Minna Hiltunen

Differing Daphnia magna assimilation efficiencies for terrestrial, bacterial, and algal carbon and fatty acids

There is considerable interest in the pathways by which carbon and growth-limiting elemental and biochemical nutrients are supplied to upper trophic levels. Fatty acids and sterols are among the most important molecules transferred across the plant-animal interface of food webs. In lake ecosystems, in addition to phytoplankton, bacteria and terrestrial organic matter are potential trophic resources for zooplankton, especially in those receiving high terrestrial organic matter inputs. We therefore tested carbon, nitrogen, and fatty acid assimilation by the crustacean Daphnia magna when consuming these resources. We fed Daphnia with monospecific diets of high-quality (Cryptomonas marssonii) and intermediate-quality (Chlamydomonas sp. and Scenedesmus gracilis) phytoplankton species, two heterotrophic bacterial strains, and particles from the globally dispersed riparian grass, Phragmites australis, representing terrestrial particulate organic carbon (t-POC). We also fed Daphnia with various mixed diets, and compared Daphnia fatty acid, carbon, and nitrogen assimilation across treatments. Our results suggest that bacteria were nutritionally inadequate diets because they lacked sterols and polyunsaturated omega-3 and omega-6 (omega-3 and omega-6) fatty acids (PUFAs). However, Daphnia were able to effectively use carbon and nitrogen from Actinobacteria, if their basal needs for essential fatty acids and sterols were met by phytoplankton. In contrast to bacteria, t-POC contained sterols and omega-6 and omega-3 fatty acids, but only at 22%, 1.4%, and 0.2% of phytoplankton levels, respectively, which indicated that t-POC food quality was especially restricted with regard to omega-3 PUFAs. Our results also showed higher assimilation of carbon than fatty acids from t-POC and bacteria into Daphnia, based on stable-isotope and fatty acids analysis, respectively. A relatively high (>20%) assimilation of carbon and fatty acids from t-POC was observed only when the proportion of t-POC was >60%, but due to low PUFA to carbon ratio, these conditions yielded poor Daphnia growth. Because of lower assimilation for carbon, nitrogen, and fatty acids from t-POC relative to diets of bacteria mixed with phytoplankton, we conclude that the microbial food web, supported by phytoplankton, and not direct t-POC consumption, may support zooplankton production. Our results suggest that terrestrial particulate organic carbon poorly supports upper trophic levels of the lakes.

Inferring phytoplankton community composition with a fatty acid mixing model

The taxon specificity of fatty acid composition in algal classes suggests that fatty acids could be used as chemotaxonomic markers for phytoplankton composition. The applicability of phospholipid-derived fatty acids as chemotaxonomic markers for phytoplankton composition was evaluated by using a Bayesian fatty acid-based mixing model. Fatty acid profiles from monocultures of chlorophytes, cyanobacteria, diatoms, euglenoids, dinoflagellates, raphidophyte, cryptophytes and chrysophytes were used as a reference library to infer phytoplankton community composition in five moderately humic, large boreal lakes in three different seasons (spring, summer and fall). The phytoplankton community composition was also estimated from microscopic counts. Both methods identified diatoms and cryptophytes as the major phytoplankton groups in the study lakes throughout the sampling period, together accounting for 54-63% of the phytoplankton. In addition, both methods revealed that the proportion of chlorophytes and cyanobacteria was lowest in the spring and increased towards the summer and fall, while dinoflagellates peaked in the spring. The proportion of euglenoids and raphidophytes was less than 8% of the phytoplankton biomass throughout the sampling period. The model estimated significantly lower proportions of chrysophytes in the seston than indicated by microscopic analyses. This is probably because the reference library for chrysophytes included too few taxa. Our results show that a fatty acid-based mixing model approach is a promising tool for estimating the phytoplankton community composition, while also providing information on the nutritional quality of the seston for consumers. Both the quantity and the quality of seston as a food source for zooplankton were high in the spring; total phytoplankton biomass was ;56 l gCL � 1 , and the physiologically important polyunsaturated fatty acids 20:5n-3 and 22:6n-3 comprised ;22% of fatty acids.

Selective transfer of polyunsaturated fatty acids from phytoplankton to planktivorous fish in large boreal lakes.

