Consequences of Consumer-Resource Stoichiometric Imbalance in Planktonic Food Webs

Consequences of consumer-resource stoichiometric imbalance in planktonic food webs

Junwen Guo

Name of Author Department of Ecology and Environmental Science Umeå 2020

This work is protected by the Swedish Copyright Legislation (Act 1960:729) Dissertation for PhD ISBN: 978-91-7855-182-8 Cover photo: Junwen Guo Electronic version available at: http://umu.diva-portal.org/ Printed by: KBC Service Center Umeå, Sweden 2020

⊙博学之，审问之，慎思之，明辨之，笃行之。

一《中庸》

Table of Contents

Abstract ...... 2 Enkel sammanfattning på svenska ...... 3 List of papers ...... 4 Author contribution ...... 5 Resource imbalance in ecological stoichiometry ...... 6 Stoichiometric composition of autotrophs, consumers and bacteria across biological levels ...... 6 Consumer-driven nutrient recycling or bacteria-driven nutrient recycling ...... 8 The unexplored role of the chemical form of resources for the bacterial-driven nutrient recycling ...... 10 Objectives of the thesis and outline ...... 11 Methodology ...... 12 An operational framework for experimental research in stoichiometry theory ...... 12 Empirical studies guided by ecological stoichiometry framework ...... 13 Materials and Methods ...... 16 Literature review ...... 16 Experimental studies ...... 16 Field survey ...... 18 Results and discussion...... 20 Effect of element chemical forms on bacteria community ammonium regeneration and net nitrogen mineralization ...... 20 Microplankton food web stoichiometric response to different forms of nitrogen...... 21 Stoichiometric response of a coastal food web to temporal and spatial gradient ...... 24 Conclusion ...... 26 Significance and further directions ...... 27 Acknowledgements ...... 29 References ...... 30 Thanks ...... 36

Abstract

Resource imbalance between consumers and their resources can come from inadequate resource quantity or quality. The ecological stoichiometry theory focuses on understanding the consequences of imbalance in elemental composition. In this thesis, I have used both resource quality (e.g., inorganic vs organic forms of nutrients) and resource quantity (e.g., terrestrial and freshwater nutrient loading to natural coastal systems) to address the consequences of consumer-resource imbalance in planktonic food webs. First, I provided a framework that summarizes how the stoichiometric imbalance is transferred from one biological level to another. The framework highlights the importance of the distribution of elements among different chemical forms and the distribution of elements among connected ecosystems. The framework then served as a guideline for the empirical work of my thesis. Second, I studied the response of bacterial community mineralization to the relative availability of different forms of nitrogen (inorganic vs. organic form) in a batch culture experiment. The study shows that different forms of nitrogen can significantly influence the growth of bacteria. More importantly, my results show that it is crucial to measure the actual bacterial carbon to nitrogen consumption ratio, rather than use classical theoretical models, to be able to make an accurate prediction of bacterial ammonium regeneration. Third, I tested the effect of different forms of nitrogen on microplankton food web dynamics in a microcosm experiment. I found that differences between nitrogen forms have a strong impact on food web dynamics that is channeled by the bacteria- phytoplankton interaction at the base of the food web. The whole microplankton food web benefits from organic forms of nitrogen as a result of increased mutualistic interactions between bacteria and phytoplankton. Hence, the form of nitrogen is an important factor to be considered in microplanktonic food web dynamics, at least on the short-term. In the final part of this thesis, I explored resource quality and quantity effects on the stoichiometric response of a natural coastal ecosystem in a field study. I expected that the relative availability of inorganic or organic forms of carbon, nitrogen and phosphorus in our sampling bays may affect organismal elemental composition both temporally and spatially. The results indicate that the stoichiometry among seston size fractions and zooplankton varied more through time than in space. However, zooplankton stoichiometry was relatively stable among species within specific months. Overall, the concentration of dissolved organic carbon and dissolved organic nitrogen in the water column were the major explanatory variables for the seston stoichiometry. In summary, this thesis uses multiple systems to elucidate how the form and input of nutrients shape the plankton food web dynamics and its stoichiometric responses.

Enkel sammanfattning på svenska

Det kan uppstå en obalans mellan konsumenter och deras resurser om födotillgången är otillräcklig eller om kvaliteten varierar. Inom ekologisk stökiometriforskning studeras vilka konsekvenser obalanser i den molekylära sammansättningen hos födoresurserna och olika trofiska nivåer kan få i ekosystemen. I min avhandling har jag undersökt hur födoresursernas kvalitet (t.ex. sammansättningen av oorganiska och organiska näringsämnen) och kvantitet påverkar (t.ex. olika älvtillföde till kustekosystem) konsument-resursbalansen i akvatiska födovävar. Först gjorde jag en litteratursammanställning över hur stökiometriska obalanser kan påverka ekosystemen på olika nivåer. Studien visar att fördelningen mellan olika ämnen är mycket betydelsefulla såväl inom som mellan olika ekosystem. Därefter utförde jag ett experiment där jag undersökte hur bakteriers mineraliseringsförmåga påverkas om kväve tillförs i oorganisk eller organisk form. Resultaten visar att de olika formerna av kväve påverkar tillväxten hos bakterierna, och att man måste mäta upptaget av både kol och kväve för att beräkna hur mycket ammonium som bakterierna kommer att mineralisera. I ett annat experiment undersökte jag hur olika former av kväve påverkar mikrobiella födovävens stuktur och dynamik. Balansen mellan autotrofi och heterotrofi hos basalproducenterna påverkades starkt av olika former av kväve, och jag kunde också observera en skillnad i energikanaliseringen uppåt i födoväven. Dock gynnades hela den mikrobiella födoväven om kvävekällan bestod av organiska föreningar, beroende på mutualistiska interaktioner mellan autotrofa organismer och heteterofa bakterier. För att förstå dynamiken hos mikrobiella födovävar i naturliga ekosystem är det därför viktigt att ta i beaktande om deras kvävekälla är i oorganisk eller organisk form. I den sista delstudien i min avhandling undersökte jag hur kol, kväve och fosforkoncentrationerna i fria vattemassan påverkar stökiometrin hos seston och djurplankton i ett kustområde i norra Östersjön. Resulten visar att stökiometrin hos seston och djurplankton påverkades mer temporalt (vår till sommar) än spatiellt (olika vikar). Djurplanktonen hade en relativt stabil CNP stökiometri jämfört med seston. Studien indikerar studien att koncentrationen av löst organiskt kol och löst organiskt kväve påverkar stökiometrin hos seston och att det även kan ha viss påverkan på djurplanktonens stökiometri. Sammansfattningsvis visar min avhandling att olika former av näringsämnen och deras balans har stor inverkan på stökiometri och funktion hos olika organismgrupper i den akvatiska födoväven.

List of papers

Chapter I. Cherif M., Faithfull C., Guo J., Meunier C., Sitters J., Uszko W. and Vasconcelos F. 2017. An operational framework for the advancement of a molecule-to-biosphere stoichiometry theory. Frontiers in Marine Science, Vol. 4.

