Exploring functional traits and functional diversity of mixotrophic phytoflagellates in small browned forest lakes—mixotrophy, phagotrophy and osmotrophy

Abstract

Mixotrophy is the combination of autotrophy and heterotrophy within a single organism. Heterotrophy in mixotrophs encompasses two main processes: ingestion of prey, termed phagotrophy, and nutrition by direct absorption and uptake of organic molecules, osmotrophy. Though osmotrophy is common in phytoplankton species it is often neglected in mixotrophic studies despite in some types of aquatic ecosystems, such as small humic lakes, obligate-osmotrophic organisms are especially important. This study was aimed at investigating the contributions of potential mixotrophs and examining the relationship between their functional traits (including osmotrophy) and functional diversity in response to environmental factors in small forest lakes. Through large-scale lake sampling, we found that light-availability and DIN concentration support potential mixotroph success. Lakes with high inputs of allochthonous organic material exhibited a greater influence of potential mixotrophs over autotrophs. This study indicates that obligate-osmotrophs may be a crucial metabolic trait in browned forest lakes, providing an adaptive advantage for mixotrophs and the inclusion of osmotrophy within mixotrophy studies appears to be promising. We also found that despite dominance, the homogeneous distribution of mixotrophs suggests functional redundancy.

INTRODUCTION

Numerous phytoplankton species have evolved and developed mechanisms to combine autotrophy and heterotrophy in the same organism, termed mixotrophy (Jones, 2000; Selosse et al., 2017). Heterotrophy in mixotrophs encompasses two main processes: phagotrophy (i.e. ingestion of prey or other food particles to acquire essential nutrients, among them carbon) (Jones, 2000) and osmotrophy (i.e. nutrition by direct absorption and uptake of dissolved organic molecules (Beardall and Raven, 2016). Phagotrophy in mixotrophs is a biogeochemically important strategy for eukaryotic phytoplankton and it is well studied and reported both in marine environments and freshwaters (Unrein et al., 2007; Hartmann et al., 2012; Gerea et al., 2018; Leles et al., 2021). In contrast, the significance of phytoplankton osmotrophy under natural conditions remains unclear and there is a need to incorporate knowledge of osmotrophy into studies of mixotrophy.

Osmotrophic pathways include direct absorption, extracellular oxidation, hydrolysis, and pinocytosis of molecules assisted by several enzymes involved in these processes (Glibert and Legrand, 2006). It is also known that, virtually all species of phytoplankton can obtain small dissolved organic compounds of low molecular weight from their surrounding environment (Nedoma et al., 2003; Flynn et al., 2019), since they all possess the necessary Krebs cycle enzymes (Beardall and Raven, 2016). Therefore, osmotrophy (Nedoma et al., 2003), might be a common trait separator in phototrophic plankton. Some phytoplankton species can use different strategies to acquire organic material from the environment. Specifically, they either have the ability to transport larger organic molecules into their cells or the ability to phagocytize cells, particles and/or large molecules (Glibert and Legrand, 2006). Moreover, mixotrophic species display a continuum of dependence on alternative pathways (Flynn et al., 2013). Facultative mixotrophs can switch from autotrophic to heterotrophic nutrition according to the prevailing environmental conditions (e.g. limiting amounts of light or nutrients) (Flynn et al., 2013; Selosse et al., 2017; Stoecker et al., 2017). Additionally, some species rely on obligate osmotrophy, indicating a dependence on their ability to absorb essential organic molecules from the environment, while others not. Martens et al. (2024) performed experiments with phytoplankton strains from an estuary revealing that organic compounds drive phytoplankton growth depending on their trophic strategy and light availability concluding that heterotrophy in phytoplankton should be considered in future research. Godrijan et al. (2020), investigated the uptake of large-sited organic molecules by osmotrophic coccolithophorids finding that they can metabolize a wide array of organic compounds in darkness, but the uptake rates are low relative to the photosynthetic potential for carbon fixation. In experiments by Carlsson et al. (1999), additions of humic materials stimulated the growth of the dinoflagellate Prorocentrum minimum. Other studies indicate that under high dissolved organic nutrient or carbon concentrations, osmotrophs should be favored over strictly photosynthetic protists. For example, in browned lakes, osmotrophy advantaged Gonyostomum semen (Glibert et al., 2001; Rohrlack, 2023). Browning may also favor phagotrophs, by increasing abundances of potential bacterial prey (Jones, 2000).

