zooplankton diversity research paper

Advertisement

Freshwater phytoplankton diversity: models, drivers and implications for ecosystem properties

• COLIN S. REYNOLDS’ LEGACY
• Review Paper
• Open access
• Published: 04 July 2020
• Volume 848 , pages 53–75, ( 2021 )

Cite this article

You have full access to this open access article

• Gábor Borics 1 , 2 ,
• András Abonyi 3 , 4 ,
• Nico Salmaso 5 &
• Robert Ptacnik 4  

16k Accesses

45 Citations

1 Altmetric

Explore all metrics

Our understanding on phytoplankton diversity has largely been progressing since the publication of Hutchinson on the paradox of the plankton. In this paper, we summarise some major steps in phytoplankton ecology in the context of mechanisms underlying phytoplankton diversity. Here, we provide a framework for phytoplankton community assembly and an overview of measures on taxonomic and functional diversity. We show how ecological theories on species competition together with modelling approaches and laboratory experiments helped understand species coexistence and maintenance of diversity in phytoplankton. The non-equilibrium nature of phytoplankton and the role of disturbances in shaping diversity are also discussed. Furthermore, we discuss the role of water body size, productivity of habitats and temperature on phytoplankton species richness, and how diversity may affect the functioning of lake ecosystems. At last, we give an insight into molecular tools that have emerged in the last decades and argue how it has broadened our perspective on microbial diversity. Besides historical backgrounds, some critical comments have also been made.

Similar content being viewed by others

Mathematical modelling of plankton–oxygen dynamics under the climate change.

Yadigar Sekerci & Sergei Petrovskii

Morpho-functional traits of phytoplankton functional groups: a review

Demtew Etisa Welbara, Demeke Kifle Gebre-Meskel & Tadesse Fetahi Hailu

Functional ecology of fish: current approaches and future challenges

Sébastien Villéger, Sébastien Brosse, … Michael J. Vanni

Avoid common mistakes on your manuscript.

Introduction

Phytoplankton is a polyphyletic group with utmost variation in size, shape, colour, type of metabolism, and life history traits. Due to the emerging knowledge in nutritional capabilities of microorganisms, our view of phytoplankton has drastically changed (Flynn et al., 2013 ). Phagotrophy is now known from all clades except diatoms and cyanobacteria. At the same time, ciliates, which have not been considered as part of ‘phytoplankton’, span a gradient in trophic modes that render the distinction between phototrophic phytoplankton and heterotrophic protozoa meaningless. This complexity has been expressed in the high diversity of natural phytoplankton assemblages. Diversity can be defined in many different ways and levels. Although the first diversity measure that encompassed the two basic components of diversity (i.e., the number of items and their relative frequencies) appeared in the early forties of the last century (Fisher et al., 1943 ), in phytoplankton ecology, taxonomic richness has been used the most often as diversity estimates. Until the widespread use of the inverted microscopes, phytoplankton ecologists did not have accurate abundance estimation methods and the net plankton served as a basis for the analyses. Richness of taxonomic groups of net samples, and their ratios were used for quality assessment (Thunmark, 1945 , Nygaard, 1949 ).

The study of phytoplankton diversity received a great impetus after Hutchinson’s ( 1961 ) seminal paper on the paradox of the plankton. The author not only contrasted Hardin’s competitive exclusion theory (Hardin, 1960 ) with the high number of co-occurring species in a seemingly homogeneous environment, but outlined possible explanations. He argued for the non-equilibrium nature of the plankton, the roles of disturbances and biotic interactions, moreover the importance of benthic habitats in the recruitment of phytoplankton. The ‘paradox of the plankton’ largely influenced the study of diversity in particular and the development of community ecology in general (Naselli-Flores & Rossetti, 2010 ). Several equilibrium and non-equilibrium mechanisms have been developed to address the question of species coexistence in pelagic waters (reviewed by Roy & Chattopadhyay, 2007 ). The paradox and the models that aimed to explain the species coexistence in the aquatic environment have been extended to terrestrial ecosystems (Wilson, 1990 ). Wilson reviewed evidences for twelve possible mechanisms that potentially could explain the paradox for indigenous New Zealand vegetation, and found that four of them, such as gradual climate change, cyclic successional processes, spatial mass effect and niche diversification, were the most important explanations. By now, the paradox has been considered as an apparent violation of the competitive exclusion principle in the entire field of ecology (Hening & Nguyen, 2020 ).

Although Hutchinson’s contribution (Hutchinson, 1961 ) has given a great impetus to research on species coexistence, the number of studies on phytoplankton diversity that time did not increase considerably (Fig.  1 ), partly because in this period, eutrophication studies dominated the hydrobiological literature.

Annual number of hits on Google Scholar for the keywords “phytoplankton diversity”

Understanding the drivers of diversity has been substantially improved from the 70 s when laboratory experiments and mathematical modelling proved that competition theory or intermediate disturbance hypothesis (IDH) provided explanations for species coexistence. Many field studies also demonstrated the role of disturbances in maintaining phytoplankton diversity, and these results were concluded by Reynolds and his co-workers (Reynolds et al., 1993 ).

From the 2000 s a rapid increase in phytoplankton research appeared (Fig.  1 ), which might be explained by theoretical and methodological improvements in ecology. The functional approaches—partly due to Colin Reynolds’s prominent contribution to this field (Reynolds et al., 2002 )—opened new perspectives in phytoplankton diversity research. Functional trait and functional ‘group’-based approaches have gained considerable popularity in recent years (Weithoff, 2003 ; Litchman & Klausmeier, 2008 ; Borics et al., 2012 ; Vallina, et al., 2017 ; Ye et al., 2019 ).

Analysis of large databases enabled to study diversity changes on larger scales in lake area, productivity or temperature (Stomp et al., 2011 ). Recent studies on phytoplankton also revealed that phytoplankton diversity was more than a single metric by which species or functional richness could be described, instead, it was an essential characteristic, which affects functioning of the ecosystems, such as resilience (Gunderson 2000 ) or resource use efficiency (Ptacnik et al., 2008 ; Abonyi et al., 2018a , b ).

The widespread use of molecular tools that reorganise phytoplankton taxonomy and reveal the presence of cryptic diversity, has changed our view of phytoplankton diversity. In this study, we aim to give an overview of the above-mentioned advancements in phytoplankton diversity. Here we focus on the following issues:

measures of diversity,

mechanisms affecting diversity,

changes of diversity along environmental gradients (area, productivity, temperature),

the functional diversity–ecosystem functioning relationship, and

phytoplankton diversity using molecular tools.

More than eight thousand studies have been published on “phytoplankton diversity” since the term first appeared in the literature in the middle of the last century (Fig.  1 ), therefore, in this review we cannot completely cover all the important developments made in recent years. Instead, we focus on the most relevant studies considered as milestones in the field, and on the latest relevant contributions. This study is a part of a Hydrobiologia special issue dedicated to the memory of Colin S. Reynolds, who was one of the most prominent and influential figures of phytoplankton ecology in the last four decades, therefore, we have placed larger emphasis on his concepts that helped our understanding of assembly and diversity of phytoplankton.

Measures of diversity

In biology, the term “diversity” encompasses two basic compositional properties of assemblages: species richness and inequalities in species abundances. Verbal definitions of diversity cannot be specific enough to describe both aspects, but these can be clearly defined by the mathematical formulas that we use as diversity measures.

Richness metrics

The simplest measure of diversity is species richness, that is, the number of species observed per sampling unit. However, this metric can only be used safely when the applied counting approach ensures high species detectability.

In case of phytoplankton, species detectability depends strongly on counting effort, therefore, measures that are standardised by the number of individuals observed, e.g. Margalef and Mehinick indices (Clifford & Stephenson, 1975 ) safeguard against biased interpretations. Ideally, standardization should take place in the process of identification. Pomati et al. ( 2015 ) gave an example how a general detection limits could be applied in retrospect to data stemming from variable counting efforts.

Species richness can also be given using richness estimators. These can be parametric curve-fitting approaches, non-parametric estimators, and extrapolation techniques using species accumulation or species-area curves (Gotelli & Colwell, 2011 ; Magurran, 2004 ). These approaches have been increasingly applied in phytoplankton ecology (Naselli-Flores et al., 2016 ; Görgényi et al., 2019 ).

Abundance-based metrics

Classical diversity metrics such as Shannon and Simpson indices combine richness and evenness into univariate vectors. Though used commonly in the literature, they are prone to misinform about the actual changes in a community, as they may reflect changes in evenness and/or richness to an unknown extent (a change in Shannon H' 1948 ) may solely be driven by a change in evenness or richness). Dominance metrics emphasise the role of the most important species (McNaughton, 1967 ). Rarity metrics, in contrast, focus on the rare elements of the assemblages (Gotelli & Colwell, 2011 ).

Species abundance distributions (SAD) and rank abundance distributions (RAD: ranking the species’ abundances from the most abundant to the least abundant) provide an alternative to diversity indices (Fisher et al., 1943 ; Magurran & Henderson, 2003 ). These parametric approaches give accurate information on community structure and are especially useful when site level data are compared. Most RADs follow lognormal distributions and allow to estimate species richness in samples (Ulrich & Ollik, 2005 ).

Mechanisms affecting diversity

• Community assembly

Understanding the processes that shape the community structure of phytoplankton requires some knowledge on the general rules of community assembly. Models and mechanisms, which have been proposed to explain the compositional patterns of biotic communities, can be linked together under one conceptual framework developed by Vellend ( 2010 , 2016 ). Vellend proposed four distinct processes that determine species composition and diversity: speciation (creation of new species, or within-species genetic modifications), selection (environmental filtering, and biotic interactions), drift (demographic stochasticity) and dispersal (movement of individuals). The four processes interact to determine community dynamics across spatial scales from global, through regional to local. The importance of the processes strongly depends on the type of community, and the studied spatial and temporal scales (Reynolds, 1993 ).

Importance of evolutionary processes in the community assembly have been demonstrated by several phylogenetic ecological studies (Cavender-Bares et al., 2009 ) and also indicated by the emergence of a new field of science called ecophylogenetics (Mouquet et al., 2012 ). As far as the phytoplankton is concerned, the role of speciation can be important when the composition and diversity of algal assemblages are studied at large (global) spatial scales. However, we may note that although microscopic analyses cannot grasp it, short-term evolutionary processes do occur locally in planktic assemblages (Balzano et al., 2011 ; Padfield et al., 2016 ; Bach et al., 2018 ).

Demographic stochasticity influences growth and extinction risk of small populations largely (Parvinen et al., 2003 ; Méndez et al., 2019 ). Similarly, it might also act on large lake phytoplankton since population size in previous years affects the success of species in the subsequent year. Small changes in initial abundances may have strong effects on seasonal development. Demographic stochasticity, however, is crucial in small isolated waters (especially in newly created ones) where the sequence of new arrivals and small differences in initial abundances likely have a strong effect on the outcome of community assembly.

Theoretical models, laboratory experiments and field studies demonstrated that the other two processes, selection and dispersal, have a pivotal role in shaping community assembly and diversity. Although this statement corresponds well with the Baas-Becking ( 1934 ) hypothesis (everything can be everywhere but environment selects), importance of selection and dispersal depends on the characteristics of the aquatic systems. Selection and dispersal can be considered as filters (Knopf, 1986 , Pearson et al., 2018 ), and using them as gradients, a two-dimensional plane can be constructed, where the positions of the relevant types of pelagic aquatic habitats can be displayed (Fig.  2 ). At high dispersal rate, the mass effect (or so-called source-sink dynamics) is the most decisive process affecting community assembly (Leibold & Chase, 2017 ). Phytoplankton of rhithral rivers is a typical example of the sink populations because its composition and diversity are strongly affected by the propagule pressure coming partly from the source populations of the benthic zone and from the limnetic habitats of the watershed (Bolgovics et al., 2017 ). The relative importance of the mass effect decreases with time and with the increasing size of the river, while the role of selection (species sorting) increases. Due to their larger size, the impact of the source-sink dynamics in potamal rivers must be smaller, and selection becomes more important in shaping community assembly. Although the role of spatial processes in lake phytoplankton assembly cannot be ignored, their importance is considerably less than that of the locally acting selection. Relevance of the spatial processes have been demonstrated for river floodplain complexes (Vanormelingen et al., 2008 ; Devercelli et al., 2016 ; Bortolini et al., 2017 ), or for the lakes of Fennoscandia (Ptacnik et al., 2010a , b ), where the large lake density facilitates the manifestation of spatially acting processes. High selection and low dispersal represent the position of phytoplankton inhabiting isolated lakes. Reviewing the literature of algal dispersal Reynolds concluded ( 2006 ) that cosmopolitan and pandemic distribution of algae is due to the fact that most of the planktic species effectively exploit the dispersal channels. However, he also noted that several species are not good dispersers, therefore, endemism might occur among algae.

Positions of the relevant types of pelagial aquatic habitats in the selection/dispersal plane

Composition and diversity of these assemblages are controlled by the locally acting environmental filtering and by biotic interactions, frequently, by competition. The environmental filtering metaphor appears in Reynolds’ habitat template approach (Reynolds, 1998 ), where the template is scaled against quantified gradients of energy and resource availability. The template represents the filter, while the habitats mean the porosity (Reynolds, 2003 ). Species that manage to pass the filter are the candidate components of the assemblages. Finally, low-level biotic interactions (Vellend, 2016 ) determine the composition and diversity of the communities.

The four mechanisms, proposed by Vellend, act differently on the various metric values of diversity. Using the special cases of Rényi’s entropy (α: → 0, 1, 2, ∞) (ESM Box 1) we can show how mechanisms influence species richness and species inequalities, and how they act on the metrics between these extremes (ESM Table 1). Drivers of functional diversity are identical with that of species diversity, but their impacts are attenuated by the functional redundancy of the assemblages.

The role of competition in the maintenance of diversity

The concept of competition and coexistence has been first proved experimentally both for artificial two-species systems (Tilman & Kilham, 1976 ; Tilman, 1977 ) and for natural phytoplankton assemblages (Sommer, 1983 ). However, limitations by different nutrients are responsible only for a small portion of diversity, even if the micronutrients are also included. Therefore, it was an important step when Sommer ( 1984 ) applying a pulsed input of one key nutrient in a flow-through culture managed to maintain the coexistence of several species; although they were competing for the same resource. Several competition experiments have been carried out in recent years demonstrating the role of inter- (Ji et al., 2017 ) and intra-specific competition (Sildever et al., 2016 ) in the coexistence of planktic algae.

The fact that one single resource added in pulses can maintain the coexistence of multiple species has been also proved by mathematical modelling (Ebenhöh, 1988 ). Using deterministic models, Huisman & Weissing ( 1999 ) showed that competition for three or more resources result in sustained species oscillations or chaotic dynamics even under constant resource supply. These oscillations in species abundance make possible the coexistence of several species on a few limiting resources (Wang et al., 2019 ).

The non-equilibrium nature of phytoplankton and the role of disturbances

One of the underlying assumptions of the classical competition theories is that species coexistence requires a stable equilibrium point (Chesson & Case, 1986 ). However, the stable equilibrium state is not a fundamental property of ecosystems (DeAngelis & Waterhouse, 1987 ; Hastings et al., 2018 ). Hutchinson put forward the idea that phytoplankton diversity could be explained by “permanent failure to achieve equilibrium” (Hutchinson, 1941 ). On a sufficiently large timescale, ecosystems seem to show transient dynamics, and do not necessarily converge to an equilibrium state (Hastings et al., 2018 ). However, the virtually static equilibrium-centred view of ecological processes cannot explain the transient behaviour of ecosystems (Holling, 1973 ; Morozov et al., 2019 ). Today, there is a broad consensus in phytoplankton ecology that composition and diversity of phytoplankton can be best explicable by non-equilibrium approaches (Naselli-Flores et al., 2003 ). The non-equilibrium theories do not reject the role of competition in community assembly but place a larger emphasis on historical effects, chance factors, spatial inequalities, environmental perturbations (Chesson & Case, 1986 ), and transient dynamics of the ecosystems (Hastings, 2004 ). The interactions among the internally driven processes and the externally imposed stochasticity of environmental variability as an explanation of community assembly have been conceptualized in the Intermediate Disturbance Hypothesis (IDH) (Connell, 1978 ). This hypothesis predicts a unimodal relationship between the intensities and frequencies of disturbances and species richness. Although this hypothesis has been developed for macroscopic sessile communities, it has become widely accepted in phytoplankton ecology (Sommer, 1999 ). It has been proposed that the frequency of disturbances has to be measured on the scale of generation times of organisms (Reynolds, 1993 ; Padisák, 1994 ). Field observation suggested that diversity peaked at disturbance frequency of 3–5 generation times (Padisák et al., 1988 ), which was also corroborated by laboratory experiments (Gaedeke & Sommer, 1986 ; Flöder & Sommer, 1999 ). The IDH, however, is not without weaknesses (Fox, 2013 ). Recognition and measurement of disturbance are among the main concerns (Sommer et al., 1993 ). Diversity changes are measured purely as responses to unmeasured events (disturbances) (Juhasz-Nagy, 1993 ), which readily leads to circular reasoning. Repeated disturbances might change the resilience of the system, which modifies the response of communities and makes the impact of disturbances on diversity unpredictable (Hughes, 2012 ).

Amalgamation of the equilibrium and non-equilibrium concepts

The existence of the equilibrium and non-equilibrium explanations of species coexistence represents a real dilemma in ecology. Being sufficiently different, and thus avoid strong competition, or sufficiently similar with ecologically irrelevant exclusion rates (as it is suggested by Hubbell’s neutral theory ( 2006 )) are both feasible strategies for species (Scheffer & van Nes, 2006 ). Coexistence of species with these different strategies is also feasible if the many sufficiently similar species create clusters along the niche axes (in accordance with Hubbel’s ( 2006 ) neutral theory), and the competitive abilities within the clusters are sufficiently large. It has been demonstrated that the so-called “lumpy coexistence” is characteristic for phytoplankton assemblages (Graco-Roza et al., 2019 ). Lumpy coexistence arises in fluctuating resource environments (Sakavara et al., 2018 ; Roelke et al., 2019 ), and show higher resilience to species invasions (Roelke & Eldridge, 2008 ) and higher resistance to allelopathy (Muhl et al., 2018 ).

The model of lumpy coexistence has its roots in mechanistic modelling of species coexistence (Scheffer & van Nes, 2006 ). Analysing lake phytoplankton data Reynolds ( 1980 , 1984 , 1988 ) demonstrated that species with similar preferences and tolerances to environmental constraints like nutrients or changes in water column stratification frequently coexist. These empirical observations were formalised later in the functional group (FG) concept (Reynolds et al., 2002 ). Despite their different theoretical backgrounds, the two approaches came to identical conclusions: species having similar positions on the niche axes form species clusters (or FGs), and in natural assemblages clusters or FGs coexist. Thus, the concept of lumpy coexistence can also be considered as a mechanistic explanation of the Reynolds’s FG concept.

The mechanisms and forces detailed above can explain how diversity is maintained at the local scale. Recent metacommunity studies, however, indicate that spatial processes have a crucial role in shaping phytoplankton diversity (Devercelli et al., 2016 ; Bortolini et al., 2017 ; Guelzow et al., 2017 ; Benito et al., 2018 ). Despite the increasing research activity in this field, spatial processes are far less studied than local ones. More in-depth knowledge on the role of connectivity of aquatic habitats and dispersal mechanisms of the phytoplankters will contribute to better understand phytoplankton diversity at regional or global scales.

Changes of diversity along environmental scales

Species–area relationships across systems.

The area dependence of species richness deserved special attention in ecology both from theoretical and practical points of view. The increase of species number with the area sampled is an empirical fact (Brown & Lomolino, 1998 ). The first model that described the so-called species–area relationship (SAR) appeared first by Arrhenius ( 1921 ) who proposed to apply power law for predicting species richness from the surveyed area. Because of the differences in the studied size scale and the studied organism groups, several other models have also been proposed such as the exponential (Gleason, 1922 ), the logistic (Archibald, 1949 ) and the linear (Connor & McCoy, 1979 ) models. However, the power-law ( S  =  c  ×  A z , where S : number of species; A : area sampled; c : the intercept, z : the exponent) is still the most widely used formula in SAR studies. The rate of change of the slope with an increasing area ( z value) depends on the studied organisms, and also on the localities. High values ( z : 0.1–0.5) were reported for macroscopic organisms (Durrett & Levin, 1996 ), while low z values characterised ( z : 0.02–0.08) the microbial systems (Azovsky, 2002 ; Green et al. 2004 ; Horner-Devine et al. 2004 ).

The phytoplankton SAR appeared first in Hutchinson’s ( 1961 ) paper, where he analysed Ruttner’s dataset on Indonesian (Ruttner, 1952 ), and Järnefelt’s ( 1956 ) data on Finnish lakes. He concluded that there was no significant relationship between the area and species richness. Hutchinson reckoned that contribution of the littoral algae to the phytoplankton might be relevant, and because the littoral/pelagic ratio decreases with lake size, this contribution also decreases. Therefore, species richness cannot increase with lake area. In a laboratory experiment, Dickerson & Robinson ( 1985 ) found that large microcosms had significantly smaller species richness values than small ones. Based on laboratory studies, published species counts from ponds lakes and oceans, Smith et al. ( 2005 ) studied phytoplankton SAR in the possible largest size scale (10 −9 to 10 7  km 2 ). They demonstrated a significant positive relationship between area and species richness. The calculated z value ( z  = 0.134) was higher than those reported in other microbial SAR studies. However, we note that this study suffers from a methodological shortcoming, because of differences in compilation of species inventories. Therefore, the results are only suggestive of possible trends that should be investigated more thoroughly.

Analysing phytoplankton monitoring data of 540 lakes in the USA Stomp et al. ( 2011 ) found only a slight increase in richness values with a considerable amount of scatter in the data. The covered size range was small in this study, and the applied counting techniques could lead to bias in richness estimation. Phytoplankton species richness showed a similar weak relationship with lake size for Scandinavian lakes (Ptacnik et al., 2010a , b ), although the counting effort was much better standardised. All the above studies suggested that species richness was not independent of water body size. However, because of the methodological differences, and differences in the covered water body size, in richness estimation or the type of the water bodies, any conclusions based on these results should be handled with caution.

Nutrients, latitudinal and altitudinal differences (Stomp et al., 2011 ) or the size of the regional species pool (Fox et al., 2000 , Ptacnik et al., 2010a , b ) also influence phytoplankton diversity. To reduce the impact of these factors, Várbíró et al. ( 2017 ) investigated phytoplankton SAR in a series of standing waters within the same ecoregion and with similar nutrient status. The water bodies covered ten orders of magnitude size range (10 −2 to 10 8  m 2 ). In this study, both the sampling effort and the sample preparation was standardised. The authors demonstrated that species richness did not scale monotonously with water body size. They managed to show the presence of the so-called Small Island Effect (SIE, Lomolino & Weiser, 2001 ), the phenomenon, when below a certain threshold area (here 10 −2 to 10 2  m 2 size range) species richness varies independently of island size. A right-skewed hump-shaped relationship was found between the area and phytoplankton species richness with a peak at 10 5 –10 6  m 2 area. This phenomenon has been called as Large Lake Effect (LLE) by the authors, and they explained it by the strong wind-induced mixing, which acts against habitat diversity in the pelagic zones of large lakes. The significance of this study is that its results help explain the controversial results of earlier phytoplankton SAR studies. The LLE explains why the species richness had not grown in the case of the Ruttner’s and Jarnefelt’s dataset. The SIE, however, explains why Dickerson & Robinson ( 1985 ) found opposite tendencies to SAR in microcosm experiments. Detailed analysis of the phytoplankton in those water bodies that produced the peak in the SAR curve in the study of Várbíró et al. ( 2017 ) demonstrated that high diversity has been caused by the intrusion of metaphytic elements to the pelagic zone (Görgényi et al., 2019 ), which can be considered as a within-lake metacommunity process.

Productivity–diversity relationships

Despite the more than half a century-long history of investigations on the productivity/diversity relationship (PDR), the shape of the relationship and the underlying mechanisms still remain a subject of debate. The models describing the PDR vary from the monotonic increasing, through the hump shaped and u-shaped to the monotonic decreasing types in the literature (Waide et al., 1999 ). In the PDR studies, there are great differences in the applied scale (local/regional/global), in the metric used to define productivity (e.g., nutrients, biomass, production rate, precipitation, evaporation), in the used diversity metrics, and also in the studied group of organisms (special phylogenetic groups, functional assemblages) (Mittelbach et al., 2001 ). PDR studies also have other methodological and statistical problems (Mittelbach et al., 2001 ). These differences in approaches may generate different patterns, which lead to confusion and inconclusive results (Whittaker & Heegaard, 2003 ; Hillebrand & Cardinale, 2010 ). Despite these uncertainties, the most general PDR patterns are the hump-shaped and positive linear relationships; the first has been observed mostly in the case of local, while the second in the case of regional scale studies (Chase & Leibold, 2002 ; Ptacnik et al., 2010a ). These patterns are so robust that they have been shown for various organisms independently from the highly different proxies applied to substitute the real productivity.

The number of studies that explicitly focus on phytoplankton PDR is few. The view that phytoplankton diversity peaks at intermediate productivity level has been demonstrated by several authors (Ogawa & Ichimura, 1984 ; Agustí et al., 1991 ; Leibold, 1999 ). This is greatly due to the fact that phytoplankton studies fortunately do not suffer from scaling problem: most studies use sample-based local α – s as diversity metrics and nutrients or biomass (Chl-a) as a surrogate measure of productivity. Unimodal relationships were found for Czech (Skácelová & Lepš, 2014 ) and Hungarian water bodies (Borics et al., 2014 ). Diversity peaked in both cases at the 10 1 –10 2  mg L −1 biovolume range, characteristic for eutrophic lakes.

It has also been demonstrated that the unimodal relationship was also true for the functional richness/productivity relationship (Borics et al., 2014 ; Török et al., 2016 ). Differences were also found between the species richness and functional richness peaks; the latter peaked at smaller biovolume range (Török et al., 2016 ). We note here that all three studies were based on monitoring data, and because of the applied sample processing, species richness values might be slightly underestimated.

Several theories have been proposed to explain this unimodal pattern. Moss ( 1973 ) reckoned that the relationship could be accounted for by that the populations of oligotrophic and eutrophic lakes overlapping at the intermediate productivity range. Rosenzweig’s ( 1971 ) paradox of enrichment hypothesis explained the unimodal relationship by the destabilized predator–prey relationship at enhanced productivity level. Tilman’s resource heterogeneity model ( 1985 ) predicts that the coexistence of competing species is enhanced when the supply of alternative resources is heterogeneous both spatially and temporally. This heterogeneity increases with resource supply together with species richness up to the point when richness declines because the correlation between spatiotemporal heterogeneity and resource supply disappeares. The resource-ratio hypothesis can also provide an explanation of the hump shaped PDR (Tilman & Pacala, 1993 ; Leibold, 1997 ). This theory suggests that relative supply of resources generates variations in species composition. Identity of the most strongly limiting resource changes, and at very high resource supply (on the descending end of the curve) only a few K-strategist specialists will prevail. The species pools overlap at intermediate productivity level, resulting in high species richness. This explanation seems to be reasonable for phytoplankton PDR studies.

Investigating the PDR in fishless ponds, Leibold ( 1999 ) found that his results could be best explained by the keystone predation hypothesis (Paine, 1966 ). This theory asserts that at low productivity exploitative competition is the main assembly rule, while with increasing productivity range the role of predator avoidance becomes more important.

The number of various explanations illustrates the complexity of processes affecting the shape of the PDR. The shifting effects of bottom-up vs. top-down control on the trophic gradient, the size of the regional species pool, that is, the number of potential colonizers, or the history of the studied water bodies (naturally eutrophic lakes are studied, or eutrophicated formerly oligotrophic ones) can considerably modify the properties of the PDRs.

With a few exceptions (Irigoien et al., 2004 ), phytoplankton PDRs have been studied almost exclusively in standing waters.

Investigating the phytoplankton PDRs in rivers Borics et al. ( 2014 ) found monotonic increasing pattern in rhithral and monotonic decreasing PDR in potamal rivers. They explained the positive linear PDR with the newly arriving species from the various adjacent habitats of the watershed, which resulted in high phytoplankton diversity even at highly eutrophic conditions. This phytoplankton is a mixture of those elements that enter the river from the connected water bodies of various types. In contrast, potamal rivers are highly selective environments in which the phytoplankton succession frequently terminates in low diversity plankton dominated by K strategist centric diatoms ( Cyclotella and S tephanodiscus spp.).

We note here that study of the regional phytoplankton PDR should be an important and challenging area of future work, which is presently hindered by the disconnected databases and by difficulties in measuring regional productivity.

Linkage between diversity and the metabolic theory of ecology

Metabolism controls patterns, processes and dynamics at each level of biological organisation from single cells to ecosystems, summarised as the metabolic theory of ecology (Brown et al., 2004 ). Metabolic theory (MTE) provides alternative explanations for observations on various fields of ecology such as in individual performance, life history, population and community dynamics, as well as in ecosystem processes. According to MTE, dynamics of metabolic processes have implications for species diversity. Metabolic processes influence population growth and interspecific competition, might accelerate evolutionary dynamics and the rate of speciation (Brown et al., 2004 ). The direct linkage between temperature and metabolic rate raises the possibility of new explanations of the well-known latitudinal dependence of species richness. Allen et al. ( 2002 ) found that for both terrestrial and aquatic environments natural logarithm of species richness should be a linear function of the mean temperature of the environment. This model has been tested both for lake and oceanic phytoplankton. Investigating more than 600 European, North and South American lakes Segura et al. ( 2015 ) found a pronounced effect of temperature on species diversity between 11 and 17 °C. Righetti et al. ( 2019 ) analysed the results of more than 500,000 phytoplankton observations from the global ocean, and also showed the relationship between temperature and species richness, but similarly to freshwater lakes the relationship was not monotonic for the whole temperature gradient. These results suggest that the MTE can be a possible explanation for the temperature dependence of diversity. However, we note that other theories emphasising longer “effective” evolutionary time (Rohde, 1992 ) or higher resource availability (Brown & Lomolino, 1998 ) can also explain this general pattern.

The functional diversity–ecosystem functioning relationship in phytoplankton

More diverse communities perform better in terms of resource use and ecosystem stability (Naeem & Li, 1997 ); known as the biodiversity-ecosystem functioning relationship (BEF). Similar to BEF relationships shown in terrestrial plant communities (Tilman et al., 1996 , 1997 ), positive BEF relationships have also been evidenced in both natural and synthetic phytoplankton communities (Ptacnik et al., 2008 ; Striebel et al., 2009 ; Stockenreiter et al., 2013 ). The BEF relationship itself, however, does not explain the mechanisms underlying the relationship. The most often recognised mechanisms are complementarity (Loreau & Hector, 2001 ) and sampling effect (Fridley, 2001 ). Complementarity means that more diverse communities complement each other in resource use in a more efficient way. Sampling effect, on the other hand, means that the chance increases for the presence of species with effective functional attributes in more diverse communities (Naeem & Wright, 2003 ).

In an attempt to get mechanistic understanding of diversity-functioning relationships, there is a growing interest in quantifying functional diversity of ecological communities (Hillebrand & Matthiessen, 2009 ). Functional diversity summarizes the values and ranges of traits that influence ecosystem functioning (Petchey & Gaston, 2006 ). By translating taxonomic into functional diversity, we may eventually also distinguish complementarity from sampling effect.

In phytoplankton ecology, two functional perspectives have been developing. First, the identification of morphological, physiological and behavioural traits (Weithoff, 2003 ; Litchman & Klausmeier, 2008 ) that affect fitness (Violle et al., 2007 ) and are, therefore, functional traits. Traits have been used in phytoplankton ecology at least since Margalef’s ‘life forms’ concept (Margalef, 1968 ; 1978 ), even if they were not referred to ‘traits’ explicitely (Weithoff & Beisner, 2019 ). Second, the recognition of characteristic functional units within phytoplankton assemblages led to the development of functional group (ecological groups) concepts (see Salmaso et al., 2015 ). These are the phytoplankton functional group concept sensu Reynolds (FG, Reynolds et al., 2002 ), the morpho-functional group concept (MFG, Salmaso & Padisák, 2007 ), and the morphological group concept (MBFG, Kruk et al., 2011 ).

