
largest share of known cyanobacterial toxins. Besides these, a few individual toxins have
been described, such as guanitoxin or lyngbyatoxins. For microcystin variants alone,
about 280 structural variants have been characterised  and new variants are still being
found. One anatoxin-a variant, dihydro-anatoxin-a has been recently recognised to be
abundant in many samples of benthic cyanobacteria , often exceeding the abundance
A tentative conclusion from several decades of cyanotoxin research may be that
chances for discovering a new group of cyanotoxins which occurs widely in aquatic (or
terrestrial) ecosystems are low, and that toxic cyanobacterial compounds yet to be discov-
ered are likely to be confined to cyanobacterial taxa that themselves are not widely dis-
tributed or that are produced by only a limited number of genotypes of more widely dis-
Water 2021, 13, 2463 26 of 41
An example of a recently discovered toxin with (yet) regionally limited occurrence is
aetokthonotoxin, a polybrominated tryptophane derivative that is suspected to be the
cause of the death of bald eagles and water birds in the United States . Brominated
metabolites are known primarily from marine cyanobacteria  while halogenated me-
tabolites in freshwater cyanobacteria are generally chlorinated . Aetokthonotoxin is
produced by Aetokthonos hydrillicola, an epiphytic cyanobacterium that morphologically
resembles Fischerella sp. (with true branching) and that is found associated with the inva-
sive water plant Hydrilla verticillata which is spreading throughout the southeast of the
United States . Many bioactive metabolites have been reported from the group of
cyanobacteria to which Aetokthonos belongs (classically termed “Stigonematales“, includ-
ing, for example Fischerella, Stigonema; ), but since most genera of this group do not
form water blooms, so far only few studies describe toxin production . At present,
aetokthonotoxin appears to be confined to a comparatively low number of aquatic ecosys-
tems and it remains to be seen whether it will reach a broader distribution.
Another cyanobacterial compound that earned considerable attention as a possible
potent toxin during the last decade is β-methylaminoalanine (BMAA), a non-proteino-
genic amino acid. However, a number of inconsistencies in respective studies, in particu-
lar with respect to accuracy of detection and quantification [217,218], led Chernoff et al.
 to the conclusion that, based on available data, BMAA does not pose a relevant health
risk for human populations. This conclusion does not deny the toxic effects observed in a
number of bioassays, where BMAA primarily causes neurological damages. The main cri-
tique of Chernoff and co-workers is that doses applied in bioassays were orders of magni-
tude higher than what could be assumed as realistic exposure. This point of view has been
challenged by Dunlop et al. , and it remains to be seen whether clear evidence can be
produced that BMAA can lead to adverse health effects at exposure levels consistent with
its occurrence in water environments (and quantified with appropriate techniques).
This highlights a general problem: the uncertainty what should be considered a toxin.
In toxicology, a toxin is not only defined by the observation of toxic effects in bioassays,
rather, the effects need to be observable when testing doses that are not too far above
realistically possible exposures . In other words, many compounds are indeed toxic
at sufficiently high dose, and only if exposure pathways exist that can cause a toxic dose
under realistic exposure scenarios—including uncertainty factors considered in the calcu-
lation of guideline values (see below)—is a compound legitimately classified as toxin.
Many compounds essential for healthy human nutrition show adverse or toxic effects at
doses substantially exceeding the usual uptake: for example, doses of some essential mi-
cronutrients such as vitamin A  or zinc  only 10 times the recommended daily
uptake may already result in adverse health effects. Labelling a cyanobacterial compound
as ‘toxin’ requires assessment by experienced toxicologists, with toxicity data preferably
evaluated by a team. Further, distinction is relevant between toxicity to mammals, includ-
ing humans, and to aquatic animals such as planktonic Crustacea.
A further aspect is that in vitro effects such as enzyme inhibition, toxicity to isolated
cells or to membranes do not allow the conclusion of toxicity to an entire animal, in which
excretion and detoxification mechanisms may well prevent the substance from reaching
the organs or cells that showed sensitivity in vitro. For this reason, derivation of guideline
values or standards is still based on whole animal studies performed following standard-
ised OECD protocols (reviewed in Lawton et al. ).
In contrast, there seems to be a recent trend to “upgrade” known cyanobacterial me-
tabolites from “bioactive” to “toxic”. For example, Sieber et al.  describe a “toxic chy-
moptrypsin inhibitor” that showed toxicity in Thamnocephalus bioassays “roughly one order
of magnitude weaker than other cyanobacterial toxins”. Although it is of substantial interest
to study cyanobacterial metabolites, many of which were isolated because they show bioac-
tivity , a labelling as toxin should be done with diffidence. Similarly, Roy-Lachapelle et
al.  wrote “Harmful algal blooms of cyanobacterial origin have the potential to generate hun-
Water 2021, 13, 2463 27 of 41
dreds of secondary metabolites referred to as cyanotoxins”, thus suggesting that any cyanobacte-
rial compound is a toxin—which is simply not true. Of the hundreds of secondary metabo-
lites found in a cyanobacterial bloom, only a small share are true toxins, although most com-
pounds or compound classes may have been shown to be bioactive in various assays
[227,228]. It is important to limit the classification of cyanobacterial metabolites as ‘toxins’
to those for which exposure can indeed potentially be harmful in order to focus public health
capacity on those hazards which indeed can cause a risk to human health.
4.3. Which Cyanotoxin Concentrations Actually Present a Threat to Human Health?
The provisional guideline value of the World Health Organisation (WHO) for Micro-
cystin-LR in drinking water of 1 µg/L has been widely used to discuss the health implica-
tions of microcystin concentrations found in waterbodies, not only by limnologists but
also by public health professionals. When concentrations of up to 2.5 µg/L were found in
the drinking water of Toledo, USA, in August 2014, for two days half a million consumers
were advised not to drink the water . Optimising drinking water treatment reduced
concentrations below 1 µg/L, and the treatment plant was later upgraded to better cope
with cyanobacterial blooms. However, besides the costs of supplying bottled water to this
population, consequences of the two days of ‘do not drink’ advisory, such as closure of
restaurants and schools, has a major impact on people’s lives and income, raising the gen-
eral question whether health risks from brief exposure to concentrations in the lower 1-
digit range justify such consequences and whether immediately focusing all available
funding on upgrading treatment would not be the better, more sustainable option. The
incident thus spotlighted the need for short-term guideline values. Additionally, the more
recent availability of toxicological data on cylindrospermopsin now allowed derivation of
guideline values for this toxin. In response, the WHO expert group on chemicals in drink-
ing water re-assessed the available information and developed further cyanotoxin guide-
line values, including for short-term exposure via drinking water and recreation. These
now cover all four major cyanotoxin groups—microcystins MCs), cylindrospermopsins
(CYNs), Anatoxin-a and congeners (ATXs) and saxitoxins (STXs). WHO also gives back-
ground documents explaining the considerations for their derivation [230–233]. In Section
4.3.1 we briefly summarise the background of these derivations and in Section 4.3.2 we
discuss their implications for threats to human health.
4.3.1. Current WHO Guideline Values for Four Groups of Cyanotoxins and How They
Chemicals in drinking water are generally regulated very conservatively. For those
with a toxicity threshold and derived from long-term animal experiments, a factor of 1000
is frequently used between the dose that caused no effects (NOAEL—No Observed Ad-
verse Effect Level) and the tolerable daily intake (TDI). This is based on a factor of 10 for
each of three possibilities: (i) higher individual sensitivity (intraspecies variability), (ii)
higher sensitivity of humans as compared to the laboratory animals (interspecies vari-
ance) and (iii) uncertainties in the toxicological data base. For (iii) the factor can be lower
if the toxicological data are comprehensive or if these are supported by epidemiological
data (as is the case for STX). These three factors are multiplied to ensure sufficient protec-
tion even if all three conditions apply simultaneously. However, the probability of all
three of these possibilities to apply to a given substance is low, and should one possibility
actually require a higher factor, this is likely to be balanced by one or both of the other
two needing only a lower factor. Thus, multiplying all three factors (10 × 10 × 10 = 1000)
gives a high margin of safety. For further information on guideline derivation, see WHO
Figure 3 illustrates the derivation of the provisional WHO guideline value for micro-
cystin-LR in drinking water from a 13-week mouse study : no treatment-related ef-
fects were observed at a dose of 40 µg/kg per day (NOAEL), while at the highest dose
tested (1000 µg/kg per day) hepatic lesions typical for MC-LR appeared in all animals, and
Water 2021, 13, 2463 28 of 41
at the lowest observed adverse effect level (LOAEL) of 200 µg/kg per day some animals
showed slight hepatic damage.
Figure 3. Schematic illustration of the derivation of tolerable daily intakes (TDI) and guideline val-
ues (GV) for lifetime daily exposure to 2 L of drinking water. The values for LOAEL and NOAEL
are from the 13-week mouse study for Microcystin-LR by Fawell et al. . The leftmost cube il-
lustrates the derivation of TDI from NOAEL by applying uncertainty factors for interspecies varia-
bility (×10), intraspecies variance (×10), and data uncertainty for less than lifetime exposure (×10).
From TDI the GV is derived by assuming a bodyweight of 60 kg, a proportion of uptake via drinking
water of P = 0.8, and a daily water consumption of C = 2 L/day.
Table 2 shows the current WHO guideline values together with the assumptions
WHO used to derive them, that is, the NOAEL, bodyweights, uncertainty factors, the vol-
umes of water to which people are likely to be exposed as well as the proportion of expo-
sure allocated to different pathways. Note that for MC and CYN these guideline values
are provisional due to limitations of the toxicological data, including data being available
only for one key congener (probably the most toxic one) for each toxin group. For short-
term and recreational exposure, extrapolation from the NOAEL uses lower uncertainty
factors for gaps in the toxicological data because uncertainties when extrapolating from
part of a lifetime in the animal experiment to a full lifetime do not apply for an exposure
of up to two weeks. For Anatoxin-a and its variants (ATXs), in an experiment designed
similarly to that for MC-LR quoted above, none of the dose levels tested indicated damage
due to ATX-a, and thus no NOAEL could be derived. However, the data do allow giving
an upper bounding level below which toxic effects are very unlikely, termed ‘health-based
reference value’. For saxitoxins (STXs), data for chronic exposure are lacking, but for acute
exposure they are available from incidents of food poisoning with marine shellfish. Note
that the acute guideline value for STXs in drinking water is derived on the basis of the
bodyweight of an infant and all recreational values are based on the bodyweight of a small
child. Refer to the WHO background documents [230–233] for more information on the
considerations leading to these derivations as well as guidance on their application, in-
cluding the understanding of ‘short-term’ exposure. Note also that while the derivation is
based on toxicological data for one main congener (due to lacking data for the others),
WHO recommends application to the sum of all congeners present in a sample.