Chapter II. Guo J., Cherif M. More than stoichiometry: the molecular composition of inorganic and organic substrates controls ammonium regeneration by bacteria. Manuscript

Chapter III. Guo J., Brugel S., Andersson A., Cherif M. Organic vs. inorganic nitrogen as promoting nutrient in aquatic microplankton food webs. Manuscript

Chapter IV. Guo J., Brugel S., Andersson A., Lau D. C. P. Stoichiometric changes in planktonic food web upon environmental fluctuations in northern oligotrophic coastal ecosystem. Manuscript

Author contribution

Chapter I. All authors contributed to the conception and design of the study and to the selection, analyses and interpretation of the relevant literature.

Chapter II. JG and MC designed the experiment, JG performed the experiment. JG conducted lab analyses and statistical analyses. JG wrote the manuscript with contribution from MC.

Chapter III. JG, SB and AA designed the experiment. JG and SB planned and performed the experiment and lab analyses. JG performed statistical analyses. JG wrote the manuscript with contribution from MC. All coauthors commented on the manuscript.

Chapter IV. JG, SB, AA and DL planned the field study. JG and SB conducted lab analyses. JG performed statistical analyses. JG wrote the manuscript with contribution from DL. All coauthors commented on the manuscript.

Author: Junwen Guo (JG), Mehdi Cherif (MC), Agneta Andersson (AA), Sonia Brugel (SB) and Danny Chun Pong Lau (DL).

Resource imbalance in ecological stoichiometry

Ecological stoichiometry is a theory that treats ecological processes as chemical reactions during which elements (e.g. carbon, nitrogen and phosphorus) are exchanged between an organism, its resource, and the environment. Ecological stoichiometry tackles the complexity of food webs by studying the balance of multiple chemical elements, e.g. carbon (C), nitrogen (N) and phosphorus (P), as they are transferred along during ecological interactions and processes (Sterner & Elser 2002). Elemental ratios (in mol:mol) are usually used to study the effects of ecological stoichiometry on ecosystem processes in food webs. Elemental ratios can be measured at any scale (from cells to ecosystems) and thus allow comparing various ecological responses across a wide range of levels. However, organisms can strongly differ in their carbon: nitrogen: phosphorus (C:N:P) ratios, from the individual to the population and community levels. They particularly differ across different trophic levels resulting in imbalance between the C:N:P ratios of consumers and their resources. Given the large variation of organism stoichiometric compositions across biological levels, the interaction between organisms and their resources have become a central research topic in the current field of ecological stoichiometry.

Stoichiometric composition of autotrophs, consumers and bacteria across biological levels Autotrophs display large plasticity in their C:N:P composition in aquatic systems. For example, in both experimental and field studies, the C:P molar ratio of phytoplankton from different taxonomic groups can vary from 28 to 290 at both intra- and inter-specific levels (see review in Finkel et al. 2010). At the community level, phytoplankton stoichiometry also greatly varies across ecosystems (Sterner & Elser 2002). Such plasticity can be attributed to multiple environmental factors, such as nutrient availability, temperature and light. These environmental factors can significantly influence the phytoplankton allocation of elements between cell components in order to optimize their growth under different conditions, and thus result in different C:N:P ratios (Elser et al. 2002; Kohler et al. 2012; Kendrick & Benstead 2013; Yvon- Durocher et al. 2017). Moreover, these factors can modify the species community composition, which will then affect the C:N:P ratios at the community level (Weber & Deutsch 2010; Moschonas et al. 2017). Hence, the various effects of environmental factors on phytoplankton C:N:P ratios involve processes at multiple scales that usually occur at the same time, making the stoichiometry of autotrophs hard to predict.

In contrast to autotrophs, consumers (e.g. herbivores) can generally maintain a relatively stable C:N:P stoichiometry (Sterner & Elser 2002; Persson et al. 2010; Golz et al. 2015). Homeostasis for a consumer is defined as the capacity to regulate their C:N:P ratios in response to variations in the C:N:P ratios of their resources. Previous studies confirmed that the variability in dietary stoichiometry had little effect on the stoichiometry of consumers within species (Andersen & Hessen 1991; McManamay et al. 2011). However, consumer stoichiometry can vary across different taxa. For instance, copepods generally have higher N:P and C:P ratios than cladocerans (Sterner & Elser 2002). Recent studies challenge the degree of consumer homeostasis, especially concerning their P content (Cross et al. 2003; DeMott & Pape 2005; Persson et al. 2010). Still, homeostasis remains an important assumption in stoichiometry modelling works that examine the role of stoichiometric imbalance between consumers and their resources (Sterner 1990; Urabe 1993; Touratier et al. 1999; Loladze et al. 2004; Cherif & Loreau 2007).

Autotrophs and consumers stoichiometry has been well documented and extensively studied since the 1990s. But another fundamental component of food webs, bacteria, has been less investigated in the field. This is probably due to limitations in sampling methods since in the natural environment, the technical separation between bacteria and other small phytoplankton is challenging. Hence, whether the C:N:P stoichiometry of bacteria is more like that of the phytoplankton or the zooplankton remained unclear for a long period. Some recent studies hold different conclusions on bacterial stoichiometry (Makino et al. 2003; Cotner et al. 2010; Hall et al. 2011; Scott et al. 2012; Stenzel et al. 2017). These studies suggest that bacteria can maintain some degree of homeostasis even if the C:P ratio shows large variability. The current consensus seems to be that bacteria stoichiometric plasticity shows a continuum of patterns somewhere in between the strict homeostasis of consumers and the plasticity of autotrophs (Persson et al. 2010). In general, whether it is homeostatic or flexible stoichiometry that prevails at the species level, the variation in elemental composition exists at the community level or even ecosystem level (Hillebrand et al. 2013; Trautwein et al. 2017). Therefore, understanding how elemental stoichiometry varies across organisms and systems helps us to understand trophic interactions and whole food web dynamics.

Differences in elemental composition result in stoichiometric imbalance between organisms and their resources, e.g., between bacteria and their dissolved organic resources, or between zooplankton and phytoplankton (Daufresne and Loreau 2001 a, b). The imbalance between elemental demand and supply of the consumer affects the physiological processes of the 7 organisms: it may limit their growth rate, alter their excretion of various elements or, at its extreme, lead to premature death (Frost et al. 2005). Eventually, these effects can feed back on the stoichiometric imbalance between organisms, since it has been shown that the stoichiometry of various species is affected by processes at the level of food webs and ecosystems, such as nutrient recycling (Vanni 2002; Atkinson et al. 2017). Hence, stoichiometric imbalance between consumers and their resources result in complex patterns at many different ecological levels.