There is a debate among researchers regarding the term “mixotrophy”. Some authors use a restrictive definition, i.e. autotrophic and phagotrophic organisms only, excluding osmotrophy (Flynn et al., 2019) and others prefer a broader approach and include osmotrophy within the term mixotrophy (Selosse et al., 2017). Nevertheless, classifying osmotrophy have remained a challenge, as it requires a deep understanding of each species’ physiology and requirements. While some species are well studied, many have been not thoroughly or not at all investigated with respect to osmotrophy and further studies are needed.

In this study we advocate the use of a broader and comprehensive approach to define the types of mixotrophy, incorporating obligate osmotrophy into the category of mixotrophs. Moreover, categorizing mixotrophic strategies poses a challenge due to the vast diversity of mixotrophic protists, compounded by incomplete or lacking data on the functional responses of most mixotrophs (Stoecker, 1998; Millette et al., 2023). Despite these challenges, attempting categorization is valuable for developing a general understanding of mixotrophy as both an evolutionary strategy and a significant driver of planktonic ecosystem dynamics (Schoonhoven, 2000).

Although mixotrophic activity is considered favorable, maintaining multiple metabolic pathways within the same cell involves associated energetic costs: the rate of use depending on conditions such as light availability, nutrient levels, and prey availability (Rothhaupt, 1996; Flöder et al., 2006). Consequently, mixotrophic organisms have evolved, as a solution, to trade-off requirements of optimal growth conditions and acquiring carbon (or other essential food items) from various sources, particularly in resource-limited environments (Selosse et al., 2017). So far, mixotrophy has been studied predominantly in marine environments (Mitra et al., 2016; Flynn et al., 2019; Leles et al., 2021), although there has been a growing interest in freshwater ecology indicated by an increasing number of field and experimental studies (Bird and Kalff, 1986; Gerea et al., 2018; Le Noac’h et al., 2024). Knowledge of distribution of mixotrophs is crucial to identify habitats where they may become dominant components of the plankton community and helping the understanding the conditions that will favor the success of mixotrophs (Millette et al., 2023). Quantifying functional diversity can help us to understand how biodiversity is distributed in a multidimensional space and to discriminate the processes shaping community structures. Therefore, the use of functional diversity metrics proved to be a suitable tool to evaluate biogeographic patterns in response to environmental changes (Stenger-Kovács et al., 2020; Al-Imari et al., 2023) and to predict and contribute to ecological “stability” of aquatic habitats (Schneider et al., 2017).

Small lakes and ponds represent the majority of freshwater ecosystems worldwide with geographically overall distribution (Wetzel, 1990; Downing et al., 2006). There are significant differences between the definition of small lakes and ponds, based on several metrics such as surface area, depth, and morphometry, which make them ecologically distinct type of aquatic habitats (Richardson et al., 2022). Ponds can be defined as small and shallow waterbodies with a maximum surface area of 5 ha, a maximum depth of 5 m, and < 30% coverage of emergent vegetation. Likewise, small lakes are more related to surface area and depth, and are not necessarily shallow (Padisák and Reynolds, 2003; Richardson et al., 2022). These water bodies interact with and impact of the surrounding environment in various ways, providing essential ecosystem services and serving as important nature-based solutions to lake restoration and conservation (Cuenca-Cambronero et al., 2023). Despite their significance for biodiversity conservation (high taxon number; Várbíró et al., 2017) and biogeochemical processes, small lakes and ponds have received relatively little attention in terms of management. Moreover, they are highly susceptible to terrestrial influences, receiving significant amounts of particulate or dissolved carbon, which leads to water brownification (Kritzberg et al., 2014; Costa et al., 2024). Brownification involves reduced light penetration and has been linked to increased persistence of mixotrophic phytoplankton species over merely phototrophic ones (Wilken et al., 2018; Calderini et al., 2022). Light availability has been identified as a primary factor influencing biovolume and success of mixotrophs (Pålsson and Granéli, 2003; Princiotta et al., 2023; Costa et al., 2024). Factors like lake morphology, temperature, trophic interactions, and nutrient levels also contribute to habitat complexity and influence mixotrophs’ success (Jost et al., 2004; Wilken et al., 2013; Koppelle et al., 2024). Exploring the mechanisms of contribution of mixotrophs to total phytoplankton biomass can enhance our understanding on how their success is related to environmental variables. It can also help to reveal their biogeographic patterns and ecological preferences.

