Chytrids enhance Daphnia fitness by selectively retained chytrid‐synthesised stearidonic acid and conversion of short‐chain to long‐chain polyunsaturated fatty acids

Abstract Chytrid fungal parasites convert dietary energy and essential dietary molecules, such as long‐chain (LC) polyunsaturated fatty acids (PUFA), from inedible algal/cyanobacteria hosts into edible zoospores. How the improved biochemical PUFA composition of chytrid‐infected diet may extend to zooplankton, linking diet quality to consumer fitness, remains unexplored. Here, we assessed the trophic role of chytrids in supporting dietary energy and PUFA requirements of the crustacean zooplankton Daphnia, when feeding on the filamentous cyanobacterium Planktothrix. Only Daphnia feeding on chytrid‐infected Planktothrix reproduced successfully and had significantly higher survival and growth rates compared with Daphnia feeding on the sole Planktothrix diet. While the presence of chytrids resulted in a two‐fold increase of carbon ingested by Daphnia, carbon assimilation increased by a factor of four, clearly indicating enhanced carbon transfer efficiency with chytrid presence. Bulk carbon (δ 13C) and nitrogen (δ 15N) stable isotopes did not indicate any treatment‐specific dietary effects on Daphnia, nor differences in trophic position among diet sources and the consumer. Compound‐specific carbon isotopes of fatty acids (δ 13CFA), however, revealed that chytrids bioconverted short‐chain to LC‐PUFA, making it available for Daphnia. Chytrids synthesised the ω‐3 PUFA stearidonic acid de novo, which was selectively retained by Daphnia. Values of δ 13CFA demonstrated that Daphnia also bioconverted short‐chain to LC‐PUFA. We provide isotopic evidence that chytrids improved the dietary provision of LC‐PUFA for Daphnia and enhanced their fitness. We argue for the existence of a positive feedback loop between enhanced Daphnia growth and herbivory in response to chytrid‐mediated improved diet quality. Chytrids upgrade carbon from the primary producer and facilitate energy and PUFA transfer to primary consumers, potentially also benefitting upper trophic levels of pelagic food webs.