Consumer-driven nutrient recycling or bacteria-driven nutrient recycling Consumer driven nutrient recycling (CNR) is one of the earliest processes considered in the field of ecological stoichiometry. It is produced by the stoichiometric imbalance between a consumer and its resource (Elser et al. 1988). Nutrients recycled by zooplankton were early on recognized as important nutrient sources for phytoplankton growth (Sterner 1990). It was also assumed (Sterner 1990), and later confirmed (Elser & Urabe 1999), that the consumers of autotrophs are strongly homeostatic, in contrast to autotrophs which are more stoichiometrically flexible. In particular, in ecosystems that are depleted in a specific nutrient (typically N or P), the autotrophs become nutrient-depleted as well, while the consumer stoichiometry stays unaltered. In parallel, non-depleted nutrients accumulate in the autotrophs above the needs of their consumers. The result is a mismatch in elemental composition between autotrophs. In order to keep its stoichiometric composition constant, primary consumers need to excrete the nutrients in excess from the autotrophs they consume, and retain the element that is depleted (Sterner 1990). The excess nutrients that are excreted by consumers, in both organic and inorganic forms, are often made available in the ecosystem again, and can directly influence the growth of the autotrophs. Vanni (2002) reviewed both direct and indirect pathways of the consumer-driven nutrient recycling in freshwater ecosystems and the potential consequences of CNR on other trophic levels (e.g. shift in the phytoplankton community composition). However, CNR will only have a strong influence on the ecosystem nutrient dynamics in the cases where the C:N:P ratios of the resources are very different from the consumers demand. Therefore, the importance of CNR varies tremendously among populations and across ecosystems. Vanni (2002) also highlighted that consumers can recycle a large proportion of the excess elements as organic nutrients. These nutrients cannot be used directly by most of the autotrophs, but are resources for bacteria. Hence, in order to better understand the role of the CNR on ecosystem function, we need to take heterotrophic bacteria into consideration since a proportion of the recycled nutrients can only be made available again for autotrophs via decomposition. Later,

Atkinson et al. (2017) reviewed studies (from 1973 to 2015) focusing on the spatial and temporal variation in the CNR dynamics and its relevance for freshwater ecosystems. They concluded that the CNR dynamics strongly depends on both abiotic (e.g. temperature, nutrient concentration) and biotic factors (e.g. trophic relationships, community composition). They also highlighted a knowledge gap concerning the links between CNR, bacteria and decomposition. Only a few studies highlight the important role of CNR on bacterial community dynamics and structure. Consumer excretion can increase microbial activity and alter bacterial community composition (Rugenski et al. 2012; Zeng et al. 2014). However, the reverse interaction, whereby bacteria influence CNR has almost only been studied theoretically (Cherif and Loreau 2009, 2013). Hence, CNR effects could be more complex than assumed by the current CNR theory, considering that phytoplankton growth is also highly dependent on the nutrients that are processed by bacteria.

Indeed, bacteria play a key role in nutrient recycling and are able to influence nutrient availability to autotrophs both directly and indirectly. Bacteria are known to compete with phytoplankton for the uptake of limiting inorganic nutrients (immobilization), while also excreting inorganic nutrients as a by-product of their consumption of organic nutrients (mineralization) (Tupas et al. 1994; Kirchman & Wheeler 1998; Glibert et al. 2016). The relative proportion of inorganic and organic nutrients available to bacteria and their stoichiometric composition determines to which side the bacteria-phytoplankton interaction leans: competition or mutualism (Bratbak & Thingstad 1985; Cotner & Biddanda 2002; Stets & Cotner 2008). Previously, the research focus of bacteria-phytoplankton stoichiometric interactions was on competition for inorganic nutrients and whether bacteria are better competitors than phytoplankton for inorganic nutrients, depending on organic carbon availability to bacteria (Kirchman 1994; Danger et al. 2007). There was little focus on the dependence of autotrophs growth on the nutrients regenerated by bacteria. Moreover, bacterial growth can be fueled by consumer-recycled organic nutrients while used simultaneously as a food source by consumers, making the role of bacteria even more complex (Cherif & Loreau 2009). The presence of bacteria will not only affect the availability of limiting nutrients for autotrophs, but also interfere with the effect of CNR on the inorganic and organic nutrients pool, which suggests that bacteria may be a major driver of aquatic food web dynamics.

The unexplored role of the chemical form of resources for the bacterial-driven nutrient recycling The important role of bacteria in nutrient recycling is determined by the resource quantity and quality. In natural aquatic systems, changes of the nutrient concentration in the water column may result from large amount of allochthonous input from e.g. terrestrial runoff and freshwater input (Kjellström & Ruosteenoja 2007; Bates et al. 2008; Meier et al. 2012; Andersson et al. 2015; Reusch et al. 2018). The significant nutrients loading to the aquatic ecosystem are known to exert strong control on food webs through various mechanisms (Chiandet & Xenopoulos 2011; De Senerpont Domis et al. 2014; Deininger et al. 2016). However, nutrients received from connected habitats vary significantly in their quality, which is determined by their chemical characteristics. To date, part of resource quality, especially the molecular configuration of resources, is largely unexplored. For instance, in aquatic ecosystems, inorganic nitrogen comes as nitrate or ammonium. The diversity of organic molecules is even greater: amino acids, nucleotides, humic substances, etc. (Creed et al. 2018). Hence, two organic resources for bacteria may have similar stoichiometry, but differ in the type of organic molecules they are made of. Some resources maybe relatively homogeneous, with essential elements such as N, equally distributed among molecules (e.g., chitin), others may contain a majority of C-rich compounds and a few N-rich compounds, such as amino acids. Different molecules are known to differ in their bioavailability to bacteria by promoting different metabolic pathways (Gale 1947; Barker 1981; Naganuma et al. 2018). The result for bacteria can be different growth rates, stoichiometry and thus mineralization and immobilization rates. Since the bacteria-phytoplankton interaction, in-between competition and mutualism, depends on the net outcome of mineralization and immobilization, so the molecular configuration of resources may eventually determine the abundance and stoichiometry of autotrophs, consumers and food webs in general. Therefore, differences in chemical forms is an additional mechanism by which the supply of nutrients can affect the dynamics of food webs.

Objectives of the thesis and outline

Given the complex nature of processes and interactions that define the stoichiometric interactions between autotrophs, bacteria and consumers, we first defined a framework to infer how the stoichiometric imbalance from one level impacts the other biological levels in Chapter I. The framework served as a guideline to the planning of the experiments that make the other three chapters of the thesis. Moreover, it allowed identifying the knowledge gaps of current stoichiometric studies.

In Chapter II, we tested experimentally the effects of the molecular composition of the supplied substrate on bacterial mineralization, by using an isolated bacterial community grown under different substrates. We compared a classical stoichiometric model to predict nitrogen mineralization, which doesn’t take molecular composition into account, with a modified version that uses measured consumption of N and C in order to better include molecular composition effects.

Next in Chapter III, we tested the effects of molecular composition on trophic interactions in an aquatic planktonic food web, by supplying N in increasing amounts, either as DIN or as DON. We used a natural microplanktonic food web from the northern Baltic Sea coast, and compared the quantitative and stoichiometric responses of different size fractions to inputs as DIN/DOC or as DON.