Water 2021, 13, 2463 29 of 41
Table 2. WHO cyanotoxin guideline values and health-based reference values, respectively, and assumptions used for
their derivation. MCs: microcystins; CYNs: cylindrospermopsins; ATXs: anatoxins; STXs: saxitoxins.
Toxin and Exposure
to Other Exposure
MCs, drinking water
lifetime 40 µg/kg
day 1000 80% 2 L 60 kg 1 µg/L
MCs, drinking water
day 100 100% 2 L 60 kg 12 µg/L
MCs, recreation 40 µg/kg
day 100 100% 0.25 L 15 kg 24 µg/L
CYNs, drinking water
lifetime 30 µg/kg
day 1000 80% 2 L 60 kg 0.7 µg/L
CYNs, drinking water
short-term * 30 µg/kg
day 300 100% 2 L 60 kg 3 µg/L
CYNs, recreation 30 µg/kg
day 300 100% 0.25 L 15 kg 6 µg/L
ATXs, drinking water
short-term * 98 µg/kg
day 100 100% 2 L 60 kg 30 µg/L **
ATXs, recreation 98 µg/kg
day 100 100% 0,25 L 15 kg 60 µg/L**
STXs, drinking water
(LOAEL) 3 100% 0.75 L 5 kg 3 µg/L
STXs, recreation 1.5 µg/kg
(LOAEL) 3 100% 0.25 L 15 kg 30 µg/L
* Short-term exposure refers to about two weeks, is not intended for multiple periods per season and is linked to informing
the public, recommending bottled water for infants and small children as well as initiation of measures to remediate the
situation. ** Health-based reference value as upper limit.
4.3.2. Cyanotoxins in the Context of Other Health Hazards in Water
Among the hazards occurring in water, the public health impact of pathogens gener-
ally far outweighs that of chemicals: for 2016, WHO  estimates almost two million
deaths worldwide from infectious diseases attributed to quantifiable effects of inadequate
water, sanitation and hygiene, amounting to 3.3% of all deaths, and in terms of disease bur-
den to 4.6% of all disability-adjusted life years (DALYs). In contrast, for chemicals in water
the WHO Guidelines for Drinking-water Quality  discuss that generally these rarely
occur in concentrations causing adverse health effects unless exposure occurs for pro-
longed periods of time, that is, in the range of years. Additionally, a scan through the
WHO background documents for many chemicals in drinking water shows that the chief
exposure pathways to most of the substances are air and food. For recreational water qual-
ity the WHO guidelines note that reports of human health impacts associated with chem-
icals in water are rare .
Acute illness attributable to exposure to chemicals in water is largely limited to inci-
dents of massive accidental contamination and even in these, there may be little exposure
through ingestion because such water is likely to be “undrinkable owing to unacceptable taste,
odour and appearance” . While bloom-containing water is indeed not palatable, drink-
ing water treatment may improve palatability by removing much of the organic matter,
and if treatment is not appropriately designed and managed, efficacy of cyanotoxin re-
moval may be limited, leading to exposure . Recreational exposure may lead to sub-
stantial cyanotoxin ingestion, and for this exposure pathway, cyanotoxins may well be the
most relevant group of chemicals . Criteria for assessing short-term exposure through
drinking water and through recreation are therefore important.
Interestingly, chronic exposure to chemicals in drinking water at levels causing
demonstratable illness is chiefly known from natural geogenic sources, in particular for
arsenic and fluoride, in some regions also uranium . While cyanotoxins are natural
substances, in most cases their occurrence in health-relevant concentrations is not natural,
Water 2021, 13, 2463 30 of 41
but due to widespread anthropogenic eutrophication, and depending on climatic and hy-
drophysical conditions, blooms may last for many months . Among the chemicals in
water to which people are exposed, cyanotoxins may therefore be those occurring most
The overall evidence for illness attributed to chemicals in water also applies to cy-
anotoxins: although some are very potent and massive bloom situations can lead to high
concentrations, there are not many cases of well-substantiated evidence for cyanotoxins
as cause of human intoxication (reviewed in Humpage et al. , for drinking water and
by Chorus and Testai  for recreational exposure). Where symptoms are unspecific
and not typical for the respective cyanotoxin(s), such as nausea, diarrhoea or skin irrita-
tion, other agents may as well have been the cause, including pathogens, possibly associ-
ated with the bloom or reaching the water together with nutrients feeding the bloom.
Prematurely dismissing these may risk not finding the real cause, and WHO discusses
criteria for attributing health complaints to cyanotoxins (see the introduction to chapter 5
in Chorus and Welker  and WHO ).
In consequence, a low one-digit number of µg/L of cyanotoxins does not imply a ‘se-
rious threat’ to human health if it occurs only for a few days, particularly if such concen-
trations are found in a waterbody and not in drinking water. However, blooms cause a
high load of organic material, and where treatment does not effectively remove this from
drinking water which then is chlorinated, they can cause elevated concentrations of chlo-
rination by-products, some of which are suspect carcinogens and cause unpleasant taste
and odour, thus causing further health concerns.
4.4. Implications of Shifts to Benthic or Metalimnetic Cyanbacteria for Human Exposure
As discussed in Section 3.3, increased water transparency through oligotrophication
opens habitats in deeper water layers and on the sediment surface, and these may be pop-
ulated by benthic, periphytic or metalimnetic cyanobacteria. Dog deaths from ingestion
of beached material at recreational sites or water in a drinking water reservoir deeply
brownish-red from Plankothrix rubescens draw substantial public attention and cause con-
cern. Such phenomena challenge water managers, particularly if they appear as water gets
clearer in consequence of investments in nutrient reduction. However, for risks of human
exposure the substantial differences in spatial distribution within the waterbody are rele-
vant: epilimnetic blooms can cause high cyanotoxin concentrations particularly if they
form scums, and where scums accumulate in bays used for recreational activities, partic-
ularly bathing, toxin concentrations can reach a range of mg/L, risking acute symptoms
through oral uptake . In contrast, Planktothrix rubescens resides in the metalimnion
during the summer, and total waterbody mixing occurs outside the main season for rec-
reational waterbody use, rendering recreational exposure risks less likely.
For lumps of detached floating or beached benthic material also, risks of oral uptake
appear unlikely. Dog deaths occur because—in contrast to humans—these animals are
attracted to this material and thus ingest substantial amounts. However, concentrations
in the surrounding water are typically low  because anatoxin-a released by lysing
cells is quickly diluted well below concentrations in the range of the WHO health-based
reference values (30 µg/L for drinking water and 60 µg/L for recreational water; see Table
2). Dilution is particularly relevant in streams in which water flowing over benthic mats
will remove released ATXs, but also in lakes and reservoirs where water volumes are large
relative to the amount of such material. Thus, while scenarios of concern cannot totally be
excluded (e.g., children playing with lumps of beached material or persons ingesting this
in near-drowning accident situations), human health risks from recreational exposure to
benthic cyanobacteria appear much lower than those from surface scums. Based on the
same considerations, this also applies to drinking water production. Wood  devel-
oped recommendations for informing the public in New Zealand to avoid contact and
particularly oral uptake of mats or lumps of benthic cyanobacteria with particular empha-
sis on small children as vulnerable group.
Water 2021, 13, 2463 31 of 41
In contrast, risks of cyanotoxins from P. rubescens in drinking water offtakes depend
on the specific situation: where thermally stratified reservoirs are the drinking water
source and drinking water offtakes are constructed with variable depths, the risk of ab-
stracting water from the metalimnetic layers of this species can be minimised. These layers
are typically quite thin and distinct and can be avoided by adjusting the offtake depth
. Where adequate systems of monitoring and management responses are in place,
cyanotoxin risks from P. rubescens can thus be more readily controlled than those from
epilimnetic blooms: removing the latter requires elaborate treatment not only addressing
toxins, but typically also the high concentration of organic substances caused by them
In consequence, where (re-)oligotrophication of waterbodies shifts species to met-
alimnetic or benthic cyanobacteria, cyanotoxin health risks are likely to be lower than
those from epilimnetic blooms during the hypertrophic or eutrophic phase of the respec-
5. Summary and Conclusions
The state of knowledge that we reviewed here is necessarily incomplete, largely omit-
ting influences of catchment conditions as well as of food-webs, and only touching upon
the complex mechanisms driving nutrient release from sediments. We do, however, be-
lieve that it sufficiently highlights the need for careful differentiation of the messages we
derive regarding the three questions we ask in the introduction. For these we propose the
1. Climate change, in particular warming, is likely to promote blooms in some water-
bodies but not in all. In waterbodies which lack the carrying capacity to sustain high
phytoplankton biomass, elevated water temperatures may chiefly affect species com-
position and dynamics rather than total biomass. Effects of climate change may be
more pronounced in shallow than in stratified waterbodies because in the former in-
direct effects such as sediment resuspension during storms are stronger. In deep strat-
ified lakes, climate change can work both ways—rendering conditions for cyanobac-
terial proliferation more favourable or less favourable. Therefore, statements gener-
ally claiming that climate change will ‘globally’ lead to ‘ever increasing’ blooms are nei-
ther scientifically sound nor helpful for management.
2. Controlling cyanobacterial blooms can definitely be successful without also reducing
N, provided the concentrations of P can be brought down to sufficiently low levels.
The outcomes of the last century’s research still hold: if one nutrient limits the carrying
capacity for biomass, excess concentrations of the other cannot promote significant
further growth. Not the N:P ratio, but the absolute concentration of the limiting nu-
trient determines the maximum possible biomass of phytoplankton and thus of cya-
nobacteria. It follows that whether N or P are currently limiting biomass can be in-
ferred from their absolute concentrations.
Focusing investment on measures to reduce one nutrient is likely to be more effective
than spreading the available funding across measures addressing both. Whether the re-
duction of P or N is more efficient is specific to the waterbody, its catchment, and socio-
We emphasise that there is no question about the need to reduce the overall emis-
sions of N into the environment in face of the dramatic increase of emissions in many
regions of the world, not only into waterbodies but also into terrestrial ecosystems and
groundwater. Strategic planning needs to include assessments of the wider environmental
impacts of local remediation measures.
3. Trophic and climatic changes will affect health risks from cyanobacteria. Trivially,
conditions leading to more cyanobacterial biomass can increase risks of exposure to
toxins. Health risks depend on the amount and time span by which cyanotoxin con-
centrations exceed WHO guideline values for the respective exposure pathway and
Water 2021, 13, 2463 32 of 41
time span. Whether trophic and climatic changes will increase health risks or not de-
pends on local conditions.