| INTRODUC TI ON
The efficiency of dietary energy transfer across the phytoplanktonzooplankton interface is crucial for upper consumers of aquatic food webs. Traditionally, phytoplankton was considered the sole functional group supplying the dietary requirements of zooplankton (Leibold, 1989) and subsequent consumers. Zooplankton growth can be limited by multiple factors such as toxic metabolites (Ger et al., 2016), but primarily by phytoplankton biomass and biochemical dietary composition (Müller-Navarra, 2008). The elemental composition of phytoplankton constrains zooplankton growth and reproduction in particular in case of low phosphorus concentration (Elser et al., 2001). In addition, dietary lipids and their fatty acids, especially long-chain (LC) polyunsaturated fatty acids (PUFA) can limit zooplankton fitness Müller-Navarra et al., 2000).
Reduced zooplankton fitness in response to the lack of dietary LC-PUFA (Gulati & Demott, 1997;Ruess & Müller-Navarra, 2019) may further constrain dietary energy and essential molecule transfer towards consumers at higher trophic levels. Accordingly, a better understanding of mechanisms that enhance PUFA transfer across the phytoplankton-zooplankton interface, when phytoplankton is limited in LC-PUFA, is important.
Cyanobacteria, such as Planktothrix agardhii, contain no LC-PUFA . Accordingly, cyanobacteria blooms can limit zooplankton growth and reproduction (Gulati & Demott, 1997;Martin-Creuzburg et al., 2009), creating an energetic bottleneck in the grazing food chain (Havens, 2008). Diet palatability can also constrain PUFA accessibility for zooplankton. Large-sized algae or cyanobacteria, irrespective of their dietary PUFA provision, are largely inaccessible for selective (Arndt, 1993) and non-selective filter-feeders (Bern, 1994;Carpenter et al., 1993). Consequently, the dominance of inedible cyanobacteria (i.e., filamentous and colonial forms) can limit zooplankton growth due to both size constraints and poor diet quality, and alternative trophic pathways become crucial.
Chytrid fungal parasites constitute a key intermediate-level trophic group between phytoplankton and zooplankton (Rasconi et al., 2014). The free-swimming chytrid zoospores are edible (<5 μm) and constitute an alternative trophic link between large inedible phytoplankton and zooplankton, known as mycoloop (Kagami et al., 2007(Kagami et al., , 2014. For example, during the presence of Asterionella, an inedible diatom rich in LC-PUFA, chytrids enhanced the somatic growth of Daphnia (Kagami et al., 2007). Similarly, chytrids infecting Planktothrix, an inedible cyanobacterium with poor nutritional quality, improved the growth of Keratella (Frenken et al., 2018) and Daphnia (Agha et al., 2016). Chytrids can improve the dietary provision for zooplankton in two non-exclusive ways. First, they fragment inedible the host into smaller, edible particles (Frenken et al., 2020;Gerphagnon et al., 2013). Second, they can enhance the biochemical dietary quality compared with the uninfected hosts Taube et al., 2019). Specifically, chytrids perform trophic upgrading by converting short-chain PUFA from inedible hosts to LC-PUFA in the edible zoospores Rasconi et al., 2020;Taube et al., 2019). Chytrids also upgrade the biochemical composition of their hosts by de novo synthesised sterols, enhancing dietary energy fluxes in pelagic food webs . Experiments suggest that: (1) the mycoloop constitutes a quantitatively important energy pathway at the phytoplanktonzooplankton interface Rasconi et al., 2020;Taube et al., 2019); and (2) chytrids enhance zooplankton consumers' fitness (Agha et al., 2016;Frenken et al., 2018). Whether and how chytrid-mediated carbon upgrading extends to zooplankton, providing a mechanistic link between improved diet quality and consumer performance, remains unexplored.
Here we aim to disentangle the chytrid-mediated diet quality effect, quantified by PUFA, from the increased diet quantity effect on zooplankton fitness. Using Daphnia magna as a herbivorous model organism, we performed a feeding experiment to compare the diets consisting of either the filamentous cyanobacterium Planktothrix rubescens, or Planktothrix infected with the obligate chytrid parasite Rhizophydium megarrhizum (Sønstebø & Rohrlack, 2011). We tested the hypothesis that PUFA-induced diet quality enhancement via chytrids would improve Daphnia fitness, independent of the positive effect of improved diet palatability. We expected that the diet consumers, potentially also benefitting upper trophic levels of pelagic food webs.

K E Y W O R D S
carbon transfer efficiency, fatty acid retention, mycoloop, PUFA, trophic upgrading including chytrid parasites would enhance the consumer survival, growth and reproduction via zoospore consumption. Our hypothesis is based on the fact that PUFA assimilated from chytrid zoospores are among the most critical dietary biomolecules transferred across the phytoplankton-zooplankton interface, whereas their lack impedes carbon assimilation of the consumer (Taipale et al., 2014).
The amount of assimilated rather than ingested carbon is the key driver of Daphnia fitness, which depends largely on its molecular composition, e.g. LC-PUFA availability. The amount of assimilated carbon is the most critical in case of limiting dietary carbon availability (Khattak et al., 2018), a case highly relevant for inedible poor diet quality phytoplankton, such as the filamentous cyanobacterium Planktothrix .
To quantify the effect of improved dietary PUFA composition via the chytrid parasite, we investigated diet ingestion and assimilation of carbon. To identify PUFA bioconversion pathways among Planktothrix, the chytrid parasite, and Daphnia, we used compoundspecific stable carbon (δ 13 C) isotope analysis (see Twining et al., 2020).
We expected Daphnia to retain specific PUFA from the chytrid diet, and thus showing improved fitness compared to Daphnia feeding on the sole cyanobacterium.
The cultures were maintained on WC Medium (Guillard & Lorenzen, 1972) in VWR cell culture flasks in a non-axenic environment at 21°C. We applied 10.9 μmol m 2 s −1 photosynthetically active radiation and a 16:8 light: dark cycle using two AquaLytic incubators (180 L, Liebherr, Germany). Prior to the experiment, the carbon content of uninfected and chytrid-infected (>50% of prevalence, i.e. % of infected filaments) Planktothrix cultures were analysed to approximate a minimum of dietary c. 0.6 mg C/L during the experiment (Lampert, 1978). The high chytrid prevalence of Planktothrix was reached by weekly 1/3 V/V% medium exchange and 1/3 V/V% dense Planktothrix culture exchange. The stock cultures of D. magna grew in pre-filtered (0.7 μm GF/F) water from Lake Lunz diluted with 10 V/V% of ADaM medium (Klüttgen et al., 1994) for multiple generations before the experiment. Daphnia magna was fed >1 mg C/L with Scenedesmus sp. and Chlamydomonas sp. (50%-50%), both grown on WC medium.