Finally, in Chapter IV, we assessed effects of spatial and temporal changes in the natrual environment on the stoichiometric composition of seston and zooplankton taxa in four bays of the northern Baltic Sea.

Methodology

An operational framework for experimental research in stoichiometry theory Thanks to ecological stoichiometry studies from the past decades, we now have great insights into stoichiometric processes from the physiological to the ecosystem level (Sardans et al. 2012; Hessen et al. 2013). However, not enough efforts have been made to integrate and organize the various stoichiometric breakthroughs into one coherent framework. Moreover, links between mechanisms at the lower and higher ends of biological organization levels are often inferred but seldom fully explicated. Therefore, the thesis started by providing a methodological framework (Chapter I) to review the ecological stoichiometry studies from the past, and to be used as a roadmap for designing the present and future work in ecological stoichiometry studies.

This framework consists of three dimensions: biological organization levels, ecological processes that link the various levels and patterns of elemental distribution within each level. By articulating these components, effects of stoichiometric imbalance at one level over all the other levels can be characterized. Instead of developing a descriptive framework, we made the framework operational so that it can be applied to explain the consequences of any stoichiometric imbalance at different biological levels, including potential future developments and integrations of other theories at all levels. By including processes from gene expression to fluxes in the biosphere, we provide a framework to connect all biological levels.

Already, we have used the framework in this thesis to organize our thinking about resource molecular composition effects on microplankton food webs. This framework predicts that differences in the distribution of elements between chemical forms in the ecosystem affects organisms via changes in ecophysiological processes, i.e., in the fluxes of carbon (C) and other elements through organisms. For example, increasing C:X ratios (where X is any essential element other than C) in food leads to less ingestion of X relative to C (Fig. 1). This change in the ratio of acquired elements should lead to lower efficiency of C accumulation and higher C:X release ratios by a homeostatic consumer (changes in physiological rates, Fig. 1). Repercussions of these physiological responses to changes in molecular composition reach higher biological levels through changes in demographic processes, ecological interactions and recycling rates (Fig. 1).

Figure1. Framework for a molecule-to-biosphere stoichiometry theory. This framework presents explicitly the biological levels and processes that need to be investigated in order to characterize the repercussions of a stoichiometric imbalance at one level over all the other levels. Each biological level should be described by the patterns of elemental distribution within it. As the framework shows, the processes investigated should go beyond the ﬂuxes and rates of elements between components that are classically considered in stoichiometric studies. Any process affecting the structure of a biological level should be considered if the changes provoked relate to changes in the distribution of elements. Numerous examples are discussed within the text (Chapter I) (Cherif et al. 2017).

Empirical studies guided by ecological stoichiometry framework The framework from Chapter I was used as a guideline for designing the empirical studies that make the rest of the thesis work. The following studies (Chapter II, III and IV) emphasize the usefulness of our framework for ecological stoichiometry studies. We started our scientific question from the ecosystem level by considering the distribution of elements among different chemical forms. We were particularly interested in exploring the effect of chemical forms of resources, e.g. inorganic vs. organic nitrogen, or nitrogen-containing vs. C-only organic molecules, on bacterial community net mineralization in Chapter II (Fig. 2), and on microplankton food web responses in Chapter III (Fig. 3) by manipulating nutrient supplies in lab-controlled experiments. We measured different ecological processes at different levels of organization. For example, we measured ammonium regeneration, net nitrogen mineralization and nitrogen use efficiency in bacteria (Chapter II) and at the food web level (Chapter III). Later,

13 we scaled up the study system to a larger-scale coastal ecosystem in Chapter IV (Fig. 4). We were not able to measure precise biological processes as in Chapter II and III because of the larger scale of the study. However, since pressing problems in ecology often exist at larger scale and on longer time scales, we can address broader questions when we study processes at a larger ecosystem scale. By scaling up through biological levels in this thesis, we linked ecological processes across different scales of biological levels. We found that processes or interactions at small scale do not necessarily hold at larger scale. By using our framework as guideline for the thesis, we have gained and extended our understanding of the effects of imbalance at the level of the molecular composition of elemental resources on food webs.

Figure 2. Framework for a molecule-to-biosphere stoichiometry theory for Chapter II. The processes and ecological components that are measured and investigated in Chapter II are highlighted in the figure. Modified from Cherif et al (2017).

Figure 3. Framework for a molecule-to-biosphere stoichiometry theory for Chapter III. The processes and ecological components that are measured and investigated in Chapter III are highlighted in the figure. Modified from Cherif et al (2017).

Figure 4. Framework for a molecule-to-biosphere stoichiometry theory for Chapter IV. The processes and ecological components that are measured and investigated in Chapter IV are highlighted in the figure. Modified from Cherif et al (2017).

Materials and Methods

Literature review To offer a methodological framework that can point to the future areas in need of focus for stoichiometric studies at different biological levels, we first conducted a literature search on ecological stoichiometry articles published up to May 2017 in Chapter I. We used the two search engines Web of Science and Scopus with broad search keywords biological stoichiometr∗ or ecological stoichiometr∗ to identify the relevant literature. The search results (> 900 articles) were selectively read depending on the relevance to the framework and were then categorized depending on their contribution to the various components of the framework (Fig. 3). We also summarized the major processes (e.g., ecological interactions, recycling rates) that link different biological levels, many of which are in need of further completion.

Experimental studies Batch culture experiment We grew an isolated bacteria community in batch cultures with different forms of carbon and nitrogen resource additions. In total, we set 3 different treatments in the experiment: 1) DON treatment: adding only dissolved organic nitrogen (DON) in which the C and N are associated in one molecular compound; 2) DIN/DOC treatment: adding both dissolved inorganic nitrogen (DIN) and dissolved organic carbon (DOC) in order to supply the C and N in a disassociated form; 3) COMBINED treatment: adding 50% of the C and N supplied as DON and another 50% as DIN and DOC treatment (Table 1). All the nutrient addition in these three treatments were repeated with two set of supply substrates (see detail in Table 1). The final N concentration for all batch units was kept the same but the C concentration changed between the two different set of supply substrates. The experiment batch units were all kept at 18°C under a 16:8 light: dark cycle for 10 days. During the experiment, nutrients concentration, bacterial abundance, bacterial stoichiometry (Table 3) were measured.

Table 1. Carbon and nitrogen substrates used in the different treatments. Each treatment was carried out with two different sets of molecules (nitrate/pyruvate/alanine, versus nitrate/-ketoglutarate/glutamate) yielding two different total C concentrations (300 µM vs. 500 µM respectively) and two different C:N ratios (S C:N =3 mol:mol vs. 5 mol:mol). N and C concentrations at the start of the experiment are indicated between parentheses.