It is well understood that the cellular up- or downregulating of the production of
toxic metabolites is of minor relevance as compared to the genotype composition of field
populations. However, we are only beginning to understand the drivers of genotype dy-
namics. Understanding these requires experiments comparing their growth rates under
tightly defined conditions (preferably in (semi-)continuous cultures), including competi-
tion experiments between strains producing different metabolites. This implies combining
approaches and techniques developed in the 20th century with the new, 21st century
The current evidence for the impact of environmental conditions on toxic genotype
occurrence is too contradictory and the understanding gleaned from the data too poor to
draw conclusions for management. Premature conclusions for management are neither
helpful nor necessary. Rather, the focus on controlling the occurrence of cyanobacteria as
such is preferable because the sheer biomass of blooms is detrimental to aquatic ecosys-
tems and compromises human use of waterbodies.
Author Contributions: Conceptualisation, I.C., J.F. and M.W.; writing—original draft preparation,
I.C., J.F. and M.W.; writing—review and editing, I.C., J.F. and M.W.; visualization, I.C. and M.W.
All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: Special thanks are due to Max Tilzer for training in the IBP-inspired fundamen-
tals of production biology in the 1980s and very recently, to Hans Paerl and Mark McCarthy for
fruitful controversial discussion of the concept of limitation. The thorough review and helpful sug-
gestions to improve the manuscript by four anonymous reviewers is greatly appreciated.
Conflicts of Interest: No conflict of interest.
1. Paerl, H.W.; Huisman, J. Climate change: A catalyst for global expansion of harmful cyanobacterial blooms. Environ. Microbiol.
Rep. 2009, 1, 27–37.
2. Moss, B.; Kosten, S.; Meerhoff, M.; Battarbee, R.W.; Jeppesen, E.; Mazzeo, N.; Havens, K.; Lacerot, G.; Liu, Z.; De Meester, L.
Allied attack: Climate change and eutrophication. Inland Waters 2011, 1, 101–105.
3. Glibert, P.M. Harmful algae at the complex nexus of eutrophication and climate change. Harmful Algae 2020, 91, 101583.
4. Reynolds, C.; Maberly, S. A simple method for approximating the supportive capacities and metabolic constraints in lakes and
reservoirs. Freshwat. Biol. 2002, 47, 1183–1188.
5. Fiore, M.F.; de Lima, S.T.; Carmichael, W.W.; McKinnie, S.M.; Chekan, J.R.; Moore, B.S. Guanitoxin, re-naming a cyanobacterial
organophosphate toxin. Harmful Algae 2020, 92, 101737.
6. Paerl, H.W. Mitigating toxic planktonic cyanobacterial blooms in aquatic ecosystems facing increasing anthropogenic and
climatic pressures. Toxins 2018, 10, 76.
7. EEA. Nutrients in Freshwater in Europe. Available online: https://www.eea.europa.eu/data-and-maps/indicators/nutrients-in-
freshwater (accessed on 25 August 2021).
8. Chorus, I.; Koehler, A.; Beulker, C.; Fastner, J.; van de Weyer, K.; Hegewald, T.; Hupfer, M. Decades needed for ecosystem
components to respond to a sharp and drastic phosphorus load reduction. Hydrobiologia 2020, 847, 4621–4651.
9. Ho, J.C.; Michalak, A.M.; Pahlevan, N. Widespread global increase in intense lake phytoplankton blooms since the 1980s. Nature
2019, 574, 667–670.
10. Chorus, I.; Welker, M. (Eds.) Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring and Management,
2nd ed.; CRC Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; p. 858.
11. Janssen, E.M.-L. Cyanobacterial peptides beyond microcystins—A review on co-occurrence, toxicity, and challenges for risk
assessment. Water Res. 2019, 151, 488–499.
Water 2021, 13, 2463 33 of 41
12. Jones, M.R.; Pinto, E.; Torres, M.A.; Dörr, F.; Mazur-Marzec, H.; Szubert, K.; Tartaglione, L.; Dell'Aversano, C.; Miles, C.O.;
Beach, D.G. CyanoMetDB, a comprehensive public database of secondary metabolites from cyanobacteria. Water Res. 2021, 196,
13. Winder, M.; Sommer, U. Phytoplankton response to a changing climate. Hydrobiologia 2012, 698, 1–12.
14. Hamilton, D.P.; Salmaso, N.; Paerl, H.W. Mitigating harmful cyanobacterial blooms: Strategies for control of nitrogen and
phosphorus loads. Aquat. Ecol. 2016, 50, 351–366.
15. Coles, J.F.; Jones, R.C. Effect of temperature on photosynthesis-light response and growth of four phytoplankton species isolated
from a tidal freshwater river. J. Phycol. 2000, 36, 7–16.
16. Reynolds, C.S. Vegetation Processes in the Pelagic: A Model for Ecosystem Theory; Ecology Institute: Oldendorf/Luhe, Germany,
1997; Volume 9.
17. Butterwick, C.; Heaney, S.; Talling, J. Diversity in the influence of temperature on the growth rates of freshwater algae, and its
ecological relevance. Freshwat. Biol. 2005, 50, 291–300.
18. Thomas, M.K.; Litchman, E. Effects of temperature and nitrogen availability on the growth of invasive and native cyanobacteria.
Hydrobiologia 2016, 763, 357–369.
19. Paerl, H.W.; Otten, T.G. Harmful cyanobacterial blooms: Causes, consequences, and controls. Microb. Ecol. 2013, 65, 995–1010.
20. de Senerpont Domis, L.N.D.S.; Mooij, W.M.; Huisman, J. Climate-induced shifts in an experimental phytoplankton community:
A mechanistic approach. In Shallow Lakes in a Changing World; Gulati, R.D., Lammens, E., De Pauw, N., Van Donk, E., Eds.;
Springer: Cham, Switzerland, 2007; pp. 403–413.
21. Mehnert, G.; Leunert, F.; Cirés, S.; Jöhnk, K.D.; Rücker, J.; Nixdorf, B.; Wiedner, C. Competitiveness of invasive and native
cyanobacteria from temperate freshwaters under various light and temperature conditions. J. Plankton Res. 2010, 32, 1009–1021.
22. Lei, L.; Li, C.; Peng, L.; Han, B.-P. Competition between toxic and non-toxic Microcystis aeruginosa and its ecological implication.
Ecotoxicology 2015, 24, 1411–1418.
23. Ninio, S.; Lupu, A.; Viner-Mozzini, Y.; Zohary, T.; Sukenik, A. Multiannual variations in Microcystis bloom episodes—
Temperature drives shift in species composition. Harmful Algae 2020, 92, 101710.
24. Cottingham, K.L.; Weathers, K.C.; Ewing, H.A.; Greer, M.L.; Carey, C.C. Predicting the effects of climate change on freshwater
cyanobacterial blooms requires consideration of the complete cyanobacterial life cycle. J. Plankton Res. 2021, 43, 10–19.
25. Van de Waal, D.B.; Verspagen, J.M.; Finke, J.F.; Vournazou, V.; Immers, A.K.; Kardinaal, W.E.A.; Tonk, L.; Becker, S.; Van Donk,
E.; Visser, P.M. Reversal in competitive dominance of a toxic versus non-toxic cyanobacterium in response to rising CO
J. 2011, 5, 1438–1450.
26. Visser, P.M.; Verspagen, J.M.; Sandrini, G.; Stal, L.J.; Matthijs, H.C.; Davis, T.W.; Paerl, H.W.; Huisman, J. How rising CO
global warming may stimulate harmful cyanobacterial blooms. Harmful Algae 2016, 54, 145–159.
27. Shuvo, A.; O’Reilly, C.M.; Blagrave, K.; Ewins, C.; Filazzola, A.; Gray, D.; Mahdiyan, O.; Moslenko, L.; Quinlan, R.; Sharma, S.
Total phosphorus and climate are equally important predictors of water quality in lakes. Aquat. Sci. 2021, 83, 1–11.
28. Reichwaldt, E.S.; Ghadouani, A. Effects of rainfall patterns on toxic cyanobacterial blooms in a changing climate: Between
simplistic scenarios and complex dynamics. Water Res. 2012, 46, 1372–1393.
29. Huisman, J.; Codd, G.A.; Paerl, H.W.; Ibelings, B.W.; Verspagen, J.M.; Visser, P.M. Cyanobacterial blooms. Nat. Rev. Microbiol.
2018, 16, 471.
30. Yang, Z.; Zhang, M.; Shi, X.; Kong, F.; Ma, R.; Yu, Y. Nutrient reduction magnifies the impact of extreme weather on
cyanobacterial bloom formation in large shallow Lake Taihu (China). Water Res. 2016, 103, 302–310.
31. Morabito, G.; Rogora, M.; Austoni, M.; Ciampittiello, M. Could the extreme meteorological events in Lake Maggiore watershed
determine a climate-driven eutrophication process? Hydrobiologia 2018, 824, 163–175.
32. Tang, Q.; Peng, L.; Yang, Y.; Lin, Q.; Qian, S.S.; Han, B.-P. Total phosphorus-precipitation and Chlorophyll a-phosphorus
relationships of lakes and reservoirs mediated by soil iron at regional scale. Water Res. 2019, 154, 136–143.
33. Litchman, E.; Klausmeier, C.A.; Schofield, O.M.; Falkowski, P.G. The role of functional traits and trade-offs in structuring
phytoplankton communities: Scaling from cellular to ecosystem level. Ecol. Lett. 2007, 10, 1170–1181.
34. Anneville, O.; Domaizon, I.; Kerimoglu, O.; Rimet, F.; Jacquet, S. Blue-green algae in a “Greenhouse Century”? New insights
from field data on climate change impacts on cyanobacteria abundance. Ecosystems 2015, 18, 441–458.
35. Rigosi, A.; Carey, C.C.; Ibelings, B.W.; Brookes, J.D. The interaction between climate warming and eutrophication to promote
cyanobacteria is dependent on trophic state and varies among taxa. Limnol. Oceanogr. 2014, 59, 99–114.
36. Mantzouki, E.; Lürling, M.; Fastner, J.; de Senerpont Domis, L.; Wilk-Woźniak, E.; Koreivienė, J.; Seelen, L.; Teurlincx, S.;
Verstijnen, Y.; Krztoń, W.; et al. Temperature effects explain continental scale distribution of cyanobacterial toxins. Toxins 2018,
37. Jöhnk, K.; Huisman, J.; Sharples, J.; Sommeijer, B.; Visser, P.; Stroom, J. Summer heatwaves promote blooms of harmful
cyanobacteria. Glob. Chang. Biol. 2008, 14, 495–512.
38. Straile, D.; Kerimoglu, O.; Peeters, F.; Jochimsen, M.C.; Kümmerlin, R.; Rinke, K.; Rothaupt, K.-O. Effects of a half a millennium
winter on a deep lake—A shape of things to come? Glob. Chang. Biol. 2010, 16, 2844–2856.