| Experimental design
Two diet systems were compared: (1) Planktothrix (cyanobacterium treatment); and (2) chytrid-infected Planktothrix (kept >50% prevalence of infection), containing the cyanobacterium, the sessile sporangia of the chytrid parasite, and free swimming chytrid zoospores (cyanobacterium-parasite-zoospore treatment). Diet treatments were run in five replicates in 1-L Corning® Square Storage plastic bottles at 18°C. Each bottle contained 22 Daphnia neonates, all born within 48 hr. The bottles were filled up to 1 L volume (600 ml pre-filtered LUS water + 400 ml diet containing c. 0.4 ± 0.1 μg C/L) and loosely covered with opaque plastic plates allowing air exchange. We applied a 16:8 light: dark cycle with 1.4 μmol m 2 s −1 photosynthetically active radiation during the light phase. Food was replaced and the residuals quantified every other day to all treatments. Daphnia were checked daily for survival and egg production.
An individual was considered dead if it was lying at the bottom of the bottle with no movement for 5 min. Dead individuals were removed and kept frozen at −20°C. The experiment was stopped when the first Daphnia neonates were observed (i.e., after 14 days).

| Life history of Daphnia
Treatment-specific difference in Daphnia growth rate was quantified Daphnia body length was directly measured at the end of the experiment on all survival individuals according to Bottrell et al. (1976) using the ocular micrometer of a Nikon SMZ 745 T binocular microscope.

| Quantitative data on ingested diet
The fresh-weight (FW) biomass of ingested diet was calculated as the difference between the diet concentration at time of feeding (hereafter diet) and after 48 hr (hereafter residual) in triplicates.
Planktothrix growth was assumed to be negligible due to the low light and water temperature compared with the culture conditions, and, therefore, not considered. Volumetric concentration of each diet source (i.e., Planktothrix, the chytrid-infected Planktothrix, and chytrid zoospores) was quantified by microscopic analysis from formaldehyde-fixed samples (4% final concentration). Cell density estimation from settled 3-ml samples using a Leica DMI 3000B inverted microscope according to Utermöhl (1958) and Lund et al. (1958), with counting in at least 2 transects. Cell biovolume were approximated based on simple geometrical forms: GR = ln DW end − ln DW start ∕d, chytrid zoospores were considered spherical (diameter of 4 μm, corresponding to 33.5 μm 3 /ind), while Planktothrix cells cylindrical (diameter of 3 μm, cell length of 5 μm, corresponding to 58.9 μm 3 / cell). The biovolume of Planktothrix filaments was calculated by multiplying the cell biovolume with the average number of cells per filaments per ml, estimated based on 20 random filaments in each sample. Ingested diet was expressed as total FW biomass assuming a density of 1. Filament lengths of Planktothrix was compared between the diet treatments, and also between the diets and residuals at each feeding occasion (after 48 hr) based on 30 (in case of very low densities, at least 20) random filaments. We did not include sporangia maturation in the calculation of diet ingestion for two reasons: (1) due to the high prevalence rate sessile chytrid sporangia were mainly in the size of zoospores; (2) it was technically impossible to separate the sporangia and the chytridinfected cyanobacterium filaments to measure their biochemical profiles separately.
Ingested Planktothrix and chytrid zoospores were converted into carbon based on gravimetrically measured DW (KERN, ABT 220-5DM) of the seston provided, the C% of diet sources and the respective % of FW biomass ingested. We also calculated ingestion efficiency for each diet source based on the ratio of the ingested and provided FW biomass (i.e. the dietary biomass % ingested).
Bacteria were also quantified in the diet sources and their residuals to estimate bacterial carbon ingestion using a Cytoflex Bacterial abundance was converted to bacteria carbon by assuming 20 fg C/cell (Agha et al., 2016;Ducklow & Carlson, 1992).