Total C:N CN ratio molecular Treatments composition DIN/DOC COMBINED DON

- - S C:N =3 N-only Nitrate, NO3 Nitrate, NO3 and (100 µM-N) (50 µM-N)

C-only Pyruvic acid, Pyr Pyruvic acid, Pyr substrates (300 µM-C) (150 µM-C) Dual C-N L-alanine, Ala L-alanine, Ala substrates (50 µM-N, 150 µM-C) (100 µM-N, 300 µM-C)

- - S C:N =5 N-only Nitrate, NO3 Nitrate, NO3 and (100 µM-N) (50 µM-N)

C-only -ketoglutarate, Ket -ketoglutarate, Ket substrates (500 µM-C) (250 µM-C) Dual C-N Glutamate, Glu Glutamate, Glu substrates (50 µM-N, 250 µM-C) (100 µM-N, 500 µM-C)

Microcosm experiment For this study, we isolated a microplankton community from the northern Baltic Sea coast under 50 µm. The microplankton community was collected near Umeå Marine Sciences Center (63° 34’N, 19° 54’E, Bothnian Sea, Sweden) and the water was pre-filtered through 50 µm-mesh filter to exclude large phytoplankton colonies and large zooplankton. In total, 12 microcosms were set up with 2 treatments 1) DON treatment: in this treatment N and C were associated in the form of DON and 2) DIN/DOC treatment: in this treatment N and C were dissociated in the form of DIN and DOC (Table 2). All microcosms were kept in a 15°C climate room under a 12:12 light: dark cycle for 10 days (Fig. 5). The nutrients (C, N, P) were added every second day, starting at day 0 (Table 2). Samples were taken for analyses of water chemistry, phytoplankton/bacteria abundance, phytoplankton/zooplankton taxonomy and C:N:P stoichiometry of different size fractions (Table 3).

Table 2. Nutrient addition in the different treatments in Chapter II.

Treatment/ DIN DOC DON NaH2PO4 · Total C Total N Total P addition (NaNO3) (C5H6O5) (C5H9NO4) H2O addition addition addition DIN/DOC + + + 500 µM 100 µM 20 µM DON + + 500 µM 100 µM 20 µM

Figure 5. Microcosm experiment set-up at the Umeå Marine Sciences Center for Chapter III. (Photo: J.Guo)

Field survey

We conducted a field survey along the northern Baltic Sea coast. Four bays, i.e. Ängerån (AN: 63°34.400N, 19°50.666E), Kalvarsskatan (KA: 63°36.072N, 19°53.140E), Stadsviken (ST: 63°33.026N, 19°47.647E), and Valviken (VA: 63°32.468N, 19°46.725E) were sampled monthly from May to September 2018, where the river inflow to these four bays covered a gradient from no river inflow to larger river inflow. Therefore, our sites covered a gradient in freshwater supply and associated inputs of terrestrial dissolved organic matter and nutrients.

Samples from each bay and month were taken for analyses of water chemistry and C:N:P stoichiometry of different size fractions (Table 3).

Table 3. Trophic composition of the different size fractions used in various chapters.

Size fraction Chapters Organism composition

0.3-0.7 µm (Size1) Chapter II, III, IV Mainly bacteria

0.7-10 µm (Size2) Chapter III, IV Picophytoplankton (picocyanobacteria, picoeukaryotes), nanophytoplankton, heterotrophic nanoflagelates, etc.

10-50 µm (Size3) Chapter III, IV Large nanophytoplankton, microphytoplankton, heterotrophic nanoflagelates, ciliates, rotifers, etc.

>50 µm (Size4) Chapter IV Large microplankton (phyto + zoo), mesozooplankton

Results and discussion

Effect of element chemical forms on bacteria community ammonium regeneration and net nitrogen mineralization In Chapter II, our experiment shows that, beyond elemental stoichiometry, the molecular composition of supply substrates plays a major role in both bacterial ammonium regeneration and net nitrogen mineralization. The substrate composition need to be taken into consideration to have a better understanding of the important role of bacteria in stoichiometric interactions.

In this experiment, we compared the ammonium regeneration to the classical model created by

Goldman (1987), which assumed that the bacteria ammonium regeneration (MN) is highly determined by the supply C:N ratio (SC:N). This model was tested in multiple lab studies examining mineralization and showing that the model is valid in many cases (Goldman et al. 1987; Goldman & Dennett 1991, 2000) and there are many other theoretical studies that assume a similar model for bacterial mineralization (Parnas 1975; Daufresne & Loreau 2001; Cherif & Loreau 2007; Manzoni et al. 2017). However, we found that bacterial ammonium regeneration could not be fully determined by the supply C:N ratio. Growing on substrates with different molecular composition but with similar C:N ratios (SC:N), bacterial communities showed significantly different ammonium regeneration (MN). We also found that the consumption C:N ratio did not reflect the supply C:N ratio as assumed by Goldman’s model, especially in the DIN/DOC and COMBINED treatments. Hence, one of the important assumption of Goldman’s model cannot be meet, which is that the SC:N should completely reflect UC:N. Therefore, we conclude that using SC:N to predict bacterial MN may in many cases not suffice.

Alternatively, we found that using the measured nutrients consumption ratio, UC:N, instead of

SC:N, improved the predictions of MN significantly (Fig. 6A). Similar results were found for prediction of net nitrogen mineralization (NNM) using either the organic substrate C:N ratio

(OC:N) or the organic consumption ratio (UC : UDON) (Fig. 6B). By correcting the model with the consumption C:N ratio, the predictions were substantially closer to measurements for all the treatments, especially for the DON treatment. Based on the findings above, we can conclude that bacteria can actively choose the resource they need when provided with a choice of substrates, instead of passively taking them up. Thus, the molecular diversity of nutrients supplied may play a major role in bacterial mineralization.

Figure 6. Measured A) ammonium regeneration (MN) and B) net nitrogen mineralization (NNM) for all treatments compared to the corresponding MN and NNM predicted from equation (7) and equation (9) using substrate C:N ratio and measured real consumption C:N ratio as estimates of the consumed substrate C:N ratio. The closer a prediction is from the 1:1 grey line, the better it matches the real measurement.

Microplankton food web stoichiometric response to different forms of nitrogen In chapter III, we scaled up a simplified version of the experimental set up of Chapter II, to study a microplankton food web, rather than an isolated bacterial community. Our study indicates that by supplying C and N associated in a single organic molecule (DON treatment) or dissociated in two different molecules (DIN/DOC treatment), the trophic interaction between bacteria and phytoplankton shifts between mutualism and competition respectively. The change in interaction at the base of the food web is then transferred to higher trophic levels, resulting in a remarkable increase in the biomass of heterotrophic grazers (Fig. 7). Further, in the microplankton food web, the structure within trophic levels was significantly different depending on the form of nutrient supply (C and N associated vs dissociated). In the middle of the experiment, the picophytoplankton dominated in both treatments, but much more in the DON treatment. Hence, the picophytoplankton benefitted more from the DON treatment, potentially because it can utilize DON directly (Berman & Bronk 2003; Tarutani et al. 2004; Bronk et al. 2007). However, at the end of experiment, in the DON treatment, the food web was dominated by bacteria, mixotrophs, and other heterotrophic plankton compared to the DIN

21 treatment. There are several possible mechanisms to explain the changes in the food web structure throughout the experiment: 1) the bacteria growth responded to nutrient enrichment with a lag phase, but still performing mineralization that supported phytoplankton growth; 2) some phytoplankton may not only take up DIN but also utilize DON directly; 3) a trophic cascade counterbalanced some of the effects of nutrient enrichment on phytoplankton near the end of the experiment. Differences in 2) and 3) between phytoplankton groups may explain the differences observed in response to the treatments. Our results indicate that available DON may be as important as DIN or total N supply in affecting microplankton food webs.