The functional trait concept has been advocated in trait-based models (Litchman et al., 2007 ) and aimed at translating biotic into functional diversity, which eventually would allow quantify functional diversity at the community level. The functional trait concept has recently been reviewed in context of measures and approaches in marine and freshwater phytoplankton (Weithoff & Beisner, 2019 ). On the other hand, the ‘functional group’ concepts have rather been developed in the context of describing characteristic functional community compositions in specific set of environment conditions (that is, the functional community–environment relationship).

The simplest functional diversity measure of phytoplankton is the number of ‘functional units’ in assemblages. That is, either the number of unique combinations of functional traits or the number of ecological groups indentified. One way to use functional units is to convert them into univariate measures corresponding to those calculated from taxonomic information (e.g., richness, evenness). Or, trait data also allow the calculation of community-level means of trait values (CWM) as an index of functional community composition (Lavorel et al., 2008 ). Second, one may consider calculate the components of functional diversity (FD) such as functional richness, functional evenness, and functional divergence (Mason et al., 2005 ); all representing independent factes of functional community compositions. The same FD concept has been developed further accounting also for the abundance of taxa within a multidimensional trait space based on functional evenness, functional divergence and functional dispersion (Laliberté & Legendre, 2010 ). The recently developed ‘FD’ R package enables one to calculated easily all the aforementioned FD measures (Laliberté & Legendre, 2010 ; Laliberté et al., 2014 ). The use of FD components in the context of BEF in phytoplankton has only started very recently (Abonyi et al., 2018a , b ; Ye et al., 2019 ). Trait-based functional diversity measures in BEF have recently been reviewed by Venail ( 2017 ).

The functional community composition–environment relationship

Functional traits can be classified as those affecting fitness via growth and reproduction (i.e., functional effect traits) and those responding to alterations in the environment (i.e., functional response traits) (Hooper et al., 2002 , 2005 ; Violle et al., 2007 ). Since many ecophysiological traits, such as nutrient and light utilization and grazer resistance, correlate with phytoplankton cell size (Litchman & Klausmeier, 2008 ), size has been recognized as a master trait. Phytoplankton cell size responds to alterations in environmental conditions, like change in water temperature (Zohary et al., 2020 ), and also affects ecosystem functioning (Abonyi et al., 2020 ). The response of freshwater phytoplankton size to water temperature changes seems to be consequent based on both the cell and colony (filament) size (Zohary et al., 2020 ). However, one may consider that cell and colony (filament) sizes are affected by multiple underlying mechanisms, and the choose of cell or colony size as functional trait might be question specific.

The functional group (ecological group) composition of phytoplankton can be predicted well by the local environment (Salmaso et al., 2015 ). However, the different functional approaches have rarely been compared in terms of how they affect the community composition–environment relationship. Kruk et al. ( 2011 ) showed that the morphological group (MBFG) composition of phytoplankton could be predicted from the local environment in a more reliable way than Reynolds’s functional groups (FG), or taxonomic composition. In a broad-scale phytoplankton dataset from Fennoscandia, Abonyi et al. ( 2018a , b ) showed that phytoplankton functional trait categories, as a community matrix, corresponded with the local environment better than Reynolds’s functional groups or the taxonomic matrix. Along the entire length of the Atlantic River Loire, Abonyi et al. ( 2014 ) showed that phytoplankton composition based on Reynolds’s FG classification provided more detailed correspondence to natural- and human-induced changes in environmental conditions than based on the morpho-functional (MFG) and morphological (MBFG) systems.

The aggregation of taxonomic information into functional units reduces data complexity that could come along with reduced ecological information (Abonyi et al., 2018a , b ). Reduced data complexity can be useful as long as it does not imply serious loss of ecological information. Information lost can happen when functional traits are not quantified adequately, cannot be identified, or when ecologically diverse taxa, such as benthic diatoms are considered similar functionally (Wang et al., 2018 ). Otherwise, the aggregation of taxonomic to functional data highlights ecological similarities among taxa (Schippers et al., 2001 ) and should lead to better correspondence between community composition and the environment (Abonyi et al., 2018a , b ).

The functional diversity–ecosystem functioning relationship

Based on taxonomic data, recent studies support a positive biodiversity–ecosystem functioning relationship in phytoplankton clearly (Naeem & Li, 1997 ; Ptacnik et al., 2008 ; Striebel et al., 2009 ). The well-known paradox of Hutchinson asking how so many species may coexist in phytoplankton (Hutchinson, 1961 ) has been reversed to how many species ensure ecosystem functioning (Ptacnik et al., 2010b ). Based on functional traits, however, almost half of the studies reported null or negative relationship between functional diversity and ecosystem functioning (Venail, 2017 ). Recently, Abonyi et al. ( 2018a , b ) argued that functional diversity based on trait categories (i.e., functional trait richness—FTR) and Reynolds’ ecological groups (i.e., functional group richness—FGR) represented different aspects of community organisation in phytoplankton. While both functional measures scaled with taxonomic richness largely, FTR suggested random or uniform occupation of niche space (Díaz & Cabido, 2001 ), while FGR more frequent niche overlaps (Ehrlich & Ehrlich, 1981 ), and therefore, enhanced functional redundancy (Díaz & Cabido, 2001 ). A key future direction will be to understand mechanisms responsible for the co-occurrence of functional units (‘functional groups’) within phytoplankton assemblages, and detail phytoplankton taxa within and among the ecological groups in a trait-based approach. This will enhance our ability to disentangle the ecological role of functional redundancy (within groups) and complementarity (among groups) in affecting ecosystem functioning in the future.

Phytoplankton diversity using molecular tools

The assessment of phytoplankton diversity in waterbodies is strongly dependent from the methods used in the taxonomic identification of species and the quantitative estimation of abundances. The adoption of different methods can strongly influence the number of taxa identified and the level of detail in the taxonomic classifications.

Premise: advantages and weaknesses of light microscopy

Traditionally, phytoplankton microorganisms have been identified using light microscopy (LM). The use of this technique was instrumental to lay the foundation of phytoplankton taxonomy. Many of the most important and well-known species of nano- (2–20 μm), micro- (20–200 μm) and macrophytoplankton (> 200 μm) have been identified by several influential papers and manuals published between the first half of the 1800 s and first half of 1900 s (e.g. (Ehrenberg, 1830 ; de Toni, 1907 ; Geitler & Pascher, 1925 ; Guiry & Guiry, 2019 ). LM is an inexpensive method providing plenty of information on the morphology and size of phytoplankton morphotypes, allowing also obtaining, if evaluated, data on abundances and community structure. Conversely, in addition to being time-consuming, the correct identification of specimens by LM requires a deep knowledge of algal taxonomy. Further, many taxa have overlapping morphological features so that the number of diacritical elements often is not enough to discriminate with certainty different species (Krienitz & Bock, 2012 ; Whitton & Potts, 2012 ; Wilmotte et al., 2017 ). The identification can be further complicated by the plasticity that characterise a number of phenotypic characteristics and their dependence from environmental conditions (Komárek & Komárková, 2003 ; Morabito et al., 2007 ; Hodoki et al., 2013 ; Soares et al., 2013 ). The adoption of electron microscopy for the study of ultra-structural details has represented an important step in the characterization of critical species (e.g. Komárek & Albertano, 1994 ) and phyla. For example, in the case of diatoms, scanning electron microscopy had a huge impact on diatom taxonomy, making traditional LM insufficient for the recognition of newly created taxa (Morales et al., 2001 ). Since aquatic samples usually contain many small, rare and cryptic species, a precise assessment of the current biodiversity is unbearable with the only use of classic LM (Lee et al., 2014 ) and electron microscopy. Nonetheless, despite its limitations, the analysis of phytoplankton by LM still continues to be the principal approach used in the monitoring of the ecological quality of waters (Hötzel & Croome, 1999 ; Lyche Solheim et al., 2014 ).

Culture-dependent approaches—classical genetic characterization of strains

Owing to the above limitations, the identification of phytoplankton species by LM has been complemented by the adoption of genetic methods. These methods are based on the isolation of single strains, their cultivation under controlled conditions, and their characterization by polymerase chain reaction (PCR) and sequencing of specific DNA markers able to discriminate among genera and species, and sometimes also between different genotypes of a same species (Wilson et al., 2000 ; D’Alelio et al., 2013 ; Capelli et al., 2017 ). After sequencing, the DNA amplicons obtained by PCR can be compared with the sequences deposited in molecular databases, e.g. those included in the International Nucleotide Sequence Database Collaboration (INSDC: DDBJ, ENA, GenBank) using dedicated tools, such as BLAST queries (Johnson et al., 2008 ). Further, the new sequences can be analysed, together with different homologous sequences, to better characterize the phylogenetic position and taxonomy of the analysed taxa in specific clades (Rajaniemi et al., 2005 ; Krienitz & Bock, 2012 ; Komárek et al., 2014 ). The phylogenetic analyses provide essential information also for evaluating the geographical distribution of species (Dyble et al., 2002 ; Capelli et al., 2017 ) and their colonization patterns (Gugger et al., 2005 ), to infer physiological traits (Bruggeman, 2011 ), and to evaluate relationships between phylogeny and sensitivity to anthropogenic stressors in freshwater phytoplankton (Larras et al., 2014 ). The selection of primers and markers, and their specificity to target precise algal groups is an essential step, which strictly depends on the objectives of investigations and availability of designated databases. For example, though 16S and 18S rRNA genes are the most represented in the INSDC databases, dedicated archives have been curated for the blast and/or phylogenetic analyses of cyanobacteria (e.g. Ribosomal Database Project; Quast et al., 2013 ; Cole et al., 2014 ) and eukaryotes (e.g. Quast et al., 2013 ; Rimet et al., 2019 ). Further, an increase in the sensitivity of the taxonomic identification based on DNA markers can be obtained through the concurrent analysis of multiple genes using Multilocus Sequence Typing (MLST) and Multilocus Sequence Analysis (MLSA) (see Wilmotte et al., 2017 , for details).

A potential issue with the single use of only microscopy or genetic methods is due to the existence of genetically almost identical different morphotypes and to the development of uncommon morphological characteristics in strains cultivated and maintained in controlled culture conditions. To solve these problems, a polyphasic approach has been proposed, which makes use of a set of complementary methods, based besides genetics, on the analysis of phenotypic traits, physiology, ecology, metabolomics and other characters relevant for the identification of species of different phyla (Vandamme et al., 1996 ; Komárek, 2016 ; Salmaso et al., 2017 ; Wilmotte et al., 2017 ).

Considering the existence of different genotypes within a single species (D’Alelio et al., 2011 ; Yarza et al., 2014 ), the genetic characterizations of phytoplankters have to be performed at the level of single strain. Excluding single cell sequencing analyses (see below), the methods have to be therefore applied to isolated and cultivated strains. This represents a huge limitation for the assessment of biodiversity, because the analyses are necessarily circumscribed only to the cultivable organisms. The rarest and the smaller ones are equally lost. Further, the genetic and/or the polyphasic approaches are time-consuming, allowing to process only one species at a time. To solve this limitation, a set of culture-independent approaches to assess biodiversity in environmental samples have been developed since the 1980s.

Culture independent approaches—traditional methods

A consistent number of molecular typing methods based on gel electrophoresis and a variety of other approaches (e.g. quantitative PCR-qPCR) have been applied since the 1980 s and 1990 s in the analysis of microbial DNA, including “phytoplankton” (for a review, see Wilmotte et al., 2017 ). These approaches are tuned to target common regions of the whole genomic DNA extracted from water samples or other substrata, providing information on the existence of specific taxonomic and toxins encoding genes (Campo et al., 2013 ; Capelli et al., 2018 ), and the taxonomic composition of the algal community without the need to isolate and cultivate individual strains. In this latter group of methods, probably one of the most used in phytoplankton ecology is the denaturing gradient gel electrophoresis (DGGE; (Strathdee & Free, 2013 ). Taking advantage of the differences in melting behaviours of double-stranded DNA in a polyacrylamide gel with a linear gradient of denaturants, DGGE allows the differential separation of DNA fragments of the same length and different nucleotide sequences (Jasser et al., 2017 ). This technique is able to discriminate differences in single-nucleotide polymorphisms without the need for DNA sequencing, providing information at level of species and genotypes. For example, analysing samples from eight lakes of different trophic status, Li et al. ( 2009 ) identified complex community fingerprints in both planktic eukaryotes (up to 52 18S rDNA bands) and prokaryotes (up to 59 16S rDNA bands). If coupled with the analyses of excised DNA bands (Callieri et al., 2007 ), or with markers composed of cyanobacterial clone libraries (Tijdens et al., 2008 ), DGGE can provide powerful indications on the diversity and taxonomic composition of phytoplankton. More recent examples of the application of this technique to phytoplankton and eukaryotic plankton are given in Dong et al. ( 2016 ), Batista & Giani ( 2019 ). A recent comparison of DGGE with other fingerprint methods (Terminal restriction fragment length polymorphism, TRFLP) was contributed by Zhang et al. ( 2018 ).

A second method that has been used in the characterization of phytoplankton from microbial DNA is fluorescence in situ hybridization (FISH), and catalysed reporter deposition (CARD)-FISH (Kubota, 2013 ). In freshwater investigations, this technique has been used especially in the evaluation of prokaryotic communities (Ramm et al., 2012 ). A third method deserving mention is cloning and sequencing (Kong et al., 2017 ).

In principle, compared to LM and traditional genetic methods, these techniques can provide an extended view of freshwater biodiversity. Nevertheless, they suffer from several limitations, due to the time, costs and expertise required for the analysis, and the incomplete characterization of biodiversity due to manifest restrictions in the methods (e.g. finite resolution of gel bands in DGGE and number and sensitivity of markers to be used in CARD-FISH). Part of these limits have been solved with the adoption of new generation methods based on the analysis of environmental and microbial DNA.

Culture independent approaches—metagenomics

The more modern methods boost the sequencing approach over the traditional constraints, allowing obtaining, without gel-based methods or cloning, hundreds of thousands of DNA sequences from environmental samples using high throughput sequencing (HTS). Under the umbrella of metagenomics, we can include a broad number of specialized techniques focused on the study of uncultured microorganisms (microbes, protists) as well as plants and animals via the tools of modern genomic analysis (Chen & Pachter, 2005 ; Fujii et al., 2019 ). The methods based on HTS analysis of microbial DNA can be classified under two broad categories, i.e. studies performing massive PCR amplification of certain genes of taxonomic or functional interest, e.g. 16S and 18S rRNA (marker gene amplification metagenomics), and the sequence-based analysis of the whole microbial genomes extracted from environmental samples (full shotgun metagenomics) (Handelsman, 2009 ; Xia et al., 2011 ). While full shotgun metagenomics techniques were used in the first global investigations of marine biodiversity (Venter et al., 2004 ; Rusch et al., 2007 ; Bork et al., 2015 ), the use of marker gene amplification metagenomics in the study of freshwater phytoplankton has shown an impressive increase in the last decade. The reasons are still due to the minor costs (a few tens of euros per sample) and the simpler bioinformatic tractability of sequences of specific genes compared to full shotgun metagenomics.

The large progress and knowledge obtained in the study of microbial communities (Bacteria and Archaea) based on the analysis of the 16S rDNA marker in the more disparate terrestrial, aquatic and host-organisms’ habitats (e.g. gut microbial communities) had a strong influence in directing the type of investigations undertaken in freshwater environments. At present, the majority of the investigations in freshwater habitats are focused on the identification of microbial (including cyanobacteria) communities, with a minority of studies focused on the photosynthetic and mixotrophic protists (phytoplankton) evaluated through deep sequencing of the 18S rDNA marker (e.g. (Mäki et al., 2017 ; Li & Morgan-Kiss, 2019 ; Salmaso et al., 2020 ).

The results obtained from the applications of HTS to freshwater samples are impressive and are unveiling a degree of diversity in biological communities previously unimaginable, including a significant presence of the new group of non-photosynthetic cyanobacteria (Shih et al., 2013 , 2017 ; Salmaso et al., 2018 ; Monchamp et al., 2019 ; Salmaso, 2019 ). Nonetheless, the application of these techniques is not free from difficulties, due to (among the others) the semiquantitative nature of data, the short DNA reads obtained by the most common HTS techniques, the variability in the copy number per cell of the most common taxonomic markers used (i.e. 16S and 18S rDNA), the incompleteness of genetic databases, which are still fed by information obtained by the isolation and cultivation approaches (Gołębiewski & Tretyn, 2020 ; Salmaso et al., 2020 ). Despite these constraints, the use of HTS techniques in the study of phytoplankton, which is just at the beginning, is contributing to revolutionize the approach we are using in the assessment of aquatic biodiversity in freshwater environments, opening the way to a next generation of investigations in phytoplankton ecology and a new improved understanding of plankton ecology.

Conclusions

In this study, we reviewed various aspects of phytoplankton diversity, including definitions and measures, mechanisms maintaining diversity, its dependence on productivity, habitat size and temperature, functional diversity in the context of ecosystem functioning, and molecular diversity.

Phytoplankton diversity cannot be explained without the understanding of mechanisms that shape assemblages. We highlighted how Vellend’s framework on community assembly (speciation, selection, drift, dispersal) could be applied to phytoplankton assemblages. Competition theories and non-equilibrium approaches fitted also well into this framework.

The available literature on phytoplankton species–area relationship contains information on isolated habitats. These studies argue that richness depends on habitat size. However, findings on eutrophic shallow water bodies suggest that habitat diversity can modify the monotonous increasing tendencies and hump-shaped relationship might occur. The literature on lake’s phytoplankton productivity–diversity relationship supports trends reported for terrestrial ecosystems, i.e. a humped shape relationship at local scale if a sufficiently large productivity range is considered. However, the shapes of the curves depend also on the types of the water bodies. In rivers, both monotonic increasing (rhithral rivers) and decreasing (potamal rivers) trends could be observed.

The aggregation of phytoplankton taxonomic data based on functional information reduces data complexity largely. The reduced biological information could come along with ecological information loss, e.g. when traits cannot be quantified adequately, or, when ecologically diverse taxa are considered similar functionally. Since pelagic phytoplankton is relatively similar functionally, the aggregation of taxonomic into functional data can highlight ecological similarities among taxa in a meaningful way. Accordingly, functional composition and diversity may help better relate phytoplankton communities to their environment and predict the effects of community changes on ecosystem functioning.

The adoption of a new generation of techniques based on the massive sequencing of selected DNA markers and planktonic genomes is beginning to change our present perception of phytoplankton diversity. Moreover, being “all-inclusive” techniques, HTS are contributing to change also the traditional concept of “phytoplankton”, providing a whole picture not only of the traditional phytoplankton groups, but of the whole microbial (including cyanobacteria) and protist (including phytoplankton) communities. The new molecular tools not only help species identification and unravel cryptic diversity, but provide information on the genetic variability of species that determine their metabolic range and unique physiological properties. These, basically influence speciation and species performances in terms of biotic interactions or colonisation success, and thus affect species assembly.

Overexploitation of ecosystems and habitat destructions coupled with global warming resulted in huge species loss on Earth. The rate of diversity loss is so high that scientists agree that the Earth’s biota entered the sixth mass extinction (Ceballos et al., 2015 ). While population shrinkage or extinction of a macroscopic animal receive large media interest (writing this sentence we have the news that the Chinese paddlefish/ Psephurus gladius/ declared extinct), extinction rate of poorly known taxa can be much higher (Régnier et al., 2015 ). Phytoplankton, invertebrates and microscopic organisms belongs to groups where extinctions do occur, but the rate of extinctions cannot be assessed. Worldwide, thousands of phytoplankton samples are investigated every day, mostly for water quality monitoring purposes. However, assessment methods focus on the identification of the dominant and subdominant taxa, because these determine mostly the values of quality metrics. Since species richness or abundance-based diversity metrics are not considered as good quality indicators (Carvallho et al., 2013 ), investigators are not forced to reveal the overall species richness of the samples. To give an accurate prediction for the species richness of a water body, an extensive sampling strategy and the use of species estimators would be required. Nevertheless, high local species richness does not necessarily mean good ecosystem health and high nature conservation value; e.g. if weak selection couples with high number of new invaders. Small water bodies with low local alpha diversity but with unique microflora can have high conservation value (Bolgovics et al., 2019 ). Preservation of large phytoplankton species diversity at the landscape or higher geographic level needs to maintain high beta diversity by the protection of unique habitats (Noss, 1983 ). Because of the multiple human impacts and global warming, small water bodies belong to the most endangered habitats whose protection is of paramount importance.

Our understanding about phytoplankton diversity has progressed in the recent decades. These were mainly motivated by elucidating mechanisms that drive diversity, and by the emergence of new approaches for analysing relationships between diversity and ecosystem functioning.

Increasing human pressure and global warming-induced latitudinal shifts in climate zones, resulting in hydrological regime shifts with serious implications for aquatic ecosystems including phytoplankton. These timely challenges will also affect near future trends in phytoplankton studies. The sound theoretical principles, together with the new molecular and statistical tools open new perspectives in diversity research, which, may let us hope that the Golden Age of studying phytoplankton diversity lies before us and not behind.

Each study in this special issue of Hydrobiologia is dedicated to the memory of the late Colin S. Reynolds, who made an outstanding contribution to aquatic science, and considered one of the most prominent phytoplankton ecologists of the last three decades. His encyclopedic work, The ecology of phytoplankton (2006) considered by many as the Bible for lake phytoplankton ecology, and serves still as a reference for many recent works. His oeuvre covers a wide range of topics within aquatic ecology, including community assembly, functional approaches, modelling of biomass production, resilience and health of aquatic ecosystems. Reynolds’s contribution to our understanding of diversity maintenance mechanisms is still relevant and served as a basis for shaping our manuscript.

Abonyi, A., M. Leitão, I. Stanković, G. Borics, G. Várbíró & J. Padisák, 2014. A large river (River Loire, France) survey to compare phytoplankton functional approaches: do they display river zones in similar ways? Ecological Indicators 46: 11–22. https://doi.org/10.1016/j.ecolind.2014.05.038 .

Article   Google Scholar  

Abonyi, A., É. Ács, A. Hidas, I. Grigorszky, G. Várbíró, G. Borics & K. T. Kiss, 2018a. 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. Freshwater Biology 63: 456–472. https://doi.org/10.1111/fwb.13084 .

Abonyi, A., Z. Horváth & R. Ptacnik, 2018b. Functional richness outperforms taxonomic richness in predicting ecosystem functioning in natural phytoplankton communities. Freshwater Biology 63: 178–186.

CAS   Google Scholar  

Abonyi, A., K. T. Kiss, A. Hidas, G. Borics, G. Várbíró & É. Ács, 2020. Cell size decrease and altered size structure of phytoplankton constrain ecosystem functioning in the middle Danube River over multiple decades. Ecosystems. https://doi.org/10.1007/s10021-019-00467-6 .

Article   PubMed   Google Scholar  

Agustí, S., Duarte, C. M. & Canfield, Jr. D. E., 1991. Biomass partitioning in Florida phytoplankton communities. Journal of Plankton Research 13: 239–245.

Google Scholar  

Allen, A. P., J. H. Brown & J. F. Gillooly, 2002. Global biodiversity, biochemical kinetics, and the energetic-equivalence rule. Science 297: 1545–1548.

Archibald, E. E. A., 1949. The specific character of plant communities: II. A quantitative approach. The Journal of Ecology 37: 274–288.

Arrhenius, O., 1921. Species and area. Journal of Ecology 9: 95–99.

Azovsky, A. I., 2002. Size-dependent species-area relationships in benthos: is the world more diverse for microbes? Ecography 25: 273–282.

Baas-Becking, L. G. M., 1934. Geobiologie of inleiding tot de milieukunde. van Stockum and Zoon, The Hague: 263.

Bach, L. T., K. T. Lohbeck, T. B. Reusch & U. Riebesell, 2018. Rapid evolution of highly variable competitive abilities in a key phytoplankton species. Nature Ecology & Evolution 2: 611–613.

Balzano, S., D. Sarno & W. H. Kooistra, 2011. Effects of salinity on the growth rate and morphology of ten Skeletonema strains. Journal of Plankton Research 33: 937–945.

Batista, A. M. M. & A. Giani, 2019. Spatiotemporal variability of cyanobacterial community in a Brazilian oligomesotrophic reservoir: the picocyanobacterial dominance. Ecohydrology & Hydrobiology 19: 566–576.

Benito, X., S. C. Fritz, M. Steinitz-Kannan, M. I. Vélez & M. M. McGlue, 2018. Lake regionalization and diatom metacommunity structuring in tropical South America. Ecology and Evolution 8: 7865–7878.

PubMed   PubMed Central   Google Scholar  

Bolgovics, Á., G. Várbíró, É. Ács, Z. Trábert, K. T. Kiss, V. Pozderka, J. Görgényi, P. Boda, B. A. Lukács, Z. Nagy-László, A. Abonyi & G. Borics, 2017. Phytoplankton of rhithral rivers: its origin, diversity and possible use for quality-assessment. Ecological Indicators 81: 587–596.

Bolgovics, Á., B. Viktória, G. Várbíró, E. Á. Krasznai-K, É. Ács, K. T. Kiss & G. Borics, 2019. Groups of small lakes maintain larger microalgal diversity than large ones. Science of The Total Environment 678: 162–172.

Borics, G., Tóthmérész, B., Lukács, B.A. and Várbíró, G., 2012. Functional groups of phytoplankton shaping diversity of shallow lake ecosystems. In Phytoplankton responses to human impacts at different scales. Springer, Dordrecht: 251–262.

Borics, G., J. Görgényi, I. Grigorszky, Z. László-Nagy, B. Tóthmérész, E. Krasznai & G. Várbíró, 2014. The role of phytoplankton diversity metrics in shallow lake and river quality assessment. Ecological Indicators 45: 28–36.

Bork, P., C. Bowler, C. de Vargas, G. Gorsky, E. Karsenti & P. Wincker, 2015. Tara Oceans Tara Oceans studies plankton at planetary scale. Introduction. Science 348: 873.

Bortolini, J. C., A. Pineda, L. C. Rodrigues, S. Jati & L. F. M. Velho, 2017. Environmental and spatial processes influencing phytoplankton biomass along a reservoirs river floodplain lakes gradient: a metacommunity approach. Freshwater Biology 62: 1756–1767.

Brown, J. H. & M. V. Lomolino, 1998. Biogeography. Sinauer, Sunderland, MA.

Brown, J. H., J. F. Gillooly, A. P. Allen, V. M. Savage & G. B. West, 2004. Toward a metabolic theory of ecology. Ecology 85: 1771–1789.

Bruggeman, J., 2011. A phylogenetic approach to the estimation of phytoplankton traits. Journal of Phycology 47: 52–65.

Callieri, C., G. Corno, E. Caravati, S. Galafassi, M. Bottinelli & R. Bertoni, 2007. Photosynthetic characteristics and diversity of freshwater Synechococcus at two depths during different mixing conditions in a deep oligotrophic lake. Journal of Limnology 66: 81–89.

Campo, E., M.-Á. Lezcano, R. Agha, S. Cirés, A. Quesada & R. El-Shehawy, 2013. First TaqMan assay to identify and quantify the cylindrospermopsin-producing cyanobacterium Aphanizomenon ovalisporum in water. Advances in Microbiology Scientific Research Publishing 03: 430–437.

Capelli, C., A. Ballot, L. Cerasino, A. Papini & N. Salmaso, 2017. Biogeography of bloom-forming microcystin producing and non-toxigenic populations of Dolichospermum lemmermannii (Cyanobacteria). Harmful Algae 67: 1–12.

Capelli, C., L. Cerasino, A. Boscaini & N. Salmaso, 2018. Molecular tools for the quantitative evaluation of potentially toxigenic Tychonema bourrellyi (Cyanobacteria, Oscillatoriales) in large lakes. Hydrobiologia 824: 109–119.

Carvalho, L., S. Poikane, A. L. Solheim, G. Phillips, G. Borics, J. Catalan, C. De Hoyos, S. Drakare, B. J. Dudley, M. Järvinen & C. Laplace-Treyture, 2013. Strength and uncertainty of phytoplankton metrics for assessing eutrophication impacts in lakes. Hydrobiologia 704: 127–140.

Cavender-Bares, J., K. H. Kozak, P. V. Fine & S. W. Kembel, 2009. The merging of community ecology and phylogenetic biology. Ecology Letters 12: 693–715.

Ceballos, G., P. R. Ehrlich, A. D. Barnosky, A. García, R. M. Pringle & T. M. Palmer, 2015. Accelerated modern human-induced species losses: entering the sixth mass extinction. Science Advances 1: e1400253.

Chase, J. M. & M. A. Leibold, 2002. Spatial scale dictates the productivity–biodiversity relationship. Nature 416: 427–430.

Chen, K. & L. Pachter, 2005. Bioinformatics for whole-genome shotgun sequencing of microbial communities. PLoS Computational Biology 1: 106–112.

Chesson, P. L. & T. J. Case, 1986. Overview: nonequilibrium community theories: chance, variability, history. In Diamond, J. & T. J. Case (eds), Community Ecology. Harper and Row Publishers Inc., New York: 229–239.

Clifford, H. T. & W. Stephenson, 1975. An Introduction to Numerical Classification. Academic Press, New York: 229.

Cole, J. R., Q. Wang, J. A. Fish, B. Chai, D. M. McGarrell, Y. Sun, C. T. Brown, A. Porras-Alfaro, C. R. Kuske & J. M. Tiedje, 2014. Ribosomal Database Project: data and tools for high throughput rRNA analysis. Nucleic Acids Research 42: D633–D642.

Connell, J. H., 1978. Diversity in tropical rain forests and coral reefs. Science 199: 1302–1310.

Connor, E. F. & E. D. McCoy, 1979. The statistics and biology of the species–area relationship. The American Naturalist 113: 791–833.

D’Alelio, D., A. Gandolfi, A. Boscaini, G. Flaim, M. Tolotti & N. Salmaso, 2011. Planktothrix populations in subalpine lakes: selection for strains with strong gas vesicles as a function of lake depth, morphometry and circulation. Freshwater Biology 56: 1481–1493.

D’Alelio, D., N. Salmaso & A. Gandolfi, 2013. Frequent recombination shapes the epidemic population structure of Planktothrix (Cyanoprokaryota) in Italian subalpine lakes. Journal of Phycology 49: 1107–1117.

de Toni, G. B., 1907. Sylloge Algarum Omnium Hucusque Cognitarum – Vol 5, Mixophyceae, Vol. 5. Sumptibus Editoris Typis Seminarii, Padova.

DeAngelis, D. L. & J. C. Waterhouse, 1987. Equilibrium and nonequilibrium concepts in ecological models. Ecological Monographs 57: 1–21.

Devercelli, M., P. Scarabotti, G. Mayora, B. Schneider & F. Giri, 2016. Unravelling the role of determinism and stochasticity in structuring the phytoplanktonic metacommunity of the Paraná River floodplain. Hydrobiologia 764: 139–156.

Díaz, S. & M. Cabido, 2001. Vive la différence: plant functional diversity matters to ecosystem processes. Trends in Ecology & Evolution 16: 646–655.

Dickerson, J. E. & J. V. Robinson, 1985. Ecology 66: 966–980.

Dong, X., W. Zhao, L. Lv, H. Zhang, F. Lv, Z. Qi, J. Huang & Q. Liu, 2016. Diversity of eukaryotic plankton of aquaculture ponds with Carassius auratus gibelio, using denaturing gradient gel electrophoresis. Iranian Journal of Fisheries Sciences 15: 1540–1555.

Durrett, R. & S. Levin, 1996. Spatial models for species–area curves. Journal of Theoretical Biology 179: 119–127.

Dyble, J., H. W. Paerl & B. A. Neilan, 2002. Genetic characterization of Cylindrospermopsis raciborskii (cyanobacteria) isolates from diverse geographic origins based on nifH and cpcBA-IGS nucleotide sequence analysis. Applied and Environmental Microbiology 68: 2567–2571.

CAS   PubMed   PubMed Central   Google Scholar  

Ebenhöh, W. 1988. Coexistence of an unlimited number of algal species in a model system. Theoretical Population Biology 34(2): 130–144.

Ehrenberg, C., 1830. Organisation, systematik und geographisches Verhältniss der Infusionsthierchen.

Ehrlich, P. & A. Ehrlich, 1981. Extinction: the causes and consequences of the disappearance of species. Random House, New York.

Fisher, R. A., A. S. Corbet & C. B. Williams, 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. The Journal of Animal Ecology 12: 42–58.