Using a large-scale sampling across 112 small, forested lakes and ponds, we investigated the contributions of potential mixotrophs and examined the relationship among their functional trait dominance (including osmotrophic pathways) and functional diversity in response to environmental factors. We expected that potential mixotrophs would exhibit differences in dominance between their autotrophic and heterotrophic nutrition (e.g. obligate osmotrophy and phagotrophy) and that their success would likely be influenced by light-limiting factors, shifting their trophic strategies. In other words, we expected a high dominance of osmotrophic flagellates in humic and browned forest lakes. Moreover, we evaluated whether there is a biogeographic distribution pattern of functional traits and diversity indices of mixotrophic organisms.

MATERIALS AND METHODS

Study area

The study was conducted in Hungary, Central Europe. A total of 112 small, forested lakes were sampled in two regions of the country: the Transdanubian Mountains and the North Hungarian Mountains (Fig. S1). The Transdanubian Mountains extend from the western edge of Lake Balaton to the Danube River, while the North Hungarian Mountains run alongside the Slovakian border (Kovács and Jakab, 2021). Hungary’s climate is continental, characterized by warm to hot summers and relatively cold, largely (especially in the past) snowy winters. The average annual air temperature is 9.7°C, with temperatures ranging from 0 to 15°C during spring. Annual precipitation averages around 600 mm (Kovács and Jakab, 2021). However, climate analysis indicates a significant decrease in annual precipitation throughout the 20^th century, particularly during the spring seasons (Kovács and Jakab, 2021).

Each lake and pond were sampled once during the spring in 2014 and 2018 (March–April). We used the approach space-for-time substitution, assuming that spatial and temporal variations are equivalent. Most of the sampled small forest lakes and ponds were characterized by shallow depth (mean around 30 cm) and small surface area. They frequently experience complete drying out during the summer months. The brown colouration of these lakes is attributed to the high content of humic materials deriving from allochthonous sources, primarily the abundant leaf litter from the surrounding deciduous forests (Al-Imari et al., 2023).

Sampling and analysis

Water samples were collected at each sampling site for chemical and phytoplankton analyses. Morphological variables such as surface area and depth of the lakes were recorded in the field. In situ measurements of water temperature (temp), pH, and conductivity (cond) were conducted using a Hach Lange HQD4 portable multimeter. In the laboratory, nutrient forms such as, nitrate (NO₃⁻), nitrite (NO₂⁻), ammonium (NH₄⁺), total phosphorous (TP), and soluble reactive phosphorous (SRP), moreover, the anions chloride (Cl⁻) and bicarbonate (HCO₃⁻); COD (chemical oxygen demand) and Pt color (Platinum-Cobalt scale, a proxy for water color discoloration) were determined. Titrimetric and spectrophotometric methods were used following international standards for the analyses (APHA, 1998; Wetzel and Likens, 2000; Cuthbert and del Giorgio, 1992). Dissolved inorganic nitrogen (DIN) was calculated as the sum of NO₃^-.-N, NO₂^–.-N, NH₄⁺-N.

The phytoplankton community was identified at the possible highest taxonomic accuracy, preferably to the species level. Quantitative analysis was conducted using a Zeiss Axiovert 100 inverted microscope, following the method outlined by (Utermöhl, 1958) by counting a minimum of 400 settling units (cells, colonies, and filaments). Algal biovolume (mm³ L⁻¹) was determined using approximate geometric models outlined by (Hillebrand et al., 1999). Phytoplankton species were grouped according to different ecological traits based on their size, shape, Reynolds functional groups (RFG; Reynolds et al., 2002; Padisák et al., 2009), movement, coloniality, trophic strategies and trophic spectrum (see details in Table I).

The groups of organisms capable of phagotrophy and/or obligate osmotrophy were considered as potential mixoplankton (some cryptophytes, dinoflagellates, chrysophytes, some flagellated chlorophytes and euglenophytes) according to the following literature: Hansson et al., 2019; Mitra et al., 2016; Stoecker et al., 2017; Schneider et al., 2020. The trophic strategies and trophic spectrum used in this study are described in more detail in Table II. A list of species with their respective trophic classification is found in Table S2.