Lake size influences various hydrological parameters, such as water retention time, circulation patterns and thermal stratification that can consequently affect the plankton community composition, benthic-pelagic coupling and the function of aquatic food webs. Although the socio-economical (particularly commercial fisheries) and ecological importance of large lakes has been widely acknowledged, little is known about the availability and trophic transfer of polyunsaturated fatty (PUFA) in large lakes. The objective of this study was to investigate trophic trajectories of PUFA in the pelagic food web (seston, zooplankton, and planktivorous fish) of six large boreal lakes in the Finnish Lake District. Docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA) and α-linolenic acid (ALA) were the most abundant PUFA in pelagic organisms, particularly in the zooplanktivorous fish. Our results show that PUFA from the n-3 family (PUFAn-3), often associated with marine food webs, are also abundant in large lakes. The proportion of DHA increased from ~4±3% in seston to ~32±6% in vendace (Coregonus albula) and smelt (Osmerus eperlanus), whereas ALA showed the opposite trophic transfer pattern with the highest values observed in seston (~11±2%) and the lowest in the opossum shrimp (Mysis relicta) and fish (~2±1%). The dominance of diatoms and cryptophytes at the base of the food web in the study lakes accounted for the high amount of PUFAn-3 in the planktonic consumers. Furthermore, the abundance of copepods in the large lakes explains the effective transfer of DHA to planktivorous fish. The plankton community composition in these lakes supports a fishery resource (vendace) that is very high nutritional quality (in terms of EPA and DHA contents) to humans.

Lake eutrophication and brownification downgrade availability and transfer of essential fatty acids for human consumption.

Fish are an important source of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) for birds, mammals and humans. In aquatic food webs, these highly unsaturated fatty acids (HUFA) are essential for many physiological processes and mainly synthetized by distinct phytoplankton taxa. Consumers at different trophic levels obtain essential fatty acids from their diet because they cannot produce these sufficiently de novo. Here, we evaluated how the increase in phosphorus concentration (eutrophication) or terrestrial organic matter inputs (brownification) change EPA and DHA content in the phytoplankton. Then, we evaluated whether these changes can be seen in the EPA and DHA content of piscivorous European perch (Perca fluviatilis), which is a widely distributed species and commonly consumed by humans. Data from 713 lakes showed statistically significant differences in the abundance of EPA- and DHA-synthesizing phytoplankton as well as in the concentrations and content of these essential fatty acids among oligo-mesotrophic, eutrophic and dystrophic lakes. The EPA and DHA content of phytoplankton biomass (mgHUFAg-1) was significantly lower in the eutrophic lakes than in the oligo-mesotrophic or dystrophic lakes. We found a strong significant correlation between the DHA content in the muscle of piscivorous perch and phytoplankton DHA content (r=0.85) as well with the contribution of DHA-synthesizing phytoplankton taxa (r=0.83). Among all DHA-synthesizing phytoplankton this correlation was the strongest with the dinoflagellates (r=0.74) and chrysophytes (r=0.70). Accordingly, the EPA+DHA content of perch muscle decreased with increasing total phosphorus (r2=0.80) and dissolved organic carbon concentration (r2=0.83) in the lakes. Our results suggest that although eutrophication generally increase biomass production across different trophic levels, the high proportion of low-quality primary producers reduce EPA and DHA content in the food web up to predatory fish. Ultimately, it seems that lake eutrophication and brownification decrease the nutritional quality of fish for human consumers.

Inferring Phytoplankton, Terrestrial Plant and Bacteria Bulk δ¹³C Values from Compound Specific Analyses of Lipids and Fatty Acids.

Stable isotope mixing models in aquatic ecology require δ13C values for food web end members such as phytoplankton and bacteria, however it is rarely possible to measure these directly. Hence there is a critical need for improved methods for estimating the δ13C ratios of phytoplankton, bacteria and terrestrial detritus from within mixed seston. We determined the δ13C values of lipids, phospholipids and biomarker fatty acids and used these to calculate isotopic differences compared to the whole-cell δ13C values for eight phytoplankton classes, five bacterial taxa, and three types of terrestrial organic matter (two trees and one grass). The lipid content was higher amongst the phytoplankton (9.5±4.0%) than bacteria (7.3±0.8%) or terrestrial matter (3.9±1.7%). Our measurements revealed that the δ13C values of lipids followed phylogenetic classification among phytoplankton (78.2% of variance was explained by class), bacteria and terrestrial matter, and there was a strong correlation between the δ13C values of total lipids, phospholipids and individual fatty acids. Amongst the phytoplankton, the isotopic difference between biomarker fatty acids and bulk biomass averaged -10.7±1.1‰ for Chlorophyceae and Cyanophyceae, and -6.1±1.7‰ for Cryptophyceae, Chrysophyceae and Diatomophyceae. For heterotrophic bacteria and for type I and type II methane-oxidizing bacteria our results showed a -1.3±1.3‰, -8.0±4.4‰, and -3.4±1.4‰ δ13C difference, respectively, between biomarker fatty acids and bulk biomass. For terrestrial matter the isotopic difference averaged -6.6±1.2‰. Based on these results, the δ13C values of total lipids and biomarker fatty acids can be used to determine the δ13C values of bulk phytoplankton, bacteria or terrestrial matter with ± 1.4‰ uncertainty (i.e., the pooled SD of the isotopic difference for all samples). We conclude that when compound-specific stable isotope analyses become more widely available, the determination of δ13C values for selected biomarker fatty acids coupled with established isotopic differences, offers a promising way to determine taxa-specific bulk δ13C values for the phytoplankton, bacteria, and terrestrial detritus embedded within mixed seston.