Flöder, S. & U. Sommer, 1999. Diversity in planktonic communities: an experimental test of the intermediate disturbance hypothesis. Limnology and Oceanography 44: 1114–1119.

Flynn, K. J., D. K. Stoecker, A. Mitra, J. A. Raven, P. M. Glibert, P. J. Hansen, E. Granéli & J. M. Burkholder, 2013. Misuse of the phytoplankton–zooplankton dichotomy: the need to assign organisms as mixotrophs within plankton functional types. Journal of Plankton Research 35: 3–11.

Fox, J. W., 2013. The intermediate disturbance hypothesis should be abandoned. Trends in Ecology & Evolution 28: 86–92.

Fox, J. W., J. McGrady-Steed & O. L. Petchey, 2000. Testing for local species saturation with nonindependent regional species pools. Ecology Letters 3: 198–206.

Fridley, J. D., 2001. The influence of species diversity on ecosystem productivity: how, where, and why? Oikos 93: 514–526.

Fujii, K., H. Doi, S. Matsuoka, M. Nagano, H. Sato & H. Yamanaka, 2019. Environmental DNA metabarcoding for fish community analysis in backwater lakes: a comparison of capture methods. PLoS ONE 14: e0210357.

Gaedeke, A. & U. Sommer, 1986. The influence of the frequency of periodic disturbances on the maintenance of phytoplankton diversity. Oecologia 71: 25–28.

Geitler, L., & A. Pascher, 1925. Cyanophyceae and Cyanochloridinae = Chlorobacteriaceae In Pascher, A. (ed), Die Süßwasserflora Deutschlands, Österreichs und der Schweiz. Verlag von Gustav Fisher, Jena: 481.

Gleason, H. A., 1922. On the relation between species and area. Ecology 3: 158–162.

Gołębiewski, M. & A. Tretyn, 2020. Generating amplicon reads for microbial community assessment with next-generation sequencing. Journal of Applied Microbiology 128: 330–354.

Görgényi, J., B. Tóthmérész, G. Várbíró, A. Abonyi, E. T-Krasznai, V. B-Béres & G. Borics, 2019. Contribution of phytoplankton functional groups to the diversity of a eutrophic oxbow lake. Hydrobiologia 830: 287–301.

Gotelli, N. J. & R. K. Colwell, 2011. Estimating species richness. Biological Diversity 12: 39–54.

Graco-Roza, C., A. M. Segura, C. Kruk, P. Domingos, J. Soininen & M. M. Marinho, 2019. Clumpy coexistence in phytoplankton: The role of functional similarity in community assembly. BioRxiv, p. 869966.

Green, J. L., A. J. Holmes, M. Westoby, I. Oliver, D. Briscoe, M. Dangerfield, M. Gillings & A. J. Beattie, 2004. Spatial scaling of microbial eukaryote diversity. Nature 432: 747–750.

Guelzow, N., F. Muijsers, R. Ptacnik & H. Hillebrand, 2017. Functional and structural stability are linked in phytoplankton metacommunities of different connectivity. Ecography 40: 719–732.

Gugger, M., R. Molica, B. Le Berre, P. Dufour, C. Bernard & J.-F. Humbert, 2005. Genetic diversity of Cylindrospermopsis strains (cyanobacteria) isolated from four continents. Applied and Environmental Microbiology 71: 1097–1100.

Guiry, M. D., & G. M. Guiry, 2019. AlgaeBase. World-wide electronic publication – National University of Ireland, Galway, http://www.algaebase.org .

Gunderson, L. H., 2000. Ecological resilience – in theory and application. Annual Review of Ecology and Systematics 31: 425–439.

Handelsman, J., 2009. Metagenetics: spending our inheritance on the future. Microbial Biotechnology 2(2): 138–139.

Hardin, G., 1960. The competitive exclusion principle. Science 131: 1292–1297.

Hastings, A., 2004. Transients: the key to long-term ecological understanding? Trends in Ecology & Evolution 19: 39–45.

Hastings, A., K. C. Abbott, K. Cuddington, T. Francis, G. Gellner, Y. C. Lai, A. Morozov, S. Petrovskii, K. Scranton & M. L. Zeeman, 2018. Transient phenomena in ecology. Science 361: eaat6412.

Hening, A. & D. H. Nguyen, 2020. The competitive exclusion principle in stochastic environments. Journal of Mathematical Biology 80: 1323–1351.

Hillebrand, H. & B. J. Cardinale, 2010. A critique for meta-analyses and the productivity–diversity relationship. Ecology 91: 2545–2549.

Hillebrand, H. & B. Matthiessen, 2009. Biodiversity in a complex world: consolidation and progress in functional biodiversity research. Ecology Letters 12: 1405–1419.

Hodoki, Y., K. Ohbayashi, Y. Kobayashi, H. Takasu, N. Okuda, S. Nakano, et al., 2013. Anatoxin-a-producing Raphidiopsis mediterranea Skuja var. grandis Hill is one ecotype of non-heterocytous Cuspidothrix issatschenkoi (Usačev) Rajaniemi et al. in Japanese lakes. Harmful Algae 21–22: 44–53.

Holling, C. S., 1973. Resilience and stability of ecological systems. Annual Review of Ecology, Evolution, and Systematics 4: 1–23.

Hooper, D. U., M. Solan, A. Symstad, S. Diaz, M. O. Gessner, N. Buchmann, V. Degrange, P. Grime, F. Hulot, F. Mermillod-Blondin, J. Roy, E. Spehn & L. van Peer, 2002. Species diversity, functional diversity, and ecosystem functioning. In Loreau, M. (ed.), Biodiversity and Ecosystem Functioning – Synthesis and Perspectives. Oxford University Press, Oxford: 195–208.

Hooper, D. U., F. S. Chapin, J. J. Ewel, A. Hector, P. Inchausti, S. Lavorel, J. H. Lawton, D. M. Lodge, M. Loreau, S. Naeem, B. Schmid, H. Setälä, A. J. Symstad, J. Vandermeer & D. A. Wardle, 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs 75: 3–35. https://doi.org/10.1890/04-0922 .

Horner-Devine, M. C., M. Lage, J. B. Hughes & B. J. M. Bohannan, 2004. A taxa–area relationship for bacteria. Nature 432: 750–753.

Hötzel, G. & R. Croome, 1999. A phytoplankton methods manual for Australian freshwaters land and water Australia. Land and Water Resources Research and Development Corporation, Canberra.

Hubbell, S. P., 2006. Neutral theory and the evolution of ecological equivalence. Ecology 87: 1387–1398.

Hughes, A., 2012. Disturbance and diversity: an ecological chicken and egg problem. Nature Education Knowledge 3: 48.

Huisman, J. & F. J. Weissing, 1999. Biodiversity of plankton by species oscillations and chaos. Nature 402: 407–410.

Hutchinson, G. E., 1941. Ecological aspects of succession in natural populations. The American Naturalist 75: 406–418.

Hutchinson, G. E., 1961. The paradox of the plankton. The American Naturalist 95: 137–145.

Irigoien, X., J. Huisman & R. P. Harris, 2004. Global biodiversity patterns of marine phytoplankton and zooplankton. Nature 429: 863–867.

Järnefelt, H., 1956, Zur Limnologie einiger Gewasser Finnlands. XVI. Mit besonderer.  

Jasser, I., A. Bukowska, J.-F. Humbert, K. Haukka & D. P. Fewer, 2017. Analysis of toxigenic cyanobacterial communities through denaturing gradient gel electrophoresis. In Kurmayer, R., K. Sivonen, A. Wilmotte & N. Salmaso (eds), Molecular Tools for the Detection and Quantification of Toxigenic Cyanobacteria. Wiley, New York: 263–269.

Ji, X., J. M. Verspagen, M. Stomp & J. Huisman, 2017. Competition between cyanobacteria and green algae at low versus elevated CO 2 : who will win, and why? Journal of Experimental Botany 68: 3815–3828.

Johnson, M., I. Zaretskaya, Y. Raytselis, Y. Merezhuk, S. McGinnis & T. L. Madden, 2008. NCBI BLAST: a better web interface. Nucleic Acids Research 36: W5–W9.

Juhasz-Nagy, P., 1993. Notes on compositional diversity. Intermediate Disturbance Hypothesis in Phytoplankton Ecology. Springer, Dordrecht: 173–182.

Knopf, F. L., 1986. Changing landscapes and the cosmopolitism of the eastern Colorado avifauna . Wildlife Society Bulletin (1973–2006) 14: 132–142.

Komárek, J., 2016. A polyphasic approach for the taxonomy of cyanobacteria: principles and applications. European Journal of Phycology 51: 1–8.

Komárek, J. & P. Albertano, 1994. Cell structure of a planktic cyanoprokaryote Tychonema bourrellyi . Algological Studies/Archiv für Hydrobiologie, Supplement Volumes Schweizerbart’sche Verlagsbuchhandlung. https://doi.org/10.1127/algol_stud/75/1995/157 .

Komárek, J. & J. Komárková, 2003. Phenotype diversity of the cyanoprokaryotic genus Cylindrospermopsis (Nostocales); review 2002. Czech Phycology, Olomouc 3: 1–30.

Komárek, J., J. Kaštovský, J. Mareš & J. R. Johansen, 2014. Taxonomic classification of cyanoprokaryotes (cyanobacterial genera) 2014, using a polyphasic approach. Preslia 86: 295–335.

Kong, P., P. Richardson & C. Hong, 2017. Diversity and community structure of cyanobacteria and other microbes in recycling irrigation reservoirs. PLoS ONE 12: e0173903.

Krienitz, L. & C. Bock, 2012. Present state of the systematics of planktonic coccoid green algae of inland waters. Hydrobiologia 698: 295–326.

Kruk, C., E. T. H. M. Peeters, E. H. Van Nes, V. L. M. Huszar, L. S. Costa & M. Scheffer, 2011. Phytoplankton community composition can be predicted best in terms of morphological groups. Limnology & Oceanography 56: 110–118.

Kubota, K., 2013. CARD-FISH for environmental microorganisms: technical advancement and future applications. Microbes and Environments 28: 3–12.

Laliberté, E. & P. Legendre, 2010. A distance-based framework for measuring functional diversity from multiple traits. Ecology 91: 299–305.

Laliberté E., P. Legendre & B. Shipley, 2014. FD: measuring functional diversity from multiple traits, and other tools for functional ecology. R package version 1.0-12.

Larras, F., F. Keck, B. Montuelle, F. Rimet & A. Bouchez, 2014. Linking diatom sensitivity to herbicides to phylogeny: a step forward for biomonitoring? Environmental Science & Technology 48: 1921–1930.

Lavorel, S., K. Grigulis, S. McIntyre, N. S. G. Williams, D. Garden, J. Dorrough, S. Berman, F. Quétier, A. Thébault & A. Bonis, 2008. Assessing functional diversity in the field – methodology matters! Functional Ecology 22: 134–147.

Lee, E., U. M. Ryan, P. Monis, G. B. McGregor, A. Bath, C. Gordon & A. Paparini, 2014. Polyphasic identification of cyanobacterial isolates from Australia. Water Research 59: 248–261.

Leibold, M. A., 1997. Do nutrient-competition models predict nutrient availabilities in limnetic ecosystems? Oecologia 110: 132–142.

Leibold, M. A., 1999. Biodiversity and nutrient enrichment in pond plankton communities. Evolutionary Ecology Research 1: 73–95.

Leibold, M. A. & J. M. Chase, 2017. Metacommunity Ecology, Vol. 59. Princeton University Press, Princeton.

Li, W. & R. M. Morgan-Kiss, 2019. Influence of environmental drivers and potential interactions on the distribution of microbial communities from three permanently stratified Antarctic lakes. Frontiers in Microbiology Frontiers 10: 1067.

Li, W., Y. Yuhe, T. Zhang, W. Feng, X. Zhang & W. Li, 2009. PCR-DGGE Fingerprinting analysis of plankton communities and its relationship to lake trophic statu. International Review of Hydrobiology 94: 528–541.

Litchman, E. & C. A. Klausmeier, 2008. Trait-based community ecology of phytoplankton. Annual Review of Ecology, Evolution, and Systematics 39: 615–639.

Litchman, E., C. A. Klausmeier, O. M. Schofield & P. G. Falkowski, 2007. The role of functional traits and trade-offs in structuring phytoplankton communities: scaling from cellular to ecosystem level. Ecology Letters 10: 1170–1181.

Lomolino, M. V. & M. D. Weiser, 2001. Towards a more general species–area relationship: diversity on all islands, great and small. Journal of Biogeography 28: 431–445.

Loreau, M. & A. Hector, 2001. Partitioning selection and complementarity in biodiversity experiments. Nature 412: 72–76.

Lyche Solheim, A., G. Phillips, S. Drakare, G. Free, M. Järvinen, B. Skjelbred, D. Tierne, W. Trodd, & S. Poikane, 2014. Water Framework Directive Intercalibration Technical Report: Northern Lake Phytoplankton ecological assessment methods.

Magurran, A., 2004. Measuring Biological Diversity. Blackwell Publishing, Oxford.

Magurran, A. E. & P. A. Henderson, 2003. Explaining the excess of rare species in natural species abundance distributions. Nature 422: 714–716.

Mäki, A., P. Salmi, A. Mikkonen, A. Kremp & M. Tiirola, 2017. Sample preservation, DNA or RNA extraction and data analysis for high-throughput phytoplankton community sequencing. Frontiers in Microbiology Frontiers 8: 1848.

Margalef, R., 1968. Perspectives in ecological theory. 111 pages. The University of Chicago Press, Chicago.

Margalef, R., 1978. Life-forms of phytoplankton as survival alternatives in an unstable environment. Oceanologica 1: 493–509.

Mason, N. W. H., D. Mouillot, W. G. Lee & J. B. Wilson, 2005. Functional richness, functional evenness and functional divergence: the primary components of functional diversity. Oikos 111: 112–118.

McNaughton, J., 1967. Relationship among functional properties of California grassland. Nature 216: 168–169.

Méndez, V., M. Assaf, A. Masó-Puigdellosas, D. Campos & W. Horsthemke, 2019. Demographic stochasticity and extinction in populations with Allee effect. Physical Review E 99: 022101.

Mittelbach, G. G., C. F. Steiner, S. M. Scheiner, K. L. Gross, H. L. Reynolds, R. B. Waide, M. R. Willig, S. I. Dodson & L. Gough, 2001. What is the observed relationship between species richness and productivity? Ecology 82: 2381–2396.

Monchamp, M. E., P. Spaak & F. Pomati, 2019. Long term diversity and distribution of non-photosynthetic cyanobacteria in peri-alpine lakes. Frontiers in Microbiology Frontiers 10: 3344.

Morabito, G., A. Oggioni, E. Caravati & P. Panzani, 2007. Seasonal morphological plasticity of phytoplankton in Lago Maggiore (.N Italy). Hydrobiologia 578: 47–57.

Morales, E. A., P. A. Siver & F. R. Trainor, 2001. Identification of diatoms (Bacillariophyceae) during ecological assessments: comparison between Light Microscopy and Scanning Electron Microscopy techniques. Proceedings of the Academy of Natural Sciences of Philadelphia Academy of Natural Sciences 151: 95–103.

Morozov, A., K. Abbott, K. Cuddington, T. Francis, G. Gellner, A. Hastings, Y. C. Lai, S. Petrovskii, K. Scranton & M. L. Zeeman, 2019. Long transients in ecology: theory and applications. Physics of Life Reviews. https://doi.org/10.1016/j.plrev.2019.09.004 .

Moss, B., 1973. Diversity in fresh-water phytoplankton. The American Midland Naturalist 90: 341–355.

Mouquet, N., V. Devictor, C. N. Meynard, F. Munoz, L. F. Bersier, J. Chave, P. Couteron, A. Dalecky, C. Fontaine, D. Gravel & O. J. Hardy, 2012. Ecophylogenetics: advances and perspectives. Biological Reviews 87: 769–785.

Muhl, R. M., D. L. Roelke, T. Zohary, M. Moustaka-Gouni, U. Sommer, G. Borics, U. Gaedke, F. G. Withrow & J. Bhattacharyya, 2018. Resisting annihilation: relationships between functional trait dissimilarity, assemblage competitive power and allelopathy. Ecology Letters 21: 1390–1400.

Naeem, S. & S. Li, 1997. Biodiversity enhances ecosystem reliability. Nature 390(6659): 507–509. https://doi.org/10.1038/37348 .

Article   CAS   Google Scholar  

Naeem, S. & J. P. Wright, 2003. Disentangling biodiversity effects on ecosystem functioning: deriving solutions to a seemingly insurmountable problem. Ecology Letters 6: 567–579.

Naselli-Flores, L. & G. Rossetti, 2010. Santa Rosalia, the icon of biodiversity. Hydrobiologia 653: 235–243.

Naselli-Flores, L., J. Padisák, M. T. Dokulil & I. Chorus, 2003. Equilibrium/steady-state concept in phytoplankton ecology. Hydrobiologia 502: 395–403.

Naselli-Flores, L., R. Termine & R. Barone, 2016. Phytoplankton colonization patterns. Is species richness depending on distance among freshwaters and on their connectivity? Hydrobiologia 764: 103–113.

Noss, R. F., 1983. A regional landscape approach to maintain diversity. BioScience 33: 700–706.

Nygaard, G., 1949. Hydrobiological studies on some Danish ponds and lakes. Pert II: The quotient hypothesis and some little known plankton organisms. Vidensk Danske. Selsk. Biol. Skr. 7: 1–293.

Ogawa, Y. & S. E. Ichimura, 1984. Phytoplankton diversity in inland waters of different trophic status. Japanese Journal of Limnology (Rikusuigaku Zasshi) 45: 173–177.

Padfield, D., G. Yvon-Durocher, A. Buckling, S. Jennings & G. Yvon-Durocher, 2016. Rapid evolution of metabolic traits explains thermal adaptation in phytoplankton. Ecology Letters 19: 133–142.

Padisák, J., 1994. Identification of relevant time-scales in nonequilibrium community dynamics, conclusions from phytoplankton surveys. New Zealand Journal of Ecology 18: 169–176.

Padisák, J., L. G. Tóth & M. Rajczy, 1988. The role of storms in the summer succession of the phytoplankton community in a shallow lake (Lake Balaton, Hungary). Journal of Plankton Research 10: 249–265.

Paine, R. T., 1966. Food web complexity and species diversity. The American Naturalist 100: 65–75.

Parvinen, K., U. Dieckmann, M. Gyllenberg & J. A. Metz, 2003. Evolution of dispersal in metapopulations with local density dependence and demographic stochasticity. Journal of Evolutionary Biology 16: 143–153.

Pearson, D. E., Y. K. Ortega, Ö. Eren & J. L. Hierro, 2018. Community assembly theory as a framework for biological invasions. Trends in Ecology & Evolution 33: 313–325.

Petchey, O. L. & K. J. Gaston, 2006. Functional diversity: back to basics and looking forward. Ecology Letters 9: 741–758.

Pomati, F., C. Tellenbach, B. Matthews, P. Venail, B. W. Ibelings & R. Ptacnik, 2015. Challenges and prospects for interpreting long-term phytoplankton diversity changes in Lake Zurich (Switzerland). Freshwater Biology 60: 1052–1059.

Ptacnik, R., A. G. Solimini, T. Andersen, T. Tamminen, P. Brettum, L. Lepistö, E. Willén & S. Rekolainen, 2008. Diversity predicts stability and resource use efficiency in natural phytoplankton communities. Proceedings of the National Academy of Sciences 105: 5134–5138.

Ptacnik, R., T. Andersen, P. Brettum, L. Lepistö & E. Willén, 2010a. Regional species pools control community saturation in lake phytoplankton. Proceedings of the Royal Society B: Biological Sciences 277: 3755–3764.

Ptacnik, R., S. D. Moorthi & H. Hillebrand, 2010b. Hutchinson reversed, or why there need to be so many species. Advances in Ecological Research 43: 1–33.

Quast, C., E. Pruesse, P. Yilmaz, J. Gerken, T. Schweer, P. Yarza, J. Peplies & F. O. Glöckner, 2013. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Research 41: D590–D596.

Rajaniemi, P., P. Hrouzek, K. Kastovská, R. Willame, A. Rantala, L. Hoffmann, J. Komárek & K. Sivonen, 2005. Phylogenetic and morphological evaluation of the genera Anabaena, Aphanizomenon, Trichormus and Nostoc (Nostocales, Cyanobacteria). International Journal of Systematic and Evolutionary Microbiology 55: 11–26.

Ramm, J., A. Lupu, O. Hadas, A. Ballot, J. Rücker, C. Wiedner & A. Sukenik, 2012. A CARD-FISH protocol for the identification and enumeration of cyanobacterial akinetes in lake sediments. FEMS Microbiology Ecology 82: 23–36.

Régnier, C., G. Achaz, A. Lambert, R. H. Cowie, P. Bouchet & B. Fontaine, 2015. Mass extinction in poorly known taxa. Proceedings of the National Academy of Sciences 112: 7761–7766.

Reynolds, C. S., 1980. Phytoplankton assemblages and their periodicity in stratifying lake systems. Ecography 3: 141–159.

Reynolds, C. S., 1984. Phytoplankton periodicity: the interactions of form, function and environmental variability. Freshwater Biology 14: 111–142.

Reynolds, C. S., 1988. The concept of biological succession applied to seasonal periodicity of phytoplankton. Verhandlungen der Internationalen  Verhandlungern für theroretische und angewandte Limnologie 23: 683–691.  

Reynolds, C. S., 1993. Scales of disturbance and their role in plankton ecology. Hydrobiologia 249: 157–172.

Reynolds, C. S., 1998. What factors influence the species composition of phytoplankton in lakes of different trophic status? Hydrobiologia 369: 11–26.

Reynolds, C. S., 2003. Pelagic community assembly and the habitat template. Bocconea 16: 323–339.

Reynolds, C. S., 2006. The Ecology of Phytoplankton. Cambridge University Press, Cambridge.

Reynolds, C. S., J. Padisák & U. Sommer, 1993. Intermediate disturbance in the ecology of phytoplankton and the maintenance of species diversity: a synthesis. Hydrobiologia 249: 183–188.

Reynolds, C. S., V. Huszar, C. Kruk, L. Naselli-Flores & S. Melo, 2002. Towards a functional classification of the freshwater phytoplankton. Journal of Plankton Research 24: 417–428.

Righetti, D., M. Vogt, N. Gruber, A. Psomas & N. E. Zimmermann, 2019. Global pattern of phytoplankton diversity driven by temperature and environmental variability. Science Advances 5: eaau6253.

Rimet, F., E. Gusev, M. Kahlert, M. G. Kelly, M. Kulikovskiy, Y. Maltsev, D. G. Mann, M. Pfannkuchen, R. Trobajo, V. Vasselon, J. Zimmermann & A. Bouchez, 2019. Diat.barcode, an open-access curated barcode library for diatoms. Scientific Reports 9: 1–12.

Roelke, D. L. & P. M. Eldridge, 2008. Mixing of supersaturated assemblages and the precipitous loss of species. The American Naturalist 171: 162–175.

Roelke, D. L., S. E. Cagle, R. M. Muhl, A. Sakavara & G. Tsirtsis, 2019. Resource fluctuation patterns influence emergent properties of phytoplankton assemblages and their resistance to harmful algal blooms. Marine and Freshwater Research 71: 56–67.

Rohde, K., 1992. Latitudinal gradients in species–diversity: the search for the primary cause. Oikos 65: 514–527.

Rosenzweig, M. L., 1971. Paradox of enrichment: destabilization of exploitation ecosystems in ecological time. Science 171(3969): 385–387.

Roy, S. & J. Chattopadhyay, 2007. Towards a resolution of ‘the paradox of the plankton’: a brief overview of the proposed mechanisms. Ecological Complexity 4: 26–33.

Rusch, D. B., A. L. Halpern, G. Sutton, K. B. Heidelberg, S. Williamson, S. Yooseph, D. Wu, J. A. Eisen, J. M. Hoffman, K. Remington, K. Beeson, B. Tran, H. Smith, H. Baden-Tillson, C. Stewart, J. Thorpe, J. Freeman, C. Andrews-Pfannkoch, J. E. Venter, K. Li, S. Kravitz, J. F. Heidelberg, T. Utterback, Y. H. Rogers, L. I. Falcón, V. Souza, G. Bonilla-Rosso, L. E. Eguiarte, D. M. Karl, S. Sathyendranath, T. Platt, E. Bermingham, V. Gallardo, G. Tamayo-Castillo, M. R. Ferrari, R. L. Strausberg, K. Nealson, R. Friedman, M. Frazier & J. C. Venter, 2007. The Sorcerer II Global Ocean Sampling expedition: Northwest Atlantic through eastern tropical Pacific. PLoS Biology Public Library of Science 5: 0398–0431.

Ruttner, F., 1952. Planktonstudien der deutschen limnologischen Sunda Expedition. Archiv fur Hydrobiologie 21: 1–274.

Sakavara, A., G. Tsirtsis, D. L. Roelke, R. Mancy & S. Spatharis, 2018. Lumpy species coexistence arises robustly in fluctuating resource environments. Proceedings of the National Academy of Sciences 115: 738–743.

Salmaso, N., 2019. Effects of habitat partitioning on the distribution of bacterioplankton in deep lakes. Frontiers in Microbiology Frontiers 10: 2257.

Salmaso, N. & J. Padisák, 2007. Morpho-Functional Groups and phytoplankton development in two deep lakes (Lake Garda, Italy and Lake Stechlin, Germany). Hydrobiologia 578: 97–112.

Salmaso, N., L. Naselli-Flores & J. Padisák, 2015. Functional classifications and their application in phytoplankton ecology. Freshwater Biology 60: 603–619.

Salmaso, N., C. Capelli, R. Rippka & A. Wilmotte, 2017. Polyphasic approach on cyanobacterial strains. In Kurmayer, R., K. Sivonen, A. Wilmotte & N. Salmaso (eds), Molecular Tools for the Detection and Quantification of Toxigenic Cyanobacteria. Wiley, New York: 125–134.

Salmaso, N., D. Albanese, C. Capelli, A. Boscaini, M. Pindo & C. Donati, 2018. Diversity and cyclical seasonal transitions in the bacterial community in a large and deep Perialpine Lake. Microbial Ecology 76: 125–143.

Salmaso, N., A. Boscaini & M. Pindo, 2020. Unraveling the diversity of eukaryotic microplankton in a large and deep perialpine lake using a high throughput sequencing approach. Frontiers in Microbiology 11: 789.

Scheffer, M. & E. H. van Nes, 2006. Self-organized similarity, the evolutionary emergence of groups of similar species. Proceedings of the National Academy of Sciences 103: 6230–6235.

Schippers, P., A. M. Verschoor, M. Vos & W. M. Mooij, 2001. Does “supersaturated coexistence” resolve the “paradox of the plankton”? Ecology Letters 4: 404–407.

Segura, A. M., D. Calliari, C. Kruk, H. Fort, I. Izaguirre, J. F. Saad & M. Arim, 2015. Metabolic dependence of phytoplankton species richness. Global Ecology and Biogeography 24: 472–482.

Shannon, C. E., 1948. A mathematical theory of communication. The Bell System Technical Journal 27: 379–423.

Shih, P. M., D. Wu, A. Latifi, S. D. Axen, D. P. Fewer, E. Talla, A. Calteau, F. Cai, N. Tandeau de Marsac, R. Rippka, M. Herdman, K. Sivonen, T. Coursin, T. Laurent, L. Goodwin, M. Nolan, K. W. Davenport, C. S. Han, E. M. Rubin, J. A. Eisen, T. Woyke, M. Gugger & C. A. Kerfeld, 2013. Improving the coverage of the cyanobacterial phylum using diversity-driven genome sequencing. Proceedings of the National Academy of Sciences of the United States of America 110: 1053–1058.

Shih, P. M., J. Hemp, L. M. Ward, N. J. Matzke & W. W. Fischer, 2017. Crown group Oxyphotobacteria postdate the rise of oxygen. Geobiology 15: 19–29.

Sildever, S., J. Sefbom, I. Lips & A. Godhe, 2016. Competitive advantage and higher fitness in native populations of genetically structured planktonic diatoms. Environmental Microbiology 18: 4403–4411.

Skácelová, O. & J. Lepš, 2014. The relationship of diversity and biomass in phytoplankton communities weakens when accounting for species proportions. Hydrobiologia 724: 67–77.

Smith, V. H., B. L. Foster, J. P. Grover, R. D. Holt, M. A. Leibold & F. de Noyelles Jr., 2005. Phytoplankton species richness scales consistently from laboratory microcosms to the world’s oceans. Proceedings of the National Academy of Sciences 102: 4393–4396.

Soares, M. C. S., M. Lürling & V. L. M. Huszar, 2013. Growth and temperature-related phenotypic plasticity in the cyanobacterium Cylindrospermopsis raciborskii . Phycological Research 61: 61–67.

Sommer, U., 1983. Nutrient competition between phytoplankton species in multispecies chemostat experiments. Archiv für Hydrobiologie 96: 399–416.

Sommer, U., 1984. The paradox of the plankton: fluctuations of phosphorus availability maintain diversity of phytoplankton in flow-through cultures 1. Limnology and Oceanography 29: 633–636.

Sommer, U., 1999. Ecology: competition and coexistence. Nature 402: 366.

Sommer, U., J. Padisák, C. S. Reynolds & P. Juhász-Nagy, 1993. Hutchinson’s heritage: the diversity–disturbance relationship in phytoplankton. Hydrobiologia 249: 1–7.

Stockenreiter, M., F. Haupt, A.-K. Graber, J. Seppälä, K. Spilling, T. Tamminen & H. Stibor, 2013. Functional group richness: implications of biodiversity for light use and lipid yield in microalgae. Journal of Phycology 49: 838–847. https://doi.org/10.1111/jpy.12092 .

Stomp, M., J. Huisman, G. G. Mittelbach, E. Litchman & C. A. Klausmeier, 2011. Large-scale biodiversity patterns in freshwater phytoplankton. Ecology 92: 2096–2107.

Strathdee, F. & A. Free, 2013. Denaturing gradient gel electrophoresis (DGGE). In Makovets, S. (ed.), DNA Electrophoresis. Methods in Molecular Biology (Methods and Protocols). Humana Press, Totowa, NJ: 145–157.

Striebel, M., S. Behl & H. Stibor, 2009. The coupling of biodiversity and productivity in phytoplankton communities: consequences for biomass stoichiometry. Ecology 90: 2025–2031. https://doi.org/10.1890/08-1409.1 .

Thunmark, S., 1945. Zur Soziologie des Süsswasserplanktons. Eine methodisch-ökologische Studie. Folia Limnologica Skandinavica 3: 1–66.

Tijdens, M., H. L. Hoogveld, M. P. Kamst-Van Agterveld, S. G. H. Simis, A. C. Baudoux, H. J. Laanbroek & H. J. Gons, 2008. Population dynamics and diversity of viruses, bacteria and phytoplankton in a shallow eutrophic lake. Microbial Ecology 56: 29–42.

Tilman, D., 1977. Resource competition between plankton algae: an experimental and theoretical approach. Ecology 58: 338–348.

Tilman, D., 1985. The resource-ratio hypothesis of plant succession. The American Naturalist 125: 827–852.

Tilman, D. & S. S. Kilham, 1976. Phosphate and silicate growth and uptake kinetics of the diatoms Asterionella formosa and Cyclotella meneghiniana in batch and in batch and semicontinuous culture 1. Journal of Phycology 12: 375–383.

Tilman, D. & S. Pacala, 1993. The maintenance of species richness in plant communities. In Ricklefs, R. & D. Schluter (eds), Species Diversity in Ecological Communities. University of Chicago Press, Chicago: 13–25.

Tilman, D., D. Wedin & J. Knops, 1996. Productivity and sustainability influenced by biodiversity in grassland ecosystems. Nature 379: 718–720. https://doi.org/10.1038/379718a0 .

Tilman, D., J. Knops, D. Wedin, P. Reich, M. Ritchie & E. Siemann, 1997. The influence of functional diversity and composition on ecosystem processes. Science 277: 1300–1302.

Török, P., E. Krasznai, V. Bácsiné Béres, I. Bácsi, G. Borics & B. Tóthmérész, 2016. Functional diversity supports the biomass–diversity humped-back relationship in phytoplankton assemblages. Functional Ecology 30: 1593–1602.