Statistical analysis

Functional diversity measurements were evaluated in relationship with the environmental variables. A functional diversity index was calculated for each lake to quantify the diversity of traits within potential mixotrophic organisms using the taxonomic matrix (species by site) and a matrix of their functional traits (i.e. RFG, size, shape, coloniality, movement by flagella, trophic strategy and trophic spectrum). Different components of functional diversity were obtained: functional richness (FRic); functional evenness (FEve); functional dispersion (FDis) and Rao’s quadratic entropy (RaoQ). FRic measures the amount of niche space occupied by species in a community, higher values of FRic indicate higher diversity of functional traits within the community. FEve assesses how evenly functional traits are distributed within the community: higher values of FEve indicate a more even distribution of functional traits (Laliberté and Legendre, 2010). Functional dispersion quantifies the extent to which species in a community are functionally distinct from each other: higher values of FDis indicate higher dispersion or dissimilarity of functional traits among species. Finally, RaoQ quantifies the total pairwise functional heterogeneity within a community: higher values of RaoQ indicate higher functional heterogeneity or dissimilarity among species within the community (Laliberté and Legendre, 2010). Multivariate analyses (i.e. variance partitioning and redundancy analysis, RDA) were performed to relate the functional diversity index to the environmental variables by sampled region (i.e. North Hungarian mountains and Transdanubian Mountains) (Supplementary material Figs S2 and S3).

We performed a non-metric multidimensional scaling (NMDS) analysis to explore and visualize the composition of the functional traits and their relationships with environmental factors of importance. Regression tree analysis (RTA) was applied to explore the potential mixotroph relative biomass (MRB) in response to multiple exploratory variables. RTA is suited to the analysis of complex ecological data to deal with nonlinear relationships and hierarchical interactions among variables and it is easy to interpret (De’ath and Fabricius, 2000). Relative biovolume (MRB) of mixotrophs was estimated as the ratio of their biovolume (i.e. the sum of osmotrophy and phagotrophy) to the total phytoplankton biovolume, as the response variable. The predictor variables included in the model were lake depth, pH, conductivity, bicarbonate, chemical oxygen demand, chloride, DIN, SRP and water color. We selected environmental variables that, based on literature, can be considered predictors of mixotrophs and which displayed high correlation (Fig. S4). To account for multicollinearity among predictor variables, a selection process was applied to eliminate variables where a high variance inflation factor was detected (VIF > 8, Zuur et al., 2007). We considered the model that provided the best trade-off between simplicity and accuracy based on RMSE (root mean square error), which is the square root of the variance of the residuals and indicates the absolute fit of the model and the R² (R² = 1—relative error), representing the percentage of variation in the dataset explained by our model. Lower RMSE values and higher R² indicated good model fits. To build the tree we used two important parameters: the minimum number of splits (minsplit) and the complexity parameter (cp). To determine the optimal minsplit, a cross-validation based methodology was used. However, due to the relatively small size (n = 112), of our dataset leave-one-out cross-validation (LOOCV) was employed to determine the appropriate minsplit. A comprehensive list of these variables, along with the reasons for their inclusion as predictors of MRB in the analysis, can be found in Table S1.

All analyses were performed in R software, version 4.4.0 (R Core Team, 2024). Functional diversity metrics were calculated in the “FD” R package (Laliberté and Legendre, 2010) using the “dbFD” function. NMDS ordination was performed in “vegan” package using “metaMDS” function (Oksanen et al., 2022), based on Bray–Curtis dissimilarity index running 999 permutations. RTA (Breiman et al., 1984)) was calculated in the Rpart’ package with the function “rpart” (Therneau and Atkinson, 2022). Spearman correlation and VIF analyses were conducted using the stats package and the “usdm” package (Naimi et al., 2014), respectively. The map was created using “ggmap” and “ggplot” packages.

RESULTS

Mixotrophs and environmental variables

The studied lakes were typically small, with an average surface area of 579 m² and shallow, with a mean depth of 33 cm, ranging from 5 cm to 2 m. Water temperature exhibited significant variability during the sampled period (spring season) ranging from 4 to 24°C (Table III). The concentration of DIN, TP and SRP, chloride, conductivity and bicarbonate displayed high variability and, on average, high concentrations along with variables related to light availability and water color (COD and water discoloration, Pt color) (Table III).