39. Posch, T.; Köster, O.; Salcher, M.M.; Pernthaler, J. Harmful filamentous cyanobacteria favoured by reduced water turnover with
lake warming. Nat. Clim. Chang. 2012, 2, 809.
40. Salmaso, N.; Boscaini, A.; Capelli, C.; Cerasino, L. Ongoing ecological shifts in a large lake are driven by climate change and
eutrophication: Evidences from a three-decade study in Lake Garda. Hydrobiologia 2018, 824, 177–195.
Water 2021, 13, 2463 34 of 41
41. Elliott, J.; Jones, I.; Thackeray, S. Testing the sensitivity of phytoplankton communities to changes in water temperature and
nutrient load, in a temperate lake. Hydrobiologia 2006, 559, 401–411.
42. Kraemer, B.M.; Mehner, T.; Adrian, R. Reconciling the opposing effects of warming on phytoplankton biomass in 188 large
lakes. Sci. Rep. 2017, 7, 1–7.
43. Gallina, N.; Anneville, O.; Beniston, M. Impacts of extreme air temperatures on cyanobacteria in five deep peri-Alpine lakes.
Limnol. Oceanogr. 2011, 70, 186–196.
44. Hennemann, M.C.; Petrucio, M.M. High chlorophyll a concentration in a low nutrient context: Discussions in a subtropical lake
dominated by Cyanobacteria. J. Limnol. 2016, 75, https://doi.org/10.4081/jlimnol.2016.1347.
45. Reinl, K.L.; Brookes, J.D.; Carey, C.C.; Harris, T.D.; Ibelings, B.W.; Morales-Williams, A.M.; de Senerpont Domis, L.N.; Atkins,
K.S.; Isles, P.D.; Mesman, J.P. Cyanobacterial blooms in oligotrophic lakes: Shifting the high-nutrient paradigm. Freshwat. Biol.
2021, 14, https://doi.org/10.1111/fwb.13791.
46. Wentzky, V.C.; Tittel, J.; Jäger, C.G.; Rinke, K. Mechanisms preventing a decrease in phytoplankton biomass after phosphorus
reductions in a German drinking water reservoir—results from more than 50 years of observation. Freshwat. Biol. 2018, 63, 1063–
47. Verburg, P.; Hecky, R.E.; Kling, H. Ecological consequences of a century of warming in Lake Tanganyika. Science 2003, 301, 505–
48. Farrell, K.J.; Ward, N.K.; Krinos, A.I.; Hanson, P.C.; Daneshmand, V.; Figueiredo, R.J.; Carey, C.C. Ecosystem-scale nutrient
cycling responses to increasing air temperatures vary with lake trophic state. Ecol. Model. 2020, 430, 109134.
49. Hupfer, M.; Lewandowski, J. Oxygen controls the phosphorus release from lake sediments—A long-lasting paradigm in
limnology. Internat. Rev. Hydrobiol. 2008, 93, 415–432.
50. Tammeorg, O.; Nürnberg, G.; Horppila, J.; Haldna, M.; Niemistö, J. Redox-related release of phosphorus from sediments in
large and shallow Lake Peipsi: Evidence from sediment studies and long-term monitoring data. J. Great Lakes Res. 2020, 46, 1595–
51. Chorus, I.; Schauser, I. Oligotrophication of Lake Tegel and Schlachtensee, Berlin-Analysis of System Components, Causalities and
Response Thresholds Compared to Responses of Other Waterbodies; Umweltbundesamt: Dessau, Germany, 2011; p. 157.
52. Kosten, S.; Huszar, V.L.; Bécares, E.; Costa, L.S.; Donk, E.; Hansson, L.A.; Jeppesen, E.; Kruk, C.; Lacerot, G.; Mazzeo, N. Warmer
climates boost cyanobacterial dominance in shallow lakes. Glob. Chang. Biol. 2012, 18, 118–126.
53. Moustaka-Gouni, M.; Sommer, U. Effects of harmful blooms of large-sized and colonial cyanobacteria on aquatic food webs.
Water 2020, 12, 1587.
54. Selmeczy, G.B.; Abonyi, A.; Krienitz, L.; Kasprzak, P.; Casper, P.; Telcs, A.; Somogyvári, Z.; Padisák, J. Old sins have long
shadows: Climate change weakens efficiency of trophic coupling of phyto-and zooplankton in a deep oligo-mesotrophic
lowland lake (Stechlin, Germany)—A causality analysis. Hydrobiologia 2019, 831, 101–117.
55. Richardson, J.; Miller, C.; Maberly, S.C.; Taylor, P.; Globevnik, L.; Hunter, P.; Jeppesen, E.; Mischke, U.; Moe, S.J.; Pasztaleniec,
A. Effects of multiple stressors on cyanobacteria abundance vary with lake type. Glob. Chang. Biol. 2018, 24, 5044–5055.
56. Mooij, W.M.; Trolle, D.; Jeppesen, E.; Arhonditsis, G.; Belolipetsky, P.V.; Chitamwebwa, D.B.; Degermendzhy, A.G.; DeAngelis,
D.L.; de Senerpont Domis, L.N.; Downing, A.S.; et al. Challenges and opportunities for integrating lake ecosystem modelling
approaches. Aquat. Ecol. 2010, 44, 633–667.
57. Vollenweider, R.A. Scientific Fundamentals of the Eutrophication of Lakes and Flowing Waters, with Particular Reference to Nitrogen
and Phosphorus as Factors in Eutrophication; Tech. Rep. DA 5/SCI/68.27; Organisation for Economic Co-operation and
Development: Paris, France, 1968; pp. 1–250.
58. Köhler, J. Detergent phosphates: An EU policy assessment. J. Business Chem. 2006, 3, 15–30.
59. Van Drecht, G.; Bouwman, A.; Harrison, J.; Knoop, J. Global nitrogen and phosphate in urban wastewater for the period 1970
to 2050. Glob. Biogeochem. Cycl. 2009, 23, https://doi.org/10.1029/2009GB003458.
60. Lee, G.F.; Jones, R.A. Detergent phosphate bans and eutrophication. Environ. Sci. Technol. 1986, 20, 330–331.
61. Bashar, R.; Gungor, K.; Karthikeyan, K.; Barak, P. Cost effectiveness of phosphorus removal processes in municipal wastewater
treatment. Chemosphere 2018, 197, 280–290.
62. McCarty, P.L. What is the best biological process for nitrogen removal: When and why? Environ. Sci. Technol. 2018, 52, 3835–
63. Winkler, M.K.; Straka, L. New directions in biological nitrogen removal and recovery from wastewater. Curr. Opin. Biotechnol.
2019, 57, 50–55.
64. Glibert, P.M.; Maranger, R.; Sobota, D.J.; Bouwman, L. Further evidence of the Haber-Bosch—harmful algal bloom (HB-HAB)
link and the risk of suggesting HAB control through phosphorus reductions only. In Just Enough Nitrogen; Sutton, M.A., Mason,
K.E., Bleeker, A., Hicks, W.K., Masso, C., Raghuram, N., Reis, S., Bekunda, M., Eds.; Springer: Cham, Switzerland, 2020; pp.
65. Paerl, H.W.; Havens, K.E.; Xu, H.; Zhu, G.; McCarthy, M.J.; Newell, S.E.; Scott, J.T.; Hall, N.S.; Otten, T.G.; Qin, B. Mitigating
eutrophication and toxic cyanobacterial blooms in large lakes: The evolution of a dual nutrient (N and P) reduction paradigm.
Hydrobiologia 2020, 847, 4359–4375.
66. Watson, S.B.; Miller, C.; Arhonditsis, G.; Boyer, G.L.; Carmichael, W.; Charlton, M.N.; Confesor, R.; Depew, D.C.; Höök, T.O.;
Ludsin, S.A. The re-eutrophication of Lake Erie: Harmful algal blooms and hypoxia. Harmful Algae 2016, 56, 44–66.
Water 2021, 13, 2463 35 of 41
67. Conley, D.J.; Paerl, H.W.; Howarth, R.W.; Boesch, D.F.; Seitzinger, S.P.; Karl, E.; Lancelot, C.; Gene, E. Controlling eutrophication:
Nitrogen and phosphorus. Science 2009, 123, 1014–1015.
68. Salmaso, N.; Tolotti, M. Phytoplankton and anthropogenic changes in pelagic environments. Hydrobiologia 2021, 848, 251–284.
69. Li, J.; Dittrich, M. Dynamic polyphosphate metabolism in cyanobacteria responding to phosphorus availability. Environ.
Microbiol. 2019, 21, 572–583.
70. Canizales, S.; Sliwszcinka, M.; Russo, A.; Bentvelzen, S.; Temmink, H.; Verschoor, A.; Wijffels, R.; Janssen, M. Cyanobacterial
growth and cyanophycin production with urea and ammonium as nitrogen source. J. Appl. Phycol. 2021, 1–13,
71. Watzer, B.; Spät, P.; Neumann, N.; Koch, M.; Sobotka, R.; Macek, B.; Hennrich, O.; Forchhammer, K. The signal transduction
protein PII controls ammonium, nitrate and urea uptake in cyanobacteria. Front. Microbiol. 2019, 10, 1428.
72. Flores, E.; Arévalo, S.; Burnat, M. Cyanophycin and arginine metabolism in cyanobacteria. Algal Res. 2019, 42, 101577.
73. Bothe, H.; Schmitz, O.; Yates, M.G.; Newton, W.E. Nitrogen fixation and hydrogen metabolism in cyanobacteria. Microbiol. Mol.
Biol. Rev. 2010, 74, 529–551.
74. Scott, J.T.; McCarthy, M.J. Nitrogen fixation may not balance the nitrogen pool in lakes over timescales relevant to
eutrophication management. Limnol. Oceanogr. 2010, 55, 1265–1270.
75. Chorus, I.; Spijkerman, E. What Colin Reynolds could tell us about nutrient limitation, N:P ratios and eutrophication control.
Hydrobiologia 2021, 848, 95-111.
76. Cáceres, C.; Spatharis, S.; Kaiserli, E.; Smeti, E.; Flowers, H.; Bonachela, J.A. Temporal phosphate gradients reveal diverse
acclimation responses in phytoplankton phosphate uptake. ISME J. 2019, 13, 2834–2845.
77. Edwards, K.F.; Thomas, M.K.; Klausmeier, C.A.; Litchman, E. Allometric scaling and taxonomic variation in nutrient utilization
traits and maximum growth rate of phytoplankton. Limnol. Oceanogr. 2012, 57, 554–566.