Tóthmérész, B., 1995. Comparison of different methods for diversity ordering. Journal of Vegetation Science 6: 283–290.

Ulrich, W. & M. Ollik, 2005. Limits to the estimation of species richness: the use of relative abundance distributions. Diversity and Distributions 11: 265–273.

Vallina, S. M., P. Cermeno, S. Dutkiewicz, M. Loreau & J. M. Montoya, 2017. Phytoplankton functional diversity increases ecosystem productivity and stability. Ecological Modelling 361: 184–196.

Vandamme, P., B. Pot, M. Gillis, P. de Vos, K. Kersters & J. Swings, 1996. Polyphasic taxonomy, a consensus approach to bacterial systematics. Microbiological Reviews 60: 407–438.

Vanormelingen, P., K. Cottenie, E. Michels, K. Muylaert, W. I. M. Vyverman & L. U. C. De Meester, 2008. The relative importance of dispersal and local processes in structuring phytoplankton communities in a set of highly interconnected ponds. Freshwater Biology 53: 2170–2183.

Várbíró, G., J. Görgényi, B. Tóthmérész, J. Padisák, É. Hajnal & G. Borics, 2017. Functional redundancy modifies species-area relationship for freshwater phytoplankton. Ecology and Evolution 7(23): 9905–9913.

Vellend, M., 2010. Conceptual synthesis in community ecology. The Quarterly Review of Biology 85: 183–206.

Vellend, M., 2016. The Theory of Ecological Communities (MPB-57). Princeton University Press, Princeton.

Venail, P., 2017. Biodiversity ecosystem functioning research in freshwater phytoplankton: a comprehensive review of trait-based studies. Advances in Oceanography and Limnology 8: 1–8.

Venter, J. C., K. Remington, J. F. Heidelberg, A. L. Halpern, D. Rusch, J. A. Eisen, D. Wu, I. Paulsen, K. E. Nelson, W. Nelson, D. E. Fouts, S. Levy, A. H. Knap, M. W. Lomas, K. Nealson, O. White, J. Peterson, J. Hoffman, R. Parsons, H. Baden-Tillson, C. Pfannkoch, Y.-H. H. Rogers & H. O. Smith, 2004. Environmental genome shotgun sequencing of the Sargasso Sea. Science 304: 66–74.

Violle, C., M.-L. Navas, D. Vile, E. Kazakou, C. Fortunel, I. Hummel & E. Garnier, 2007. Let the concept of trait be functional! Oikos 116: 882–892.

Waide, R. B., M. R. Willig, C. F. Steiner, G. G. Mittelbach, L. Gough, S. I. Dodson, G. P. Juday & R. Parmenter, 1999. The relationship between primary productivity and species richness. Annual Review of Ecology and Systematics 30: 257–300.

Wang, C., V.-B. Béres, C. C. Stenger-Kovács, X. Li & A. Abonyi, 2018. Enhanced ecological indication based on combined planktic and benthic functional approaches in large river phytoplankton ecology. Hydrobiologia 818: 163–175.

Wang, L., Y. Tang, R. W. Wang & X. Y. Shang, 2019. Re-evaluating the ‘plankton paradox’using an interlinked empirical data and a food web model. Ecological Modelling 407: 108721.

Weithoff, G., 2003. The concepts of ‘plant functional types’ and ‘functional diversity’ in lake phytoplankton – a new understanding of phytoplankton ecology? Freshwater Biology 48: 1669–1675.

Weithoff, G. & B. E. Beisner, 2019. Measures and approaches in trait-based phytoplankton community ecology – from freshwater to marine ecosystems. Frontiers in Marine Science. https://doi.org/10.3389/fmars.2019.00040 .

Whittaker, R. J. & E. Heegaard, 2003. What is the observed relationship between species richness and productivity? Comment Ecology 84: 3384–3390.

Whitton, B. A. & M. Potts, 2012. Introduction to the cyanobacteria. In Whitton, B. A. (ed.), Ecology of Cyanobacteria II. Springer, Dordrecht: 1–13.

Wilmotte, A., H. D. I. Laughinghouse, C. Capelli, R. Rippka & N. Salmaso, 2017. Taxonomic identification of cyanobacteria by a polyphasic approach. In Kurmayer, R., K. Sivonen, A. Wilmotte & N. Salmaso (eds), Molecular Tools for the Detection and Quantification of Toxigenic Cyanobacteria. Wiley, New York: 79–119.

Wilson, J. B., 1990. Mechanisms of species coexistence: twelve explanations for Hutchinson’s ‘paradox of the plankton’: evidence from New Zealand plant communities. New Zealand Journal of Ecology 13: 17–42.

Wilson, K. M., M. A. Schembri, P. D. Baker & C. P. Saint, 2000. Molecular characterization of the toxic cyanobacterium Cylindrospermopsis raciborskii and design of a species-specific PCR. Applied and Environmental Microbiology 66: 332–338.

Xia, L. C., J. A. Cram, T. Chen, J. A. Fuhrman & F. Sun, 2011. Accurate genome relative abundance estimation based on shotgun metagenomic reads. PLoS ONE. https://doi.org/10.1371/journal.pone.0027992 .

Article   PubMed   PubMed Central   Google Scholar  

Bericksichtigung des Planktons. Annals of Zoological Society “Vancimo” 17: 1–201.

Yarza, P., P. Yilmaz, E. Pruesse, F. O. Glöckner, W. Ludwig, K. H. Schleifer, W. B. Whitman, J. Euzéby, R. Amann & R. Rosselló-Móra, 2014. Uniting the classification of cultured and uncultured bacteria and archaea using 16S rRNA gene sequences. Nature Reviews Microbiology 12: 635–645.

Ye, L., C.-W. Chang, S.-I. S. Matsuzaki, N. Takamura, C. E. Widdicombe & C.-H. Hsieh, 2019. Functional diversity promotes phytoplankton resource use efficiency. Journal of Ecology 107: 2353–2363. https://doi.org/10.1111/1365-2745.13192 .

Zhang, W., Y. Mo, J. Yang, J. Zhou, Y. Lin, A. Isabwe, J. Zhang, X. Gao & Z. Yu, 2018. Genetic diversity pattern of microeukaryotic communities and its relationship with the environment based on PCR-DGGE and T-RFLP techniques in Dongshan Bay, southeast China. Continental Shelf Research 164: 1–9.

Zohary, T., G. Flaim & U. Sommer, 2020. Temperature and the size of freshwater phytoplankton. Hydrobiologia. https://doi.org/10.1007/s10750-020-04246-6 .

Download references

Acknowledgements

Open access funding provided by ELKH Centre for Ecological Research. BG was supported by the GINOP-2.3.2-15-2016-00019 project and by the NKFIH OTKA K-132150 Grant. NS was supported by the co-financing of the European Regional Development Fund through the Interreg Alpine Space programme, project Eco-AlpsWater (Innovative Ecological Assessment and Water Management Strategy for the Protection of Ecosystem Services in Alpine Lakes and Rivers - https://www.alpine-space.eu/projects/eco-alpswater ). AA was supported by the National Research, Development and Innovation Office, Hungary (NKFIH, PD 124681).

Author information

Authors and affiliations.

Department of Tisza Research, Centre for Ecological Research, Danube Research Institute, Bem tér 18/c, 4026, Debrecen, Hungary

Gábor Borics

GINOP Sustainable Ecosystems Group, Centre for Ecological Research, Klebelsberg Kuno u. 3, 8237, Tihany, Hungary

Centre for Ecological Research, Institute of Ecology and Botany, Alkotmány u. 2-4, 2163, Vácrátót, Hungary

András Abonyi

WasserCluster Lunz – Biologische Station GmbH, Dr. Carl Kupelwieser-Promenade 5, 3293, Lunz am See, Austria

András Abonyi & Robert Ptacnik

Research and Innovation Centre, Fondazione Edmund Mach, Via E. Mach 1, 38010, San Michele all’Adige, Italy

Nico Salmaso

You can also search for this author in PubMed   Google Scholar

Contributions

BG wrote ‘Introduction’, ‘Mechanisms affecting diversity’, ‘Diversity measures’, Changes of diversity along environmental scales , ‘Conclusions’ and ‘Outlook’ with substantial contribution from RP. AA, RP wrote ‘The functional diversity–ecosystem functioning relationship in phytoplankton’, NS wrote ‘Phytoplankton diversity using molecular tools’ chapters.

Corresponding author

Correspondence to Gábor Borics .

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Guest editors: Judit Padisák, J. Alex Elliott, Martin T. Dokulil & Luigi Naselli-Flores / New, old and evergreen frontiers in freshwater phytoplankton ecology: the legacy of Colin S. Reynolds

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (DOCX 25 kb)

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Borics, G., Abonyi, A., Salmaso, N. et al. Freshwater phytoplankton diversity: models, drivers and implications for ecosystem properties. Hydrobiologia 848 , 53–75 (2021). https://doi.org/10.1007/s10750-020-04332-9

Download citation

Received : 25 February 2020

Revised : 09 June 2020

Accepted : 13 June 2020

Published : 04 July 2020

Issue Date : January 2021

DOI : https://doi.org/10.1007/s10750-020-04332-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Diversity maintenance
• Ecosystem functioning
• Functional diversity
• Molecular approaches
• Taxonomic diversity
• Find a journal
• Publish with us
• Track your research

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Zooplankton-phytoplankton biomass and diversity relationships in the Great Lakes

Roles Conceptualization, Formal analysis, Resources, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected]

Affiliation Natural Resources Research Institute, University of Minnesota, Duluth, MN, United States of America

Roles Conceptualization, Data curation, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – review & editing

Roles Conceptualization, Data curation, Formal analysis, Methodology, Resources, Validation, Writing – review & editing

Affiliation Department of Natural Resources and Cornell Biological Field Station, Cornell University, Ithaca, NY, United States of America

Roles Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing – review & editing

Roles Conceptualization, Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Writing – review & editing

Roles Conceptualization, Data curation, Investigation, Methodology, Project administration, Resources, Writing – review & editing

Affiliation U.S. EPA Great Lakes National Program Office, Chicago, IL, United States of America

Roles Conceptualization, Methodology, Writing – review & editing

• Katya E. Kovalenko, 
• Euan D. Reavie, 
• Stephanie Figary, 
• Lars G. Rudstam, 
• James M. Watkins, 
• Anne Scofield, 
• Christopher T. Filstrup

• Published: October 26, 2023
• https://doi.org/10.1371/journal.pone.0292988
• Peer Review
• Reader Comments

Quantifying the relationship between phytoplankton and zooplankton may offer insight into zooplankton sensitivity to shifting phytoplankton assemblages and the potential impacts of producer-consumer decoupling on the rest of the food web. We analyzed 18 years (2001–2018) of paired phytoplankton and zooplankton samples collected as part of the United States Environmental Protection Agency (U.S. EPA) Great Lakes Biology Monitoring Program to examine both the long-term and seasonal relationships between zooplankton and phytoplankton across all five Laurentian Great Lakes. We also analyzed effects of phytoplankton diversity on zooplankton biomass, diversity, and predator-prey (zooplanktivore/grazer) ratios. Across the Great Lakes, there was a weak positive correlation between total algal biovolume and zooplankton biomass in both spring and summer. The relationship was weaker and not consistently positive within individual lakes. These trends were consistent over time, providing no evidence of increasing decoupling over the study period. Zooplankton biomass was weakly negatively correlated with algal diversity across lakes, whereas zooplankton diversity was unaffected. These relationships did not change when we considered only the edible phytoplankton fraction, possibly due to the high correlation between total and edible phytoplankton biovolume in most of these lakes. Lack of strong coupling between these producer and consumer assemblages may be related to lagging responses by the consumers, top-down effects from higher-level consumers, or other confounding factors. These results underscore the difficulty in predicting higher trophic level responses, including zooplankton, from changes in phytoplankton assemblages.

Citation: Kovalenko KE, Reavie ED, Figary S, Rudstam LG, Watkins JM, Scofield A, et al. (2023) Zooplankton-phytoplankton biomass and diversity relationships in the Great Lakes. PLoS ONE 18(10): e0292988. https://doi.org/10.1371/journal.pone.0292988

Editor: Hans G. Dam, University of Connecticut, UNITED STATES

Received: February 2, 2023; Accepted: October 3, 2023; Published: October 26, 2023

This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

Data Availability: All relevant data are available within the manuscript and its Supporting Information files.

Funding: These data were collected as part of the U.S. Environmental Protection Agency’s (EPA’s) Great Lakes Biology Monitoring Program. Thus, the study design for sample collection and taxonomic analysis to evaluate phytoplankton and zooplankton communities was determined by the EPA, and followed methods specified by the standard operating procedures associated with this program. The funder did not determine the data analysis method, decision to publish, or assist with preparation of the manuscript beyond the scope of the contributing author affiliated with EPA.

Competing interests: The authors have declared that no competing interests exist.

Introduction

With a few rare exceptions, aquatic ecosystems in the Anthropocene have experienced changes in temperature and nutrient concentrations, which can lead to shifts in phytoplankton assemblages [ 1 – 3 ]. In many cases, these compositional changes can alter the seasonal timing and amplitude of primary productivity [ 4 , 5 ] and functional attributes of phytoplankton [ 6 , 7 ]. Changes in phytoplankton assemblage composition and dynamics can lead to decoupling of primary producers and consumers, which may destabilize planktonic food webs with cascading effects on tertiary consumers [ 8 – 10 ].

Theory predicts and observational studies have shown that greater phytoplankton diversity is linked to increased phytoplankton resource use efficiency (horizontal diversity effects within trophic levels, [ 11 ]) and to increased zooplankton growth rate, diversity, and abundance (vertical diversity effects across trophic levels [ 12 ]). Phytoplankton diversity can also directly influence consumers via biochemical diversity in food resources, which should increase zooplankton diversity [ 13 ], and these diversity effects may produce direct and indirect feedbacks to buffer primary consumer populations and entire food webs from abrupt shifts in their resource base. Because phytoplankton diversity can decrease variability in zooplankton productivity [ 12 ], greater algal diversity may support more zooplankton predators and therefore greater predator-prey ratios within the zooplankton community. However, diversity effects are not consistent across systems [ 14 ] and different measures of phytoplankton diversity can have opposing influences on horizontal and vertical diversity effects [ 15 ]. For example, communities dominated by cyanobacteria may have larger proportions of inedible taxa [ 16 ], which might limit zooplankton biomass [ 17 , 18 ] or have no impact [ 19 ]. Predator-prey biomass ratios can respond to environmental stressors when predators take longer to recover from perturbations, e.g., in isolated environments [ 20 ]; however, other studies show remarkable consistency in predator-prey ratios across a wide range of taxa and systems [ 21 , 22 ].

The structure of large lake food webs is less understood than that of smaller lakes [ 23 , 24 ], and previous vertical diversity studies have largely focused on smaller ecosystems. In the Laurentian Great Lakes, several attributes of phytoplankton assemblages, including total biovolume, cell densities, average cell sizes, and species composition, have fluctuated considerably in the last few decades, with likely causes being changes in nutrient availability, invasive species, and climate change [ 5 , 25 – 27 ]. Decreasing algal cell sizes in particular [ 27 ] could have repercussions for the entire aquatic food web, consistent with a climate change signal linked to decreasing organism sizes at community, species, and population levels across a range of ecosystems [ 28 ]. In the Great Lakes, zooplankton shifted to greater dominance by calanoid copepods, particularly Limnocalanus macrurus [ 29 ], abundances of the predatory invasive cladoceran Bythotrephes increased in some lakes [ 30 ], causing declines in some species [ 31 – 33 ] and changes vertical distribution in others due to migration to greater depths as an anti-predatory response to Bythotrephes [ 31 ]. With a wealth of long-term historical data, there have been multiple detailed analyses of trends in specific assemblages [ 25 , 34 , 35 ] and concurrent trends [ 36 , 37 ]; however, the degree of zooplankton and phytoplankton coupling, vertical diversity effects, and detailed associations between specific groups of taxa are less well understood.

Ideally, investigations of the relationships between primary producers and consumers should use high-resolution productivity data and information on feeding selectivity [ 38 ]. However, long-term high-resolution in situ productivity data are relatively sparse and often limited to smaller geographic areas (e.g., [ 39 ]), and landscape-scale analyses often rely on standing biomass. Controlled studies of feeding selectivity, usually conducted in laboratory settings, are similarly difficult to extrapolate to diverse and dynamic natural settings. We used nearly 20 years of paired zooplankton and phytoplankton data from the U.S. EPA Great Lakes Biology Monitoring Program to examine ecological associations, long-term and seasonal dynamics of zooplankton-phytoplankton coupling, and effects of phytoplankton diversity on zooplankton biomass and diversity. We predicted that there would be a positive correlation between algal biovolume and zooplankton biomass, and that the slope of this relationship would decrease over time because of increasing decoupling of the two trophic levels associated with changes in phytoplankton assemblages. We also tested relationships between algal diversity and total zooplankton biomass, zooplankton diversity, and zooplanktivore-grazer ratios, and explored group-level associations between the major types of zooplankton and algae.

Materials and methods

We used data collected as part of the U.S. Environmental Protection Agency (EPA) Great Lakes Biology and Water Quality Monitoring Programs in the pelagic Laurentian Great Lakes of North America, focusing on years which had matching phytoplankton and zooplankton data (2001–2018). Samples are collected twice per year in the spring (usually April) and summer (usually August) from 72 sites across the five Great Lakes: Lakes Erie, Ontario, Huron, Michigan, and Superior ( S1 Table ). For phytoplankton, equal volumes of water were collected by a rosette sampler from multiple depths (0, 5, 10, 20 m) at each station representing the upper 20 m of the isothermal water column in the spring or the epilimnion in the summer [ 25 ]. Four spring samples from individual depths were composited to form an integrated sample; in summer, a minimum of two and maximum of four depths (typically 0, 5, 10 m, and lower epilimnion, but fewer taken when the mixed layer is shallow) were composited to form a representative sample from the epilimnion [ 40 ]. Samples were preserved with Lugol’s iodine solution and analyzed as described in U.S. EPA Great Lakes National Program Office (GLNPO) standard operating procedure [ 41 ]. Briefly, we used the Utermöhl method [ 42 ] for soft-bodied algal identification. Subsamples were processed for detailed diatom assessment by acid digestion, slide-mounting and high-resolution microscopy. Algal specimens were also measured to allow for biovolume calculations [ 43 ].

Phytoplankton taxa were characterized as edible or inedible based on a combination of entity shape and nutritional quality. Characterization of edibility in freshwater phytoplankton has been considered previously [ 44 ], and we followed similar methods. We assumed that cyanobacteria are less desirable food organisms due to their poor nutritional quality [ 45 ]. Further, we considered a prevailing size and shape of entities (as single cells, filaments, globular colonies) greater than 50 μm to be inedible. Therefore, algae such as filamentous diatoms are considered problematic as food for zooplankton despite their high nutritional value. We acknowledge that previously published assumptions around edibility are overly simplistic, and that edibility of a given phytoplankton taxon is likely grazer-specific. For instance, some larger zooplankton taxa may be equipped to disaggregate large, filamentous diatoms into edible sizes, as noted in a limited set of species-specific studies from marine systems (e.g., [ 46 ]). Such nuances should be considered in the future, but we treat our analyses as a first attempt to evaluate this phenomenon in the Great Lakes. Using these edibility criteria, we filtered out all phytoplankton taxa with low nutritional and low shape edibility ( S2 Table ), and recalculated biovolume of remaining phytoplankton at each site.

Crustacean zooplankton and rotifers were collected by vertical tows taken across the same depth range, at the same time and stations as the phytoplankton data. All samples were collected according to U.S. EPA GLNPO standard operating procedure LG402 [ 47 ] and analyzed following LG403 [ 48 ]. Samples used here were collected using a 63 μm mesh net towed from 20 m or 1 m above the bottom, whichever was shallower, to the surface, at a rate of 0.5 m/s. As with phytoplankton, zooplankton sample collection for this program occurs 24 hours a day, and some stations are sampled during the day and some at night. Zooplankton samples from 20 m were not available for 2007 (both seasons) and for the spring season 2008–2011, and fewer stations had matching data for the two assemblages earlier in the time series. Plankton were narcotized with soda water and preserved with sucrose formalin. Separate counts with different subsampling approaches were done for crustaceans and microzooplankton (rotifers, nauplii) and data combined to densities (numbers/ m 3 ). A minimum of 400 individuals for each of the two counts were identified to the smallest practical taxonomic unit (mostly species) and up to 20 individuals in each taxonomic unit were measured for length in mm using a computerized drawing tablet [ 48 ]. Dreissenid veligers were not included in the total biomass calculations because they have not been measured consistently across the years (sensitivity analysis demonstrates that < 2% of the site-years are affected by this bias). Dry weight individual biomass (μg) was calculated from taxa-specific length-weight regressions available in the standard operating procedures [ 29 , 48 ]. Some rotifer equations used width measurements.

Statistical analyses

We used simple linear models to test for correlations between phytoplankton biovolume and zooplankton biomass, and correlations between phytoplankton and zooplankton diversity (Shannon H). All biovolume analyses were repeated with total and edible phytoplankton biovolume. In addition, we tested the relationships between zooplankton excluding predatory cladoceran ( Bythotrephes , Cercopagis , Leptodora and Polyphemus ) and Limnocalanus biomass, and edible algal biovolume and diversity, although Limnocalanus varies in its degree of zooplanktivory across the Great Lakes [ 49 ]. Data distribution was checked using qqnorm function in R and log 10 -transformation was applied to reduce skewness when warranted (biovolume and biomass data). Additionally, Spearman cross-correlation analyses were used to visualize the relationships between key groups of phytoplankton and zooplankton ( S2 Table ). Zooplankton predator ratios were calculated using the sum of predatory cladocerans ( Bythotrephes , Cercopagis , Leptodora and Polyphemus ) and Limnocalanus biomass relative to other zooplankton. Generalized additive models were fitted to visualize zooplankton predator-prey relationships with algal community metrics; model parameters were set to default as passed on to geom_smooth function in ggplot2. Time of day analyses were used to understand the relative importance of sampling time on zooplankton biomass-edible algal biovolume correlations within lakes. All analyses were done in R [ 50 ].

Across the 20 years of data and all of the lakes, there was a weak positive correlation between total algal biovolume and zooplankton biomass (P < 0.0001, R 2 = 0.19). This Great Lakes-wide correlation was season-dependent, with the overall trend driven primarily by the summer (P < 0.0001, R 2 = 0.15, vs. spring R 2 = 0.06, Fig 1 ). The relationship was also scale-dependent and varied across individual lakes, with a positive relationship in Lake Erie in both seasons and in Lake Huron in the spring, but a lack of significant correlations in lakes Ontario, Superior and Michigan in either season ( Fig 1 ). The slopes of biomass-biovolume relationship (evidence of decoupling) did not change uniformly with time (P > 0.05, Fig 2 ).

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

Data are presented for all lakes and for individual Laurentian Great Lakes by season.

https://doi.org/10.1371/journal.pone.0292988.g001

Fewer stations had matching data for the two assemblages earlier in the time series and spring data was unavailable for zooplankton between 2008–2011 (see S1 Table for complete summary of stations sampled by year).

https://doi.org/10.1371/journal.pone.0292988.g002

Total zooplankton biomass was very weakly negatively correlated with phytoplankton Shannon diversity (P = 0.001, R 2 < 0.01) and this relationship was similarly weak in both spring and summer across all lakes ( Fig 3 ). This weak negative effect was driven largely by Lake Erie, which spanned the longest gradient of both biomass and diversity, and was less pronounced in other lakes. Zooplankton diversity was likewise very weakly correlated with phytoplankton diversity (R 2 < 0.02, S1 Fig ). Most of the biomass of different zooplankton groups was unrelated or weakly negatively related to overall algal richness and diversity, with the exception of a stronger positive relationship for Limnocalanus (R 2 = 0.15, Fig 4 ). The majority of zooplankton groups had closer associations with other zooplankton groups (e.g., predatory cladoceran and rotifers), followed by biovolumes of Cyanophyta, Chlorophyta, and total algal biovolume ( Fig 4 ). Some variation in zooplankton predator-prey ratios was explained by algal diversity (P < 0.0001), whereas algal richness and biovolume did not have a strong linear effect ( S2 Fig ).

https://doi.org/10.1371/journal.pone.0292988.g003

Spearman correlation coefficients color-coded by shade intensity; all biovolume and biomass metrics have been log 10 -transformed. Relationships with visible R have P < 0.0001, whereas relationships with R < 0.10 are displayed as white text on light background.

https://doi.org/10.1371/journal.pone.0292988.g004

Edible algal biovolume was closely correlated with the overall algal biovolume (across lakes R 2 = 0.78, P < 0.0001 in each lake), with the largest discrepancy observed for Lake Erie ( Fig 5 ), where cyanobacteria are abundant in the summer. Our edibility criteria excluded algae with low nutritional value as well as those with difficult to manipulate shapes; we did not consider the two types of edibility filters separately, because even in the extreme scenario, there was a close relationship with total algal biovolume. Because of this relatively high correlation, most of the zooplankton-phytoplankton relationships were not greatly affected when considering only edible phytoplankton biovolume ( S3 and S4 Figs). Results of analyses excluding predatory cladocerans and Limnocalanus detected similarly weak trends to those for total zooplankton biomass ( S5 and S6 Figs). Examining zooplankton-phytoplankton relationship by the time of sampling demonstrated relatively minor effects of time of day on the shape of the biovolume-biomass relationship in individual lakes ( S7 Fig ). The relationship between total and edible biovolume did not exhibit directional changes over time ( S8 Fig ).

White line indicates the 1:1 ratio, the degree of departure from this line illustrates decreasing relative biovolume of edible algal taxa.

https://doi.org/10.1371/journal.pone.0292988.g005

There was a statistically significant but weak correlation between phytoplankton biovolume and zooplankton biomass across this long-term, large-scale dataset; however, it only held across the entire basin, and not individual lakes, and only in the summer. The weak correlation between phytoplankton biovolume and zooplankton biomass on a lake by lake basis could result from a lag in the response of zooplankton consumers to algal changes or variable top-down forcing on zooplankton across the lakes. If a lag in consumer response is present, we would expect the relationship to be stronger in the summer, which was generally the case, even though the correlation was still very poor in terms of predictive power and not statistically significant for most individual lakes. It is not surprising that the large trophic gradient of these lakes, from oligotrophic to meso-eutrophic, was also reflected by the gradient in zooplankton biomass and phytoplankton biovolume across the entire basin. Similarly, in other lakes the coupling between phytoplankton biomass and zooplankton biomass was limited beyond a certain productivity level [ 51 , 52 for Lakes Balaton and Lake Constance].

The slope and strength of the relationship between phytoplankton and zooplankton did not vary significantly with time, despite considerable shifts in algal and zooplankton community composition and productivity [ 5 , 25 , 29 ], providing little additional evidence for a disruption in coupling of producers and consumers. The match/mismatch hypothesis focuses on the consequences of inter-specific differences in response to climate change leading to potentially non-linear responses in the patterns of synchrony [ 9 ]. Such decoupling has already been observed in other systems as a result of a mismatch between trophic levels responding primarily to photoperiod vs. those responding to temperature [ 53 ]. In temperate lakes, the timing of thermal stratification affects the spring diatom blooms which are increasingly mismatched with keystone consumer dynamics [ 54 ]. In the Great Lakes, decreasing diatom cell sizes due to accelerated loss of larger individuals during summer stratification [ 27 ], for example, could make consumers rely on less energetically optimal smaller-sized algae. Longer ice-free periods in Lake Superior have resulted in longer stratification and increased primary production [ 5 ] and could lead to a timing mismatch between the peak of the spring bloom and zooplankton reproduction. The relationships of zooplankton biomass and diversity with edible phytoplankton were similar to those with total phytoplankton biovolume, likely because edible and total phytoplankton biovolume were closely correlated in all lakes with exception of Lake Erie, the most productive lake with a greater incidence of harmful algal blooms. Although other studies have shown that the proportion of inedible phytoplankton, particularly Cyanobacteria, increases in higher productivity lakes [ 16 , 55 , 56 ], cyanobacteria can also be abundant in oligotrophic systems [ 57 ] and can constitute a considerable part of the total biomass across large total phosphorus gradients [ 58 ]. Increasing biomass of less-edible phytoplankton, such as Cyanobacteria, has been observed to limit zooplankton resource use efficiency and the structure of trophic interactions [ 16 ]. However, the relationship between cyanobacterial blooms and zooplankton is variable, and previous studies have observed positive correlations between cyanobacteria concentrations and several groups of zooplankton [ 19 ].

Bottom-up forcing was demonstrated to be important in Lakes Michigan and Huron [ 59 ], where declines in zooplankton biomass and particularly herbivorous cladocerans were associated with simultaneous declines in spring chlorophyll indicating potential grazer limitation [ 36 , 59 , 60 ]. In other cases, changes in zooplankton are better explained by top-down forcing through increased invertebrate or fish predation [ 30 , 33 , 39 ], including changes in vertical distribution [ 61 ]. It is likely that the relative importance of these forces varies across the large spatial and trophic gradient and with season, contributing to the overall uncertainty in the zooplankton-phytoplankton relationship.

Zooplankton biomass was weakly negatively correlated with algal diversity, and it is possible that counteractive effects of algal diversity can be manifested through improved chances of balanced nutrition vs. dilution of the most nutritious taxa [ 13 ]. This effect sign was the opposite of the one we expected based on prior studies [ 12 , 13 ] possibly because pelagic Great Lakes do not include highly eutrophic waters, where extreme cyanobacterial dominance (and therefore decreased overall algal diversity) is more likely to reduce availability and diversity of preferred algal resources to the extent detrimental to consumers. Zooplankton and phytoplankton Shannon diversity were not significantly correlated in our study, providing additional evidence for inconsistent vertical diversity effects across aquatic ecosystems. Positive vertical diversity effects have been observed between bacterial and nanoflaggelate assemblages [ 62 ]; however, zooplankton diversity was not predicted by phytoplankton diversity across a wide range of marine systems [ 63 ], tropical streams [ 64 ], or temperate lakes [ 65 ].

We observed stronger correlations between the different zooplankton groups (with a particularly high correlation between predatory cladocerans and rotifers) than between zooplankton and phytoplankton. This may indirectly suggest a lack of strong feeding selectivity for zooplankton feeding on phytoplankton, at least at the division level, as well as a lack of general avoidance by zooplankton of Cyanobacteria [ 45 ], ability to adapt [ 66 ], or masking of feeding selectivity by other confounding factors. One of those factors could be availability of picoplankton, which could make an important contribution to the diets of smaller zooplankton. The predator-prey ratio of the zooplankton assemblage was weakly positively predicted by algal diversity, providing marginal support for our hypothesis that more diverse algal assemblages may support greater predator densities, which may not be surprising in the light of the overall weak links between zooplankton and phytoplankton in this system.

It is important to note that over these time scales, our dataset has temporal sampling limitations (only 2 sampling events/station/year) and lower number of stations sampled in the earlier years. Integrated samples are collected from the isothermal upper layer of the water column to favor even sampling of the phytoplankton assemblage. Although we did not see time of sampling explaining additional variation, other studies have shown that many zooplankton species have pronounced vertical migration [ 67 – 69 ] which could further contribute to the observed uncertainty about zooplankton-phytoplankton relationships. All of these factors may limit our ability to draw conclusions about the strength of temporal trends across the entire study period.

Understanding the relationships between phytoplankton and zooplankton is important for predicting the effects of climate change and nutrient loading on food web structure and higher trophic level [ 54 , 59 ]. A close correspondence between primary producer and consumer assemblages, indicative of bottom-up regulation, can make consumer populations more vulnerable to changing algal phenology and decreased overall lake productivity. However, we did not observe a close correspondence in the Great Lakes, making it more difficult to predict how the higher trophic levels would be affected by the continued changes in phytoplankton assemblages.

Supporting information

S1 fig. vertical diversity effects, or correlations between phytoplankton and zooplankton diversity..

https://doi.org/10.1371/journal.pone.0292988.s001

S2 Fig. Zooplankton predator-prey (i.e., zooplanktivore-grazer) ratios as a function of attributes of phytoplankton assemblage.