A total of 203 phytoplankton taxa were identified in the samples of which 68 taxa were classified as potential mixotrophs. A high proportion of mixotrophic organisms was observed (Fig. 1A). The majority of mixotroph species (80%) belonged to nanophytoplankton size class (2–20 μm) and no mixotrophic species were found in macrophytoplankton size class (> 200 μm) (Fig. 1B). All mixotrophic species were flagellated, and the biovolume of colonial species exhibited high variability across the lakes (Fig. 1C).

The results of the multivariate analysis showed no significant relationships among FD and the factors analyzed in the study. The residuals of the VPA also demonstrated that other factors not analyzed in study could be explaining the mixotrophs FD (Supplementary material, Fig. S2). The RDA was not significant and the axis explained little variation, i.e. 7.5% for axis 1 and 2% for axis 2 (Supplementary material, Fig. S3). The value of functional diversity index on a 0–1 scale ranged from 0.0006 to 0.75, reflecting the distribution of functional traits within the mixotroph community in the studied lakes (Table III and Fig. 1D). Lower values of RaoQ indicates a lower diversity of trait combinations within the mixotroph assemblage. The mean value of FRic above 0.5 also suggests a low variety of traits present in the community (Fig. 1D). Moreover, FDis indicates that the mixotroph species were functionally similar, and FEve shows that mixotroph traits were well distributed within the community in these lakes (Fig. 1D).

The low stress value (0.11) associated with the NMDS indicated a good fit and well represented the ordination. The NMDS showed no apparent clustering of sampling sites (Fig. 2). Trait composition revealed that the size class of macrophytoplankton exhibited high dissimilarity, while microphytoplankton and photo-osmotrophic phytoplankton (autotrophy) were clustered (Fig. 2). The proportions of mixotrophs (i.e. obligate-osmotrophs and phagotrophs) was similar to that of nanophytoplankton (Fig. 2). Environmental variables related to water salinity (conductivity and chloride) and alkalinity (HCO₃⁻) displayed similarities. Variables related to nutrients (DIN and SRP) and water color (pt color and COD) also showed similar patterns and were grouped in relation to the nano-mixotrophic species (Fig. 2).

The RTA model selected COD, DIN, pH, HCO₃⁻, water color, and Cl⁻ as the main predictor of MRB across the studied lakes (Fig. 3) and explained 51% of the variance of the dataset (R² = 1—relative error, RMSE = 0.22). On average, MRB represented a high proportion of the total phytoplankton biovolume (75%), with COD being the main predictor of MRB across the forest lakes (Fig. 3, layer 1). In cases of COD < 5.6 mg O₂ L⁻¹, lower MRB values (0.31) (left side of the tree dendrogram) were found showing the importance of dissolved nitrogen concentration for MRB (Fig. 3, layer 2). Lakes with elevated DIN (left side of the tree dendrogram) were conditioned by pH (Fig. 3, layer 3), with higher pH leading to higher MRB (0.79). Light availability (represented by COD) determined the MRB when pH is lower than 7. Among lakes with relatively low values of COD (< 36 mg O₂ L⁻¹) and pH < 7, the lowest average of MRB was observed (0.27). Among lakes with COD > 36 mg O₂ L⁻¹, a higher proportion of MRB (0.77) was exhibited (Fig. 3).

Lakes with combination of relatively high COD and low DIN were conditioned by bicarbonate (HCO₃⁻), on the right side of the tree dendrogram (Fig. 3, layer 3). In other words, in alkaline waters, MRB was driven by water color. The subset of lakes with dark waters (n = 8) had half of the MRB values (0.47) compared to alkaline waters with water color < 201 mg L⁻¹ Pt (mean 118.3 mg L⁻¹ Pt), where MRB = 0.82 (Fig. 3, layer 4). Lastly, lakes with mean HCO₃⁻ ≤ 68 mg L⁻¹ exhibited, on average, high MRB values (0.80 to 0.94) according to chloride concentration (Fig. 3, layer 4).