Bacterial and phytoplankton responses to nutrient amendments in a boreal lake differ according to season and to taxonomic resolution.

Nutrient limitation and resource competition in bacterial and phytoplankton communities may appear different when considering different levels of taxonomic resolution. Nutrient amendment experiments conducted in a boreal lake on three occasions during one open water season revealed complex responses in overall bacterioplankton and phytoplankton abundance and biovolume. In general, bacteria were dominant in spring, while phytoplankton was clearly the predominant group in autumn. Seasonal differences in the community composition of bacteria and phytoplankton were mainly related to changes in observed taxa, while the differences across nutrient treatments within an experiment were due to changes in relative contributions of certain higher- and lower-level phylogenetic groups. Of the main bacterioplankton phyla, only Actinobacteria had a treatment response that was visible even at the phylum level throughout the season. With increasing resolution (from 75 to 99% sequence similarity) major responses to nutrient amendments appeared using 454 pyrosequencing data of 16S rRNA amplicons. This further revealed that OTUs (defined by 97% sequence similarity) annotated to the same highly resolved freshwater groups appeared to occur during different seasons and were showing treatment-dependent differentiation, indicating that OTUs within these groups were not ecologically coherent. Similarly, phytoplankton species from the same genera responded differently to nutrient amendments even though biovolumes of the majority of taxa increased when both nitrogen and phosphorus were added simultaneously. The bacterioplankton and phytoplankton community compositions showed concurrent trajectories that could be seen in synchronous succession patterns over the season. Overall, our data revealed that the response of both communities to nutrient changes was highly dependent on season and that contradictory results may be obtained when using different taxonomic resolutions.

Suitability of Phytosterols Alongside Fatty Acids as Chemotaxonomic Biomarkers for Phytoplankton.

The composition and abundance of phytoplankton is an important factor defining ecological status of marine and freshwater ecosystems. Chemotaxonomic markers (e.g., pigments and fatty acids) are needed for monitoring changes in a phytoplankton community and to know the nutritional quality of seston for herbivorous zooplankton. Here we investigated the suitability of sterols along with fatty acids as chemotaxonomic markers using multivariate statistics, by analyzing the sterol and fatty acid composition of 10 different phytoplankton classes including altogether 37 strains isolated from freshwater lakes. We were able to detect a total of 47 fatty acids and 29 sterols in our phytoplankton samples, which both differed statistically significantly between phytoplankton classes. Due to the high variation of fatty acid composition among Cyanophyceae, taxonomical differentiation increased when Cyanophyceae were excluded from statistical analysis. Sterol composition was more heterogeneous within class than fatty acids and did not improve separation of phytoplankton classes when used alongside fatty acids. However, we conclude that sterols can provide additional information on the abundance of specific genera within a class which can be generated by using fatty acids. For example, whereas high C16 ω-3 PUFA (polyunsaturated fatty acid) indicates the presence of Chlorophyceae, a simultaneous high amount of ergosterol could specify the presence of Chlamydomonas spp. (Chlorophyceae). Additionally, we found specific 4α-methyl sterols for distinct Dinophyceae genera, suggesting that 4α-methyl sterols can potentially separate freshwater dinoflagellates from each other.