78. Dolman, A.M.; Rücker, J.; Pick, F.R.; Fastner, J.; Rohrlack, T.; Mischke, U.; Wiedner, C. Cyanobacteria and cyanotoxins: The
influence of nitrogen versus phosphorus. PLoS ONE 2012, 7, e38757.
79. Kolzau, S.; Wiedner, C.; Rücker, J.; Köhler, J.; Köhler, A.; Dolman, A.M. Seasonal patterns of nitrogen and phosphorus limitation
in four German lakes and the predictability of limitation status from ambient nutrient concentrations. PLoS ONE 2014, 9, e96065.
80. Reynolds, C. What factors influence the species composition of phytoplankton in lakes of different trophic status? Hydrobiologia
1998, 369, 11–26.
81. Reynolds, C.S. Non-determinism to probability, or N:P in the community ecology of phytoplankton: Nutrient ratios. Arch.
Hydrobiol. 1999, 146, 23–35.
82. Reynolds, C.; Bellinger, E. Patterns of abundance and dominance of the phytoplankton of Rostherne Mere, England: Evidence
from an 18-year data set. Aquat. Sci. 1992, 54, 10–36.
83. Lampert, W.; Sommer, U. Limnoecology: The Ecology of Lakes and Streams; Oxford University Press: New York, NY, USA, 2007.
84. Vollenweider, R.; Kerekes, J. Loading concept as basis for controlling eutrophication philosophy and preliminary results of the
OECD programme on eutrophication. Prog. Water Technol. 1980, 12, 5–38.
85. Phillips, G.; Pietiläinen, O.-P.; Carvalho, L.; Solimini, A.; Solheim, A.L.; Cardoso, A. Chlorophyll–nutrient relationships of
different lake types using a large European dataset. Aquat. Ecol. 2008, 42, 213–226.
86. Carvalho, L.; McDonald, C.; de Hoyos, C.; Mischke, U.; Phillips, G.; Borics, G.; Poikane, S.; Skjelbred, B.; Solheim, A.L.; Van
Wichelen, J. Sustaining recreational quality of European lakes: Minimizing the health risks from algal blooms through
phosphorus control. J. Appl. Ecol. 2013, 50, 315–323.
87. Fastner, J.; Abella, S.; Litt, A.; Morabito, G.; Vörös, L.; Pálffy, K.; Straile, D.; Kümmerlin, R.; Matthews, D.; Phillips, M.G.; et al.
Combating cyanobacterial proliferation by avoiding or treating inflows with high P load—Experiences from eight case studies.
Aquat. Ecol. 2016, 50, 367–383.
88. Salonen, K.; Sarvala, J.; Horppila, J.; Keto, J.; Malin, I.; Malinen, T.; Niemistö, J.; Ruuhijärvi, J. Development of Lake Vesijärvi
through four decades of remediation efforts. Hydrobiologia 2020, 847, 4601–4619.
89. Quinlan, R.; Filazzola, A.; Mahdiyan, O.; Shuvo, A.; Blagrave, K.; Ewins, C.; Moslenko, L.; Gray, D.K.; O'Reilly, C.M.; Sharma,
S. Relationships of total phosphorus and chlorophyll in lakes worldwide. Limnol. Oceanogr. 2021, 66, 392–404.
90. Redfield, A. On the proportions of organic derivations in seawater and their relation to the composition of plankton. In James
Johnstone Memorial Volume; Daniel, R., Ed.; University Press of Liverpool: Liverpool, UK, 1934; pp. 176–192.
91. Redfield, A.C. The biological control of chemical factors in the environment. Am. Sci. 1958, 46, 205–221.
92. Klausmeier, C.A.; Litchman, E.; Daufresne, T.; Levin, S.A. Optimal nitrogen-to-phosphorus stoichiometry of phytoplankton.
Nature 2004, 429, 171–174.
93. Søndergaard, M.; Lauridsen, T.L.; Johansson, L.S.; Jeppesen, E. Nitrogen or phosphorus limitation in lakes and its impact on
phytoplankton biomass and submerged macrophyte cover. Hydrobiologia 2017, 795, 35–48.
94. Mur, L.R.; Gons, H.J.; van Liere, L. Competition of the green alga Scenedesmus and the blue-green alga Oscillatoria: With 5 figures
in the text. Int. Ver. Theor. Angew. Limnol. Mitt. 1978, 21, 473–479.
95. Downing, J.A.; Watson, S.B.; McCauley, E. Predicting cyanobacteria dominance in lakes. Can. J. Fish. Aquat. Sci. 2001, 58, 1905–
96. Jeppesen, E.; Søndergaard, M.; Jensen, J.P.; Havens, K.E.; Anneville, O.; Carvalho, L.; Coveney, M.F.; Deneke, R.; Dokulil, M.T.;
Foy, B.; et al. Lake responses to reduced nutrient loading–an analysis of contemporary long-term data from 35 case studies.
Freshwat. Biol. 2005, 50, 1747–1771.
Water 2021, 13, 2463 36 of 41
97. Vuorio, K.; Järvinen, M.; Kotamäki, N. Phosphorus thresholds for bloom-forming cyanobacterial taxa in boreal lakes.
Hydrobiologia 2020, 847, 4389–4400.
98. González, E.J.; Quirós, R. Eutrophication of reservoirs in Venezuela: Relationships between nitrogen, phosphorus and
phytoplankton biomass. Oecologia Aust. 2011, 15, 458–475.
99. Welker, M.; Walz, N. Plankton dynamics in a river lake system—on continuity and discontinuity. Hydrobiologia 1999, 408/409,
100. Smith, V.H. Low nitrogen to phosphorus ratios favor dominance by blue-green algae in lake phytoplankton. Science 1983, 221,
101. Trimbee, A.M.; Prepas, E. Evaluation of total phosphorus as a predictor of the relative biomass of blue-green algae with
emphasis on Alberta lakes. Can. J. Fish. Aquat. Sci. 1987, 44, 1337–1342.
102. Davidson, E.A.; Howarth, R.W. Nutrients in synergy. Nature 2007, 449, 1000–1001.
103. Saito, M.A.; Goepfert, T.J.; Ritt, J.T. Some thoughts on the concept of colimitation: Three definitions and the importance of
bioavailability. Limnol. Oceanogr. 2008, 53, 276–290.
104. Harpole, W.S.; Ngai, J.T.; Cleland, E.E.; Seabloom, E.W.; Borer, E.T.; Bracken, M.E.; Elser, J.J.; Gruner, D.S.; Hillebrand, H.;
Shurin, J.B. Nutrient co-limitation of primary producer communities. Ecol. Lett. 2011, 14, 852–862.
105. De Baar, H. von Liebig’s law of the minimum and plankton ecology (1899–1991). Prog. Oceanogr. 1994, 33, 347–386.
106. Dolman, A.M.; Wiedner, C. Predicting phytoplankton biomass and estimating critical N: P ratios with piecewise models that
conform to Liebig's law of the minimum. Freshwat. Biol. 2015, 60, 686–697.
107. Lewis, W.M.; Wurtsbaugh, W.A. Control of lacustrine phytoplankton by nutrients: Erosion of the phosphorus paradigm. Int.
Rev. Hydrobiol. 2008, 93, 446–465.
108. Tilman, D.; Kilham, S.S.; Kilham, P. Phytoplankton community ecology: The role of limiting nutrients. Ann. Rev. Ecol. Syst. 1982,
109. Sommer, U. The paradox of the plankton: Fluctuations of phosphorus availability maintain diversity of phytoplankton in flow-
through cultures. Limnol. Oceanogr. 1984, 29, 633–636.
110. Maberly, S.C.; Pitt, J.-A.; Davies, P.S.; Carvalho, L. Nitrogen and phosphorus limitation and the management of small
productive lakes. Inland Waters 2020, 10, 159–172.
111. Müller, S.; Mitrovic, S.M. Phytoplankton co-limitation by nitrogen and phosphorus in a shallow reservoir: Progressing from the
phosphorus limitation paradigm. Hydrobiologia 2015, 744, 255–269.
112. Lewis, W.M.; Wurtsbaugh, W.A.; Paerl, H.W. Rationale for control of anthropogenic nitrogen and phosphorus to reduce
eutrophication of inland waters. Environ. Sci. Technol. 2011, 45, 10300–10305.
113. Cao, S.; Du, R.; Peng, Y.; Li, B.; Wang, S. Novel two stage partial denitrification (PD)-Anammox process for tertiary nitrogen
removal from low carbon/nitrogen (C/N) municipal sewage. Chem. Eng. J. 2019, 362, 107–115.
114. Shatwell, T.; Köhler, J. Decreased nitrogen loading controls summer cyanobacterial blooms without promoting nitrogen-fixing
taxa: Long-term response of a shallow lake. Limnol. Oceanogr. 2019, 64, S166–S178.
115. Steinman, A.D.; Spears, B.M. (Eds.) Internal Phosphorus Loading in Lakes: Causes, Case Studies, and Management; J. Ross Publishing:
Plantation, FL, USA, 2020; p. 464.
116. Andersen, J.M. Effect of nitrate concentration in lake water on phosphate release from the sediment. Water Res. 1982, 16, 1119–
117. Jensen, H.S.; Andersen, F.O. Importance of temperature, nitrate, and pH for phosphate release from aerobic sediments of four
shallow, eutrophic lakes. Limnol. Oceanogr. 1992, 37, 577–589.
118. Schauser, I.; Chorus, I.; Lewandowski, J. Effects of nitrate on phosphorus release: Comparison of two Berlin lakes. Acta
Hydrochim. Hydrobiol. 2006, 34, 325–332.
119. Wauer, G.; Gonsiorczyk, T.; Kretschmer, K.; Casper, P.; Koschel, R. Sediment treatment with a nitrate-storing compound to
reduce phosphorus release. Water Res. 2005, 39, 494–500.
120. Paerl, H.W.; Valdes, L.M.; Joyner, A.R.; Piehler, M.F.; Lebo, M.E. Solving problems resulting from solutions: Evolution of a dual
nutrient management strategy for the eutrophying Neuse River Estuary, North Carolina. Environ. Sci. Technol. 2004, 38, 3068–
121. Howarth, R.W.; Marino, R. Nitrogen as the limiting nutrient for eutrophication in coastal marine ecosystems: Evolving views
over three decades. Limnol. Oceanogr. 2006, 51, 364–376.
122. Malone, T.C.; Newton, A. The globalization of cultural eutrophication in the coastal ocean: Causes and consequences. Front.
Mar. Sci. 2020, 7, 670, doi:10.3389/fmars.2020.00670.