Regarding the trophic strategy, we observed a shift in trophic strategy according to differences in light availability, nutrient concentration, and alkalinity or acidity of the lakes. High biovolume of photo-osmotrophic phytoplankton (strictly autotrophs) can be observed on the left side of the RTA dendrogram, where COD values are the lowest. On the other hand, the proportion of mixotrophs, represented by obligate-osmotrophs and phagotrophs, is higher in lakes with high organic matter, and MRB is mediated by DIN and secondarily by the ions in the water. Osmotrophs represent, on average, the highest biovolume values among mixotrophs (light purple in pie chart, Fig. 3). However, there is an observed increase in the biovolume of potential phagotrophs in lakes with high COD and DIN, as well as in lakes with low HCO₃⁻ and high chloride (Fig. 3).

DISCUSSION

Despite a great deal of recent interest in phytoplankton mixotrophy, there are several open questions regarding the mechanisms and ecological relevance of mixotrophy in aquatic ecosystems that still need to be addressed. We investigated the contribution of the potential mixotroph functional traits and functional diversity indices in response to environmental factors in humic small lakes and ponds and found that environmental variables related to water color and dissolved inorganic nitrogen concentration supported dominance and success of potential mixotrophs. Moreover, aside from the classification challenges, we intended to incorporate obligate-osmotrophy in phytoplankton as an important trophic trait into the studies of mixotrophy, especially in light-limited habitats such as humic lakes. This study revealed an interesting pattern in the small forested lakes: high input of allochthonous organic material influenced more the mixotrophs than the strictly autotrophs, suggesting that in these habitats mixotrophy is a metabolic trait that provides an adaptive advantage, making mixotrophs stronger competitors. However, the presence and biomass of potential mixotrophs were homogeneously distributed across the studied lakes leading to a low functional diversity index and suggesting that the mixotrophic species exhibited similar ecological functions, possibly a case of functional redundancy (Várbíró et al., 2017). Such a uniform ecological structure could potentially affect the ability of these communities to respond to environmental variation and maintain ecosystem stability.

Osmotrophy is possible in all phytoplankton species, however to differing extents (Nedoma et al., 2003). Thus, classifying and quantifying osmotrophic activity has remained a great challenge and is one of the reasons why the majority of the mixotrophic studies only considers phagotrophic organisms. Though osmotrophy is often neglected some species of phytoplankton rely on obligate osmotrophy, indicating their dependence on absorbing dissolved organic matter for nutritional requirements and growth (Table II, i.e. obligate-osmotrophy). These species are often used in ecological status assessment as for example, Fistulifera (formerly Navicula) saprophyla (Van Dam et al., 1994) and euglenids that can indicate lakes favored by waterfowl (Padisák, 1993). Recognizing the importance of obligate osmotrophy as a metabolic trait for categorizing mixotrophy and elucidating its role in aquatic ecosystems, particularly in light-limited humic browned lakes, we employed a rather broad and comprehensive definition to classify the types of mixotrophy, incorporating obligate-osmotrophy into the category of mixotrophs. Mixotrophic organisms thrive in various environments and despite costs they benefit from synergies between photosynthesis and nutrient acquisition. Model predictions show that high light levels favor mixotrophy due to enhanced carbon assimilation, while increased nutrients strengthen their competitive edge against heterotrophs (Edwards, 2019). However, in our study we observed a high proportion of MRB over strict autotrophs suggesting a strong adaptive response through environmental filters.