Blue line indicates a Generalized Additive Model (GAM) fit. Algal biovolume is in μm 3 /L, log 10 transformed; other metrics are diversity and richness.

https://doi.org/10.1371/journal.pone.0292988.s002

S3 Fig. Edible phytoplankton biovolume and zooplankton biomass correlations by season.

https://doi.org/10.1371/journal.pone.0292988.s003

S4 Fig. Total zooplankton biomass as a function of only edible phytoplankton diversity.

https://doi.org/10.1371/journal.pone.0292988.s004

S5 Fig. Herbivorous zooplankton biomass as a function of only edible phytoplankton biovolume.

https://doi.org/10.1371/journal.pone.0292988.s005

S6 Fig. Herbivorous zooplankton biomass as a function of only edible phytoplankton Shannon diversity.

https://doi.org/10.1371/journal.pone.0292988.s006

S7 Fig. Effects of sampling time on zooplankton biomass-edible algal biovolume correlations within lakes.

https://doi.org/10.1371/journal.pone.0292988.s007

S8 Fig. Temporal dynamics of the edible algal biovolume as a function of total algal biovolume.

Data are presented across all Great Lakes.

https://doi.org/10.1371/journal.pone.0292988.s008

S1 Table. Total number of stations sampled by year, lake and season.

Lakes: ER–Erie, HU–Huron, MI–Michigan, ON–Ontario, SU–Superior and seasons: Spr–spring, Sum–summer.

https://doi.org/10.1371/journal.pone.0292988.s009

S2 Table. Summary of phytoplankton data with edibility rankings by shape and nutritional content.

SPECCODE–standard species code; maxRelBiov–maximum relative biovolume within a sample (indicator of relative importance combined with frequency), frequency–number of samples in which the taxon was detected; DIV–division; SPECIES–species name, nutrition edibility and shape edibility–categorical rankings.

https://doi.org/10.1371/journal.pone.0292988.s010

https://doi.org/10.1371/journal.pone.0292988.s011

• View Article
• Google Scholar
• PubMed/NCBI
• 38. Wetzel RG (2001) Limnology: Lake and River Ecosystems. 3rd ed. Academic Press, San Diego. 1024 pp.
• 40. U.S. EPA (2022) Standard Operating Procedure. Great Lakes National Program Office, U.S. Environmental Protection Agency, Chicago, IL.
• 41. U.S. EPA (2010) SOP LG401, Standard Operating Procedure for Phytoplankton Analysis. Revision 05, February 2010. Great Lakes National Program Office, U.S. Environmental Protection Agency, Chicago, IL.
• 47. U.S. EPA (2017). SOP LG402. Standard Operating Procedure for Zooplankton Sample Collection and Preservation and Secchi Depth Measurement Field Procedures. Revision 12, February 2017. Great Lakes National Program Office, U.S. Environmental Protection Agency, Chicago, IL.
• 48. U.S. EPA (2017) SOP LG403, Standard Operating Procedure for Zooplankton Analysis. Revision 08, February 2017. Great Lakes National Program Office, U.S. Environmental Protection Agency, Chicago, IL.
• 50. R Core Team (2022) R: A language and environment for statistical computing. R Foundation for statistical Computing, Vienna, Austria. https://www.R-project.org/ .

• Open supplemental data
• Reference Manager
• Simple TEXT file

People also looked at

Original research article, zooplankton dominance shift in response to climate-driven salinity change: a mesocosm study.

• Tvärminne Zoological Station, University of Helsinki, Hanko, Finland

Climate change predictions indicate global changes in salinity with negative implications for plankton food webs; an important baseline for functioning of marine ecosystems. Current understanding of how salinity change will impact plankton communities is mostly limited to the salinization of freshwater environments, with little known about the effects of changing salinity in marine systems. In this study, we investigate the effect of salinity change on zooplankton communities under different salinity change scenarios of the Baltic Sea. Projections for future salinity change derived from regional physical-biogeochemical models were used to set-up an outdoor mesocosm experiment in the coastal area of the Gulf of Finland. Each mesocosm was inoculated with natural plankton using a mixture of both marine and freshwater communities, mimicking the natural influx of freshwater species from rivers into the Baltic Sea. Zooplankton diversity and composition changed possibly due to different salinity tolerances among the species. Among zooplankton, rotifers dominated in low salinities (74%) and cladocerans and copepods (69%) in high salinities. Our results suggest that the zooplankton community will shift to a rotifer dominated community in areas with declining salinity due to the intolerance of other zooplankton groups to freshening.

Introduction

The combined effects of climate change, habitat loss and degradation has put substantial strain on aquatic ecosystems ( Travis, 2003 ). One specific pressure on both marine and freshwater environments globally includes changing salinity, as a result of direct human activity such as ‘salting’ roads ( Schuler et al., 2017 ; Hintz et al., 2021 ) as well as climate change; predicted to affect precipitation and ice melt ( Helm et al., 2010 ; Ross et al., 2021 ). Salinity change can have a profound impact on aquatic organisms, including altering the distribution, phenology, abundance, composition and trophic interactions of plankton ( Intergovernmental Panel Climate Change, 2018 ).

Zooplankton play a key role in the pelagic food web. They link primary producers (i.e. phytoplankton) to higher trophic levels (such as fish), thus their abundance and community structure directly affect the dynamics of fisheries resources ( Sommer et al., 2002 ) as well as the biodiversity of higher trophic levels such as birds and mammals. How zooplankton will be affected by climate warming has been the subject of great debate ( Gyllström et al., 2005 ; Mackas et al., 2007 ; Lewandowska et al., 2014 ; Šorf et al., 2015 ). However, current understanding of how salinity change in aquatic environments will impact zooplankton communities is mostly limited to the salinization of freshwater environments (e.g. Hintz et al., 2017 ; Lin et al., 2017 ; Moffett et al., 2020 ), estuaries (e.g. Gao et al., 2008 ) or single-species responses (e.g. Cervetto et al., 1999 ), while community-level changes and consequences for food web interactions in marine environments remain poorly understood. Studies on the effects of salinization of freshwater environments, have reported a range of responses including, dominance shifts and disruption of trophic interactions ( Lin et al., 2017 ; Gutierrez et al., 2018 ). These results imply that a similar response to salinity change could be expected in marine environments.

Changes in salinity may directly or indirectly influence zooplankton community composition, leading to the local extinction of some species and the appearance of others. Based on existing literature, zooplankton may be directly affected by salinity depending on their physiology and tolerance to salinity change ( Nielsen et al., 2003 ; Schallenberg et al., 2003 ; Hall, 2004 ). Direct responses of zooplankton to stressful salinity conditions include disturbances in reproduction, development and growth ( Hart et al., 2003 ; Santangelo et al., 2014 ; Bashevkin and Pechenik, 2015 ). Indirect impacts through trophic interactions are expected because salinity can influence the composition and physiology of phytoplankton ( Flameling and Kromkamp, 1994 ; Bisson and Kirst, 1995 ; Ayadi, 2004 ; Chakraborty et al., 2011 ), affecting the nutritional food quality of zooplankton (van de de Waal et al., 2010 ) and causing or contributing to food shortages ( Perumal et al., 2009 ). Alterations in zooplankton community composition as a response to salinity change are likely to have negative consequences on food availability and growth of planktivorous fish, such as herring and sprat ( Möllmann et al., 2004 ). This in turn could affect higher trophic levels, such as cod, seabirds and mammals ( Alvarez-Fernandez et al., 2015 ).

Trophic interactions in plankton communities often differ between marine and freshwater environments and thus also their response to environmental perturbations. A recent study by Murphy et al. (2020) showed that warming conditions had no effect on zooplankton density in marine environments, whereas it had a negative effect on zooplankton density and increased grazing by herbaceous zooplankton in freshwaters. Stronger predator-herbivore interactions were also found in freshwater plankton compared to marine plankton communities ( Shurin et al., 2002 ). Sommer and Sommer (2006) argue the difference in the strength of trophic cascades between lentic and marine environments is caused by differences in the zooplankton-phytoplankton link. Cladocerans dominate in freshwaters, whereas copepods dominate in marine ecosystems. Contrary to filter feeders such as cladocerans, copepods can select their food particles based on prey size ( Tiselius and Jonsson, 1990 ), enabling copepod-dominated marine communities to be more resilient to disturbances of trophic cascades due to their selective feeding.

Climate change projections indicate that an increase in precipitation in northern Europe coupled with faster rates of ice-melt will lead to a decline in salinity in northern marine environments ( HELCOM, 2013 ). This includes the Baltic Sea, where salinity is predicted to decrease by 1.5 to 2 psu by the end of the century ( Meier et al., 2006 ). These salinity changes have been shown to be a bottleneck for both marine and freshwater species distribution and diversity, which could have negative impacts on marine ecology, monitoring, modelling and fisheries ( Vuorinen et al., 2015 ). However, these salinity predictions are often uncertain with large variations depending on which general circulation models (GCMs) are used for predictions ( Saraiva et al., 2019 ; Blenckner et al., 2021 ). Differences in observed data can also depend on region and vertical stratification ( Lehmann et al., 2021 ), such as in the Gulf of Finland, where surface salinity increased by 0.5 psu between 1927-2021 ( Merkouriadi and Leppäranta, 2014 ). Other areas of the world, including the Chesapeake Bay and the Baltic Sea, are predicted to show increasing fluctuations in salinity with climate change ( Muhling et al., 2017 ; Blenckner et al., 2021 ).

Following these predictions, we designed a salinity change experiment using floating mesocosm platform deployed in the Baltic Sea to investigate the direct and indirect effects of salinity change on zooplankton communities. The use of mesocosms is a well-established tool to explore plankton community responses to environmental change owing to the advantage of manipulation and replication ( Benton et al., 2007 ; Woodward et al., 2010 ). The brackish environment of the Baltic Sea, where freshwater and marine plankton communities can interact ( Telesh et al., 2008 ), is an ideal study location to investigate the effect of salinity change on natural plankton communities. The general aim of the study was to examine interactions of zooplankton communities under different scenarios of salinity change of the Baltic Sea. We hypothesized that:

(1) Zooplankton abundance will be negatively affected by physiological intolerance to changing salinity,

(2) Salinity change will reduce phytoplankton biomass due to oxidative stress leading to food limitation for zooplankton and altered trophic interactions, and

(3) Salinity change will alter community composition and diversity of zooplankton species depending on their tolerance to salinity; marine species will dominate in higher salinities and brackish and freshwater species in lower ones.

Insights from these findings are important to understand how zooplankton communities may differ under climate driven salinity change and the possible effects it may have on trophic transfer efficiency and overall ecosystem functioning.

Materials and Methods

Experimental design and sampling.

We conducted a salinity change experiment during summer 2019 (July-September) using outdoor mesocosms to investigate the effect of freshening on phytoplankton abundance and zooplankton community composition in pelagic ecosystems. The experiment was set up offshore the Tvärminne Zoological Station, Finland (59° 50’ 40” N, 23° 14’ 57” E). It consisted of 12 transparent, manually mixed plastic (200 µm thick LDPE) enclosures (1600L volume, diameter 0.9 m, depth 2 m, conical bottom) and combined four salinity scenarios [3.5 (control, no salt addition); 5.5; 7.5; 9.5 psu] with three replicates. Salinity levels were based on predicted salinity change scenarios in the Baltic Sea from Meier et al. (2012) and downscaled to local conditions using data from the MONICOAST monitoring buoy deployed offshore the Tvärminne Zoological Station. On 29 July 2019, we partially filled each mesocosm with 800 L of water taken from Gennarbyviken reservoir (59° 55’ 23” N, 23° 12’ 20” E) due to its low salinity levels (0 psu) and presence of freshwater plankton species despite marine origin (the reservoir was enclosed from the sea in 1957). We then added 800 L of water to each mesocosm taken from the bay in front of the Tvärminne Zoological Station due to its marine conditions (5.5 psu) and associated marine plankton species. The water in the mesocosms was assessed using light microscope to make sure the plankton community had survived the water transport and the initial community composition was similar in all experimental units. On 212 and 213 day of the year (DOY) we added sea salt (Aquarium System Instant Ocean) in batches to the mesocosms to make up four different salinity treatments. We added the salt to each mesocosm by extracting 10 L of water and mixing the salt with the water until it was dissolved. We added the salt to each mesocosm in a slow, circular movement using a Secchi disk to ensure that the mixture was evenly dispersed. The control mesocosm was also disturbed using a Secchi disk to make sure all treatments have received the same agitation. The next day, we measured the salinity to ensure we had reached our salinity goals for each mesocosm. To prevent rainfall and allochthonous nutrient inputs (mainly bird droppings), transparent plastic lids were installed above each mesocosm with gaps to allow air circulation in the mesocosm.

Samples for bacterial and phytoplankton as well as nutrient concentrations were taken every two days. In this article, however, we concentrate on the effects of salinity on chlorophyll -a concentration and zooplankton community composition. We measured all abiotic and biotic parameters at 1m depth from the middle of the mesocosm. Temperature (°C), salinity, dissolved oxygen (mg O2/L) were measured every two days using a portable calibrated digital water meter (MU 6100 H, VWR). 4L of water from each mesocosm was taken every 2 days using a water sampler and placed into a 10L plastic container for transportation to the laboratory. The canister was stored in a cold room before gentle mixing was applied and samples for different uses were syphoned off from the container (e.g. chlorophyll- a , dissolved nutrients, bacteria, etc.). For zooplankton, samples were taken every 6 days as to not over-deplete zooplankton numbers compared to natural conditions. A 50μm mesh-size plankton net was dragged from 1m depth from the centre of each mesocosm and then flushed into 250 ml plastic bottles with 30% ethanol for transportation to the laboratory. Additional in situ measurements were taken to be able to compare experimental data to local conditions and assess the effect of enclosure (data not shown). The experiment was terminated after 30 days due to apparent damage to some of the mesocosm bags.

Response Variables

Chlorophyll- a concentration (μgL -1 ) was used as a surrogate of phytoplankton biomass. We vacuum-filtered 100-150 ml of water sample for each mesocosm though glass fibre filters (GF/F, Whatman, Inc., Massachusetts, USA). Filters were stored in the dark in -20°C to prevent chlorophyll breakdown. We later measured chlorophyll- a concentration in each filter using a fluorometer (Varian Inc., Cary Eclipse) after ethanol extraction.

To quantify zooplankton abundance, the sample was settled in 50 ml sedimentation chambers and counted using an Olympus CK30 at x10 and x40 magnification using an inverted microscope technique ( Lund et al., 1958 ). Depending on the density of the sample, a Forsblom plankton splitter was used to split the sample into ¼, ½ or ¾ of the original sample. Identification to genus level (and species where possible) was made using Telesh et al. (2008) . Abundance was measured as individuals per litre (ind.L -1 ). Few adult copepods were recorded, therefore we grouped adults and copepodites together in analyses. Nauplii were analysed separately as copepods exhibit different feeding modes during their life cycle ( Brandl, 2005 ) and therefore may differ in their response to treatment effects. Zooplankton richness was quantified as number of species and zooplankton evenness (as opposite of dominance) was quantified using Pielou’s Index ( Pielou, 1966 ). We measured how the community richness changed immediately after salt addition to see if there was an immediate effect of ‘salt shock’ on zooplankton community structure. No statistically significant change in richness was observed (treatment, F= 0.87, p= 0.4963, time, F= 4.20, p= 0.0650). All subsequent analyses are computed from the first sampling after salt addition (DOY 217).

Statistical Analysis

We conducted mixed effects models to test the effect of salinity change on chlorophyll- a concentration, zooplankton community structure (rotifers, cladocerans and copepods) and diversity indices (richness, Pielou’s evenness). Chlorophyll- a concentration and zooplankton group abundances were log transformed to fulfil the normality and variance homogeneity assumptions. A linear mixed effects model fitted using restricted maximum likelihood estimation (REML) was used to determine effect of salinity change over time on response variables. Model selection was made by using Akaike information criteria (AIC) and autocovariance estimates (ACF) (see Supplementary Material ). The response variable was x and the fixed effects were salinity treatment, time and their interaction. Mesocosm ID was included as a random factor. Significance levels were set at p<0.05. Model residuals were checked for homogeneity of variances and normality. All analyses were performed in R v4.0.3 ( R Core Team, 2020 ) using package nlme ( Pinheiro et al., 2020 ). Graphs were produced using R package ggplot2 ( Wickham, 2016 ) with a Local Regression (loess) smoother set at the default smoothing span of 0.75.

Food Availability for Zooplankton

Salinity was an important factor for determining chlorophyll -a concentration (a proxy for phytoplankton biomass and therefore also zooplankton food availability) during the experiment (P<0.01, Table 1 ). Overall, the lowest and highest salinities (3.5 and 9.5 psu) displayed significantly more chlorophyll- a than the intermediate salinities ( Figure 1 and Table 1 ). Chlorophyll- a concentration also varied over time (P=< 0.01, Table 1 ). In the extreme salinity treatments (3.5 and 9.5 psu) chlorophyll- a concentration peaked in initial stage of the experiment (DOY 213), with a secondary peak during the second half of the experiment (DOY 227 and DOY 229). Chlorophyll then declined for the rest of the experiment (DOY 241). In the 5.5 and 7.5 psu treatments chlorophyll- a was highest in the initial stage of the experiment (DOY 213) before declining for the rest of the experiment (DOY 241). Overall, chlorophyll- a concentration was highest during the first 3 days of the experiment for all treatments.

Table 1 Results of the mixed effect models for response variables with salinity, time, and their interaction.

Figure 1 Chlorophyll a concentration (dashed green line) and abundance of zooplankton groups (rotifer, cladoceran and copepod) over time for each treatment. The dashed grey line at DOY 211 indicates salt addition.

Zooplankton Abundance and Community Composition

Total zooplankton abundance was not affected by salinity (P=0.251, Table 1 ) but varied over time (p<0.001, Table 1 ). Abundance of zooplankton increased from the start of the experiment and peaked between DOY 217 and 223. Abundance then declined in the last two weeks of the experiment.

Rotifers were the most abundant zooplankton group during the whole course of the experiment. Rotifer assemblages were dominated by Keratella quadrata and Keratella cochlearis in all treatments. A significant effect of salinity treatment over time was found for rotifer abundance (P=<0.05, Table 1 ). Rotifers were most abundant in 3.5 psu and least abundant in 9.5 psu salinity treatment. Overall, abundance of rotifers increased during the first two weeks of the experiment before declining for the rest of the experiment in all treatments.

The most abundant cladoceran species was Bosmina longirostris in all treatments. Abundance of cladocerans was not affected by salinity (P=0.4891, Table 1 ) but varied over time (P<0.001, Table 1 ). Abundance was initially low at the start of the experiment before a rapid increase during the second and third weeks of the experiment.

The most dominant copepod species in all samples were Acartia tonsa and Eurytemora affinis . Total copepod abundance (sum of adults and copepodites) was not affected by salinity (P=0.8371, Table 1 ) but varied over time (P=<.001, Table 1 ). Abundance was low at the start of the experiment before exhibiting a rapid increase during the first week of the experiment. The population then declined during the second and third week. Nauplii abundance was not affected by salinity (P=0.8131, Table 1 ) but varied over time (P<0.01, Table 1 ). Initially there was a low abundance of nauplii before increasing during the first week of the experiment. Abundance then declined in the second and third weeks of the experiment before increasing again during the fourth and final week of the experiment. Among the two dominant copepod species, E. affinis abundance was affected by salinity change (P=<0.05, Table 1 ) as well as time (P<0.01, Table 1 ). They were most abundant in 9.5 psu salinity treatment and least abundant in 3.5 psu. Abundance increased over the first four weeks of the experiment, until the final week when abundance rapidly decreased in all treatments. A. tonsa was not affected by changing salinity but its abundance varied over time (P<.001, Table 1 ) with a large spike in abundance during the first week of the experiment and population decline thereafter.

Zooplankton Richness and Evenness

A total of 18 zooplankton taxa (9 rotifer, 4 cladoceran, 3 copepod, 1 barnacle larvae and 1 chironomidea larvae) were recorded in the experiment ( Supplementary Material , Figure S1 ). Zooplankton richness was not affected by salinity treatment and did not vary over time ( Table 1 ).

A significant treatment-by-time interaction was found for zooplankton evenness (P=<0.05, Table 1 and Figure 2 ). In the 3.5 psu treatment, initial low zooplankton evenness was associated with a community dominated by rotifer K. quadrata (72%) and K. cochlearis (18%) ( Figure 3 ). Evenness increased over time when other rotifer species become more abundant and reached the maximum at the end of the experiment ( Figure 2 ) with the greatest contribution of K. cochliaris (48%) followed by Brachionous (18%) and an 8% contribution of K. quadrata .

Figure 2 Zooplankton community evenness for each treatment over time. The dashed blue line at DOY 211 indicates salt addition.

Figure 3 Proportional abundance of taxa on DOY 211, 217, 229 and 241 of the experiment for all salinity treatments. Taxa under 5% abundance are excluded for simplicity. Note that DOY 211 is before salt was added to the treatments and is included to show similarity in community at the start of the experiment.

In the 5.5 psu salinity treatment evenness was initially low and heavily dominated by K. quadrata (74%) and K. cochliaris (19%) ( Figure 3 ). A steady increase in evenness was observed until the end of the experiment with a reduction in the dominance of K. quadrata (10%) and increase in dominance of B. longirostris (36%).

In the 7.5 psu salinity treatment, there was initially low evenness in the community dominated by K. quadrata (73%) and K. cochliaris (21%) ( Figure 3 ). Rapid increase of zooplankton evenness in the third week of the experiment was related to the dominance shift towards B. longirostris (40%) and E. affinis (16%). At the end of the experiment the 7.5 psu treatment was comprised of B. longirostris (31%), K. cochliaris (21%), Euchlanis (18%), and K. quadrata (10%).

In the 9.5 psu salinity treatment, evenness was initially low. The community was dominated by K. quadrata (68%) and K. cochlearis (24%) ( Figure 3 ). Evenness rapidly increased in the first two weeks with a reduction in dominance of K. quadrata (39%) and an increase in B. longirostris (25%), E. affinis (14%), Keratella testudo (10%) and A. tonsa (9%). At the end of the experiment, evenness decreased again after dominance shift to B. longirostris (56%), followed by K. cochliaris (24%) and E. affinis (13%).

Our results imply that changing salinity will lead to a re-organisation of the zooplankton community structure (evenness and community composition). Salinity change had a direct impact on zooplankton evenness and caused a dominance shift from rotifers to cladocerans and later to copepods in high salinity treatments ( Figure 3 ). No such shift was observed in low salinity mesocosms, suggesting that rotifers will dominate zooplankton communities in low salinity environments in the future, such as with declining salinity of the Baltic Sea. Rotifers make up a significant proportion of the mesozooplankton community ( Telesh and Heerkloss, 2002 ), are mostly found in freshwater environments and are generally intolerant to salinity increase ( Sarma et al., 2006 ; Medeiros et al., 2010 ). Jansson et al. (2020) reported that rotifer abundance has been increasing in the Baltic Sea since the 1960s, but this increase was linked to warming and eutrophication and not declining salinity. This may have been due to the relatively narrow salinity range (5.2-6.4 psu) where their study was conducted (Gulf of Riga). In contrast, our experiment, which simulated future salinity change scenarios ( Meier et al., 2012 ), confirms that rotifers are directly affected by salinity change ( Table 1 ) and dominate in low salinities, accounting for 82% of the community composition ( Figure 3 ).

In contrast to the lowest salinity treatment (3.5 psu), where different rotifer species dominated zooplankton community during the whole experimental period, in the higher salinity treatments (7.5 and 9.5 psu) rotifers were outcompeted by cladocerans and copepods ( B. longirostris and E. affinis , Figure 3 ). Hence, we see a reduction in zooplankton evenness at the end of the experiment in 9.5 psu salinity treatment when some zooplankton groups are reduced or lost (e.g. K. quadrata , K. testudo and Brachionus) the shift is complete and the majority of the community is dominated by B. longirostris ( Figure 2 ). In the 7.5 psu salinity treatment the same shift in dominance was found as in 9.5 psu, but it took longer to reach a cladoceran-dominated community ( Figure 1 ). B. longirostris has been found to adapt to a wide range of salinities ( Jeppesen et al., 1994 ; Deasley et al., 2012 ), thus it was able to outcompete rotifers in 7.5 and 9.5 psu treatments. However, as filter feeders, cladocerans can be negatively affected by toxic cyanobacteria blooms expected to proliferate with projected freshening and warming of the Baltic Sea ( Meier et al., 2011 ) promoting rotifer and copepod abundance. Some rotifers were able to persist in high salinities however, namely K. cochlearis. This could be due to K. cochlaris being able to tolerate fluctuations in salinity ( Paturej and Gutkowska, 2015 ).

Zooplankton richness increased directly after salt addition in all experimental units, with dominant rotifers suffering and copepods benefiting from the disturbance ( Figure 3 ). Copepods dominated by A. tonsa and E. affinis were most abundant in the 9.5 psu salinity treatment. Over the course of the experiment population of A. tonsa declined and was replaced by E. affinis ( Figure S1 , supplementary online material ) , indicating competitive exclusion, as both species have similar feeding strategies and prey size preferences ( Richman et al., 1977 ; Engstrom, 2000 ). A. tonsa exhibit a broad tolerance to salinity, being able to survive salinities as low as 0.5 psu, but with an optimum salinity of between 10 and 20 psu ( Cervetto et al., 1999 ; Calliari et al., 2006 ). E. affinis , which is generally regarded as a brackish species ( Lee and Petersen, 2003 ), has salinity optimum of 10 psu ( Karlsson et al., 2018 ), which is close to our highest salinity treatment and therefore may be better adapted to this salinity range than A. tonsa.

An alternative reason for the increased abundance of E. affinis found in high salinities could be a potential hormetic response to salt addition whereby, in rapidly producing more nauplii, the population of E. affinis is able to adapt to stressful conditions by guaranteeing the survival of at least a few individuals. Further evidence of this is given by an increase in juvenile and adult copepods found in the second week of the experiment, suggesting the nauplii produced in the first week were able to survive the high salinity environment. A hormetic response is when an environmental stressor (e.g. toxins, herbicides) stimulates an adaptive response that increases the resistance of the organism to a level of stress ( Calabrese et al., 2007 ). This phenomenon has been studied elsewhere in nature ( Calabrese and Baldwin, 1998 ; Beckers et al., 2009 ; Alyokhin et al., 2013 ; Vargas-Hernandez et al., 2017 ).

In intermediate salinities (5.5 and 7.5 psu) copepods were outcompeted by cladocerans, which reflects their different salinity optima, copepods preferring marine and cladocerans freshwater conditions ( Sommer and Sommer, 2006 ). In the lowest salinity treatment (3.5 psu), rotifer abundance did not affect the population dynamics of cladocerans or copepods ( Figure 1 ), with groups tracking the pattern of food availability without apparent competition. This result has also been found elsewhere: a study by MacIsaac and Gilbert (1989) showed that a high abundance of rotifers can be maintained in the presence of small cladocerans such as B. longirostris . We found that in the 3.5 and 9.5 psu salinity treatments an increase in community evenness over the first two weeks of the experiment ( Figure 2 ) corresponded to a reduction in chlorophyll- a concentration. Conversely, when evenness declined in those treatments in the third and fourth week, we saw a recovery in chlorophyll- a concentration. We suggest this could be due to a more diverse zooplankton community using their resources more efficiently and therefore being able to reduce producer biomass (phytoplankton) more greatly ( Duffy, 2002 ). In contrast, constant grazing from multiple zooplankton groups keeps phytoplankton concentrations low in the second half of the experiment in the intermediate salinity treatments ( Figure 1 ). Rotifer dominated communities have also been found to be less efficient in controlling phytoplankton biomass than other zooplankton groups ( Jakobsen et al., 2003 ), which might be a reason why we observed recovery of phytoplankton biomass (measured as chlorophyll a concentration) in low salinity treatments by the end of the experiment ( Figure 1 ). This suggests with declining salinity of the Baltic Sea a shift to a rotifer dominated community may lead to less efficient grazing of phytoplankton and the increased risk of intense algal blooms in summer months.

In conclusion, we show that salinity change results in the restructuring of the zooplankton community. In low salinities, the dominance of rotifers throughout the experiment suggests they are most well adapted to freshwater environments, though some brackish species that are able to tolerate fluctuations in salinity (e.g. A. tonsa copepods and B. longirostris cladocerans) can persist in low abundance. With environments predicted to decline in salinity, such as the Baltic Sea, shifts from marine to brackish conditions will result in reduced food availability and increased resource competition between copepods and cladocerans, benefitting the latter due to their better adaptation to brackish conditions. This will therefore lead to a shift from a community comprised of complex organisms (copepods) to simpler ones (cladocerans and rotifers), potentially strengthening the negative effects of warming and eutrophication on zooplankton ( Daufresne et al., 2009 ; Jansson et al., 2020 ), and consequently altering the functioning of pelagic food web. Although this study focusses on salinity change, in nature environmental variables are rarely acting in isolation to one another. Other stressors such as warming can affect the metabolic activity of zooplankton ( Mayor et al., 2015 ) and lead to multistressor effects ( Souissi et al., 2021 ). Trophic interactions may also be affected by salinity change and effects on other trophic levels, especially with regards to toxic cyanobacteria and phytoplankton should be considered in future studies ( Viitasalo and Bonsdorff, 2022 ).

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author Contributions

Study conception and design: CH and AL. Data collection: CH. Data analysis and interpretation of results: CH and AL. Draft manuscript preparation: CH and AL. All authors have reviewed and approved the final version of this manuscript.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

This research was funded by the Onni Talas Foundation Finland and the University of Helsinki start-up grant of AL. The experimental mesocosm platform is a part of the AQUACOSM (project No 731065) funded by the European Commission EU H2020-INFRAIA.

Acknowledgments

We are grateful for the help of students and technicians at Tvärminne Zoological Station that participated in field and laboratory work supporting this publication. Local firefighters are acknowledged for the water transport.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmars.2022.861297/full#supplementary-material

Supplementary Figure 1 | Zooplankton abundance with salinity treatment over time for all zooplankton groups. Rare species excluded from analysis are not shown. Statistical tests were performed as described in the main text, methods section.

Alvarez-Fernandez S., Licandro P., van Damme C. J. G., Hufnagl M. (2015). Effect of Zooplankton on Fish Larval Abundance and Distribution: A Long-Term Study on North Sea Herring (Clupea Harengus). ICES J. Mar. Sci. 72 (9), 2569–2577. doi: 10.1093/icesjms/fsv140

CrossRef Full Text | Google Scholar

Alyokhin A., Vincent C., Giordanengo P. (Eds.) (2013). “Insect Pests of Potato,” in Global Perspectives on Biology and Management (Amsterdam; Boston: Academic Press, Elsevier).