123. Statham, P.J. Nutrients in estuaries—An overview and the potential impacts of climate change. Sci. Total Environ. 2012, 434, 213–
124. Carstensen, J.; Conley, D.J.; Almroth-Rosell, E.; Asmala, E.; Bonsdorff, E.; Fleming-Lehtinen, V.; Gustafsson, B.G.; Gustafsson,
C.; Heiskanen, A.-S.; Janas, U. Factors regulating the coastal nutrient filter in the Baltic Sea. Ambio 2020, 49, 1194–1210.
125. Walve, J.; Sandberg, M.; Larsson, U.; Lännergren, C. A Baltic Sea estuary as a phosphorus source and sink after drastic load
reduction: Seasonal and long-term mass balances for the Stockholm inner archipelago for 1968–2015. Biogeosciences 2018, 15,
126. Chen, Q.; Christiansen, G.; Deng, L.; Kurmayer, R. Emergence of nontoxic mutants as revealed by single filament analysis in
bloom-forming cyanobacteria of the genus Planktothrix. BMC Microbiol. 2016, 16, 23.
Water 2021, 13, 2463 37 of 41
127. Haruštiaková, D.; Welker, M. Chemotype diversity in Planktothrix rubescens (cyanobacteria) populations is correlated to lake
depth. Environ. Microbiol. Rep. 2017, 9, 158–168.
128. Dokulil, M.T.; Teubner, K. Deep living Planktothrix rubescens modulated by environmental constraints and climate forcing.
Hydrobiologia 2012, 698, 29–46.
129. Jacquet, S.; Briand, J.-F.; Leboulanger, C.; Avois-Jacquet, C.; Oberhaus, L.; Tassin, B.; Vinçon-Leite, B.; Paolini, G.; Druart, J.-C.;
Anneville, O. The proliferation of the toxic cyanobacterium Planktothrix rubescens following restoration of the largest natural
French lake (Lac du Bourget). Harmful Algae 2005, 4, 651–672.
130. Oberhaus, L.; Briand, J.F.; Leboulanger, C.; Jacquet, S.; Humbert, J.F. Comparative effects of the quality and quantity of light
and temperature on the growth of Planktothrix agardhii and P. rubescens. J. Phycol. 2007, 43, 1191–1199.
131. Gammeter, S.; Forster, R.; Zimmermann, U. Limnologische Untersuchung des Zürichsees 1972–1996. Gas Wasser Abwasser 1998,
132. Salmaso, N. Factors affecting the seasonality and distribution of cyanobacteria and chlorophytes: A case study from the large
lakes south of the Alps, with special reference to Lake Garda. Hydrobiologia 2000, 438, 43–63.
133. Bürgi, H.; Stadelmann, P. Change of phytoplankton composition and biodiversity in Lake Sempach before and during
restoration. Hydrobiologia 2002, 469, 33–48.
134. Schanz, F.; Thomas, E. Reversal of eutrophication in Lake Zürich. Water Qual. Bull. 1981, 6, 108–112.
135. Wöbbecke, K.; Klett, G.; Rechenberk, B. Wasserbeschaffenheit der Wichtigsten Seen in der Bundesrepublik Deutschland;
Umweltbundesamt: Berlin, Germany, 2003; p. 158.
136. Jacquet, S.; Kerimoglu, O.; Rimet, F.; Paolini, G.; Anneville, O. Cyanobacterial bloom termination: The disappearance of
Planktothrix rubescens from Lake Bourget (France) after restoration. Freshwat. Biol. 2014, 59, 2472–2487.
137. Feuillade, J. The cyanobacterium (blue-green alga) Oscillatoria rubescens D.C. Adv. Limnol. 1994, 41, 77–93.
138. Ruggiu, D.; Morabito, G.; Panzani, P.; Pugnetti, A. Trends and relations among basic phytoplankton characteristics in the course
of the long-term oligotrophication of Lake Maggiore (Italy). Hydrobiologia 1998, 369/370, 243-257.
139. Walsby, A.E.; Schanz, F.; Schmid, M. The Burgundy-blood phenomenon: A model of buoyancy change explains autumnal
waterblooms by Planktothrix rubescens in Lake Zurich. New Phytol. 2006, 169, 109–122.
140. Naselli-Flores, L.; Barone, R.; Chorus, I.; Kurmayer, R. Toxic cyanobacterial blooms in reservoirs under a semiarid
mediterranean climate: The magnification of a problem. Environ. Toxicol. 2007, 22, 399–404.
141. Boscaini, A.; Brescancin, F.; Cerasino, L.; Fedrigotti, C.; Fano, E.A.; Salmaso, N. Vertical and horizontal distribution of the
microcystin producer Planktothrix rubescens (Cyanobacteria) in a small perialpine reservoir. Adv. Oceanogr. Limnol. 2017, 8, 208–
142. Fastner, J.; Humpage, A.R. Hepatotoxic cyclic peptides—Microcystins and nodularins. In Toxic Cyanobacteria in Water, 2nd ed.;
Chorus, I., Welker, M., Eds.; CRC Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; pp. 21–52.
143. Ernst, B.; Hoeger, S.J.; O'brien, E.; Dietrich, D.R. Abundance and toxicity of Planktothrix rubescens in the pre-alpine Lake
Ammersee, Germany. Harmful Algae 2009, 8, 329–342.
144. Yankova, Y.; Villiger, J.; Pernthaler, J.; Schanz, F.; Posch, T. Prolongation, deepening and warming of the metalimnion change
habitat conditions of the harmful filamentous cyanobacterium Planktothrix rubescens in a prealpine lake. Hydrobiologia 2016, 776,
145. Gallina, N.; Beniston, M.; Jacquet, S. Estimating future cyanobacterial occurrence and importance in lakes: A case study with
Planktothrix rubescens in Lake Geneva. Aquat. Sci. 2017, 79, 249–263.
146. Derot, J.; Yajima, H.; Jacquet, S. Advances in forecasting harmful algal blooms using machine learning models: A case study
with Planktothrix rubescens in Lake Geneva. Harmful Algae 2020, 99, 101906.
147. Wood, S.A.; Kelly, L.T.; Bouma-Gregson, K.; Humbert, J.F.; Laughinghouse IV, H.D.; Lazorchak, J.; McAllister, T.G.; McQueen,
A.; Pokrzywinski, K.; Puddick, J. Toxic benthic freshwater cyanobacterial proliferations: Challenges and solutions for enhancing
knowledge and improving monitoring and mitigation. Freshwat. Biol. 2020, 65, 1824–1842.
148. Quiblier, C.; Wood, S.A.; Echenique-Subiabre, I.; Heath, M.; Villeneuve, A.; Humbert, J.F. A review of current knowledge on
toxic benthic freshwater cyanobacteria—Ecology, toxin production and risk management. Water Res. 2013, 47, 5464–5479.
149. Floury, M.; Usseglio-Polatera, P.; Delattre, C.; Souchon, Y. Assessing long-term effects of multiple, potentially confounded
drivers in ecosystems from species traits. Glob. Chang. Biol. 2017, 23, 2297–2307.
150. Abonyi, A.; Ács, É.; Hidas, A.; Grigorszky, I.; Várbíró, G.; Borics, G.; Kiss, K.T. Functional diversity of phytoplankton highlights
long-term gradual regime shift in the middle section of the Danube River due to global warming, human impacts and
oligotrophication. Freshwat. Biol. 2018, 63, 456–472.
151. Minaudo, C.; Abonyi, A.; Leitão, M.; Lançon, A.M.; Floury, M.; Descy, J.-P.; Moatar, F. Long-term impacts of nutrient control,
climate change, and invasive clams on phytoplankton and cyanobacteria biomass in a large temperate river. Sci. Total Environ.
2021, 756, 144074.
152. Vidal, L.; Ballot, A.; Azevedo, S.M.F.O.; Padisák, J.; Welker, M. Introduction to cyanobacteria. In Toxic Cyanobacteria in Water,
2nd ed.; Chorus, I., Welker, M., Eds.; CRC Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; pp. 163–
153. Bouma-Gregson, K.; Kudela, R.M.; Power, M.E. Widespread anatoxin-a detection in benthic cyanobacterial mats throughout a
river network. PLoS ONE 2018, 13, e0197669.
Water 2021, 13, 2463 38 of 41
154. Gaget, V.; Humpage, A.R.; Huang, Q.; Monis, P.; Brookes, J.D. Benthic cyanobacteria: A source of cylindrospermopsin and
microcystin in Australian drinking water reservoirs. Water Res. 2017, 124, 454–464.
155. Cantoral Uriza, E.; Asencio, A.; Aboal, M. Are we underestimating benthic cyanotoxins? Extensive sampling results from Spain.
Toxins 2017, 9, 385.
156. Smith, Z.J.; Martin, R.M.; Wei, B.; Wilhelm, S.W.; Boyer, G.L. Spatial and temporal variation in paralytic shellfish toxin
production by benthic Microseira (Lyngbya) wollei in a freshwater New York lake. Toxins 2019, 11, 44.
157. Lowe, R. Periphyton patterns in lakes. In Algal Ecology: Freshwater Benthic Ecosystems; Stevenson, R.J., Bothwell, M.L., Lowe, R.L.,
Eds.; Academic Press: San Diego, CA, USA, 1996.
158. Jähnichen, S.; Jäschke, K.; Wieland, F.; Packroff, G.; Benndorf, J. Spatio-temporal distribution of cell-bound and dissolved
geosmin in Wahnbach Reservoir: Causes and potential odour nuisances in raw water. Water Res. 2011, 45, 4973–4982.
159. Fastner, J.; Beulker, C.; Geiser, B.; Hoffmann, A.; Kröger, R.; Teske, K.; Hoppe, J.; Mundhenk, L.; Neurath, H.; Sagebiel, D. Fatal
neurotoxicosis in dogs associated with tychoplanktic, anatoxin-a producing Tychonema sp. in mesotrophic Lake Tegel, Berlin.
Toxins 2018, 10, 60.
160. Reynolds, C.S. The Ecology of Phytoplankton; Cambridge University Press: Cambridge, UK, 2006.
161. Padisák, J.; Crossetti, L.O.; Naselli-Flores, L. Use and misuse in the application of the phytoplankton functional classification:
A critical review with updates. Hydrobiologia 2009, 621, 1–19.
162. Humpage, A.R.; Fastner, J. Cylindrospermopsins. In Toxic Cyanobacteria in Water, 2nd ed.; Chorus, I., Welker, M., Eds.; CRC
Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; pp. 53–71.
163. Testai, E. Anatoxin-A and analogues. In Toxic Cyanobacteria in Water, 2nd ed.; Chorus, I., Welker, M., Eds.; CRC Press on Behalf
of World Health Organisation: Geneva, Switzerland, 2021; pp. 72–93.
164. Testai, E. Saxitoxins or paralytic shellfish poisons. In Toxic Cyanobacteria in Water, 2nd ed.; Chorus, I., Welker, M., Eds.; CRC
Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; pp. 94–108.