Contrary to the results from Chapter II, bacterial C:N stoichiometry was significantly different between the two treatments at the end of the experiment, suggesting not only the importance of the form of nutrients supplied but also of trophic interactions in controlling bacterial stoichiometry. We also found that the whole food web net nitrogen use efficiency (NUE) was much higher in the DON treatment than in the DIN treatment (Fig. 8). In the DIN/DOC treatment, bacteria were taking up C and N independently, while C and N uptake was associated in the DON treatment. As bacteria had lower C:N ratio in the DON treatment, bacteria immobilised more N per fixed C. More interestingly, there was no significant difference between treatments in the stoichiometry of the other size fractions, implying that food web NUE might be higher only because N content is higher in bacteria (Del Giorgio & Cole 1998; Fonte et al. 2013). Hence, the role of bacteria for food web nutrient use efficiency might possibly be more central than what was expected.

Figure 7. Food web structure over the experiment in the original water, and the DIN and DON treatments at day 5 and 9. The average carbon biomass of different functional groups is presented within boxes (mgC l-1), the size of the circle is proportional to the carbon biomass. The grey area represents autotrophs and the white area represents heterotrophs. BA=bacteria, Pico=picophytoplankton, AU=autotrophs, MX=mixotrophs, HT= heterotrophs, ZP= microzooplankton.

Figure 8: The whole food web net resource use efficiency (RUE) (mean ± se) between treatments. The black bar represents the DON treatment and the grey bar represents the DIN treatment. NUE= nitrogen use efficiency, PUE= phosphorus use efficiency.

Stoichiometric response of a coastal food web to temporal and spatial gradient In Chapter IV, our field data showed that the stoichiometry of the seston and the zooplankton varied temporally more than spatially. The spatiotemporal stoichiometric changes of the seston were size fraction-specific and especially significant during the transition from May to June. Compared to a C:N ratio that was relatively stable, the temporal variability in C:P and N:P ratios of different seston size fractions was much higher (Fig. 9). This variation could result from differences in terrestrial and freshwater nutrient inputs during the spring. However, we found that the zooplankton C:N:P composition leaned towards homeostasis within specific months, despite changes in taxonomic composition through time. We hypothesize that the spatiotemporal variation in the C:N:P ratio in the seston may have induced resource mismatch for the zooplankton, and could likely have constrained the transfer and recycling of nutrients in this coastal microplankton food web.

Figure 9. Molar C:N (A), C:P (B), and N:P (C) ratios (mean ± se) of different seston size fractions and zooplankton from the study bays in May-September. All ratios were log-transformed prior to plotting.

Moreover, our results suggest that DOC and DON were important drivers for the seston C:N stoichiometric changes (Fig. 10). DOC mainly drove the stoichiometry of seston fraction Size 3 (10-50 µm), and DON mainly drove the stoichiometry of seston fractions Size 2 (0.7-10 µm) and Size 4 (>50 µm). In our study sites, the freshwater inputs highly increase the DOC concentration in the water column, which has strong effect on seston C:nutrients ratios of Size 3 fraction.

Figure 10. Results of redundancy analyses (RDAs) on seston C:N, C:P and N:P with spatiotemporal changes in physicochemical characteristics of the study sites in June-September 2018. The RDAs were performed on centred- reduced variables. The numbers after the ratios indicate the specific seston size fractions.

Conclusion

In this thesis, we have investigated one of the knowledge gap in the ecological stoichiometry theory, the role of the molecular composition of supplied substrates on food web interactions and stoichiometric responses. I found that the molecular composition of C and N in the supplied substrates can significantly affect the bacteria community ammonium regeneration and net N mineralization (Chapter II). Hence the response of bacteria to molecular composition effects at the basis of food webs is important. As a result, when C and N are supplied in one molecule, the bacteria-autotrophs interaction is shifted towards mutualism, while, when C and N are supplied as two separate molecules, the interaction is pushed towards competition. Changes in this interaction at the basis of the food web then significantly influence the whole food web structure and the production of higher trophic levels (Chapter III). However, in natural systems, where there are multiple resources with various degrees of association between C and N, the effects of molecular composition become harder to predict, especially given that the organic resources from terrestrial runoff are not fully bioavailable to organisms (Chapter IV). Besides, other important processes in natural ecosystems, listed in the framework (Fig. 1) (Chapter I), are likely to affect the stoichiometric responses of food web components through changes in the imbalance between consumers and their resources.

Significance and further directions

To adequately predict stoichiometric responses in a changing aquatic environment, improved methodology is needed in order to acquire data from food webs that are appropriate to determine trophic interactions. To our knowledge, Chapter II, III and IV are the first few studies collecting data on bacterial stoichiometry in the size range of 0.3-0.7 µm both in lab experiment and field survey. In previous studies, the range of sizes for bacterial stoichiometric analyses collected in lab or field studies was usually between 0.7-1 µm due to the filtration method used (Makino et al. 2003; Cotner et al. 2010; Hall et al. 2011; Scott et al. 2012; Stenzel et al. 2017). However, in natural waters, the size range 0.7-1 µm commonly contains a high proportion of picophytoplankton (generally in a size range from 0.5 to 2 µm) and make such bacteria stoichiometric measurement possibly incorrect. By using the new size range 0.3-0.7 µm, we captured CNP composition of bacteria out of line with other studies, especially the ones that studied natural bacterial assemblage (Scott et al. 2012; Stenzel et al. 2017). For example, our results from field study (Chapter IV) showed that bacteria C:P and N:P ratios are much lower than Redfield ratio (C:N:P=106:16:1), but other studies in both freshwater and marine ecosystem found that the C:P and N:P ratio were higher than Redfield ratio.

Another obstacle to stoichiometric studies of natural aquatic systems is that not enough stoichiometric data are collected at the species and individual levels. In Chapter III, we picked individuals of the zooplankton species present (e.g. Eurytemora sp., Acartia sp., Bosmina sp.) under microscope and measured their stoichiometry separately. Most field studies of zooplankton stoichiometry only analyze bulk zooplankton samples. Our method may give more insight into the stoichiometric interactions involving zooplankton.