Trophic upgrading via the microbial food web may link terrestrial dissolved organic matter to Daphnia

36 Direct consumption of allochthonous resources generally yields poor growth and reproduction in 37 zooplankton, but it is still unclear how trophic upgrading of terrestrial dissolved organic matter 38 (tDOM) via microbial food web may support zooplankton. We compared survival, somatic growth, 39 and reproduction of Daphnia magna fed with heterotrophic flagellate Paraphysomonas vestita and 40 three algal diets. Paraphysomonas was fed lake bacteria that used tDOM as a substrate to simulate 41 an allochthonous diet that zooplankton encounter in lakes. The highest survival, growth, and 42 reproduction was achieved with a diet of Cryptomonas, while Daphnia performance was the worst 43 when fed Microcystis. Paraphysomonas and Scenedesmus diets lead to intermediate growth and 44 reproduction. Cryptomonas contained high amounts of essential polyunsaturated fatty acids (PUFA) 45 and phytosterols that supported high somatic growth and reproduction, whereas poor performance 46 of Daphnia on cyanobacterial diet was most likely due to lack of sterols. Paraphysomonas 47 contained some phytosterols, but not in sufficient amounts, and also essential PUFA 48 (eicosapentaenoic and arachidonic acid) that enhance zooplankton growth and reproduction. Our 49 results indicate that tDOM-based microbial food web supports Daphnia performance even as a sole 50 food source, and may be important in providing zooplankton with essential biochemical 51 components when phytoplankton quantity or quality is low. 52 Introduction 53 The recognition that consumers in aquatic ecosystems may be fueled by terrestrial food 54 sources in addition to in-lake phytoplankton production has sparked vast amount of research during 55 the past three decades. The extent and possible pathways of consumer allochthony have been 56 studied both under laboratory conditions and in the field (e.g. Grey et al. 2001; Pace et al., 2004; 57 Berggren et al., 2014; Taipale et al., 2014). Field studies utilizing stable isotope ratios (C, N, H) 58 have found that large fraction of consumer biomass can be traced to allochthonous sources (Pace et 59 al., 2004; Berggren et al., 2014; Tanentzap et al. 2017, and references therein). Zooplankton 60 allochthony may also vary seasonally following the relative availability of phytoplankton and 61 allochthonous food sources (Grey et al. 2001). However, laboratory feeding experiments have 62 questioned the feasibility of high zooplankton allochthony. Although zooplankton (mainly 63 Daphnia) can survive on purely allochthonous diets, their growth efficiency, somatic growth rate 64 and reproductive output is very low on allochthonous compared to phytoplankton diets (Brett et al., 65 2009; Wenzel et al., 2012; Taipale et al., 2014). Consequently, high inputs of terrestrial carbon and 66 high consumer allochthony have been linked to low production of wild zooplankton (Kelly et al., 67 2014) and fish (Rask et al., 2014; Karlsson et al., 2015). 68 Most laboratory feeding trials testing consumer allochthony have been conducted using 69 terrestrial particulate organic matter (tPOM) as the food source. More than 90% of terrestrial 70 organic matter in lakes is in the dissolved form (DOM) (Kortelainen et al., 1993; Mattsson et al., 71 2005), and tPOM entering the lake in the shoreline or via river flow may rapidly sediment out of the 72 water column. Thus, pelagic consumers especially in large lakes may have limited access to tPOM. 73 Terrestrial DOM (tDOM) can be used as a substrate by bacteria, which can be grazed by 74 heterotrophic protists including flagellates and ciliates (the microbial loop) or directly by 75 zooplankton (Tranvik, 1992; Weisse 2004). Daphnia have been shown to benefit from tDOM 76 directly or via tDOM-supported bacteria when algae is limiting (McMeans et al., 2015). Previous 77 studies (Wenzel et al., 2012; Taipale et al., 2014) have found that Daphnia performance is better 78 when feeding on mixtures of phytoplankton and bacteria than on mixtures of phytoplankton and 79 tPOM, suggesting that DOM may be the more probable pathway for allochthonous organic matter 80 to enter the grazer food web. According to feeding experiments, bacteria alone cannot support 81 Daphnia growth and some taxa may even be toxic to Daphnia as a sole food source (Taipale et al., 82 2012; Freese and Martin-Creuzburg 2012). Few studies have been conducted on Daphnia 83 performance on diets of heterotrophic flagellates, but results have been variable (Sanders et al., 84 1996; Bec et al., 2003; 2006). 