| Elemental C, N, P, and stable C and N isotope analyses
The elemental C and N contents as well as the stable isotopes of carbon (δ 13 C) and nitrogen (δ 15 N) of diet sources were analysed during three occasions in triplicates (n = 9): before the experiment, immediately after the experiment, and 1 week after the end of the experiment. Previous data on the same cultures showed that the sole Planktothrix varied less in C%, N%, and lipid content compared with the chytrid zoospores (Rasconi et al., 2020). In a preliminary experiment, a large variation in C% and N% of the chytrid-infected Planktothrix and chytrid zoospores (Appendix S1) was also observed.
Here the sole Planktothrix diet was analysed only once in triplicates (n = 3) right before the experiment. Daphnia were analysed for C, N, and P in triplicates, both at the beginning (Daphnia neonates; time point zero, T 0 ) and at the end of the experiment (time point end, T E ).
To separate the chytrid zoospores from Planktothrix filaments, c. 100 ml culture was filtered five times through a 20μm mesh followed by a single filtration step through a 5μm mesh. Absence of Planktothrix in chytrid zoospore samples was evaluated using a Nikon Eclipse TS100 inverted microscope at 200× magnification.
Individual diet sources were collected on muffled and pre- To rule out a potential dietary effect of P over PUFA in Daphnia, we quantified POP in Daphnia at T 0 and T E based on the ascorbic acid colorimetric method following persulfate digestion (American Public Health Association, 1998). P was not measured in diet sources during the experiment, but P was never limiting in culture conditions (i.e. C:P < 50 in all diet sources, independent measure of the experiment, data not shown). Elemental ratios expressed as molar ratios were C:N for diet sources, and C:N, C:P and N:P for Daphnia.

| Fatty acid and compound-specific carbon isotope analyses
The fatty acid composition and their compound-specific carbon isotopes (δ 13 C FA ) were analysed in the diet treatments in line with the elemental composition. Individual diet sources (c. 400 ml of culture, c. 1 mg seston; DW) were collected on muffled and pre-weighed GF/F Whatman™ filters (47 mm, 0.7 μm pore size). Daphnia were analysed for FA and δ 13 C FA profiles at the beginning (T 0 ) and at the end (T E ) of the experiment in triplicates. Daphnia were freeze-dried (see above) and c. 1 mg DW was transferred into tin caps.
Lipids were extracted from the freeze-dried, homogenised samples using chloroform-methanol mix (2:1), following the protocol described in Heissenberger et al. (2010). FA were derivatised to methyl esters by incubation with 1% H 2 SO 4 in methanol at 50°C for 16 hr. FA methyl esters (FAMEs) were dried under N 2 and dissolved in hexane. For quantification via flame ionisation, FAMEs were separated using a GC (Trace™ 1,310 Thermo Specific, Italy) equipped with a Supelco™ SP-2560 column (100 m × 0.25 mm × 0.2 μm). 13 C isotope analysis was performed using a Thermo Trace 1,310 GC (ThermoFisher Scientific), connected via a ConFlo IV (Thermo Co.) to an isotope ratio mass spectrometer (DELTA V Advantage, Thermo Co.). FAMEs were separated using a VF-WAXms 60-m column, 0.25 mm ID, film thickness 0.25 μm following oxidation to CO 2 in a combustion reactor, filled with Ni, Pt, and Cu wires, at a temperature of 1,000°C. The temperature started at 80°C, which was kept for 2 min, after which the temperature was raised by 30°C/min to 175°C, by 5°C/min to 200°C and finally by 2.4°C/min to 250°C, which was maintained for 30 min. Samples were run against certified Me-C20:0 standards (USGS70: δ 13 C = −30.53‰, USGS71: δ 13 C = −10.5‰, and USGS72: δ 13 C = −1.54‰), which were used for drift and linear correction. All peaks were validated and corrected manually if necessary.