As bacteria are important DIN regenerators for phytoplankton, and zooplankton are important food source for upper trophic levels, e.g. fish, a focus on these trophic levels will allow us to better understand the stoichiometric responses of these food webs to a changing environment. More important, in this thesis, we combined multiple community ecology theories and environmental factors in order to improve our understanding of stoichiometric response in natural assemblages (Chapter II, III and IV), filling some of the knowledge gaps identified in Chapter I, where integration of stoichiometric with processes at the population and community levels appear to be the most urgently needed.

Finally, we have tried to use the mechanisms that we found in the controlled lab experiments to explain what happens in natural systems. However, how to measure the stoichiometry of specific levels or ecological processes (e.g. ammonium regeneration, nutrients consumption) in the field is not obvious. Therefore, in the future, new methodologies (e.g., X-ray microanalysis and secondary mass spectrometry, isotope dilution) should be used to target more specifically the components of food web trophic interactions that will help us linking different biological theories to ecological stoichiometry, in order to gain a better understanding of natural ecosystems.

Acknowledgements

I thank Sonia Brugel, Sebastian Diehl, Mehdi Cherif for providing useful comments on this kappa.

References

Andersen, T. & Hessen, D.O. (1991). Carbon, nitrogen, and phosphorus content of freshwater zooplankton. Limnol. Oceanogr., 36, 807–814. Andersson, A., Meier, H.E.M., Ripszam, M., Rowe, O., Wikner, J., Haglund, P., et al. (2015). Projected future climate change and Baltic Sea ecosystem management. Ambio, 44, 345– 356. Atkinson, C.L., Capps, K.A., Rugenski, A.T. & Vanni, M.J. (2017). Consumer-driven nutrient dynamics in freshwater ecosystems: from individuals to ecosystems. Biol. Rev., 92, 2003– 2023. Barker, H.A. (1981). Amino acid degradation by anaerobic bacteria. Annu. Rev. Biochem., 50, 23–40. Bates, B.C., Z.W. Kundzewicz, S., J.P., W. & Palutikof, E. (2008). Climate change and water. IPCC Secretariat, Geneva. Berman, T. & Bronk, D. (2003). Dissolved organic nitrogen: a dynamic participant in aquatic ecosystems. Aquat. Microb. Ecol., 31, 279–305. Bratbak, G. & Thingstad, T.F. (1985). Phytoplankton-bacteria interactions - an apparent paradox - Analysis of a model system with both competition and commensalism. Mar. Ecol. Prog. Ser., 25, 23–30. Bronk, D.A., See, J.H., Bradley, P., Killberg, L., Bronk, D.A., See, J.H., et al. (2007). DON as a source of bioavailable nitrogen for phytoplankton. Biogeosciences, Eur. Geosci. Union, 4, 283–296. Cherif, M., Faithfull, C., Guo, J., Meunier, C.L., Sitters, J., Uszko, W., et al. (2017). An operational framework for the advancement of a molecule-to-biosphere stoichiometry theory. Front. Mar. Sci., 4. Cherif, M. & Loreau, M. (2007). Stoichiometric constraints on resource use, competitive interactions, and elemental cycling in microbial decomposers. Am. Nat., 169, 709–724. Cherif, M. & Loreau, M. (2009). When microbes and consumers determine the limiting nutrient of autotrophs: a theoretical analysis. Proceedings. Biol. Sci., 276, 487–497. Chiandet, A.S. & Xenopoulos, M.A. (2011). Landscape controls on seston stoichiometry in urban stormwater management ponds. Freshw. Biol., 56, 519–529. Cotner, J.B. & Biddanda, B.A. (2002). Small players, large role: Microbial influence on biogeochemical processes in pelagic aquatic ecosystems. Ecosystems. Cotner, J.B., Hall, E.K., Scott, J.T. & Heldal, M. (2010). Freshwater bacteria are

stoichiometrically flexible with a nutrient composition similar to seston. Front. Microbiol., 1, 132. Creed, I.F., Bergström, A.K., Trick, C.G., Grimm, N.B., Hessen, D.O., Karlsson, J., et al. (2018). Global change-driven effects on dissolved organic matter composition: Implications for food webs of northern lakes. Glob. Chang. Biol., 24, 3692–3714. Cross, W.F., Benstead, J.P., Rosemond, A.D. & Wallace, J.B. (2003). Consumer-resource stoichiometry in detritus-based streams. Ecol. Lett., 6, 721–732. Danger, M., Leflaive, J., Oumarou, C., Ten-Hage, L. & Lacroix, G. (2007). Control of phytoplankton-bacteria interactions by stoichiometric constraints. Oikos, 116, 1079–1086. Daufresne, T. & Loreau, M. (2001). Ecological stoichiometry, primary producer-decomposer interactions, and ecosystem persistence. Ecology, 82, 3069–3082. Deininger, A., Faithfull, C.L., Lange, K., Bayer, T., Vidussi, F. & Liess, A. (2016). Simulated terrestrial runoff triggered a phytoplankton succession and changed seston stoichiometry in coastal lagoon mesocosms. Mar. Environ. Res., 119, 40–50. DeMott, W.R. & Pape, B.J. (2005). Stoichiometry in an ecological context: Testing for links between Daphnia P-content, growth rate and habitat preference. Oecologia, 142, 20–27. Elser, J.J., Elser, M.M., MacKay, N.A. & Carpenter, S.R. (1988). Zooplankton-mediated transitions between N- and P-limited algal growth1. Limnol. Oceanogr., 33, 1–14. Elser, J.J., Frost, P., Kyle, M., Urabe, J. & Andersen, T. (2002). Effects of light and nutrients on plankton stoichiometry and biomass in a P-limited lake. Hydrobiologia, 481, 101–112. Elser, J.J. & Urabe, J. (1999). The stoichiometry of consumer-driven nutrient recycling: Theory, observations, and consequences. Ecology, 80, 735–751. Finkel, Z. V., Beardall, J., Flynn, K.J., Quigg, A., Rees, T.A. V. & Raven, J.A. (2010). Phytoplankton in a changing world: Cell size and elemental stoichiometry. J. Plankton Res. Fonte, E.S., Amado, A.M., Meirelles-Pereira, F., Esteves, F.A., Rosado, A.S. & Farjalla, V.F. (2013). The combination of different carbon sources enhances bacterial growth efficiency in aquatic ecosystems. Microb. Ecol., 66, 871–878. Frost, P.C., Evans-White, M.A., Finkel, Z. V., Jensen, T.C. & Matzek, V. (2005). Are you what you eat? Physiological constraints on organismal stoichiometry in an elementally imbalanced world. In: Oikos. pp. 18–28. Gale, E.F. (1947). The assimilation of amino-acids by bacteria; the passage of certain amino- acids across the cell wall and their concentration in the internal environment of Streptococcus faecalis. J. Gen. Microbiol., 1, 53–76.