85 One of the reasons proposed why Daphnia has poor growth on allochthonous diets is their 86 lack of essential biomolecules, especially polyunsaturated fatty acids (PUFA) and sterols (Brett et 87 al., 2009; Taipale et al., 2014). Compared to many algae, tPOM contains very little PUFA, while 88 bacteria contain none (Lechevalier and Lechevalier 1988; Taipale et al., 2014). Some studies have 89 found that the fatty acid composition of heterotrophic flagellates depends on whether they feed on 90 algae or bacteria (Zhukova and Kharlamenko, 1999; Véra et al., 2001) while others conclude that 91 biosynthesis of lipids produces a consistent fatty acid (and sterol) composition in flagellates 92 irrespective of diet (Bec et al., 2010; Parrish et al., 2012). Bacteria, including cyanobacteria, also 93 lack sterols while phytoplankton contain various sterols in composition that is species-specific 94 (Taipale et al., 2016). The sterol composition of flagellates is poorly studied, but so far studies have 95 indicated that heterotrophic flagellates are capable of sterol synthesis (Klein Breteler et al., 1999; 96 Bec et al., 2006). In addition to concentrating PUFA and sterols present in their food e.g. by 97 selective retention, heterotrophic flagellates may enhance low quality bacterial or cyanobacterial 98 food for Daphnia by either biosynthesizing PUFA and sterols de novo or modifying dietary short99 chain PUFA to eicosapentaenoic acid (EPA, 20:5ω3) and docosahexaenoic acid (DHA, 22:6ω3). 100 This so called ‘trophic upgrading’ by heterotrophic flagellates has been observed in several studies 101 (Klein Breteler et al., 1999; Veloza et al., 2006; Bec et al., 2006; 2010). Also, indirect evidence of 102 trophic upgrading was obtained when increased abundance of Paraphysomonas vestita in a 103 decaying Microcystis culture was associated with rising EPA and DHA concentrations with a 104 concurrent decrease in short-chain PUFA prominent in Microcystis (Park et al., 2003). 105 Previous studies on Daphnia performance on allochthonous diets have used tPOM, single 106 strains of bacteria grown in artificial growth media, or heterotrophic flagellates growing on these 107 bacteria as a diet source (but see McMeans et al., 2015). We conducted a feeding experiment where 108 we constructed a simple microbial food web of tDOM (peat extract)-natural lake bacteria109 Paraphysomonas vestita to better simulate the pathway for allochthonous carbon to enter 110 zooplankton diets in lakes. We compared Daphnia somatic growth and reproduction on this 111 allochthonous diet to diets of three phytoplankton taxa (Cryptomonas, Scenedesmus, Microcystis) 112 known to vary in their quality as food for Daphnia. Our hypothesis was that 1) Daphnia survival 113 would be better on tDOM-based microbial diet than on pure bacterial diets (as seen in other studies) 114 and 2) Daphnia somatic growth and reproduction would be lower than on the algal diets. 115 116 Method 117 Experimental set up 118 We compared survival, growth and reproduction of Daphnia feeding on either diets of 119 algae or a diet of a heterotrophic flagellate that was grown on bacteria utilizing tDOM as a 120 substrate. For the experiment, we used Daphnia magna clone (DK-35-9), that originated from a 121 pond in North Germany and has been raised successfully in laboratory for several years. Prior to the 122 experiment Daphnia were reared several generations on Scenedesmus. Daphnia neonates (<24h 123 old) of multiple moms were pooled and randomly distributed among treatments (20 ind./treatment) 124 and some were used to determine Daphnia mean initial body weight. During the feeding 125 experiment, individual Daphnia were raised in 40mL vials in ADaM medium (Klüttgen et al., 126 1994). Daphnia were maintained on one of five different diets: no food, Cryptomonas marssonii, 127 Scenedesmus gracilis, Microcystis sp. (strain 130, unicellular, non-toxic) or the heterotrophic 128 flagellate Paraphysomonas vestita. The three algae were cultured in growth media optimal for each 129 of them (Table 1) in 14h:10h light:dark cycle at 20oC. The heterotrophic flagellate was grown in a 130 culture medium containing tDOM extracted from unfertilized garden peat (Kekkilä luonnonturve) 131 which was inoculated with lake bacteria (1 mL of 0.2 μm filtered lake water) a few days prior to 132 addition of the flagellate. Paraphysomonas was concentrated with gentle centrifugation, but the diet 133 given to Daphnia likely contained also bacteria. 134 The media was changed and the Daphnia fed every other day. We offered the food at non135 limiting concentration: 1.5 mg C L-1 on days 0-2, 2 mg C L-1 on day 4 and 5 mg C L-1 from day 6 136 onwards. Every day Daphnia were inspected and dead animals, and the number of offspring were 137 recorded. Sampling was conducted in the middle and in the end of the experiment (days 7 and 14), 138 and Daphnia (10 ind.) in each treatment were collected for measurements of length, weight and 139 subjected to fatty acid analysis. Due to difficulties in culturing the heterotrophic flagellate, we 140 ended the Paraphysomonas treatment already after 12 days. To facilitate the comparison with 141 Daphnia in algal diets that lasted 14 days, the eggs and embryos in Daphnia brood pouch in the 142 Paraphysomonas treatment on day 12 were included as “potential neonates” for day 14. 143 144 Fatty acid and sterol analysis 145 Prior to analy