Cell size and shape are considered main traits in phytoplankton ecology (Naselli-Flores et al., 2007). Recent models suggest that small organisms are mostly autotrophic due to their high surface to volume ratio that facilitates nutrient acquisition (Leles et al., 2021). Conversely, large organisms tend towards heterotrophy, prioritizing nutrient acquisition through phagotrophy over diffuse absorption of organic carbon sources (e.g. Nedoma et al., 2003). Finally, intermediate-sized organisms exhibit a range of phago-mixotrophic behaviors, balancing phototrophic affinity for surface area with phagotrophic affinity for cell volume (Leles et al., 2021; Millette et al., 2023). Since these findings originate from theoretical modeling, further fieldwork and laboratory investigations are necessary to confirm whether mixotrophy indeed maintains a functional association with cell size. On the other hand, it has been proposed that a mixotrophic cell should be larger to support two nutritional systems, but to date, there is no supporting evidence in the literature (Mitra et al., 2024). Our in-situ findings confirm this suggestion showing nanoplankton as the main representatives of mixotrophic phytoflagellates (both obligate-osmotrophic and phagotrophic organisms), thus suggesting that the energetic costs for this trophic spectrum exceed those of autotrophs and that the energetic demands of these organisms probably come from their prey or are absorbed obligatorily from the environment. Additionally, motility is common trait among plankton (by flagella or buoyance, for instance) and is crucial for various functions including nutrient acquisition, avoiding predators and finding optimal light conditions for photosynthesis (Litchman and Klausmeier, 2008). Trait trade-offs faced by mixotrophs versus autotrophs include the necessity of motility for predatory activities. This requirement potentially increases encounters with prey but also incurs some energetic costs, although perhaps not very high costs (Sommer, 1988). Recent research does not support this argument, as motility is widespread among plankton and can be facilitated by self-propelled flagella or a combination of buoyancy and turbulence (Naselli-Flores et al., 2021; Naselli-Flores and Padisák, 2024). However, we found a strong relation between flagella and buoyancy (colonial species), suggesting that in these small lakes, both morphological traits are fundamental for mixotroph success.

Low functional diversity within the mixotrophic community indicates that the species present in the community exhibit similar functional traits. Limited variation in the functional characteristics or strategies among the mixotrophic species, could lead to functional redundancy and potentially affect ecosystem resilience (Várbíró et al., 2017). These findings hold significant importance due to the prevalence of small lakes and ponds globally, which are acknowledged as vital sources of biodiversity and ecosystem services (Borthagaray et al., 2023; Cuenca-Cambronero et al., 2023), providing breeding habitats for amphibians, drinking water for birds and mammals, among others. Given the importance of small lakes and ponds, there is a compelling need for their conservation, highlighting their critical role in maintaining ecological integrity (Cuenca-Cambronero et al., 2023).

Chemical Oxygen Demand (COD) was the primary descriptive variable in the RTA, meaning a positive relationship with MRB, and representing the lower amount of oxygen required to chemically oxidize organic matter in the water. COD levels in browned forested lakes can be interconnected and influenced by various factors associated with the habitat characteristics (Aoki et al., 2004). The brown colouration in these lakes caused by dissolved organic matter (DOM) derived from leaf-litter or soil runoff and indicates the presence of organic compounds that are susceptible to oxidation (Pace and Cole, 2002) potentially leading to elevated COD levels. As a result, COD can serve as a proxy for light availability in these aquatic habitats. This is because in highly productive ecosystems, light availability may influence the rate of photosynthesis, thereby affecting organic matter production and decomposition rates. In turn, this can impact COD levels, as higher rates of photosynthesis may lead to higher oxygen production and lower COD values, as we can observe in subsets of lakes with higher proportion of autotrophy (Fig. 3). On the other hand, in ecosystems with limited light penetration due to factors such as high turbidity or shading from vegetation, photosynthesis may be reduced, potentially leading to lower rates of organic matter production and slower decomposition rates. In such cases, COD levels may be influenced more by the input of allochthonous organic matter (e.g. from terrestrial sources) than by autochthonous production.

Nutrient availability represented by dissolved inorganic nitrogen (DIN), was the second most important variable for MRB in our study. We observed two contrasting situations: the last node of the tree (Fig. 3) shows that lakes with high productivity (DIN > 122 μM) and intensive brown colouration (highest COD values) exhibited high mixotrophic dominance (MRB = 94%), suggesting that mixotrophy is advantageous to obtain organic molecules for growth in low light environments. This has been also observed previously in laboratory (Costa et al., 2022), in in situ experiments (Princiotta et al., 2023), and in natural environments (Saad et al., 2016; Costa et al., 2019). On the other hand, a high proportion of mixotrophy (MRB between 80–94%) was found where DIN was relatively low (< 34 μM). Mixotrophic organisms can have reduced energetic costs associated with nitrogen assimilation compared to autotrophic organisms, achieving this directly from prey in organic forms such as amino acids and nucleic acids, instead of relying solely on DIN (Skovgaard, 2000). Additionally, they recycle ammonium released during prey assimilation, further reducing costs. This efficiency results in a 20% reduction in photoreductant cost for nitrate assimilation (Flynn and Hipkin, 1999).