| Carbon transfer efficiency
Carbon transfer efficiency was calculated as the ratio of carbon accrual by Daphnia (i.e., the total weight gained in carbon) to the total ingested carbon (Müller-Navarra et al., 2000), and expressed as %: where DDWC is the average DW of Daphnia expressed in carbon in each replicate at the end (T E ) and start of the experiment (T 0 ), FDWC TOT is the total DW of ingested diet in carbon. Total ingested carbon was calculated from ingestion rates, standardised by the number of surviving Daphnia individuals at each feeding occasion (i.e. accounting for the change in abundances per replicate). Compoundspecific 13 C PUFA pathways. We quantified each δ 13 C PUFA value to follow dietary PUFA sources for Daphnia by calculating the difference in δ 13 C of each FA between the consumer (δ 13 C C ) and its respective diet sources (δ 13 C D ), i.e. Δ 13 C FA , following Chiapella et al. (2021): In a similar way, we calculated Δ 13 C FA differences within Daphnia between LIN ➔ ARA, ALA ➔ SDA, ALA ➔ EPA, and SDA ➔ EPA conversion pathways.

| Data analyses
Treatment-specific differences in Daphnia growth rates, egg production rates, survival, and carbon transfer efficiency were tested with one-way ANOVA, or Wilcoxon in the case of unequal variance.
Ingested diet among sources, as well as the C:N, C:P and N:P ratios
F I G U R E 1 (a) Survival of Daphnia (in %) during the experiment; (b) somatic growth rates of Daphnia based on dry weight; (c) egg production rates of Daphnia. D_CY: Daphnia feeding on the cyanobacterium (Planktothrix), D_CPZ: Daphnia feeding on the chytrid-infected cyanobacterium (Planktothrix, chytrid parasites and chytrid zoospores). Letters indicate the level of significance between diet treatments based on ANOVA (n = 5, at p < 0.05).

| Filament length of Planktothrix
The length of Planktothrix did not differ significantly between uninfected (range: 30-1,000 μm, median: 530 μm) and infected The prevalence of chytrids in the chytrid-infected Planktothrix diet stayed >50% during the experiment, and it decreased linearly with time in the residual of diet collected after 48 hr (Appendix S1).
The difference in the prevalence of chytrids between the diet provided and its residual increased over time, i.e. a decreasing negative trend in ∆ (LM, R 2 Adj = 0.715; p < 0.001, Appendix S1).

| Ingestion rates
The ingestion rate of FW diets per Daphnia differed significantly among diet sources (Figure 2a).

| Stoichiometry of diet sources and Daphnia
Chytrid zoospores had significantly higher C content than the chytrid-infected Planktothrix, while C in chytrid-infected and un- In Daphnia, the C, N, and P contents did not differ significantly among Daphnia at T 0 and T E , nor at T E between diet treatments (Appendix S1, multiple comparisons of means, Tukey, p > 0.05, in all cases). The C:N ratio did not differ significantly between

| Bulk δ 13 C and δ 15 N values of diet sources and Daphnia
The δ
The fatty acid composition of diet sources, weighted by their respective DW ingested, explained 99.1% variance in the FA profile