Google Scholar

Ayadi H. (2004). “Structure of the Phytoplankton Communities in Two Lagoons of Different Salinity in the Sfax Saltern (Tunisia)”. J. Plankton Res. 26 (6), 669–679. doi: 10.1093/plankt/fbh047

Bashevkin S. M., Pechenik J. A. (2015). The Interactive Influence Ssof Temperature and Salinity on Larval and Juvenile Growth in the Gastropod Crepidula Fornicata (L.). J. Exp. Mar. Biol. Ecol. 470, 78–91. doi: 10.1016/j.jembe.2015.05.004

Beckers G. J. M., Jaskiewicz M., Liu Y., Underwood W. R., He S. Y., Zhang S., et al. (2009). Mitogen-Activated Protein Kinases 3 and 6 Are Required for Full Priming of Stress Responses in Arabidopsis Thaliana. Plant Cell 21 (3), 944–953. doi: 10.1105/tpc.108.062158

PubMed Abstract | CrossRef Full Text | Google Scholar

Benton T. G., Solan M., Travis J. M. J., Sait S. M. (2007). Microcosm Experiments can Inform Global Ecological Problems. Trends Ecol. Evol. 22 (10), 516–521. doi: 10.1016/j.tree.2007.08.003

Bisson M., Kirst G. (1995). Osmotic Acclimation and Turgor Pressure Regulation in Algae. Natur Wissenschaften Aufsätze 82, 461–471. doi: 10.1007/BF01131597

Blenckner T., Ammar Y., Müller-Karulis B., Niiranen S., Arneborg L., Li Q. (2021). The Risk for Novel and Disappearing Environmental Conditions in the Baltic Sea. Front. Mar. Sci. 8, 745722. doi: 10.3389/fmars.2021.745722

Brandl Z. (2005). Freshwater Copepods and Rotifers: Predators and Their Prey. Hydrobiologia 546 (1), 475–489. doi: 10.1007/s10750-005-4290-3

Calabrese E. J., Bachmann K. A., Bailer A. J., Bolger P. M., Borak J., Cai L., et al. (2007). Biological Stress Response Terminology: Integrating the Concepts of Adaptive Response and Preconditioning Stress Within a Hormetic Dose–Response Framework. Toxicol. Appl. Pharmacol. 222 (1), 122–128. doi: 10.1016/j.taap.2007.02.015

Calabrese E. J., Baldwin L. A. (1998). Hormesis as a Biological Hypothesis. Environ. Health Perspect. 106, 6. doi: 10.1289/ehp.98106s1357

Calliari D., Andersen CM, Thor P., Gorokhova E., Tiselius P. (2006). Salinity Modulates the Energy Balance and Reproductive Success of Co-Occurring Copepods Acartia Tonsa and A. Clausi in Different Ways. Mar. Ecol. Prog. Ser. 312, 177–188. doi: 10.3354/meps312177

Cervetto G., Gaudy R., Pagano M. (1999). Influence of Salinity on the Distribution of Acartia Tonsa (Copepoda, Calanoida). J. Of Exp. Mar. Biol. Ecol. 239, 33–45. doi: 10.1016/S0022-0981(99)00023-4

Chakraborty P., Acharyya T., Raghunadh Babu P. V., Bandyopadhyay D. (2011). Impact of Salinity and PH on Phytoplankton Communities in a Tropical Freshwater System: An Investigation With Pigment Analysis by HPLC. J. Environ. Monitoring 13 (3), 614. doi: 10.1039/c0em00333f

Daufresne M., Lengfellner K., Sommer U. (2009). Global Warming Benefits the Small in Aquatic Ecosystems. Proc. Natl. Acad. Sci. 106 (31), 12788–12793. doi: 10.1073/pnas.0902080106

Deasley K., Korosi J. B., Thienpont J. R., Kokelj S. V., Pisaric M. F. J., Smol J. P. (2012). Investigating the Response of Cladocera to a Major Saltwater Intrusion Event in an Arctic Lake From the Outer Mackenzie Delta (Nt, Canada). J. Paleolimnol. 48 (2), 287–296. doi: 10.1007/s10933-012-9577-6

Duffy J. E. (2002). Biodiversity and Ecosystem Function: The Consumer Connection. Oikos 99 (2), 201–219. doi: 10.1034/j.1600-0706.2002.990201.x

Engstrom J. (2000). Feeding Interactions of the Copepods Eurytemora Affinis and Acartia Bifilosa With the Cyanobacteria Nodularia Sp. J. Plankton Res. 22 (7), 1403–1409. doi: 10.1093/plankt/22.7.1403

Flameling I. A., Kromkamp J. (1994). Responses of Respiration and Photosynthesis of Scenedesmus Protuberans Fritsch to Gradual and Steep Salinity Increases. J. Plankton Res. 16 (12), 1781–1791. doi: 10.1093/plankt/16.12.1781

Gao Q., Xu Z., Zhuang P. (2008). The Relation Between Distribution of Zooplankton and Salinity in the Changjiang Estuary. Chin. J. Oceanol. Limnol. 26 (2), 178–185. doi: 10.1007/s00343-008-0178-1

Gutierrez M. F., Tavşanoğlu ÜN., Vidal N., Yu J., Mello F.T.-d., Çakiroglu A. I., et al. (2018). Salinity Shapes Zooplankton Communities and Functional Diversity and Has Complex Effects on Size Structure in Lakes. Hydrobiologia 813 (1), 237–255. doi: 10.1007/s10750-018-3529-8

Gyllström M., Hansson L. A., Jeppesen E., García Criado F., Gross E., Irvine K., et al. (2005). The Role of Climate in Shaping Zooplankton Communities of Shallow Lakes. Limnol. Oceanogr. 50 (6), 2008–2021. doi: 10.4319/lo.2005.50.6.2008

Hall C. J. (2004). Effects of Salinity and Temperature on Survival and Reproduction of Boeckella Hamata (Copepoda: Calanoida) from a Periodically Brackish Lake. J. Plankton Res. 23 (1), 97–104. doi: 10.1093/plankt/23.1.97

Hart B. T., Lake P. S., Angus Webb J., Grace M. R. (2003). Ecological Risk to Aquatic Systems From Salinity Increases. Aust. J. Bot. 51 (6), 689. doi: 10.1071/BT02111

HELCOM (2013). “Climate Change in the Baltic Sea Area: HELCOM Thematic Assessment in 2013,” in Balt. Sea Environ. Proc. No. 137 . (Katajanokanlaituri 6 B FI-00160 Helsinki: Helsinki Commission). Available at: http://www.helcom.fi .

Helm K. P., Bindoff N. L., Church J. A. (2010). Changes in the Global Hydrological-Cycle Inferred From Ocean Salinity: HYDROLOGICAL Cycle AND Ocean Salinity. Geophysical Res. Lett. 37 (18), 0094–8276. doi: 10.1029/2010GL044222

Hintz W. D., Fay L., Relyea R. A. (2021). Road Salts, Human Safety, and the Rising Salinity of Our Fresh Waters Frontiers in Ecology and the Environment. Front. Ecol. Environ. 20 (1), 22–30. doi: 10.1002/fee.2433

Hintz W. D., Mattes B. M., Schuler M. S., Jones D. K., Stoler A. B., Lind L., et al. (2017). Salinization Triggers a Trophic Cascade in Experimental Freshwater Communities With Varying Food-Chain Length. Ecol. Appl. 27 (3), 833–844. doi: 10.1002/eap.1487

Intergovernmental Panel Climate Change. (2018). “Global Warming of 1.5 °C,” in An IPCC Special Report on the Impacts of Global Warming of 1.5 °C Above Pre-Industrial Levels and Related Global Greenhouse Gas Emission Pathways, in the Context of Strengthening the Global Response to the Threat of Climate Change, Sustainable Development, and Efforts to Eradicate Poverty (Geneva: World Meteorological Organization).

Jakobsen T. S., Hansen P. B., Jeppesen E., Grønkjær P., Søndergaard M. (2003). Impact of Three-Spined Stickleback Gasterosteus Aculeatus on Zooplankton and Chl a in Shallow, Eutrophic, Brackish Lakes. Mar. Ecol. Prog. Ser. 262, 277–284. doi: 10.3354/meps262277

Jansson A., Klais-Peets R., Grinienė E., Rubene G., Semenova A., Lewandowska A., et al. (2020). Functional Shifts in Estuarine Zooplankton in Response to Climate Variability. Ecol. Evol. 10 (20), 11591–11606. doi: 10.1002/ece3.6793

Jeppesen E., Søndergaard M., Kanstrup E., Petersen B., Eriksen R. B., Hammershøj M., et al. (1994). Does the Impact of Nutrients on the Biological Structure and Function of Brackish and Freshwater Lakes Differ? Hydrobiologia 275 (1), 15–30. doi: 10.1007/BF00026696

Karlsson K., Puiac S., Winder M. (2018). Life-History Responses to Changing Temperature and Salinity of the Baltic Sea Copepod Eurytemora Affinis. Mar. Biol. 165 (2), 30. doi: 10.1007/s00227-017-3279-6

Lee C. E., Petersen C. H. (2003). Effects of Developmental Acclimation on Adult Salinity Tolerance in the Freshwater-Invading Copepod Eurytemora Affinis. Physiol. Biochem. Zool. 76 (3), 296–301. doi: 10.1086/375433

Lehmann A., Myrberg K., Post P., Chubarenko I., Dailidiene I., Hinrichsen H.-H., et al. (2021). Salinity Dynamics of the Baltic Sea. Preprint. Dynamics of the Earth System: Interactions. Earth. Syst. Dyn. 13 (1), 373–392. doi: 10.5194/esd-2021-15

Lewandowska A. M., Boyce D. G., Hofmann M., Matthiessen B., Sommer U., Worm B. (2014). Effects of Sea Surface Warming on Marine Plankton. Ecol. Lett. 17 (5), 614–623. doi: 10.1111/ele.12265

Lin Q., Xu L., Hou J., Liu Z., Jeppesen E., Han B.-P. (2017). Responses of Trophic Structure and Zooplankton Community to Salinity and Temperature in Tibetan Lakes: Implication for the Effect of Climate Warming”. Water Res. 124, 618–629. doi: 10.1016/j.watres.2017.07.078

Lund J. W. G., Kipling C., Le Cren E. D. (1958). “The Inverted Microscope Method of Estimating Algal Numbers and the Statistical Basis of Estimations by Counting”. Hydrobiologia 11 (2), 143–170. doi: 10.1007/BF00007865

MacIsaac H. J., Gilbert J. J. (1989). Competition Between Rotifers and Cladocerans of Different Body Sizes. Oecologia 81 (3), 295–301. doi: 10.1007/BF00377074

Mackas D. L., Batten S., Trudel M. (2007). Effects on Zooplankton of a Warmer Ocean: Recent Evidence From the Northeast Pacific. Prog. Oceanogr. 75 (2), 223–252. doi: 10.1016/j.pocean.2007.08.010

Mayor D. J., Sommer U., Cook K. B., Viant M. R. (2015). The Metabolic Response of Marine Copepods to Environmental Warming and Ocean Acidification in the Absence of Food. Sci. Rep. 5 (1), 13690. doi: 10.1038/srep13690

Medeiros A. M. A., Barbosa J. E. L., Medeiros P. R., Rocha R. M., Silva L. F. (2010). Salinity and Freshwater Discharge Determine Rotifer Distribution at the Mossoró River Estuary (Semiarid Region of Brazil). Braz. J. Biol. 70 (3), 551–557. doi: 10.1590/S1519-69842010000300011

Meier H., Eilola K., Almroth E. (2011). Climate-Related Changes in Marine Ecosystems Simulated With a 3-Dimensional Coupled Physical-Biogeochemical Model of the Baltic Sea. Climate Res. 48 (1), 31–55. doi: 10.3354/cr00968

Meier H. E. M., Hordoir R., Andersson H. C., Dieterich C., Eilola K., Gustafsson B. G., et al. (2012). Modeling the Combined Impact of Changing Climate and Changing Nutrient Loads on the Baltic Sea Environment in an Ensemble of Transient Simulations for 1961–2099. Clim. Dyn. 39 (9), 2421–2441. doi: 10.1007/s00382-012-1339-7

Meier H. E. M., Kjellström E., Graham L. P. (2006). Estimating Uncertainties of Projected Baltic Sea Salinity in the Late 21st Century. Geophysical Res. Lett. 33 (15), 0094–8276. doi: 10.1029/2006GL026488

Merkouriadi I., Leppäranta M. (2014). Long-Term Analysis of Hydrography and Sea-Ice Data in Tvärminne, Gulf of Finland, Baltic Sea. Clim. Change 124 (4), 849–859. doi: 10.1007/s10584-014-1130-3

Moffett E. R., Baker H. K., Bonadonna C. C., Shurin J. B., Symons C. C. (2020). Cascading Effects of Freshwater Salinization on Plankton Communities in the Sierra Nevada. Limnol. Oceanogr. Lett . doi: 10.1002/lol2.10177

Möllmann C., Kornilovs G., Fetter M., Koster F. W. (2004). Feeding Ecology of Central Baltic Sea Herring and Sprat. J. Fish Biol. 65, 1563–1581. doi: 10.1111/j.0022-1112.2004.00566.x

Muhling B. A., Jacobs J., Stock C. A., Gaitan G. F., Saba V. S. (2017). Projections of the Future Occurrence, Distribution, and Seasonality of Three Vibrio Species in the Chesapeake Bay under a High‐emission Climate Change Scenario. GeoHealth 1 (7), 278–296. doi: 10.1002/2017GH000089

Murphy G. E. P., Romanuk T. N., Worm B. (2020). Cascading Effects of Climate Change on Plankton Community Structure. Ecol. Evol. 10 (4), 2170–2181. doi: 10.1002/ece3.6055

Nielsen D. L., Brock M. A., Crossle K., Harris K., Healey M., Jarosinski I. (2003). The Effects of Salinity on Aquatic Plant Germination and Zooplankton Hatching From Two Wetland Sediments. Freshwater Biol. 48 (12), 2214–2223. doi: 10.1046/j.1365-2427.2003.01146.x

Paturej E., Gutkowska A. (2015). The Effect of Salinity Levels on the Structure of Zooplankton Communities. Arch. Biol. Sci. 67 (2), 483–492. doi: 10.2298/ABS140910012P

Perumal N. V., Rajkumar M., Perumal P., Rajasekar K. T. (2009). Seasonal Variations of Plankton Diversity in the Kaduviyar Estuary, Nagapattinam, Southeast Coast of India. J. Environ. Biol. 30 (6), 1035–1046.

PubMed Abstract | Google Scholar

Pielou E. C. (1966). “The Measurement of Diversity in Different Types of Biological Collections”. J. Theor. Biol. 13, 131–144. doi: 10.1016/0022-5193(66)90013-0

Pinheiro J., Bates D., DebRoy S., Sarkar D., R Core Team (2020) R Package Version 3.1-150 . Available at: https://CRAN.R-project.org/package=nlme .

R Core Team (2020). R: A Language and Environment for Statistical Computing (Vienna, Austria. URL https://www.R-project.org/: R Foundation for Statistical Computing).

Richman S., Heinle D. R., Huff R. (1977). “Grazing by Adult Estuarine Calanoid Copepods of the Chesapeake Bay”. Mar. Biol. 42 (1), 69–84. doi: 10.1007/BF00392015

Roddie B. D., Leakey R. J. G., Berry A. J. (1984). “Salinity-Temperature Tolerance and Osmoregulation in Eurytemora Affinis (Poppe) (Copepoda : Calanoida) in Relation to Its Distribution in the Zooplankton of the Upper Reaches of the Forth Estuary”. J. Exp. Mar. Biol. Ecol. 79 (2), 191–211. doi: 10.1016/0022-0981(84)90219-3

Ross A. C., Najjar R. G., Li M. (2021). A Metamodel-Based Analysis of the Sensitivity and Uncertainty of the Response of Chesapeake Bay Salinity and Circulation to Projected Climate Change. Estuaries Coasts 44 (1), 70–87. doi: 10.1007/s12237-020-00761-w

Santangelo J. M., Francisco de A E., Manca M., Bozelli R. L. (2014). Disturbances Due to Increased Salinity and the Resilience of Zooplankton Communities: The Potential Role of the Resting Egg Bank. Hydrobiologia 722 (1), 103–113. doi: 10.1007/s10750-013-1683-6

Saraiva S., Markus Meier H. E., Andersson H., Höglund A., Dieterich C., Gröger M., et al. (2019). Uncertainties in Projections of the Baltic Sea Ecosystem Driven by an Ensemble of Global Climate Models. Front. Earth Sci. 0. doi: 10.3389/feart.2018.00244

Sarma S. S. S., Nandini S., Morales-Ventura J., Delgado-Martínez I., González-Valverde L. (2006). Effects of NaCl Salinity on the Population Dynamics of Freshwater Zooplankton (Rotifers and Cladocerans). Aquat. Ecol. 40 (3), 349. doi: 10.1007/s10452-006-9039-1

Schallenberg M., Hall C. J., Burns C. W. (2003). Consequences of Climate-Induced Salinity Increases on Zooplankton Abundance and Diversity in Coastal Lakes. Mar. Ecol. Prog. Ser. 251, 181–189. doi: 10.3354/meps251181

Schuler M. S., Hintz W. D., Jones D. K., Lind L. A., Mattes B. M., Stoler A. B., et al. (2017). How Common Road Salts and Organic Additives Alter Freshwater Food Webs: In Search of Safer Alternatives. J. Appl. Ecol. 54 (5), 1353–1361. doi: 10.1111/1365-2664.12877

Shurin J. B., Borer E. T., Seabloom E. W., Anderson K., Blanchette C. A., Broitman B., et al. (2002). A Cross-Ecosystem Comparison of the Strength of Trophic Cascades: Strength of Cascades. Ecol. Lett. 5 (6), 785–791. doi: 10.1046/j.1461-0248.2002.00381.x

Sommer U., Sommer F. (2006). Cladocerans Versus Copepods: The Cause of Contrasting Top–Down Controls on Freshwater and Marine Phytoplankton. Oecologia 147 (2), 183–194. doi: 10.1007/s00442-005-0320-0

Sommer U., Stibor H., Katechakis A., Sommer F., Hansen T. (2002). Pelagic Food Web Configurations at Different Levels of Nutrient Richness and Their Implications for the Ratio Fish Production. Hydrobiologia 484, 11–20. doi: 10.1023/A:1021340601986

Souissi A., Hwang J-S., Souissi S. (2021). “Reproductive Trade-Offs of the Estuarine Copepod Eurytemora Affinis under Different Thermal and Haline Regimes.” Sci. Rep. 11 (1), 20139. doi: 10.1038/s41598-021-99703-0.

Šorf M., Davidson T. A., Brucet S., Menezes R. F., Søndergaard M., Lauridsen T. L., et al. (2015). Zooplankton Response to Climate Warming: A Mesocosm Experiment at Contrasting Temperatures and Nutrient Levels. Hydrobiologia 742 (1), 185–203. doi: 10.1007/s10750-014-1985-3

Telesh I. V., Heerkloss R. (2002). Atlas of Estuarine Zooplankton of the Southern and Eastern Baltic Sea. Part I: Rotifera (Hamburg: Verlag Dr. Kovač), 89 pp.

Telesh I., Postel L., Heerkloss R., Mironova K., Skarlato S. (2008). Zooplankton of the Open Baltic Sea: Atlas. Meereswissenschaftliche Berichte No 73 2008 - Marine Science Reports No 73 2008 The Baltic Marine Biologists Publication No.21(76), 1–290. doi: 10.12754/MSR-2008-0073

Tiselius P., Jonsson P. R. (1990). Foraging Behaviour of Six Calanoid Copepods: Observations and Hydrodynamic Analysis”. Mar. Ecol. Prog. Ser. 66, 23–33. doi: 10.3354/meps066023

Travis J. M. J. (2003). “Climate Change and Habitat Destruction: A Deadly Anthropogenic Cocktail. Proc. R. Soc. London Ser. B: Biol. Sci. 270 (1514), 467–473. doi: 10.1098/rspb.2002.2246

Vargas-Hernandez M., Macias-Bobadilla I., Guevara-Gonzalez R. G., de Romero-Gomez S.J., Rico-Garcia E., Ocampo-Velazquez R. V., et al. (2017). Plant Hormesis Management With Biostimulants of Biotic Origin in Agriculture. Front. Plant Sci. 8, 1762. doi: 10.3389/fpls.2017.01762

Viitasalo M., Bonsdorff E. (2022). Global Climate Change and the Baltic Sea Ecosystem: Direct and Indirect Effects on Species, Communities and Ecosystem Functioning. Earth System Dynamics 13 (2), 711–747. doi: 10.5194/esd-13-711-2022

de Waal D.B.v., Verschoor A. M., Verspagen J. M. H., van Donk E., Huisman J. (2010). Climate-Driven Changes in the Ecological Stoichiometry of Aquatic Ecosystems. Front. Ecol. Environ. 8 (3), 145–152. doi: 10.1890/080178

Vuorinen I., Hänninen J., Rajasilta M., Laine P., Eklund J., Montesino-Pouzols F., et al. (2015). Scenario Simulations of Future Salinity and Ecological Consequences in the Baltic Sea and Adjacent North Sea Areas–Implications for Environmental Monitoring. Ecol. Indic. 50, 196–205. doi: 10.1016/j.ecolind.2014.10.019.

Wickham H. (2016). Ggplot2: Elegant Graphics for Data Analysis (New York: Springer-Verlag). Available at: https://ggplot2.tidyverse.org .

Woodward G., Perkins D. M., Brown L. E. (2010). climate Change and Freshwater Ecosystems: Impacts Across Multiple Levels of Organization. Philos. Trans. R. Soc. B: Biol. Sci. 365 (1549), 2093–2106. doi: 10.1098/rstb.2010.0055

Keywords: salinity, plankton, diversity, Baltic Sea, marine, brackish, habitat loss, climate change

Citation: Hall CAM and Lewandowska AM (2022) Zooplankton Dominance Shift in Response to Climate-Driven Salinity Change: A Mesocosm Study. Front. Mar. Sci. 9:861297. doi: 10.3389/fmars.2022.861297

Received: 24 January 2022; Accepted: 25 April 2022; Published: 03 June 2022.

Reviewed by:

Copyright © 2022 Hall and Lewandowska. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Clio A. M. Hall, [email protected]

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• PMC10602272

Zooplankton-phytoplankton biomass and diversity relationships in the Great Lakes

Katya E. Kovalenko

1 Natural Resources Research Institute, University of Minnesota, Duluth, MN, United States of America

Euan D. Reavie

Stephanie figary.

2 Department of Natural Resources and Cornell Biological Field Station, Cornell University, Ithaca, NY, United States of America

Lars G. Rudstam

James m. watkins, anne scofield.

3 U.S. EPA Great Lakes National Program Office, Chicago, IL, United States of America

Christopher T. Filstrup

Associated data.

All relevant data are available within the manuscript and its Supporting Information files.

Quantifying the relationship between phytoplankton and zooplankton may offer insight into zooplankton sensitivity to shifting phytoplankton assemblages and the potential impacts of producer-consumer decoupling on the rest of the food web. We analyzed 18 years (2001–2018) of paired phytoplankton and zooplankton samples collected as part of the United States Environmental Protection Agency (U.S. EPA) Great Lakes Biology Monitoring Program to examine both the long-term and seasonal relationships between zooplankton and phytoplankton across all five Laurentian Great Lakes. We also analyzed effects of phytoplankton diversity on zooplankton biomass, diversity, and predator-prey (zooplanktivore/grazer) ratios. Across the Great Lakes, there was a weak positive correlation between total algal biovolume and zooplankton biomass in both spring and summer. The relationship was weaker and not consistently positive within individual lakes. These trends were consistent over time, providing no evidence of increasing decoupling over the study period. Zooplankton biomass was weakly negatively correlated with algal diversity across lakes, whereas zooplankton diversity was unaffected. These relationships did not change when we considered only the edible phytoplankton fraction, possibly due to the high correlation between total and edible phytoplankton biovolume in most of these lakes. Lack of strong coupling between these producer and consumer assemblages may be related to lagging responses by the consumers, top-down effects from higher-level consumers, or other confounding factors. These results underscore the difficulty in predicting higher trophic level responses, including zooplankton, from changes in phytoplankton assemblages.

Introduction

With a few rare exceptions, aquatic ecosystems in the Anthropocene have experienced changes in temperature and nutrient concentrations, which can lead to shifts in phytoplankton assemblages [ 1 – 3 ]. In many cases, these compositional changes can alter the seasonal timing and amplitude of primary productivity [ 4 , 5 ] and functional attributes of phytoplankton [ 6 , 7 ]. Changes in phytoplankton assemblage composition and dynamics can lead to decoupling of primary producers and consumers, which may destabilize planktonic food webs with cascading effects on tertiary consumers [ 8 – 10 ].

Theory predicts and observational studies have shown that greater phytoplankton diversity is linked to increased phytoplankton resource use efficiency (horizontal diversity effects within trophic levels, [ 11 ]) and to increased zooplankton growth rate, diversity, and abundance (vertical diversity effects across trophic levels [ 12 ]). Phytoplankton diversity can also directly influence consumers via biochemical diversity in food resources, which should increase zooplankton diversity [ 13 ], and these diversity effects may produce direct and indirect feedbacks to buffer primary consumer populations and entire food webs from abrupt shifts in their resource base. Because phytoplankton diversity can decrease variability in zooplankton productivity [ 12 ], greater algal diversity may support more zooplankton predators and therefore greater predator-prey ratios within the zooplankton community. However, diversity effects are not consistent across systems [ 14 ] and different measures of phytoplankton diversity can have opposing influences on horizontal and vertical diversity effects [ 15 ]. For example, communities dominated by cyanobacteria may have larger proportions of inedible taxa [ 16 ], which might limit zooplankton biomass [ 17 , 18 ] or have no impact [ 19 ]. Predator-prey biomass ratios can respond to environmental stressors when predators take longer to recover from perturbations, e.g., in isolated environments [ 20 ]; however, other studies show remarkable consistency in predator-prey ratios across a wide range of taxa and systems [ 21 , 22 ].

The structure of large lake food webs is less understood than that of smaller lakes [ 23 , 24 ], and previous vertical diversity studies have largely focused on smaller ecosystems. In the Laurentian Great Lakes, several attributes of phytoplankton assemblages, including total biovolume, cell densities, average cell sizes, and species composition, have fluctuated considerably in the last few decades, with likely causes being changes in nutrient availability, invasive species, and climate change [ 5 , 25 – 27 ]. Decreasing algal cell sizes in particular [ 27 ] could have repercussions for the entire aquatic food web, consistent with a climate change signal linked to decreasing organism sizes at community, species, and population levels across a range of ecosystems [ 28 ]. In the Great Lakes, zooplankton shifted to greater dominance by calanoid copepods, particularly Limnocalanus macrurus [ 29 ], abundances of the predatory invasive cladoceran Bythotrephes increased in some lakes [ 30 ], causing declines in some species [ 31 – 33 ] and changes vertical distribution in others due to migration to greater depths as an anti-predatory response to Bythotrephes [ 31 ]. With a wealth of long-term historical data, there have been multiple detailed analyses of trends in specific assemblages [ 25 , 34 , 35 ] and concurrent trends [ 36 , 37 ]; however, the degree of zooplankton and phytoplankton coupling, vertical diversity effects, and detailed associations between specific groups of taxa are less well understood.

Ideally, investigations of the relationships between primary producers and consumers should use high-resolution productivity data and information on feeding selectivity [ 38 ]. However, long-term high-resolution in situ productivity data are relatively sparse and often limited to smaller geographic areas (e.g., [ 39 ]), and landscape-scale analyses often rely on standing biomass. Controlled studies of feeding selectivity, usually conducted in laboratory settings, are similarly difficult to extrapolate to diverse and dynamic natural settings. We used nearly 20 years of paired zooplankton and phytoplankton data from the U.S. EPA Great Lakes Biology Monitoring Program to examine ecological associations, long-term and seasonal dynamics of zooplankton-phytoplankton coupling, and effects of phytoplankton diversity on zooplankton biomass and diversity. We predicted that there would be a positive correlation between algal biovolume and zooplankton biomass, and that the slope of this relationship would decrease over time because of increasing decoupling of the two trophic levels associated with changes in phytoplankton assemblages. We also tested relationships between algal diversity and total zooplankton biomass, zooplankton diversity, and zooplanktivore-grazer ratios, and explored group-level associations between the major types of zooplankton and algae.

Materials and methods

We used data collected as part of the U.S. Environmental Protection Agency (EPA) Great Lakes Biology and Water Quality Monitoring Programs in the pelagic Laurentian Great Lakes of North America, focusing on years which had matching phytoplankton and zooplankton data (2001–2018). Samples are collected twice per year in the spring (usually April) and summer (usually August) from 72 sites across the five Great Lakes: Lakes Erie, Ontario, Huron, Michigan, and Superior ( S1 Table ). For phytoplankton, equal volumes of water were collected by a rosette sampler from multiple depths (0, 5, 10, 20 m) at each station representing the upper 20 m of the isothermal water column in the spring or the epilimnion in the summer [ 25 ]. Four spring samples from individual depths were composited to form an integrated sample; in summer, a minimum of two and maximum of four depths (typically 0, 5, 10 m, and lower epilimnion, but fewer taken when the mixed layer is shallow) were composited to form a representative sample from the epilimnion [ 40 ]. Samples were preserved with Lugol’s iodine solution and analyzed as described in U.S. EPA Great Lakes National Program Office (GLNPO) standard operating procedure [ 41 ]. Briefly, we used the Utermöhl method [ 42 ] for soft-bodied algal identification. Subsamples were processed for detailed diatom assessment by acid digestion, slide-mounting and high-resolution microscopy. Algal specimens were also measured to allow for biovolume calculations [ 43 ].

Phytoplankton taxa were characterized as edible or inedible based on a combination of entity shape and nutritional quality. Characterization of edibility in freshwater phytoplankton has been considered previously [ 44 ], and we followed similar methods. We assumed that cyanobacteria are less desirable food organisms due to their poor nutritional quality [ 45 ]. Further, we considered a prevailing size and shape of entities (as single cells, filaments, globular colonies) greater than 50 μm to be inedible. Therefore, algae such as filamentous diatoms are considered problematic as food for zooplankton despite their high nutritional value. We acknowledge that previously published assumptions around edibility are overly simplistic, and that edibility of a given phytoplankton taxon is likely grazer-specific. For instance, some larger zooplankton taxa may be equipped to disaggregate large, filamentous diatoms into edible sizes, as noted in a limited set of species-specific studies from marine systems (e.g., [ 46 ]). Such nuances should be considered in the future, but we treat our analyses as a first attempt to evaluate this phenomenon in the Great Lakes. Using these edibility criteria, we filtered out all phytoplankton taxa with low nutritional and low shape edibility ( S2 Table ), and recalculated biovolume of remaining phytoplankton at each site.

Crustacean zooplankton and rotifers were collected by vertical tows taken across the same depth range, at the same time and stations as the phytoplankton data. All samples were collected according to U.S. EPA GLNPO standard operating procedure LG402 [ 47 ] and analyzed following LG403 [ 48 ]. Samples used here were collected using a 63 μm mesh net towed from 20 m or 1 m above the bottom, whichever was shallower, to the surface, at a rate of 0.5 m/s. As with phytoplankton, zooplankton sample collection for this program occurs 24 hours a day, and some stations are sampled during the day and some at night. Zooplankton samples from 20 m were not available for 2007 (both seasons) and for the spring season 2008–2011, and fewer stations had matching data for the two assemblages earlier in the time series. Plankton were narcotized with soda water and preserved with sucrose formalin. Separate counts with different subsampling approaches were done for crustaceans and microzooplankton (rotifers, nauplii) and data combined to densities (numbers/ m 3 ). A minimum of 400 individuals for each of the two counts were identified to the smallest practical taxonomic unit (mostly species) and up to 20 individuals in each taxonomic unit were measured for length in mm using a computerized drawing tablet [ 48 ]. Dreissenid veligers were not included in the total biomass calculations because they have not been measured consistently across the years (sensitivity analysis demonstrates that < 2% of the site-years are affected by this bias). Dry weight individual biomass (μg) was calculated from taxa-specific length-weight regressions available in the standard operating procedures [ 29 , 48 ]. Some rotifer equations used width measurements.

Statistical analyses

We used simple linear models to test for correlations between phytoplankton biovolume and zooplankton biomass, and correlations between phytoplankton and zooplankton diversity (Shannon H). All biovolume analyses were repeated with total and edible phytoplankton biovolume. In addition, we tested the relationships between zooplankton excluding predatory cladoceran ( Bythotrephes , Cercopagis , Leptodora and Polyphemus ) and Limnocalanus biomass, and edible algal biovolume and diversity, although Limnocalanus varies in its degree of zooplanktivory across the Great Lakes [ 49 ]. Data distribution was checked using qqnorm function in R and log 10 -transformation was applied to reduce skewness when warranted (biovolume and biomass data). Additionally, Spearman cross-correlation analyses were used to visualize the relationships between key groups of phytoplankton and zooplankton ( S2 Table ). Zooplankton predator ratios were calculated using the sum of predatory cladocerans ( Bythotrephes , Cercopagis , Leptodora and Polyphemus ) and Limnocalanus biomass relative to other zooplankton. Generalized additive models were fitted to visualize zooplankton predator-prey relationships with algal community metrics; model parameters were set to default as passed on to geom_smooth function in ggplot2. Time of day analyses were used to understand the relative importance of sampling time on zooplankton biomass-edible algal biovolume correlations within lakes. All analyses were done in R [ 50 ].

Across the 20 years of data and all of the lakes, there was a weak positive correlation between total algal biovolume and zooplankton biomass (P < 0.0001, R 2 = 0.19). This Great Lakes-wide correlation was season-dependent, with the overall trend driven primarily by the summer (P < 0.0001, R 2 = 0.15, vs. spring R 2 = 0.06, Fig 1 ). The relationship was also scale-dependent and varied across individual lakes, with a positive relationship in Lake Erie in both seasons and in Lake Huron in the spring, but a lack of significant correlations in lakes Ontario, Superior and Michigan in either season ( Fig 1 ). The slopes of biomass-biovolume relationship (evidence of decoupling) did not change uniformly with time (P > 0.05, Fig 2 ).