Bicarbonate and pH also affected MRB (biovolume dominance) in the RTA. Generally, in aquatic environments, most dissolved inorganic C is bicarbonate, while CO₂ is the substrate for RuBisCO (Beardall and Raven, 2016). Mixotrophic organisms can use bicarbonate through carbon concentrating mechanisms to convert it into CO₂ for use in photosynthesis if they possess the necessary biochemical apparatus (the enzyme: carbon-anhydrase). Some groups, like chrysoflagellates, in theory, do not have such a biochemical pathway (Maberly et al., 2009) though field studies present contradictory observations (Reynolds et al., 1993). The ability to use bicarbonate expands the potential carbon sources available to mixotrophic organisms, enhancing their ecological flexibility and allowing them to thrive in diverse environmental conditions. Additionally, the pH of water can impact the activity of enzymes engaged in both photosynthesis and organic matter digestion. During autotrophy, pH tends to increase, whereas during heterotrophy, pH typically decreases. Moreover, extreme changes in pH can have serious deleterious effects (Hansen, 2002). For instance, at low pH, the functioning of the respiratory electron transport system is limited (G.-Tóth, 1993). In the browned lakes, which often contain high levels of dissolved organic compounds, pH may fluctuate due to the activity of microbial DOM decomposition (Bastviken et al., 2004). However, bicarbonate buffering can help mitigate these fluctuations, maintaining a relatively stable pH, which may allow mixotrophy to thrive. Recent studies indicate that even under neutral hardwater conditions, browning can overturn cyanobacterial dominance and biomass, favoring mixotrophic cryptophytes instead (Lyche Solheim et al., 2024). Therefore, the pathways to acquire carbon may differ within mixotroph trophic strategies, highlighting the need to separate obligate-osmotrophy and phagotrophy mixotrophic nutritional modes.

A lot of variation remained unexplained in the NMDS, indicating unidentified but influential variables not considered in this analysis. Potential candidate variables include those related to herbivory, which is particularly pertinent in these small lakes. Our small forested lakes are generally fishless and thus there is likely considerable zooplankton grazing pressure on the phytoplankton (Scheffer et al., 2006). Furthermore, some mixotrophic organisms, such as cryptophytes, are recognized for their high nutritional value to zooplankton (Peltomaa et al., 2017). However, the nutritional quality of mixotrophs is related to several other factors, such as biogeochemical and stoichiometric profiles, and may vary according to the composition of the mixotroph assemblage (Millette et al., 2023). Moreover, environmental and demographic stochasticity could be influencing phytoplankton trait patterns and community structure (Seltmann et al., 2019). Our study highlights osmotrophy as an important trait in mixotrophy dominance in humic and browned lakes. Future research on species physiology and osmotrophic pathways are needed to better elucidate the role and classification of obligately osmotrophic organisms.

CONCLUSION

This study indicates that mixotrophy might be a crucial metabolic trait in small forested lakes, providing an adaptive advantage to potential mixotrophs (including obligate osmotrophs) over strictly photosynthetic organisms. Thus, the inclusion of osmotrophy within mixotrophy studies appears to be promising, especially in light-limited aquatic environments with high organic matter content. Further field and laboratory studies are needed to understand patterns and to classify species with greater confidence. We found that, mixotrophic species were homogeneously distributed across our 112 study lakes, exhibiting similar functional traits, possibly indicating high functional redundancy.

ACKNOWLEDGEMENTS

We acknowledge the contribution of Edit Király and Rita Zsuga-Bíró for their technical assistance in the field-, and laboratory- and microscopy analyses.

AUTHOR CONTRIBUTIONS

MRA Costa and J. Padisák conceptualized and designed the study. E. Lengyel, and GB Selmeczy conducted/supervised the fieldwork and laboratory analysis for water physical and chemical parameters. E. Lengyel, provided the map figure and assisted in the statistical analyses. MRA Costa performed the statistical analysis and wrote first draft of the manuscript. All authors contributed to the interpretation of the results and to the writing and editing of the manuscript and gave final approval for publication.

FUNDING

This study was financed by the Hungarian National Research, Development and Innovation Office, KKP 144068 (JP, EL, GBS) and K 137950 (EL); National Laboratory for Water Science and Water Safety program (RRF-2.3.1-21-2022-00008).

DATA AVAILABILITY

The authors do not have permission to share data.

References