165. Rantala, A.; Fewer, D.; Hisbergues, M.; Rouhiainen, L.; Vaitomaa, J.; Börner, T.; Sivonen, K. Phylogenetic evidence for the early
evolution of microcystin synthesis. Proc. Natl. Acad. Sci. USA 2004, 101, 568–573.
166. Yoshida, M.; Yoshida, T.; Takashima, Y.; Hosoda, N.; Hiroishi, S. Dynamics of microcystin-producing and non-microcystin-
producing Microcystis populations is correlated with nitrate concentration in a Japanese lake. FEMS Microbiol. Lett. 2007, 266,
167. Rinta-Kanto, J.M.; Konopko, E.A.; DeBruyn, J.M.; Bourbonniere, R.A.; Boyer, G.L.; Wilhelm, S.W. Lake Erie Microcystis:
Relationship between microcystin production, dynamics of genotypes and environmental parameters in a large lake. Harmful
Algae 2009, 8, 665–673.
168. Sabart, M.; Pobel, D.; Briand, E.; Combourieu, B.; Salencon, M.J.; Humbert, J.F.; Latour, D. Spatiotemporal variations in
microcystin concentrations and in the proportions of microcystin-producing cells in several Microcystis aeruginosa populations.
Appl. Environ. Microbiol. 2010, 76, 4750–4759.
169. Kardinaal, W.E.A.; Janse, I.; Kamst-van Agterveld, M.P.; Meima, M.; Snoek, J.; Mur, L.R.; Huisman, J.; Zwart, G.; Visser, P.M.
Microcystis genotype succession in relation to microcystin concentrations in freshwater lakes. Aquat. Microb. Ecol. 2007, 48, 1–12.
170. Welker, M.; Šejnohová, L.; von Döhren, H.; Nemethova, D.; Jarkovsky, J.; Maršálek, B. Seasonal shifts in chemotype composition
of Microcystis sp. communities in the pelagial and the sediment of a shallow reservoir. Limnol. Oceanogr. 2007, 52, 609–619.
171. Rohrlack, T.; Edvardsen, B.; Skulberg, R.; Halstvedt, C.B.; Utkilen, H.C.; Ptacnik, R.; Skulberg, O.M. Oligopeptide chemotypes
of the toxic freshwater cyanobacterium Planktothrix can form subpopulations with dissimilar ecological traits. Limnol. Oceanogr.
2008, 53, 1279–1293.
172. Kim, S.G.; Joung, S.H.; Ahn, C.Y.; Ko, S.R.; Boo, S.M.; Oh, H.M. Annual variation of Microcystis genotypes and their potential
toxicity in water and sediment from a eutrophic reservoir. FEMS Microbiol. Ecol. 2010, 74, 93–102.
173. Agha, R.; Lezcano, M.A.; del Mar Labrador, M.; Cirés, S.; Quesada, A. Seasonal dynamics and sedimentation patterns of
Microcystis oligopeptide-based chemotypes reveal subpopulations with different ecological traits. Limnol. Oceanogr. 2014, 59,
174. Baunach, M.; Chowdhury, S.; Stallforth, P.; Dittmann, E. The landscape of recombination events that create nonribosomal
peptide diversity. Mol. Biol. Evol. 2021, 38, 2116–2130.
175. Kaebernick, M.; Neilan, B.A.; Börner, T.; Dittmann, E. Light and the transcriptional response of the microcystin biosynthesis
gene cluster. Appl. Environ. Microbiol. 2000, 66, 3387–3392.
176. Sevilla, E.; Martin-Luna, B.; Vela, L.; Bes, M.T.; Fillat, M.F.; Peleato, M.L. Iron availability affects mcyD expression and
microcystin-LR synthesis in Microcystis aeruginosa PCC7806. Environ. Microbiol. 2008, 10, 2476–2483.
177. Pimentel, J.S.; Giani, A. Microcystin production and regulation under nutrient stress conditions in toxic Microcystis strains. Appl.
Environ. Microbiol. 2014, 80, 5836–5843.
178. Orr, P.T.; Jones, G.J. Relationship between microcystin production and cell division rates in nitrogen-limited Microcystis
aeruginosa cultures. Limnol. Oceanogr. 1998, 43, 1604–1614.
179. Brakhage, A.A. Regulation of fungal secondary metabolism. Nat. Rev. Microbiol. 2013, 11, 21–32.
180. Keller, N.P. Fungal secondary metabolism: Regulation, function and drug discovery. Nat. Rev. Microbiol. 2019, 17, 167–180.
181. Briand, E.; Bormans, M.; Quiblier, C.; Salençon, M.-J.; Humbert, J.-F. Evidence of the cost of the production of microcystins by
Microcystis aeruginosa under differing light and nitrate environmental conditions. PLoS ONE 2012, 7, e29981.
Water 2021, 13, 2463 39 of 41
182. Gobler, C.J.; Burkholder, J.M.; Davis, T.W.; Harke, M.J.; Johengen, T.; Stow, C.A.; Van de Waal, D.B. The dual role of nitrogen
supply in controlling the growth and toxicity of cyanobacterial blooms. Harmful Algae 2016, 54, 87–97.
183. Peng, G.; Martin, R.M.; Dearth, S.P.; Sun, X.; Boyer, G.L.; Campagna, S.R.; Lin, S.; Wilhelm, S.W. Seasonally relevant cool
temperatures interact with N chemistry to increase microcystins produced in lab cultures of Microcystis aeruginosa NIES-843.
Environ. Sci. Technol. 2018, 52, 4127–4136.
184. Brandenburg, K.; Siebers, L.; Keuskamp, J.; Jephcott, T.G.; Van de Waal, D.B. Effects of nutrient limitation on the synthesis of
N-rich phytoplankton toxins: A meta-analysis. Toxins 2020, 12, 221.
185. Welker, M.; von Döhren, H. Cyanobacterial peptides—Nature’s own combinatorial biosynthesis. FEMS Microbiol. Rev. 2006, 30,
186. Fastner, J.; Erhard, M.; von Döhren, H. Determination of oligopeptide diversity within a natural population of Microcystis spp.
(Cyanobacteria) by typing single colonies by matrix-assisted laser desorption ionization-time of flight mass spectrometry. Appl.
Environ. Microbiol. 2001, 67, 5069–5076.
187. Welker, M.; Christiansen, G.; von Döhren, H. Diversity of coexisting Planktothrix (Cyanobacteria) chemotypes deduced by mass
spectral analysis of microcystins and other oligopeptides. Arch. Microbiol. 2004, 182, 288–298.
188. Welker, M.; Maršálek, B.; Šejnohová, L.; von Döhren, H. Detection and identification of oligopeptides in Microcystis
(cyanobacteria) colonies: Toward an understanding of metabolic diversity. Peptides 2006, 27, 2090–2103.
189. Agha, R.; Quesada, A. Oligopeptides as biomarkers of cyanobacterial subpopulations. Toward an understanding of their
biological role. Toxins 2014, 6, 1929–1950.
190. Natumi, R.; Janssen, E.M.-L. Cyanopeptide co-production dynamics beyond mirocystins and effects of growth stages and
nutrient availability. Environ. Sci. Technol. 2020, 54, 6063–6072.
191. Suominen, S.; Brauer, V.S.; Rantala-Ylinen, A.; Sivonen, K.; Hiltunen, T. Competition between a toxic and a non-toxic
Microcystis strain under constant and pulsed nitrogen and phosphorus supply. Aquat. Ecol. 2017, 51, 117–130.
192. Ibelings, B.W.; Kurmayer, R.; Azevedo, S.M.F.O.; Wood, S.A.; Chorus, I.; Welker, M. Understanding the occurrence of
cyanobacteria and cyanotoxins. In Toxic Cyanobacteria in Water, 2nd ed.; Chorus, I., Welker, M., Eds.; CRC Press on Behalf of
World Health Organisation: Geneva, Switzerland, 2021; pp. 213–294.
193. Hayes, P.K.; Barker, G.L.A. Genetic diversity within baltic sea populations of Nodularia (cyanobacteria). J. Phycol. 1997, 33, 919–
194. Cadel-Six, S.; Peyraud-Thomas, C.; Brient, L.; de Marsac, N.T.; Rippka, R.; Méjean, A. Different genotypes of anatoxin-producing
cyanobacteria coexist in the Tarn River, France. Appl. Environ. Microbiol. 2007, 73, 7605–7614.
195. Tanabe, Y.; Kasai, F.; Watanabe, M.M. Multilocus sequence typing (MLST) reveals high genetic diversity and clonal population
structure of the toxic cyanobacterium Microcystis aeruginosa. Microbiology 2007, 153, 3695–3703.
196. Yilmaz, M.; Foss, A.J.; Selwood, A.I.; Özen, M.; Boundy, M. Paralytic shellfish toxin producing Aphanizomenon gracile strains
isolated from Lake Iznik, Turkey. Toxicon 2018, 148, 132–142.
197. Kurmayer, R.; Blom, J.F.; Deng, L.; Pernthaler, J. Integrating phylogeny, geographic niche partitioning and secondary metabolite
synthesis in bloom-forming Planktothrix. ISME J. 2014, 9, 909–921.
198. Briand, E.; Escoffier, N.; Straub, C.; Sabart, M.; Quiblier, C.; Humbert, J.F. Spatiotemporal changes in the genetic diversity of a
bloom-forming Microcystis aeruginosa (cyanobacteria) population. ISME J. 2009, 3, 419–429.
199. Litchman, E.; Klausmeier, C.A. Trait-based community ecology of phytoplankton. Ann. Rev. Ecol. Evol. Syst. 2008, 39, 615–639.
200. Finkel, Z.V.; Beardall, J.; Flynn, K.J.; Quigg, A.; Rees, T.A.V.; Raven, J.A. Phytoplankton in a changing world: Cell size and
elemental stoichiometry. J. Plankton Res. 2010, 32, 119–137.
201. Hesse, K.; Kohl, J.-G. Effects of light and nutrients supply on growth and microcystin content of different strains of Microcystis
aeruginosa. In Cyanotoxins: Occurrence, Causes, Consequences; Chorus, I., Ed.; Springer: Berlin/Heidelberg, Germany, 2001; pp.
202. Briand, E.; Gugger, M.; Francois, J.F.; Bernard, C.; Humbert, J.F.; Quiblier, C. Temporal variations in the dynamics of potentially
microcystin-producing strains in a bloom-forming Planktothrix agardhii (cyanobacterium) population. Appl. Environ. Microbiol.
2008, 74, 3839–3848.
203. Kardinaal, W.E.A.; Tonk, L.; Janse, I.; Hol, S.; Slot, P.; Huisman, J.; Visser, P.M. Competition for light between toxic and nontoxic
strains of the harmful cyanobacterium Microcystis. Appl. Microbiol. Biotechnol. 2007, 73, 2939–2946.