Data are presented for all lakes and for individual Laurentian Great Lakes by season.

Fewer stations had matching data for the two assemblages earlier in the time series and spring data was unavailable for zooplankton between 2008–2011 (see S1 Table for complete summary of stations sampled by year).

Total zooplankton biomass was very weakly negatively correlated with phytoplankton Shannon diversity (P = 0.001, R 2 < 0.01) and this relationship was similarly weak in both spring and summer across all lakes ( Fig 3 ). This weak negative effect was driven largely by Lake Erie, which spanned the longest gradient of both biomass and diversity, and was less pronounced in other lakes. Zooplankton diversity was likewise very weakly correlated with phytoplankton diversity (R 2 < 0.02, S1 Fig ). Most of the biomass of different zooplankton groups was unrelated or weakly negatively related to overall algal richness and diversity, with the exception of a stronger positive relationship for Limnocalanus (R 2 = 0.15, Fig 4 ). The majority of zooplankton groups had closer associations with other zooplankton groups (e.g., predatory cladoceran and rotifers), followed by biovolumes of Cyanophyta, Chlorophyta, and total algal biovolume ( Fig 4 ). Some variation in zooplankton predator-prey ratios was explained by algal diversity (P < 0.0001), whereas algal richness and biovolume did not have a strong linear effect ( S2 Fig ).

Spearman correlation coefficients color-coded by shade intensity; all biovolume and biomass metrics have been log 10 -transformed. Relationships with visible R have P < 0.0001, whereas relationships with R < 0.10 are displayed as white text on light background.

Edible algal biovolume was closely correlated with the overall algal biovolume (across lakes R 2 = 0.78, P < 0.0001 in each lake), with the largest discrepancy observed for Lake Erie ( Fig 5 ), where cyanobacteria are abundant in the summer. Our edibility criteria excluded algae with low nutritional value as well as those with difficult to manipulate shapes; we did not consider the two types of edibility filters separately, because even in the extreme scenario, there was a close relationship with total algal biovolume. Because of this relatively high correlation, most of the zooplankton-phytoplankton relationships were not greatly affected when considering only edible phytoplankton biovolume ( S3 and S4 Figs). Results of analyses excluding predatory cladocerans and Limnocalanus detected similarly weak trends to those for total zooplankton biomass ( S5 and S6 Figs). Examining zooplankton-phytoplankton relationship by the time of sampling demonstrated relatively minor effects of time of day on the shape of the biovolume-biomass relationship in individual lakes ( S7 Fig ). The relationship between total and edible biovolume did not exhibit directional changes over time ( S8 Fig ).

White line indicates the 1:1 ratio, the degree of departure from this line illustrates decreasing relative biovolume of edible algal taxa.

There was a statistically significant but weak correlation between phytoplankton biovolume and zooplankton biomass across this long-term, large-scale dataset; however, it only held across the entire basin, and not individual lakes, and only in the summer. The weak correlation between phytoplankton biovolume and zooplankton biomass on a lake by lake basis could result from a lag in the response of zooplankton consumers to algal changes or variable top-down forcing on zooplankton across the lakes. If a lag in consumer response is present, we would expect the relationship to be stronger in the summer, which was generally the case, even though the correlation was still very poor in terms of predictive power and not statistically significant for most individual lakes. It is not surprising that the large trophic gradient of these lakes, from oligotrophic to meso-eutrophic, was also reflected by the gradient in zooplankton biomass and phytoplankton biovolume across the entire basin. Similarly, in other lakes the coupling between phytoplankton biomass and zooplankton biomass was limited beyond a certain productivity level [ 51 , 52 for Lakes Balaton and Lake Constance].

The slope and strength of the relationship between phytoplankton and zooplankton did not vary significantly with time, despite considerable shifts in algal and zooplankton community composition and productivity [ 5 , 25 , 29 ], providing little additional evidence for a disruption in coupling of producers and consumers. The match/mismatch hypothesis focuses on the consequences of inter-specific differences in response to climate change leading to potentially non-linear responses in the patterns of synchrony [ 9 ]. Such decoupling has already been observed in other systems as a result of a mismatch between trophic levels responding primarily to photoperiod vs. those responding to temperature [ 53 ]. In temperate lakes, the timing of thermal stratification affects the spring diatom blooms which are increasingly mismatched with keystone consumer dynamics [ 54 ]. In the Great Lakes, decreasing diatom cell sizes due to accelerated loss of larger individuals during summer stratification [ 27 ], for example, could make consumers rely on less energetically optimal smaller-sized algae. Longer ice-free periods in Lake Superior have resulted in longer stratification and increased primary production [ 5 ] and could lead to a timing mismatch between the peak of the spring bloom and zooplankton reproduction. The relationships of zooplankton biomass and diversity with edible phytoplankton were similar to those with total phytoplankton biovolume, likely because edible and total phytoplankton biovolume were closely correlated in all lakes with exception of Lake Erie, the most productive lake with a greater incidence of harmful algal blooms. Although other studies have shown that the proportion of inedible phytoplankton, particularly Cyanobacteria, increases in higher productivity lakes [ 16 , 55 , 56 ], cyanobacteria can also be abundant in oligotrophic systems [ 57 ] and can constitute a considerable part of the total biomass across large total phosphorus gradients [ 58 ]. Increasing biomass of less-edible phytoplankton, such as Cyanobacteria, has been observed to limit zooplankton resource use efficiency and the structure of trophic interactions [ 16 ]. However, the relationship between cyanobacterial blooms and zooplankton is variable, and previous studies have observed positive correlations between cyanobacteria concentrations and several groups of zooplankton [ 19 ].

Bottom-up forcing was demonstrated to be important in Lakes Michigan and Huron [ 59 ], where declines in zooplankton biomass and particularly herbivorous cladocerans were associated with simultaneous declines in spring chlorophyll indicating potential grazer limitation [ 36 , 59 , 60 ]. In other cases, changes in zooplankton are better explained by top-down forcing through increased invertebrate or fish predation [ 30 , 33 , 39 ], including changes in vertical distribution [ 61 ]. It is likely that the relative importance of these forces varies across the large spatial and trophic gradient and with season, contributing to the overall uncertainty in the zooplankton-phytoplankton relationship.

Zooplankton biomass was weakly negatively correlated with algal diversity, and it is possible that counteractive effects of algal diversity can be manifested through improved chances of balanced nutrition vs. dilution of the most nutritious taxa [ 13 ]. This effect sign was the opposite of the one we expected based on prior studies [ 12 , 13 ] possibly because pelagic Great Lakes do not include highly eutrophic waters, where extreme cyanobacterial dominance (and therefore decreased overall algal diversity) is more likely to reduce availability and diversity of preferred algal resources to the extent detrimental to consumers. Zooplankton and phytoplankton Shannon diversity were not significantly correlated in our study, providing additional evidence for inconsistent vertical diversity effects across aquatic ecosystems. Positive vertical diversity effects have been observed between bacterial and nanoflaggelate assemblages [ 62 ]; however, zooplankton diversity was not predicted by phytoplankton diversity across a wide range of marine systems [ 63 ], tropical streams [ 64 ], or temperate lakes [ 65 ].

We observed stronger correlations between the different zooplankton groups (with a particularly high correlation between predatory cladocerans and rotifers) than between zooplankton and phytoplankton. This may indirectly suggest a lack of strong feeding selectivity for zooplankton feeding on phytoplankton, at least at the division level, as well as a lack of general avoidance by zooplankton of Cyanobacteria [ 45 ], ability to adapt [ 66 ], or masking of feeding selectivity by other confounding factors. One of those factors could be availability of picoplankton, which could make an important contribution to the diets of smaller zooplankton. The predator-prey ratio of the zooplankton assemblage was weakly positively predicted by algal diversity, providing marginal support for our hypothesis that more diverse algal assemblages may support greater predator densities, which may not be surprising in the light of the overall weak links between zooplankton and phytoplankton in this system.

It is important to note that over these time scales, our dataset has temporal sampling limitations (only 2 sampling events/station/year) and lower number of stations sampled in the earlier years. Integrated samples are collected from the isothermal upper layer of the water column to favor even sampling of the phytoplankton assemblage. Although we did not see time of sampling explaining additional variation, other studies have shown that many zooplankton species have pronounced vertical migration [ 67 – 69 ] which could further contribute to the observed uncertainty about zooplankton-phytoplankton relationships. All of these factors may limit our ability to draw conclusions about the strength of temporal trends across the entire study period.

Understanding the relationships between phytoplankton and zooplankton is important for predicting the effects of climate change and nutrient loading on food web structure and higher trophic level [ 54 , 59 ]. A close correspondence between primary producer and consumer assemblages, indicative of bottom-up regulation, can make consumer populations more vulnerable to changing algal phenology and decreased overall lake productivity. However, we did not observe a close correspondence in the Great Lakes, making it more difficult to predict how the higher trophic levels would be affected by the continued changes in phytoplankton assemblages.

Supporting information

Blue line indicates a Generalized Additive Model (GAM) fit. Algal biovolume is in μm 3 /L, log 10 transformed; other metrics are diversity and richness.

Data are presented across all Great Lakes.

Lakes: ER–Erie, HU–Huron, MI–Michigan, ON–Ontario, SU–Superior and seasons: Spr–spring, Sum–summer.

SPECCODE–standard species code; maxRelBiov–maximum relative biovolume within a sample (indicator of relative importance combined with frequency), frequency–number of samples in which the taxon was detected; DIV–division; SPECIES–species name, nutrition edibility and shape edibility–categorical rankings.

Funding Statement

These data were collected as part of the U.S. Environmental Protection Agency’s (EPA’s) Great Lakes Biology Monitoring Program. Thus, the study design for sample collection and taxonomic analysis to evaluate phytoplankton and zooplankton communities was determined by the EPA, and followed methods specified by the standard operating procedures associated with this program. The funder did not determine the data analysis method, decision to publish, or assist with preparation of the manuscript beyond the scope of the contributing author affiliated with EPA.

Data Availability

• PLoS One. 2023; 18(10): e0292988.

Decision Letter 0

11 Apr 2023

PONE-D-23-03077Zooplankton-phytoplankton biomass and diversity relationships in the Great LakesPLOS ONE

Dear Dr. Kovalenko,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Two reviewers see merit in the manuscript, but both suggest major revision. Reviewer 1 emphasizes a lack of detail in the methods, and the need to temper some statements in the light of tenuous correlations.  This reviewer provided a marked up copy of the manuscript with comments and edits. I urge you to look carefully at that document. The second reviewer is concerned that you are missing recent relevant literature and that your discussion misses some mechanistic explanations.  Please, carefully consider both reviewers' comments and suggestions if you decide to submit a revised version of the manuscript. 

Please submit your revised manuscript by May 26 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at  gro.solp@enosolp . When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Hans G. Dam, Ph. D.

Academic Editor

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information .

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

2. Has the statistical analysis been performed appropriately and rigorously?

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: • The weak correlations presented here should be treated as such and conclusions based on these relationships should be considered with caution to not over emphasize their significance.

• The paper is well written, but limited in the details included in the methods section. Additional details should be included, particularly in the statistical analysis section as how the analysis was accomplished (Time to process, diversity index, spearman correlation, etc)

• Some limitations of the data were brought up (edibility), however all limitations (integrated samples, net speed and size, day/night, only up to 2 sample points per station per year, minimum of 2 stations per lake) need to be discussed and the implications considered.

• Several figure and table captions contained methods or analysis results not discussed in the body of the paper. The authors should review all captions to ensure that all appropriate content is included in the body of the text.

• Additional comments and edit suggestions made directly in the attached PDF.

Reviewer #2: Dear Author,

your study based on the ratio of predator-to-prey biomass as a key element of trophic structure of lake ecosystems. It is typically investigated from a food chain perspective, ignoring channels of energy transfer that may govern community structure where you study try to approach those processes across space and time.

Line 42-43: I would suggest to add some important recent studies on the predator-prey biomass ratio that have shown that can be stable. In general I would suggest to ass more literature on paper that have been testing this ration and explain what they have been observing.

literature example: Perkins et al. 2022 Consistent predator-prey biomass scaling in

complex food webs

Line 50-53: I think there are studies that show changes of size spectra, as well as biomass across time and space. I would suggest to include those studies because there are closely related with what you are testing in this study.

Line 195-196: How do you explain this? This a pattern that have been observed or not by other studies you should include them and try to explain why you studies confirm on not what other paper have found. You describe the patterns but you haven't tried to explain them.

Line 198-199: I don't think that your have tested the size spectra, maybe you mean something else. Please explain.

Line 208-210:Are all the zooplankton species included in the study herbivores. If not then any you have to separate them cause the carnivore zooplankton is most probably not going to respond on the phytoplankton biovolume changes. Also the presence of carnivore zooplankton may affect the size of the herbivore zooplankton that may in turn be affected by the phytoplankton volume. so it may be a combined affect of those two. This is maybe something to include in the discussion session.

I think in general the discussion session is missing important literature about their field as well as some possible mechanistic explanation about the patterns you are observing. But the pattern you observe are interesting!

6. PLOS authors have the option to publish the peer review history of their article ( what does this mean? ). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy .

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool,  https://pacev2.apexcovantage.com/ . PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at  gro.solp@serugif . Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-23-03077_reviewer_comments.pdf

Author response to Decision Letter 0

16 Jun 2023

please see the letter of responses (attached) for detailed responses to reviewer comments (it is easier to read in that format)

Submitted filename: responsesReviewCommentsKovalenko2.docx

Decision Letter 1

15 Aug 2023

PONE-D-23-03077R1Zooplankton-phytoplankton biomass and diversity relationships in the Great LakesPLOS ONE

I have now received  feedback from the two original reviewers of your manuscript on the revised version. One reviewer stated directly to me: " The revised manuscript addresses all the comments/suggestions made by the reviewers and I'm very pleased that you have extended your analysis. The present revised manuscript makes an excellent an timing contribution to PLOS ONE." The other review is copied below, and you can see that it contains stylistic suggestions and comments to improve the final version of the paper. Once you make those, i can move to accept the paper. Like the reviewer, I was unable to open the attached figures in the manuscript. Please ensure they are in a format that the journal accepts.

Please submit your revised manuscript by Sep 29 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at  gro.solp@enosolp . When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

2. Is the manuscript technically sound, and do the data support the conclusions?

3. Has the statistical analysis been performed appropriately and rigorously?

4. Have the authors made all data underlying the findings in their manuscript fully available?

5. Is the manuscript presented in an intelligible fashion and written in standard English?

6. Review Comments to the Author

Reviewer #1: Thank you to the authors for their revisions.

Please see my comments on the following minor concerns:

- Line 12 (and lines 73 and 83), EPA should defined at first use. If not in the abstract then in the introduction.

- Lines 48-49, this sentence is awkward. the authors should revise the sentence to avoid the double use of "can" and clarify this statement.

- Lines 200-202 (Figure Caption for Figure 5), the authors should include full lake names in the figures for consistency with the other figures included in the text, (not supplemental), or define the acronyms in the caption.

- Line 267, what do the authors mean by "dilution of this pattern". Perhaps this can be reworded for clarity.

- Figure 1, the units should be properly formatted as superscript in the figure axes.

- Figure 2, the units should be properly formatted as superscript in the figure axes.

- Figure 3, the units should be properly formatted as superscript in the figure axis.

- Figure 4, The colors for this figure, particularly for Limnocalanus, and algal diversity are difficult to read on a computer screen and near impossible to read when printed. I strongly recommend the authors choose a different text color as the white text on white correlation is bad.

- Figure 5, the units should be properly formatted as superscript in the figure axis. Additionally, the acronyms as described above.

- FigS2, I can't open this file. Is this file corrupt?

- FigS3, the units should be properly formatted as superscript in the figure axes.

- FigS4, the units should be properly formatted as superscript in the figure axis.

- FigS5, the units should be properly formatted as superscript in the figure axes.

- FigS6, the units should be properly formatted as superscript in the figure axis.

- FigS7, the units should be properly formatted as superscript in the figure axes. Additionally, 'Spr' and 'Sum' do not need to be shortened here, they could easily be written out without impacting the figure.

7. PLOS authors have the option to publish the peer review history of their article ( what does this mean? ). If published, this will include your full peer review and any attached files.

Author response to Decision Letter 1

29 Sep 2023

Reviewer #1: Thank you to the authors for their revisions.

______ Defined as suggested.

______ Thank you for pointing this out, rephrased as “Predator-prey biomass ratios can respond to environmental stressors when predators take longer to recover from perturbations, e.g., in isolated environments [20]; however, other studies show remarkable consistency in predator-prey ratios across a wide range of taxa and systems [21,22].”

_______ Done, full lake names included.

______ Clarified as “masking of feeding selectivity by other confounding factors”, meaning that we cannot exclude presence of feeding selectivity however it may be masked by other, stronger drivers of phytoplankton abundance.

_______ Done.

______In this case, we adjusted the color of the text and background specifically so that non-significant correlations (with very low R values) do not show up. Because this was intentional, we have not changed the figure but clarified it better in the legend.>>>

_______ All supplemental figures have been redone to address reviewer comments. Fig S2 was redone to ensure it is not corrupt. The current version opens fine, but all .eps figures (as required by the journal) need a postscript reader or need to be converted to a .pdf.

Submitted filename: responsesReviewCommentsKovalenkoRound2.docx

Decision Letter 2

PONE-D-23-03077R2

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/ , click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at gro.solp@gnillibrohtua .

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact gro.solp@sserpeno .

Additional Editor Comments (optional):

Acceptance letter

19 Oct 2023

Dear Dr. Kovalenko:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact gro.solp@sserpeno .

If we can help with anything else, please email us at gro.solp@enosolp .

Thank you for submitting your work to PLOS ONE and supporting open access.

PLOS ONE Editorial Office Staff

on behalf of

Dr. Hans G. Dam

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 16 August 2021

Decline in plankton diversity and carbon flux with reduced sea ice extent along the Western Antarctic Peninsula

• Yajuan Lin   ORCID: orcid.org/0000-0002-9057-9321 1 , 2 , 3 ,
• Carly Moreno   ORCID: orcid.org/0000-0002-3046-1014 4 ,
• Adrian Marchetti   ORCID: orcid.org/0000-0003-4608-4775 4 ,
• Hugh Ducklow   ORCID: orcid.org/0000-0001-9480-2183 5 ,
• Oscar Schofield 6 ,
• Erwan Delage 7 ,
• Michael Meredith 8 ,
• Zuchuan Li 1 , 9 ,
• Damien Eveillard   ORCID: orcid.org/0000-0002-8162-7360 7 , 10 ,
• Samuel Chaffron   ORCID: orcid.org/0000-0001-5903-617X 7 , 10 &
• Nicolas Cassar   ORCID: orcid.org/0000-0003-0100-3783 1 , 2  

Nature Communications volume  12 , Article number:  4948 ( 2021 ) Cite this article

7641 Accesses

24 Citations

66 Altmetric

Metrics details

• Carbon cycle
• Marine biology
• Water microbiology

Since the middle of the past century, the Western Antarctic Peninsula has warmed rapidly with a significant loss of sea ice but the impacts on plankton biodiversity and carbon cycling remain an open question. Here, using a 5-year dataset of eukaryotic plankton DNA metabarcoding, we assess changes in biodiversity and net community production in this region. Our results show that sea-ice extent is a dominant factor influencing eukaryotic plankton community composition, biodiversity, and net community production. Species richness and evenness decline with an increase in sea surface temperature (SST). In regions with low SST and shallow mixed layers, the community was dominated by a diverse assemblage of diatoms and dinoflagellates. Conversely, less diverse plankton assemblages were observed in waters with higher SST and/or deep mixed layers when sea ice extent was lower. A genetic programming machine-learning model explained up to 80% of the net community production variability at the Western Antarctic Peninsula. Among the biological explanatory variables, the sea-ice environment associated plankton assemblage is the best predictor of net community production. We conclude that eukaryotic plankton diversity and carbon cycling at the Western Antarctic Peninsula are strongly linked to sea-ice conditions.

Similar content being viewed by others

Genome-scale community modelling reveals conserved metabolic cross-feedings in epipelagic bacterioplankton communities

Nils Giordano, Marinna Gaudin, … Samuel Chaffron

Hidden diversity and potential ecological function of phosphorus acquisition genes in widespread terrestrial bacteriophages

Jie-Liang Liang, Shi-wei Feng, … Jin-tian Li

The microbial carbon pump and climate change

Nianzhi Jiao, Tingwei Luo, … Carol Robinson

Introduction

The Southern Ocean disproportionally contributes to the global climate system, accounting for almost half of the anthropogenic CO 2 and 75% of the heat uptake by the oceans 1 , 2 . The Western Antarctic Peninsula (WAP) system has exhibited some of the most significant changes in the Southern Ocean 3 , 4 , with rising air temperature up to 7 °C since 1950 5 , warming and freshening of the upper ocean 6 , warming of the deeper ocean 6 , deepening of the mixed layer depth (MLD) 7 , and the fastest sea ice decrease in Antarctica (Fig.  1 ) 8 , 9 . Whilst atmospheric warming trends at the Antarctic Peninsula have paused or even reversed in places since the end of the twentieth century, this is understood as natural interannual climate variability that is superposed on the longer-term trends 10 . There have been observed ecosystem changes throughout the entire Antarctic marine food web 7 , 11 , 12 , 13 , 14 . At the base of the food web, WAP eukaryotic plankton including phytoplankton and microzooplankton support higher trophic levels ranging from krill to penguins and whales 13 , drive biogeochemical cycles 15 , 16 , 17 , and regulate oceanic carbon uptake 7 . Thus, given the fundamental importance of eukaryotic plankton at the WAP, it is imperative to understand and predict the changes in plankton community structure, biodiversity, and carbon flux in this rapidly changing environment 18 .

a Location of the study area (red box) and the Palmer LTER sampling grid with hydrostations (blue dots). b Time-series of monthly averaged sea-ice area (SIA) anomalies for Palmer LTER sampling area from 1979 to 2017. Blue (red) bars represent negative (positive) SIA compared to a 39-year climatology for a particular month. SIA data were downloaded from Palmer LTER DataZoo ( http://pal.lternet.edu/data ). Cyan arrows highlight the sampling periods in this study.

The WAP system is characterized by a short but highly productive growing season during austral spring and summer 19 . The net community production (NCP) represents the balance between gross primary production and community respiration. When the organic carbon pool at the mixed layer is under a steady state, the net carbon flux in (i.e., NCP) equals the net carbon flux out (i.e., carbon export). Therefore, NCP reflects the amount of organic carbon available for export out of the surface MLD.

Here, we analyze five years of high-resolution NCP and high-throughput DNA sequencing data to explore the contribution of polar eukaryotic plankton to biological carbon fluxes. We show that among the considered environmental factors (iron not included), SST and sea-ice condition are strong predictors for community structure and NCP. We find that biodiversity is reduced when SST is high at the WAP. Finally, in order to improve NCP predictions, we build machine-learning models including in-depth community structure, community co-occurrence patterns, and physical conditions. Among the top-performing NCP models, a sea-ice associated plankton assemblage is a key predictor, with central (i.e., most connected) taxa identified as Thalassiosira , Odontella , Porosira , Actinocyclus , Proboscia , Chaetoceros, and Gyrodinium . The combination of biogeochemical tools and DNA metabarcoding sheds a unique insight into environmental forcings, plankton diversity, community structure and interaction, and biological carbon flux variability in a rapidly changing polar environment.

Results and discussion

O 2 /Ar-based in situ NCP observations at the WAP in austral summer from 2012 to 2016 demonstrate substantial spatial heterogeneity and interannual variability (Fig.  2 ). This was a period of moderately positive sea-ice area (SIA) anomalies following a more prolonged period of anomalously low SIA (Fig.  1 ). During the summer, NCP was highest in the shelf zone and decreased offshore, which is consistent with previous ship-based 15 , 20 and satellite-based observations 19 . In addition, the observed NCP exhibited marked interannual variability related to ice conditions. The two years that feature late sea-ice retreat (2014 and 2016) were associated with abnormally high summer NCP (t-test, p < 0.0001) (Fig.  2 ). In previous studies, elevated NCP or primary production under high sea-ice conditions were attributed to ice-melt enhanced water column stability, thus higher light availability, and potential iron supplied by sea ice 7 , 20 . From a decadal study considering climate oscillations 13 , bloom-favorable conditions at the WAP have been linked to negative winter and spring phases of the Southern Annular Mode (SAM), the dominant mode of extra-tropical climate variability in the Southern Hemisphere 21 . Negative SAM leads to increased ice extent in winter, restricting deep mixing, and then enhanced ice-melt in spring/summer, resulting in intensified stratification.

January averaged sea-ice concentrations derived from passive microwave satellite measurements ( top ). Red contours represent the biologically relevant ‘ice-edge’ defined as an ice concentration threshold of 5%. Underway estimates of NCP using O 2 /Ar method from the annual PAL-LTER sampling cruises along the WAP grids ( bottom ).

Eukaryotic plankton, including phytoplankton and microzooplankton, are key drivers of carbon fluxes at the WAP 16 . Based on a five-year WAP DNA sampling, we explored the plankton community structure and diversity via sequencing of the 18S rRNA gene marker. At the phylum level, four eukaryotic plankton dominated the WAP surface water, including diatoms (25.0%), cryptophytes (23.0%), dinoflagellates (19.6%), and haptophytes (11.3%) (Fig.  S3 ). Other eukaryotic plankton groups, mostly heterotrophic protists, contributed less than 5% of the 18S reads. Community composition differed substantially between years with high and low sea ice extent (Fig.  S4 ). For the years 2014 and 2016 with high sea ice, eukaryotic plankton communities comprised on average 39.5 ± 3.0% diatoms, 20.5 ± 1.3% dinoflagellates, 15.1 ± 6.1% cryptophytes, and 7.1 ± 2.4% haptophytes. In contrast, for warm years with less sea ice in 2012, 2013, and 2015, eukaryotic plankton communities comprised on average 28.9 ± 8.3% cryptophytes, 19.0 ± 1.6% dinoflagellates, 14.4 ± 4.4% haptophytes, and 14.0 ± 1.7% diatoms, all significantly different from cold years (two-sided t-test, p < 0.0001).

At the finest taxonomic resolution, 2480 amplicon sequence variants (ASVs) were identified from the five-year amplicon dataset (119 samples). Canonical correspondence analysis (CCA) illustrates that either ice conditions or SST is the dominant driver on community structure at the ASV level (Fig.  3 ). The first axis CCA1 (17.5% of the variance) separates samples from late (2014 and 2016) and early (2012, 2013, and 2015) ice-retreat years. The most substantial abiotic factor associated with CCA1 is SST (negatively correlated), and the most substantial biotic factors associated with CCA1 are Chl and biological O 2 (both positively correlated), consistent with ice-melt enhanced biomass and productivity. Freshwater inputs were estimated from oxygen isotope signatures. Low salinity, as well as high fractions of sea-ice melt and meteoric water, are also associated with CCA1 but to a lesser extent than SST. CCA2 (7.3% of the variance) separates mainly the offshore and nearshore samples, with distance to coast (X grid) and MLD being the top two associated environmental factors. CCA3 (5.7% of the variance) indicates community differentiation along the north-to-south gradient (Y grid), potentially reflecting a long-term ice retreat impact on communities and/or a north-to-south climate gradient along the WAP 11 , 19 ; CCA3 could also reflect interannual variability in sea ice extent. Overall, sea-ice conditions and associated environmental parameters, such as low SST (Fig.  S1 and S2 ) and low salinity, are the primary drivers of community differentiation at the ASV level.

Each point represents the eukaryotic plankton community composition from a surface ocean sample with year indicated by color and station type (i.e., offshore, nearshore north, or nearshore south) indicated by shape. CCA1 and 2 are depicted ( a ) as well as CCA1 and 3 ( b ). Vectors indicate stepwise selected environmental constrains, both biotic and abiotic, with factor names marked at the end. Acronyms for selected environmental factors: SST – sea surface temperature, SiO4 – silicate concentration, mldst – mixed layer depth defined by potential density, XgridCal – X grid or grid station calculated from GPS, YgridCal – Y grid or grid line calculated from GPS, f sim – fraction of sea-ice melt estimated from δ 18 O, f met – fraction of meteoric water estimated from δ 18 O, Chl – chlorophyll concentration, o2ar – biological oxygen supersaturation. Source data are provided as a Source Data file.

To further investigate the effect of temperature as one of the major abiotic factors influencing polar plankton composition 22 , we examined the temperature effect on biodiversity (ASV based) using three diversity indices, Chao1, Pielou’s evenness, and Shannon. Chao1, a measure for species richness, demonstrated an evident decline towards higher SST (Fig.  4 ), with a 40% decrease in the index for a 4 °C rise in SST. Pielou’s evenness and Shannon, which consider both species richness and evenness, also decreased significantly with increasing SST. It indicates that communities in warmer WAP waters show lower richness and lower evenness, i.e., that a few taxa dominate. Interestingly, in a recent global analysis on plankton biodiversity from Tara Oceans 23 , the temperature was also identified as the major explanatory factor for global-scale eukaryotic plankton biodiversity estimated by the Shannon index, but with the opposite trend, i.e., a decreased diversity towards higher latitude or lower temperature. We note that whilst the Tara Oceans dataset represents the most comprehensive oceanic DNA sampling efforts to date, it featured only limited sampling in the Southern Ocean (three data points), and it did not include a longitudinal survey that captures the mesoscale effect of changing temperature on a given community; our five-year WAP sample collection hence complements well the Tara Oceans observations for the previously under-sampled Southern Ocean. One explanation for the unexpected high biodiversity observed under low temperature at the WAP is that the ice-associated plankton communities consist mainly of diatoms (Fig.  S4 ), which are highly diverse and can thrive at lower temperatures compared to other phytoplankton 24 . As a unique longitudinal test case (e.g., as documented in 25 ), our results suggest that global warming may decrease plankton diversity in coastal Antarctica.

Three diversity indices, including Chao1 ( a ), Pielou’s evenness ( b ), and Shannon ( c ), plotted against SST, with year indicated by color and station type indicated by shape. Reads from each sample were rarefied to an even depth. Solid red lines represent linear fittings and the gray bands represent 95% confidence intervals. All three diversity indices show significant negative correlations with SST (two-sided t-test, p < 0.01). Source data are provided as a Source Data file.

Due to the high dimensionality of the ASV dataset, it is challenging to model NCP based on community structure. Thus, we applied a weighted gene correlation network analysis (WGCNA) approach to delineate clusters of 18S ASVs into subnetworks or modules (Fig.  S5 ) 26 , 27 . This approach allows us to reduce the total number of variables while preserving information on ASV abundances and potential interactions 28 . In total, we identified 12 modules from the five-year global community structure (Fig.  S5 ). Each module represents an assemblage of predicted highly interconnected plankton community members, potentially indicating a group of organisms with strong ecological overlap and/or interactions 29 . The eigenvalue of each module represents the overall abundance of the assemblage. In addition, in order to investigate the niche partitioning of the different community assemblages or modules, correlation analysis was performed in WGCNA to link them to different abiotic and biotic factors (Fig.  S5 ). Next, we applied Genetic Programming (GP), a machine learning approach based on evolution computation 30 , 31 , to generate and parameterize statistical models that predict NCP based on the WGCNA generated bio-assemblages ( n = 12) and physical factors ( n = 6) (see “Methods” for a detailed list). The relationships between carbon-based plankton biomass, their physiology (indirectly modeled as functions of environmental factors), and biogeochemical rates are often non-linear and could involve multiple layers of interactions. GP allows us to capture the complex and non-linear relationships between these different factors to predict NCP without an a priori assumption. The overall idea of this modeling approach is that biogeochemical rates are a function of (i) the community composition and abundance; and (ii) the specific metabolic rates regulated by environmental factors, such as the photosynthesis-irradiance curve and the productivity−temperature ( Q 10 ) relationship. The top four GP solutions (Table  S2 , ranked by mean square error (MSE)) provide good predictions on NCP with R 2 ranging from 0.70 to 0.80. Among the explanatory variables selected by the models, the top two physical factors ranked by selection frequency are SST and MLD, followed by surface photosynthetically active radiation (PAR). This suggests that temperature and light are likely the primary physiological limiting factors on NCP in the WAP system. As an alternative, SST could be an indirect proxy for time since ice-retreat, i.e., higher SST indicates a longer time after the initial ice retreat.