| Compound specific carbon isotopes of PUFA in Daphnia
The δ 13 C LIN values differed significantly among diets and Daphnia The Δ 13 C discrimination factor of individual PUFA indicated that Daphnia feeding on Planktothrix were significantly depleted in δ 13 C LIN (−7.8‰), while slightly enriched in δ 13 C ALA (+0.2‰). Daphnia feeding on the chytrid-infected Planktothrix were enriched in both δ 13 C LIN (+6.9‰) and δ 13 C ALA (+2.8‰) , although the difference was only significant for LIN. Comparing Daphnia feeding on chytrid zoospores, the consumer was slightly enriched in all δ 13 C LIN (+1.2‰), δ 13 C ALA (+0.8‰) and δ 13 C SDA (+1.2‰) from the zoospores, but the differences were not significant.
Since SDA was missing in the Planktothrix diet, as well as ARA and EPA in Daphnia neonates, we calculated the Δ 13 C discrimination factor to compare δ 13 C of SDA, ARA, AND EPA with δ 13 C of their precursors: LIN, ALA, and SDA (Table 1). Δ 13 C ARA indicated slightly depleted δ 13 C in the consumer compared with its precursor LIN, while the difference was not significant for either diet treatments.
Consumers' SDA was enriched with δ 13 C compared to δ 13 C ALA in both diet treatments, while it was only significant in Daphnia feeding on Planktothrix. Daphnia feeding on the Planktothrix diet was largely depleted in δ 13 C EPA compared to δ 13 C SDA , while compared to both δ 13 C ALA and δ 13 C SDA in the chytrid-infected diet treatment.

| DISCUSS ION
The results supported our hypothesis that chytrids, as an Chytrids ensuring Daphnia reproduction under a cyanobacterium bloom (Planktothrix) have already been suggested experimentally. Agha et al. (2016) showed that c. 2,000-2,500 chytrid zoospores per ml enhanced Daphnia fitness and reproduction. Our chytrid zoospore density was considerably lower (c. 600 ind/ml) but could already enhance Daphnia fitness substantially. This may also have direct relevance to observational data where natural chytrid zoospore density can reach c. 3,000 ind/ml (see Rasconi et al., 2012;Sime-Ngando, 2012). Sustained cladoceran growth has been evidenced during cyanobacteria blooms (Davis et al., 2012;Moustaka-Gouni et al., 2006), suggesting the existence and importance of alternative energy pathways. As trophic transfer efficiency between phytoplankton and zooplankton decreases with cyanobacteria dominance (McCauley & Kalff, 1981;Selmeczy et al., 2019) and also the dominance of other large-sized colonial and filamentous algae (Kagami et al., 2007(Kagami et al., , 2014, our results also support the mycoloop hypothesis of parasitic chytrids providing the energetic needs of zooplankton for survival, growth, and reproduction.