204. LeBlanc Renaud, S.; Pick, F.R.; Fortin, N. Effect of light intensity on the relative dominance of toxigenic and nontoxigenic strains
of Microcystis aeruginosa. Appl. Environ. Microbiol. 2011, 77, 7016–7022.
205. Sommer, U.; Kilham, S.S. Phytoplankton natural community experiments: A reinterpretation. Limnol. Oceanogr. 1985, 30, 436–
206. Briand, E.; Bormans, M.; Gugger, M.; Dorrestein, P.C.; Gerwick, W.H. Changes in secondary metabolic profiles of Microcystis
aeruginosa strains in response to intraspecific interactions. Environ. Microbiol. 2016, 18, 384–400.
207. Rohrlack, T.; Christiansen, G.; Kurmayer, R. Putative antiparasite defensive system involving ribosomal and nonribosomal
oligopeptides in cyanobacteria of the genus Planktothrix. Appl. Environ. Microbiol. 2013, 79, 2642–2647.
208. Agha, R.; Gross, A.; Gerphagnon, M.; Rohrlack, T.; Wolinska, J. Fitness and eco-physiological response of a chytrid fungal
parasite infecting planktonic cyanobacteria to thermal and host genotype variation. Parasitology 2018, 145, 1279–1286.
209. Bouaïcha, N.; Miles, C.O.; Beach, D.G.; Labidi, Z.; Djabri, A.; Benayache, N.Y.; Nguyen-Quang, T. Structural diversity,
characterization and toxicology of microcystins. Toxins 2019, 11, 714.
Water 2021, 13, 2463 40 of 41
210. Puddick, J.; van Ginkel, R.; Page, C.D.; Murray, J.S.; Greenhough, H.E.; Bowater, J.; Selwood, A.I.; Wood, S.A.; Prinsep, M.R.;
Truman, P. Acute toxicity of dihydroanatoxin-a from Microcoleus autumnalis in comparison to anatoxin-a. Chemosphere 2021, 263,
211. Breinlinger, S.; Phillips, T.J.; Haram, B.N.; Mareš, J.; Yerena, J.A.M.; Hrouzek, P.; Sobotka, R.; Henderson, W.M.; Schmieder, P.;
Williams, S.M. Hunting the eagle killer: A cyanobacterial neurotoxin causes vacuolar myelinopathy. Science 2021, 371, 6536.
212. Thapa, H.R.; Agarwal, V. Obligate brominating enzymes underlie bromoform production by marine cyanobacteria. J. Phycol.
2021, (in press).
213. Cadel-Six, S.; Dauga, C.; Castets, A.M.; Rippka, R.; Tandeau de Marsac, N.; Welker, M. Halogenase genes in two non-ribosomal
peptide synthetase gene clusters of Microcystis (Cyanobacteria): Sporadic distribution and evolution. Mol. Biol. Evol. 2008, 25,
214. Wilde, S.B.; Johansen, J.R.; Wilde, H.D.; Jiang, P.; Bartelme, B.; Haynie, R.S. Aetokthonos hydrillicola gen. et sp. nov.: Epiphytic
cyanobacteria on invasive aquatic plants implicated in Avian Vacuolar Myelinopathy. Phytotaxa 2014, 181, 243–260.
215. Walton, K.; Berry, J.P. Indole alkaloids of the Stigonematales (Cyanophyta): Chemical diversity, biosynthesis and biological
activity. Mar. Drugs 2016, 14, 73.
216. Cirés, S.; Alvarez-Roa, C.; Wood, S.A.; Puddick, J.; Loza, V.; Heimann, K. First report of microcystin-producing Fischerella sp.
(Stigonematales, Cyanobacteria) in tropical Australia. Toxicon 2014, 88, 62–66.
217. Faassen, E.J. Presence of the neurotoxin BMAA in aquatic ecosystems: What do we really know? Toxins 2014, 6, 1109–1138.
218. Lance, E.; Arnich, N.; Maignien, T.; Biré, R. Occurrence of β-N-methylamino-l-alanine (BMAA) and isomers in aquatic
environments and aquatic food sources for humans. Toxins 2018, 10, 83.
219. Chernoff, N.; Faassen, E.J.; Hill, D.J. β-Methylamino-L-alanine (BMAA). In Toxic Cyanobacteria in Water, 2nd ed.; Chorus, I.,
Welker, M., Eds.; CRC Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; pp. 123–136.
220. Dunlop, R.; Banack, S.; Bishop, S.; Metcalf, J.; Murch, S.; Davis, D.; Stommel, E.; Karlsson, O.; Brittebo, E.; Chatziefthimiou, A.
Is exposure to BMAA a risk factor for neurodegenerative diseases? A response to a critical review of the BMAA hypothesis.
Neurotox. Res. 2021, 1–26, doi:10.1007/s12640-020-00302-0.
221. Lawton, L.A.; Metcalf, J.S.; Žegura, B.; Junek, R.; Welker, M.; Törökné, A.K.; Bláha, L. Laboratory analysis of cyanobacterial
toxins and bioassays. In Toxic Cyanobacteria in Water, 2nd ed.; Chorus, I., Welker, M., Eds.; CRC Press on Behalf of World Health
Organisation: Geneva, Switzerland, 2021; pp. 745–800.
222. Bastos Maia, S.; Rolland Souza, A.S.; Costa Caminha, M.d.F.; Lins da Silva, S.; Callou da Cruz, R.S.B.L.; Carvalho dos Santos,
C.; Batista Filho, M. Vitamin A and pregnancy: A narrative review. Nutrients 2019, 11, 681.
223. Fosmire, G.J. Zinc toxicity. Am. J. Clin. Nutr. 1990, 51, 225–227.
224. Sieber, S.; Grendelmeier, S.M.; Harris, L.A.; Mitchell, D.A.; Gademann, K. Microviridin 1777: A toxic chymotrypsin inhibitor
discovered by a metabologenomic approach. J. Nat. Prod. 2020, 83, 438–446.
225. Welker, M.; Dittmann, E.; von Döhren, H. Cyanobacteria as a source of natural products. In Natural Product Biosynthesis by
Microorganisms and Plants, Part C, 517th ed.; Hopwood, D.A., Ed.; 2012; Elsevier, Amsterdam. pp. 23–46.
226. Roy-Lachapelle, A.; Duy, S.V.; Munoz, G.; Dinh, Q.T.; Bahl, E.; Simon, D.F.; Sauvé, S. Analysis of multiclass cyanotoxins
(microcystins, anabaenopeptins, cylindrospermopsin and anatoxins) in lake waters using on-line SPE liquid chromatography
high-resolution Orbitrap mass spectrometry. Anal. Meth. 2019, 11, 5289–5300.
227. Demay, J.; Bernard, C.; Reinhardt, A.; Marie, B. Natural products from cyanobacteria: Focus on beneficial activities. Mar. Drugs
2019, 17, 320.
228. Vijayakumar, S.; Menakha, M. Pharmaceutical applications of cyanobacteria—A review. J. Acute Med. 2015, 5, 15–23.
229. US EPA. Drinking Water Health Advisory for the Cyanobacterial Microcystin Toxins; United States Environmental Protection Agency:
Washington, DC, USA, 2015.
230. WHO. Cyanobacterial Toxins: Microcystin. Background Document for Development of WHO Guidelines for Drinking-Water Quality and
Guidelines for Safe Recreational Water Environments; World Health Organization: Geneva, Switzerland, 2020.
231. WHO. Cyanobacterial Toxins: Cylindrospermopsin. Background Document for Development of WHO Guidelines for Drinking-Water
Quality and Guidelines for Safe Recreational Water Environments; World Health Organization: Geneva, Switzerland, 2020.
232. WHO. Cyanobacterial Toxins: Anatoxin-a. Background Document for Development of WHO Guidelines for Drinking-Water Quality and
Guidelines for Safe Recreational Water Environments; World Health Organization: Geneva, Switzerland, 2020.
233. WHO. Cyanobacterial Toxins: Saxitoxins. Background Document for Development of WHO Guidelines for Drinking-Water Quality and
Guidelines for Safe Recreational Water Environments; World Health Organization: Geneva, Switzerland, 2020.
234. WHO. Guidelines for Drinking-Water Quality, 4th ed.; Incorporating the 1st Addendum ed.; World Health Organization: Geneva,
235. WHO. Developing Drinking-Water Quality Regulations and Standards; World Health Organization: Geneva, Switzerland, 2018; p.
236. Fawell, J.K.; Mitchell, R.E.; Everett, D.J.; Hill, R.E. The toxicity of cyanobacterial toxins in the mouse: I. Microcystin-LR. Fawell
JK, Mitchell RE, Everett DJ, Hill RE (1999). The toxicity of cyanobacterial toxins in the mouse: I. Microcystin-LR. Hum Exper Toxicol.
18:162–7. 1999, 1, :162–167.
237. WHO. Safer Water, Better Health. 2019 Update; World Health Organization: Geneva, Switzerland, 2019; 67p.
238. WHO. Guidelines for Safe Recreational Water Environments; Coastal and Fresh Waters; World Health Organisation: Geneva,
Switzerland, 2021; Volume 1, 138p.
Water 2021, 13, 2463 41 of 41
239. Newcombe, G.; Ho, L.; Capelo-Neto, J. Controlling cyanotoxin occurrence: Drinking-water treatment. In Toxic Cyanobacteria in
Water, 2nd ed.; Chorus, I., Welker, M., Eds.; CRC Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; pp.
240. Humpage, A.R.; Cunliffe, D.A.; Chorus, I. Understanding Exposure: Drinking Water. In Toxic Cyanobacteria in Water, 2nd ed.;
Chorus, I., Welker, M., Eds.; CRC Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; pp. 305–332.
241. Chorus, I.; Testai, E. Understanding Exposure: Recreation and occupational activities. In Toxic Cyanobacteria in Water, 2nd ed.;
Chorus, I., Welker, M., Eds.; CRC Press on Behalf of World Health Organisation: Geneva, Switzerland, 2021; pp. 333–367.
242. Wood, S.A. Advice on Benthic Cyanobacteria Health Risks and Communication Strategies in the Southland Region; Report No. 3075;
Cawthorn Institute: Nelson, New Zealand, 2017; 19 pp.
243. Brookes, J.D.; Burch, M.D.; Grützmacher, G.; Klitzke, S. Managing cyanotoxin risks at the drinking-water offtake. In Toxic
Cyanobacteria in Water, 2nd ed.; Chorus, I., Welker, M., Eds.; CRC Press on Behalf of World Health Organisation: Geneva,
Switzerland, 2021; pp. 564–589.