Among the community-assemblage factors in GP solutions, module turquoise (MET) is the most important predictor for NCP. MET is also the largest module identified, which consists of 126 ASVs, mainly representing diverse groups of diatoms and dinoflagellates (Fig.  5a ; Supplementary Data  2 ). The central nodes in the MET network, which represent the top-10 most connected ASVs or the central ASVs for the network structure 32 , include the diatom genera Thalassiosira , Odontella , Porosira , Actinocyclus , Proboscia , Chaetoceros , and the dinoflagellate Gyrodinium . MET appears in all top-four performed GP solutions, and it is the sole biological factor in solution 1 explaining a majority of the NCP variability ( R 2 = 0.70) (Table  S2 ). The overall spatial distributions of MET are consistent with the January averaged sea-ice distribution estimated by satellite (Fig.  5c ). In the WGCNA correlation analysis (Fig.  S5 ), MET is significantly correlated with low SST ( R = −0.51, p = 1 × 10 −8 ), low salinity ( R = −0.48, p = 8 × 10 −8 ), shallow MLD ( R = −0.36, p = 1 × 10 −4 ), elevated sea-ice melt ( R = 0.33, p = 3 × 10 −4 ) and meteoric freshwater ( R = 0.34, p = 3 × 10 −4 ). Although sea-ice melt and meteoric freshwater both show positive effects on MET, they could exert this through different mechanisms. Because iron concentrations were not included in this analysis, we cannot discern whether sea ice and/or glacial ice melt impact productivity through altering light and/or iron levels by increased stratification or fertilization. However, according to a recent study at the eastern Antarctic Peninsula, sea-ice melt could mostly influence carbon fixation through water column stabilization, while the effect of glacial melt could be through providing a significant amount of iron to the system 33 . In the Southern Ocean, photosynthetic efficiency (Fv/Fm) varies with an iron status where lower values suggest iron stress and higher values of iron sufficiency 34 . Previous studies have found NCP to be positively correlated with Fv/Fm 35 . Although no direct iron measurements were made in this study, iron availability being a first-order factor regulating NCP at the WAP cannot be ruled out. Furthermore, besides NCP ( R = 0.57, p = 8 × 10 −11 ) ship-based observations of primary production (PP) and bacterial production (BP) are also both positively correlated with MET with R = 0.47, p = 1 × 10 −7 and R = 0.61, p = 2 × 10 −12 , respectively. This indicates that MET-dominated regions have high biological activities. The high NCP values associated with MET are largely driven by autotrophs; otherwise, we would expect a negative correlation between MET/NCP and BP.

a The MET subnetwork presents a diverse group of correlated ASVs or nodes, with each node colored by its centrality (i.e., darker color for higher centrality). The hub nodes of the network, i.e., the top-10 ASVs with the highest connectivity or the central members for the network, were identified at the genus level as Thalassiosira , Odontella , Porosira , Actinocyclus , Proboscia , Chaetoceros, and Gyrodinium . b ASVs with higher module membership, i.e., higher intramodular connectivity, are more correlated with volumetric NCP ( R = 0.7, two-sided t-test p = 9.6 × 10 −16 ). c The biogeography of MET at the WAP, with color indicating the eigenvalue in the relative unit. Black points show the DNA sampling locations. Source data are provided in Supplementary Data  2 and 4 .

Two other community-assemblages, MER (module red, 27 ASVs) and MEG (module green, 29 ASVs), also contribute to NCP models but to a lesser extent (Table  S2 ). In the GP solutions (1) and (2), MER contributes to NCP positively, and the addition of MER marginally improves NCP prediction ( R 2 from 0.70 to 0.72, MSE from 1.18 to 1.12). The MER assemblage is dominated by the cryptophyte Geminigera that appear in warmer waters (SST, R = 0.45, p = 6 × 10 −7 ) and towards the north WAP (Y grid, R = 0.4, p = 2 × 10 −5 ) (Figs.  S5 and S6 ). MEG contributes to NCP negatively in GP solutions (2) and (3). It represents a group of heterotrophic protists, dominated by Picomonas and Telonema . The top two environmental factors correlated with MEG are MLD ( R = 0.5, p = 2 × 10 −8 ) and distance to shore (X grid, R = 0.49, p = 6 × 10 −8 ).

Based on our observations and analyses, we hypothesize that the summer plankton community—NCP system at the WAP mainly follows three broad patterns: (i) large centric diatoms associated with ice-melt form intensive blooms and fuel a short food chain from krill to other top predators 13 . In particular, the spring melt of sea ice and glacial discharge could work in concert to stabilize the water column and provide a source of iron. This high productivity combined with small losses through trophic transfer results in high export production. (ii) In warmer water, small cryptophytes dominate. Compared to large diatoms, their growth could be more efficiently checked by small microzooplankton grazers 36 , thus resulting in lower biomass for export 16 . Moreover, the food chain starting from small phytoplankton is longer due to more trophic level transfers, and the organic matter could be more subject to remineralization 37 . (iii) With deep mixing, primary production in the water column is low due to light limitation. Because of the limited food resource, heterotrophic protists feeding on bacteria and detritus dominate the microzooplankton system. Compared to scenarios (i) and (ii), more organic carbon may be recycled through the microbial loop, which further reduces carbon export and air−sea CO 2 fluxes 7 . The last pattern displays the lowest NCP. Previous WAP studies using an inverse food web model illustrated that microzooplankton grazing and the microbial loop could consume a significant amount of carbon 38 , 39 . With climate change, the WAP region is projected to have a significant loss in summer sea ice, a rise in sea surface temperatures, and deeper mixing associated with more open water and stronger winds. Consequently, the latter two scenarios may become more prevalent in the upcoming years to decades.

Although our study represents the longest record of eukaryotic DNA-based community structure and NCP in coastal Antarctica, our observations are limited to seasonal snapshots of the (summer) WAP system. These observations need to be expanded to larger spatial and temporal scales in the Southern Ocean. In the future, correlation-based analyses and statistical models will need to be further validated with field incubations and lab experiments. Non-targeted omics-based surveys (e.g., metagenomic, metatranscriptomic, and proteomic studies) will provide additional insights into the microbial metabolic pathways, which are directly linked to the biogeochemical rates and associated ecosystem functions. Moreover, they need to be coupled with high-resolution time-series studies to help us unravel changes in phytoplankton phenology and predator-prey dynamics. Despite the methodological limitations and uncertainties, our results indicate that temperature and sea ice extent are two important environmental factors regulating the summer WAP eukaryotic plankton community structure, biodiversity, productivity, and associated carbon export potential. To the extent that the observed interannual variability in the influence of sea ice extent on ecosystem structure and functioning serves as a proxy for broader, longer-term ecological consequences associated with climate change, the WAP and other coastal Antarctica regions could be destined for reduced biodiversity and biological carbon drawdown. However, a longer time series will be needed to confirm the pattern.

Environmental data and DNA sampling

Environmental data from the Palmer Long-Term Ecological Research (LTER) cruises can be accessed from the online data repository Palmer LTER DataZoo ( http://pal.lternet.edu/data ). The detailed sampling methods and in situ biological rates measurements were described previously in 40 . In brief, each year in January, a research vessel conducted intensive oceanographic and biological surveys across the shelf-transects and a north−south gradient at the West Antarctic Peninsula (WAP). During the annual LTER cruises, underway measurements and surface water sampling were conducted from the ship’s continuous flow-through system. Discrete water samples in-depth profiles were collected using a Conductivity−Temperature−Depth (CTD) rosette. Mixed layer depth (MLD) was estimated from the ship’s CTD profiles by Δσ θ = 0.03 kg m −3 using a threshold method 41 .

PAR above the water was continuously recorded from the mast PAR sensor of the ship. It was converted to PAR just beneath the water surface using a constant of 0.92. Average PAR in the mixed layer (PAR_mld) was then calculated following the method described in 42 .

Freshwater fractions were estimated from salinity and oxygen isotope signatures (δ 18 O) in seawater detailed in 43 . In brief, sampled seawater was assumed to be a mixture of ice-melt, meteoric meltwater, and Circumpolar Deep Water (CDW). A three-end member mass balance method was used to calculate the fractions, with salinity and δ 18 O values in 7 and 2.1‰ for sea-ice melt, 0 and −16‰ for meteoric meltwater, and 34.73 and 0.1‰ for CDW.

In order to collect eukaryotic plankton DNA, surface seawater from the ship’s underway flow-through system was gently vacuum-filtered onto a 47 mm, 0.45 µm Supor filter (Pall Corporation, New York, NY, USA) for years 2012 and 2013, or a 47 mm, 0.2 µm Supor filter for years 2014, 2015, and 2016. The filtration volumes were about 4 L or less at high biomass stations. For each filtration, the exact filtrate volume was recorded for later quantitative microbiome profiling (QMP). The filters were immediately stored at −80 °C until further analysis.

Remote sensing data

January sea ice concentrations from 1979 to 2020 were downloaded from the National Snow and Ice Data Center website https://nsidc.org/ . The data are in the polar stereographic projection, with each grid representing a 25 × 25 km area. January sea surface temperature (SST) data from 1982 to 2012 were acquired by the AVHRR and downloaded from NOAA website https://www.ncei.noaa.gov/ . January SST data from 2013 to 2020 were acquired by MODIS-Aqua and downloaded from NASA ocean color website https://oceancolor.gsfc.nasa.gov/ . SST data have a spatial resolution of 4 × 4 km in the equatorial region. Finally, we extracted January SST and sea ice concentrations in the Palmer grid from lines 0 to 900 and stations 0 to 220 (Figs.  S1 and S2 ).

Underway O 2 /Ar—NCP measurements

The O 2 concentration in the mixed layer is influenced by physical and biological processes. Using Ar, an inert gas with similar solubility properties as O 2 , we decomposed total O 2 into physical and biological components. Seawater O 2 /Ar ratios were measured underway from the ship’s flow-through system, using an equilibrator inlet mass spectrometer (EIMS) 44 . Biological O 2 supersaturation was estimated as

High-resolution NCP in units of mmol O 2 m −2 day −1 , was then derived from Δ(O 2 /Ar) and NCEP reanalysis winds as previously described in 45 , except for a modification to the gas exchange weighting following 46 . NCP estimation can be expressed as the equation below,

Where k denotes the gas transfer velocity for O 2 (estimated based on 47 ) and [O 2 ] sat is the equilibrium saturation concentration of O 2 (calculated based on 48 ). According to 45 , the ship-based O 2 /Ar NCP estimates are highly correlated with NCP calculated from the seasonal DIC drawdown in this region ( R 2 = 0.83). Note that our O 2 /Ar-NCP measurements in this study only reflect the mixed layer carbon fluxes and we do not assess the sequestration timescales.

DNA extraction and metabarcoding

DNA extraction and PCR were conducted as previously described 40 . In brief, cells were lysed by bead-beating at 4800 rpm for 1 min with 0.2 g of 0.1 mm Zr beads in 400 µl of Qiagen lysis buffer AP1. DNA was then extracted using DNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA) following the manufacturer’s instructions. rRNA gene amplicon libraries were constructed using dual indexed 18S rRNA gene V4 primer set 16 , EukF (5′–CCAGCASCYGCGGTAATTCC–3′) and EukR (5′–ACTTTCGTTCTTGAT–3′). For each sample, PCR amplifications were conducted in triplicates with one blank as a control for contamination. The PCR reactions followed a 30-cycle program with annealing temperature at 57 °C. The resulting PCR products were purified using QIAquick PCR Purification Kit (Qiagen), and were pooled in equimolar concentration to 10 ng/µl approximately. The pooled amplicon libraries were sequenced at Duke Center for Genomic and Computational Biology in three MiSeq 300PE runs.

Sequence processing

Paired-end reads with dual indices were assembled using VSEARCH v2.3.4 (Rognes et al. 2016) following the algorithm described in 49 . The merged reads were (i) demultiplexed in QIIME 1 50 ; (ii) trimmed to remove Illumina adapters, primers, and barcodes, using BBDuk (v38.29) ( http://jgi.doe.gov/data-and-tools/bb-tools/ ); and (iii) processed following the DADA2 pipeline (version 1.10.1) to infer ASVs 51 . Quality filtering and denoising with chimeras removal were performed using the incorporated functions in the DADA2 package. In total, 2480 ASVs were identified from the five-year amplicon dataset. The sequence counts per sample (after quality filtering) were reported in Supplementary Data  1 , with median = 67,547 reads per sample.

ASVs were then classified by the ‘assignTaxnonomy’ function in DADA2 following the naïve Bayesian classifier method 52 , using a DADA2 formatted Silva 132 reference database (DOI 10.5281/zenodo.1172783) 53 . The resulted classification for each ASV is presented in Supplementary Data  3 .

Alpha diversity

After discarding two samples with the lowest counts 2016S33 and 1016S34, the libraries ( n = 117) were rarified to even depth. Alpha diversity indices Chao1 and Shannon (H′) were calculated for each sample using R package Phyloseq v1.26.1 54 . Pielou’s evenness was calculated as J = H ′/ln( S ), where S is the total number of ASVs observed in a rarified sample.

Canonical correspondence analysis (CCA)

CCA was conducted to investigate the relationships between community composition changes and environmental constrains 55 using R package vegan (v2.5-4) 56 . In total, 14 environmental variables were initially examined, including XgridCal, YgridCal, PAR, Salinity, SST, Chl, mixed layer depth, volumetric NCP, biological oxygen supersaturation, PO 4 , SiO 4 , N+N, fsim, and fmet . Bacterial production (BP) and primary production (PP) were not included in this analysis due to a large number of missing values. After a stepwise variable selection based on Akaike Information Criterion (permutation = 1000 per step), the constrained community CCA was conducted with selected environmental variables and the results were presented in Fig.  3 .

Construct statistical models for NCP

Below we describe a three-step procedure: (i) ASV counts were normalized to generate QMP; (ii) in order to reduce the model dimension, a WGCNA was applied to QMP to generate modules or bio-assemblages, and (iii) the resulting WGCNA modules and environmental variables were fed to the genetic programming (GP) algorithm to construct predictive models for NCP.

Quantitative microbiome profiling (QMP)

All samples from the year 2014 and four samples from the year 2013 (2013SA, 2013SB, 2013SC, and 2013SD) were processed first, and 0.88 ng of Schizosaccharomyces pombe gDNA (ATCC #24843D-5, Manassas, VA, USA) in single-use aliquot was spiked into to each sample as an internal standard before DNA extraction. The S. pombe reads did not turn out high enough for normalization in the resulting library, i.e., ≤0.1%. For samples from the years 2012, 2013 (except for the previous four samples), and 2015, we increased the amount of S. pombe gDNA to 16.0 ng per sample and the internal standard proportion turned out appropriate for detection (0.7−5.7% of the total 18S counts). In the third batch, samples from the year 2016 were extracted with no internal standard due to a logistic issue in the lab.

For samples from the second batch, we normalized the ASV counts to QMP (in unit of 18S gene copy numbers L −1 ) using internal standards and recorded filtration volumes as described in 40 .

For samples from the first and third batches, reads were normalized to QMP using an empirical linear relationship 40 ( R 2 = 0.94) between x —cryptophyte Alloxanthin concentrations in μg/L, and y —cryptophyte 18S rRNA gene counts in copies/mL: y = 2.05 × 10 6 x . The Alloxanthin concentrations in the linear calibration range from 0.01 to 6.22 μg/L. Although Alloxanthin concentrations for all samples used in this calculation are above 0.01 μg/L, we note that there is higher uncertainty towards the lower concentration end. The resulted Phaeocystis 18S QMP in years 2014 and 2016 were strongly correlated with Phaeocystis CHEMTAX abundances ( R 2 = 0.62), except for two outliner samples 2014S17 and 2014S44, likely due to low Alloxanthin concentrations in these two samples. QMP for these two samples was then recalculated using an empirical linear relationship derived from Phaeocystis CHEMTAX abundance 40 .

As a complementary analysis, we recalculated QMP (QMP_recal) using the empirical HPLC-CHEMTAX normalization for samples from 2013 to 2016 (no HPCL data for the year 2012). The resulted QMP_recal is highly similar to the internal standard method QMP ( y = 0.99 x , R 2 = 0.87), except for one sample 2015S13.

Weighted gene correlation network analysis

ASVs which were not observed more than three times in at least 20% of the samples were removed from the count table. WGCNA was conducted to identify inter-connected plankton bio-assemblages (modules) and correlate them with environmental variables using R package WGCNA v 1.66 27 . The calculated QMP for 112 samples were included in one WGCNA run with all samples considered independent of each other. The QMP matrix was log-transformed. The detailed R codes for each step of this analysis are presented as a supplementary file. Soft thresholding power was set at 4, which was the minimum value for the scale-free topology fit reaching R 2 = 0.9. In module identification using dynamic tree cut, the minimum module size was set at 10 in order to generate medium to large modules. The analysis resulted in 12 WGCNA modules from thousands of ASVs, thereby significantly reducing the number of input variables for the NCP model. The co-occurrence network of each module was visualized using an open-source tool Cytoscape (v3.7.0).

Genetic programming to build NCP models

Based on community structure (modules) and abiotic environmental variables, GP was used to construct statistical models predicting volumetric NCP. GP is a machine-learning approach based on evolutionary computation and it has been successfully used to construct a NCP algorithm based on satellite observations in a previous study 31 . In this study, the input factors for GP are, (i) the eigenvalues for 12 WGCNA modules, representing the biological/community factors with resolution at ASV level, and (ii) a list of physical factors, including MLD, PAR, PAR_mld, Salinity, SST, f sim , f met , X grid, and Y grid, which may directly or indirectly influence plankton physiology. The combined dataset ( n = 112) was randomly split into even training and validation datasets. GP was then conducted using Eureqa (v1.24.0) following the recommendations by 57 . The candidate solutions with varying complexity were ranked by mean squared error  (Table  S2 ). In order to reduce the risk of overfitting, the complexity of the candidate solutions was kept to a minimum.

Reporting summary

Further information on research design is available in the  Nature Research Reporting Summary linked to this article.

Data availability

DNA sequencing data generated in this study have been deposited in the National Center for Biotechnology Information (NCBI) under accession number PRJNA508517. Palmer LTER data are available through Datazoo ( http://pal.lternet.edu/data ). Silva 132 reference database used for taxonomy classification was downloaded from ( https://doi.org/10.5281/zenodo.1172783 ).  Source data are provided with this paper.

Code availability

The R codes for WGCNA analysis and the MATLAB codes for NCP calculation can be accessed from GitHub ( https://github.com/nicolascassar/WGCNA-Analyses and https://github.com/nicolascassar/O2Ar_calculations ).

Frölicher, T. L. et al. Dominance of the Southern Ocean in anthropogenic carbon and heat uptake in CMIP5 models. J. Clim. 28 , 862–886 (2015).

Article   ADS   Google Scholar  

Meredith, M. et al. Polar Regions. Chapter 3, IPCC Special Report on the Ocean and Cryosphere in a Changing Climate. https://www.ipcc.ch/srocc/chapter/chapter-3-2/ (2019).

Vaughan, D. G. et al. Recent rapid regional climate warming on the Antarctic Peninsula. Clim. Change 60 , 243–274 (2003).

Article   Google Scholar  

Metzl, N. A canary in the Southern Ocean. Nat. Clim. Chang. 9 , 651–652 (2019).

Turner, J. et al. Antarctic climate change and the environment: an update. Polar Rec. 50 , 237–259 (2014).

Meredith, M. P. & King, J. C. Rapid climate change in the ocean west of the Antarctic Peninsula during the second half of the 20th century. Geophys. Res. Lett . 32 , L19604 (2005).

Brown, M. S. et al. Enhanced oceanic CO2 uptake along the rapidly changing West Antarctic Peninsula. Nat. Clim. Chang. 9 , 678–683 (2019).

Article   ADS   CAS   Google Scholar  

Stammerjohn, S., Massom, R., Rind, D. & Martinson, D. Regions of rapid sea ice change: an inter-hemispheric seasonal comparison. Geophys. Res. Lett . 39 , L06501 (2012).

Stammerjohn, S. E., Martinson, D. G., Smith, R. C. & Iannuzzi, R. A. Sea ice in the western Antarctic Peninsula region: Spatio-temporal variability from ecological and climate change perspectives. Deep Sea Res. Part II Top. Stud. Oceanogr. 55 , 2041–2058 (2008).

Turner, J. et al. Absence of 21st century warming on Antarctic Peninsula consistent with natural variability. Nature 535 , 411–415 (2016).

Article   ADS   CAS   PubMed   Google Scholar  

Montes-Hugo, M. et al. Recent changes in phytoplankton communities associated with rapid regional climate change along the western Antarctic Peninsula. Science 323 , 1470–1473 (2009).

Schofield, O. et al. Decadal variability in coastal phytoplankton community composition in a changing West Antarctic Peninsula. Deep Sea Res. Part I Oceanogr. Res. Pap. 124 , 42–54 (2017).

Saba, G. K. et al. Winter and spring controls on the summer food web of the coastal West Antarctic Peninsula. Nat. Commun . 5 , 4318 (2014).

Fraser, W. R. & Hofmann, E. E. A predator 1 s perspective on causal links between climate change, physical forcing, and ecosystem response. Mar. Ecol. Prog. Ser. 265 , 1–15 (2003).

Huang, K., Ducklow, H., Vernet, M., Cassar, N. & Bender, M. L. Export production and its regulating factors in the West Antarctica Peninsula region of the Southern Ocean. Global Biogeochem. Cycles 26 , GB2005 (2012).

Lin, Y. et al. Specific eukaryotic plankton are good predictors of net community production in the Western Antarctic Peninsula. Sci. Rep . 7 , 14845 (2017).

Swann, G. E. A., Pike, J., Leng, M. J., Sloane, H. J. & Snelling, A. M. Temporal controls on silicic acid utilisation along the West Antarctic Peninsula. Nat. Commun. 8 , 14645 (2017).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Ducklow, H. W. et al. Spring–summer net community production, new production, particle export and related water column biogeochemical processes in the marginal sea ice zone of the Western Antarctic Peninsula 2012–2014. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 376 , 20170177 (2018).

Li, Z., Cassar, N., Huang, K., Ducklow, H. & Schofield, O. Interannual variability in net community production at the Western Antarctic Peninsula region (1997–2014). J. Geophys. Res. Ocean. 121 , 4748–4762 (2016).

Eveleth, R. et al. Ice melt influence on summertime net community production along the Western Antarctic Peninsula. Deep Sea Res. Part II Top. Stud. Oceanogr . 139 , 89−102 (2016).

Thompson, D. W. J. & Wallace, J. M. Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Clim. 13 , 1000–1016 (2000).

Salazar, G. et al. Gene expression changes and community turnover differentially shape the global ocean metatranscriptome. Cell 179 , 1068–1083 (2019).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Ibarbalz, F. M. et al. Global trends in marine plankton diversity across kingdoms of life. Cell 179 , 1084–1097 (2019).

Brun, P. et al. Ecological niches of open ocean phytoplankton taxa. Limnol. Oceanogr. 60 , 1020–1038 (2015).

García, F. C., Bestion, E., Warfield, R. & Yvon-Durocher, G. Changes in temperature alter the relationship between biodiversity and ecosystem functioning. Proc. Natl Acad. Sci. USA 115 , 10989–10994 (2018).

Article   PubMed   PubMed Central   CAS   Google Scholar  

Zhang, B. & Horvath, S. A general framework for weighted gene co-expression network analysis. Stat. Appl. Genet. Mol. Biol . 4 , 17 (2005).

Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinform. 9 , 559 (2008).

Article   CAS   Google Scholar  

Guidi, L. et al. Plankton networks driving carbon export in the oligotrophic ocean. Nature 532 , 465−470 (2016).

Faust, K. & Raes, J. Microbial interactions: from networks to models. Nat. Rev. Microbiol. 10 , 538 (2012).

Article   CAS   PubMed   Google Scholar  

Koza, J. R. & Koza, J. R. Genetic Programming: on the Programming of Computers by Means of Natural Selection Vol. 1 (MIT Press, 1992).

Li, Z. & Cassar, N. Satellite estimates of net community production based on O2/Ar observations and comparison to other estimates. Glob. Biogeochem. Cycles 30 , 735–752 (2016).

Banerjee, S., Schlaeppi, K. & van der Heijden, M. G. A. Keystone taxa as drivers of microbiome structure and functioning. Nat. Rev. Microbiol. 16 , 567–576 (2018).

Wang, B. et al. Meteoric water promotes phytoplankton carbon fixation and iron uptake off the eastern tip of the Antarctic Peninsula (eAP). Prog. Oceanogr . 185 , 102347 (2020).

Suggett, D. J., Moore, C. M., Hickman, A. E. & Geider, R. J. Interpretation of fast repetition rate (FRR) fluorescence: signatures of phytoplankton community structure versus physiological state. Mar. Ecol. Prog. Ser. 376 , 1–19 (2009).

Cassar, N. et al. The influence of iron and light on net community production in the Subantarctic and Polar Frontal Zones. Biogeosciences 8 , 227–237 (2011).

Calbet, A. & Landry, M. R. Phytoplankton growth, microzooplankton grazing, and carbon cycling in marine systems. Limnol. Oceanogr. 49 , 51–57 (2004).

Dickman, E. M., Newell, J. M., González, M. J. & Vanni, M. J. Light, nutrients, and food-chain length constrain planktonic energy transfer efficiency across multiple trophic levels. Proc. Natl Acad. Sci. USA 105 , 18408–18412 (2008).

Sailley, S. F. et al. Carbon fluxes and pelagic ecosystem dynamics near two western Antarctic Peninsula Adélie penguin colonies: an inverse model approach. Mar. Ecol. Prog. Ser. 492 , 253–272 (2013).

Ducklow, H. W., Doney, S. C. & Sailley, S. F. Ecological controls on biogeochemical fluxes in the western Antarctic Peninsula studied with an inverse foodweb model. Polar Sci. 26 , 122–139 (2015).

Google Scholar  

Lin, Y., Gifford, S., Ducklow, H., Schofield, O. & Cassar, N. Towards quantitative microbiome community profiling using internal standards. Appl. Environ. Microbiol. 85 , e02634–18 (2019).

de Boyer Montégut, C., Madec, G., Fischer, A. S., Lazar, A. & Iudicone, D. Mixed layer depth over the global ocean: an examination of profile data and a profile-based climatology. J. Geophys. Res. Ocean . 109 , C12003 (2004).

Li, Z. & Cassar, N. A mechanistic model of an upper bound on oceanic carbon export as a function of mixed layer depth and temperature. Biogeosci. Discuss. 2017 , 1–27 (2017).

CAS   Google Scholar  

Meredith, M. P. et al. Changing distributions of sea ice melt and meteoric water west of the Antarctic Peninsula. Deep Sea Res. Part II Top. Stud. Oceanogr. 139 , 40–57 (2017).

Cassar, N. et al. Continuous high-frequency dissolved O2/Ar measurements by equilibrator inlet mass spectrometry. Anal. Chem. 81 , 1855–1864 (2009).

Eveleth, R. et al. Ice melt influence on summertime net community production along the Western Antarctic Peninsula. Deep Sea Res. Part II Top. Stud. Oceanogr. 139 , 89–102 (2017).

Teeter, L., Hamme, R. C., Ianson, D. & Bianucci, L. Accurate estimation of net community production from O2/Ar measurements. Glob. Biogeochem. Cycles 32 , 1163–1181 (2018).

ADS   CAS   Google Scholar  

Sweeney, C. et al. Constraining global air‐sea gas exchange for CO2 with recent bomb 14C measurements. Global Biogeochem. Cycles 21 , GB2015 (2007).

Garcia, H. E. & Gordon, L. I. Oxygen solubility in seawater: better fitting equations. Limnol. Oceanogr. 37 , 1307–1312 (1992).

Edgar, R. C. & Flyvbjerg, H. Error filtering, pair assembly and error correction for next-generation sequencing reads. Bioinformatics 31 , 3476–3482 (2015).

Caporaso, J. G. et al. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7 , 335–336 (2010).

Callahan, B. J. et al. DADA2: high-resolution sample inference from Illumina amplicon data. Nat. Methods 13 , 581 (2016).

Wang, Q., Garrity, G. M., Tiedje, J. M. & Cole, J. R. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl. Environ. Microbiol. 73 , 5261–5267 (2007).

Callahan, B. Silva taxonomic training data formatted for DADA2 (Silva version 132). Zenodo. https://doi.org/10.5281/zenodo.1172783 (2018).

McMurdie, P. J. & Holmes, S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One 8 , e61217 (2013).

Ter Braak, C. J. F. & Verdonschot, P. F. M. Canonical correspondence analysis and related multivariate methods in aquatic ecology. Aquat. Sci. 57 , 255–289 (1995).

Oksanen, J. et al. Package ‘vegan’. Community Ecol. Packag. Version 2 , 1–295 (2013).

Schmidt, M. & Lipson, H. Distilling free-form natural laws from experimental data. Science. 324 , 81–85 (2009).

Download references

Acknowledgements

We thank the scientists, scientific support personnel, the crew on R/V Laurence M. Gould, and the PAL-LTER team for their assistance in sample collection and underway measurements. We are grateful to Hans Gabathuler for instrument support and Naomi Shelton for field sampling. This work is supported by NSF OPP-1643534 to N.C., NSF OPP-1341479 to A.M., and NSF PLR-1440435 to H.D. and O.S. (Palmer LTER). M.M. was supported by the UK Natural Environment Research Council via the BAS Polar Oceans program.

Author information

Authors and affiliations.

Division of Earth and Ocean Sciences, Nicholas School of the Environment, Duke University, Durham, NC, USA

Yajuan Lin, Zuchuan Li & Nicolas Cassar

Université de Brest—UMR 6539 CNRS/UBO/IRD/Ifremer, Laboratoire des sciences de l’environnement marin—IUEM, Rue Dumont D’Urville, Plouzané, France

Yajuan Lin & Nicolas Cassar

Environmental Research Center, Duke Kunshan University, Kunshan, China

Department of Marine Sciences, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

Carly Moreno & Adrian Marchetti

Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY, USA

Hugh Ducklow

Rutgers University’s Center for Ocean Observing Leadership (RU COOL), Department of Marine and Coastal Sciences, School of Environmental and Biological Sciences, Rutgers University, New Brunswick, NJ, USA

Oscar Schofield

Université de Nantes, CNRS UMR 6004, LS2N, Nantes, France

Erwan Delage, Damien Eveillard & Samuel Chaffron

British Antarctic Survey, Cambridge, United Kingdom

Michael Meredith

Applied Ocean Physics and Engineering, Woods Hole Oceanographic Institution, Woods Hole, MA, USA

Research Federation for the Study of Global Ocean Systems Ecology and Evolution, FR2022/Tara Oceans GOSEE, Paris, France

Damien Eveillard & Samuel Chaffron

You can also search for this author in PubMed   Google Scholar

Contributions

N.C. and Y.L. conceived the study. Y.L., C.M., H.D., M.M., and O.S. collected the field samples and underway data. Y.L., C.M., and A.M. processed the DNA samples. Y.L., S.C., D.E., E.D. and Z.L. analyzed the data. Y.L. and N.C. wrote the papers with inputs from all other authors.

Corresponding authors

Correspondence to Yajuan Lin or Nicolas Cassar .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks Harriet Alexander and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Description of additional supplementary files, supplementary data 1, supplementary data 2, supplementary data 3, supplementary data 4, reporting summary, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Lin, Y., Moreno, C., Marchetti, A. et al. Decline in plankton diversity and carbon flux with reduced sea ice extent along the Western Antarctic Peninsula. Nat Commun 12 , 4948 (2021). https://doi.org/10.1038/s41467-021-25235-w

Download citation

Received : 13 July 2020

Accepted : 30 July 2021

Published : 16 August 2021

DOI : https://doi.org/10.1038/s41467-021-25235-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Water masses shape pico-nano eukaryotic communities of the weddell sea.

• Olga Flegontova
• Pavel Flegontov

Communications Biology (2023)

Spatiotemporal high-resolution mapping of biological production in the Southern Ocean

• Xianliang L. Pan
• Xiangxing Lai
• Yutaka W. Watanabe

Communications Earth & Environment (2023)

Spatial distribution of planktonic ciliates in waters around the northeastern Antarctic Peninsula and the South Orkney Plateau

• Wuchang Zhang

Polar Biology (2023)

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.