| Qualitative trophic transfer
Dietary phosphorus could have potentially driven the treatmentspecific differences in life history traits of Daphnia. Enhanced Daphnia fitness, however, may not be explained by the mere elemental composition of the diet sources since Daphnia did not differ significantly in their respective C:P, N:P, and C:N ratios between diet treatments. Furthermore, the C:P ratio of Daphnia remain wellbelow the potentially limiting C:P c. 155 (Khattak et al., 2018), further emphasising that Daphnia were not P limited in any of the diet treatments. The N:P ratio of chytrids is expected to follow the ratio of the host, which affects zoospore production (Frenken et al., 2017).
We did not measure dietary N:P, but lower N:P in Daphnia feeding on the chytrid-infected diet might indicate higher P availability relative to N, yet no significant differences in N availability between diet treatments. The C:N ratios reported for diet sources (c. 5) vary little and are similar to those reported by Frenken et al. (2017). In line with Barranco et al. (2020), chytrids might have upgraded diet quality in terms of N, resulting in significantly lower C:N ratios in chytrid zoospores compared with the host. However, lowering C:N did not extend further to Daphnia, suggesting that Daphnia adjusted its C:N ratio irrespective of dietary supply. Organic matter leakage linked to chytrid infection and related higher bacterial abundance could also lead to lower C:N ratio and so diet quality increase for zooplankton (Barranco et al., 2020). Our results confirmed a higher bacterial abundance linked to chytrid-infection (Agha et al., 2016;Klawonn et al., 2021;Senga et al., 2018). Ingested bacterial carbon, however, remained only a minor fraction of total dietary carbon.
The fact that the Daphnia fatty acid profile did change irrespective of the diet treatments is in line with the observation that ingested bacterial carbon does not provide dietary energy for growth and certainly not for reproduction (Taipale et al., 2012). Using a similar Planktothrix-chytrid-Daphnia experimental setup, Agha et al. (2016) showed that bacterial diet reduced somatic growth, fecundity and offspring size compared with the sole Planktothrix and chytridinfected Planktothrix diets. Accordingly, functional carbon that affected Daphnia fitness may have been originated from the cyanobacteria and chytrids, while the diet response observed did not depend on bacteria. Furthermore, the significant C:N decrease in Daphnia compared with the neonates may suggest that dietary carbon rather than nitrogen availability was constrained, especially for Daphnia feeding on the sole Planktothrix diet (i.e. C:N < 5). The eventual C limitation of Daphnia therefore emphasises the paramount importance of the molecular composition of diet, and thus, the quality of the C transferred to zooplankton.
No or slight differences in bulk δ 13 C and δ 15 N among the host, chytrid-infected host, and Daphnia have been reported before, suggesting a large variation in isotopic enrichment across host-parasitic systems (Barranco et al., 2020, and references therein). Despite the high variation in our data, the results may highlight the limitations of bulk stable isotopes to estimate trophic positions and basal energy sources. As reported on a diatom-chytrid-Keratella system (Barranco et al., 2020), here we also showed that the host together with the chytrids tended to be at the highest trophic position, while the zooplankton at a lower level. Grey (2006)

| Compound specific 13 C sheds light on carbon assimilation
The The bioconversion of ALA to EPA by Daphnia was supported by CSI data that suggested that the chytrid parasite also synthesised LIN, an essential PUFA known to be synthesised mainly by algae (Taipale et al., 2013). The isotopically significantly heavier LIN (and also ALA) in chytrid zoospores, as well as in Daphnia relative to the chytrid-infected Planktothrix, further suggests the metabolic use of these PUFA by which isotopically heavier LIN and ALA were retained by Daphnia. The isotopically lighter δ 13 C SDA values also suggest SDA synthesis in the chytrid zoospore, which was subsequently selectively retained by Daphnia. While we did not find treatmentspecific differences in ARA retention of Daphnia, the isotopically lighter ARA compared to LIN also suggests the synthesis of these essential PUFA, irrespective of the two diet treatments. Similarly, the isotopically lighter EPA compared to ALA also suggests the endogenous EPA synthesis by Daphnia. The quantitative role of chytrids in supporting the innate conversion of short chain to LC-PUFA in Daphnia remains to be further elucidated, e.g. via 13 C-labelling of key functional PUFA. We note that in addition to PUFA, chytrids also produce sterols de novo Kagami et al., 2007), mediating dietary sterol provision and thus positive dietary effects of chytrids on consumers. We conclude that chytrid-mediated improved dietary PUFA extends to Daphnia, which enhances its performance when exposed to poor nutritional quality and hardly palatable cyanobacteria.

ACK N OWLED G M ENTS
We thank Samuel-Karl Kämmer, Katharina Winter and Gertraud Steniczka, WasserCluster Lunz, for their help with preparations for the experiment and laboratory analyses. We thank Dr Thomas Rohrlack for providing the Planktothrix and chytrid isolates.

FU N D I N G I N FO R M ATI O N
This work was supported by the Austrian Science Fund (FWF Project P 30419-B29).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.