Del Giorgio, P.A. & Cole, J.J. (1998). Bacterial growth efficiency in natural aquatic systems. Annu. Rev. Ecol. Syst., 29, 503–541. Glibert, P.M., Wilkerson, F.P., Dugdale, R.C., Raven, J.A., Dupont, C.L., Leavitt, P.R., et al. (2016). Pluses and minuses of ammonium and nitrate uptake and assimilation by phytoplankton and implications for productivity and community composition, with emphasis on nitrogen-enriched conditions. Limnol. Oceanogr., 61, 165–197. Goldman, J.C., Caron, D.A. & Dennett, M.R. (1987). Regulation of gross growth efficiency and ammonium regeneration in bacteria by substrate C: N ratio. Limnol. Oceanogr., 32, 1239–1252. Goldman, J.C. & Dennett, M.R. (1991). Ammonium regeneration and carbon utilization by marine bacteria grown on mixed substrates. Mar. Biol., 109, 369–378. Goldman, J.C. & Dennett, M.R. (2000). Growth of marine bacteria in batch and continuous culture under carbon and nitrogen limitation. Limnol. Oceanogr., 45, 789–800. Golz, A.L., Burian, A. & Winder, M. (2015). Stoichiometric regulation in micro- and mesozooplankton. J. Plankton Res., 37, 293–305. Hall, E.K., Maixner, F., Franklin, O., Daims, H., Richter, A. & Battin, T. (2011). Linking Microbial and Ecosystem Ecology Using Ecological Stoichiometry: A Synthesis of Conceptual and Empirical Approaches. Ecosystems, 14, 261–273. Hessen, D.O., Elser, J.J., Sterner, R.W. & Urabe, J. (2013). Ecological stoichiometry: An elementary approach using basic principles. Limnol. Oceanogr., 58, 2219–2236. Hillebrand, H., Steinert, G., Boersma, M., Malzahn, A., Meunier, C.L., Plum, C., et al. (2013). Goldman revisited: Faster-growing phytoplankton has lower N: P and lower stoichiometric flexibility. Limnol. Oceanogr., 58, 2076–2088. Kendrick, M.R. & Benstead, J.P. (2013). Temperature and nutrient availability interact to mediate growth and body stoichiometry in a detritivorous stream insect. Freshw. Biol., 58, 1820–1830. Kirchman, D.L. (1994). The uptake of inorganic nutrients by heterotrophic bacteria. Microb. Ecol., 28, 255–271. Kirchman, D.L. & Wheeler, P.A. (1998). Uptake of ammonium and nitrate by heterotrophic bacteria and phytoplankton in the sub-Arctic Pacific. Deep. Res. Part I Oceanogr. Res. Pap., 45, 347–365. Kjellström, E. & Ruosteenoja, K. (2007). Present-day and future precipitation in the Baltic Sea region as simulated in a suite of regional climate models. Clim. Change, 81, 281–291. Kohler, T.J., Heatherly, T.N., El-Sabaawi, R.W., Zandonà, E., Marshall, M.C., Flecker, A.S.,

et al. (2012). Flow, nutrients, and light availability influence Neotropical epilithon biomass and stoichiometry. Freshw. Sci., 31, 1019–1034. Loladze, I., Kuang, Y., Elser, J.J. & Fagan, W.F. (2004). Competition and stoichiometry: coexistence of two predators on one prey. Theor. Popul. Biol., 65, 1–15. Makino, W., Cotner, J.B., Sterner, R.W. & Elser, J.J. (2003). Are bacteria more like plants or animals? Growth rate and resource dependence of bacterial C:N:P stoichiometry. Funct. Ecol., 17, 121–130. Manzoni, S., Čapek, P., Mooshammer, M., Lindahl, B.D., Richter, A. & Šantrůčková, H. (2017). Optimal metabolic regulation along resource stoichiometry gradients. Ecol. Lett., 20, 1182–1191. McManamay, R.A., Webster, J.R., Valett, H.M. & Dolloff, C.A. (2011). Does diet influence consumer nutrient cycling? Macroinvertebrate and fish excretion in streams. J. North Am. Benthol. Soc., 30, 84–102. Meier, H.E.M., Hordoir, R., Andersson, H.C., Dieterich, C., Eilola, K., Gustafsson, B.G., et al. (2012). Modeling the combined impact of changing climate and changing nutrient loads on the Baltic Sea environment in an ensemble of transient simulations for 1961-2099. Clim. Dyn., 39, 2421–2441. Moschonas, G., Gowen, R.J., Paterson, R.F., Mitchell, E., Stewart, B.M., McNeill, S., et al. (2017). Nitrogen dynamics and phytoplankton community structure: the role of organic nutrients. Biogeochemistry, 134, 125–145. Naganuma, T., Iinuma, Y., Nishiwaki, H., Murase, R., Masaki, K. & Nakai, R. (2018). Enhanced bacterial growth and gene expression of D-amino acid dehydrogenase With D- glutamate as the sole carbon source. Front. Microbiol., 9. Parnas, H. (1975). Model for decomposition of organic material by microorganisms. Soil Biol. Biochem., 7, 161–169. Persson, J., Fink, P., Goto, A., Hood, J.M., Jonas, J. & Kato, S. (2010). To be or not to be what you eat: Regulation of stoichiometric homeostasis among autotrophs and heterotrophs. Oikos, 119, 741–751. Reusch, T.B.H., Dierking, J., Andersson, H.C., Bonsdorff, E., Carstensen, J., Casini, M., et al. (2018). The Baltic Sea as a time machine for the future coastal ocean. Sci. Adv., 4, eaar8195. Rugenski, A.T., Múrria, C. & Whiles, M.R. (2012). Tadpoles enhance microbial activity and leaf decomposition in a neotropical headwater stream. Freshw. Biol., 57, 1904–1913. Sardans, J., Rivas-Ubach, A. & Peñuelas, J. (2012). The elemental stoichiometry of aquatic and

terrestrial ecosystems and its relationships with organismic lifestyle and ecosystem structure and function: A review and perspectives. Biogeochemistry, 111, 1–39. Scott, J.T., Cotner, J.B. & LaPara, T.M. (2012). Variable stoichiometry and homeostatic regulation of bacterial biomass elemental composition. Front. Microbiol., 3, 42. De Senerpont Domis, L.N., Van De Waal, D.B., Helmsing, N.R., Van Donk, E. & Mooij, W.M. (2014). Community stoichiometry in a changing world: Combined effects of warming and eutrophication on phytoplankton dynamics. Ecology, 95, 1485–1495. Stenzel, B., Rofner, C., Pérez, M.T. & Sommaruga, R. (2017). Stoichiometry of natural bacterial assemblages from lakes located across an elevational gradient. Sci. Rep., 7, 5875. Sterner, R.. & Elser, J.. (2002). Ecological stoichiometry: the biology of elements from molecules to the biosphere. Sterner, R.W. (1990). The ratio of nitrogen to phosphorus resupplied by herbivores - zooplankton and the algal competitive arena. Am. Nat., 136, 209–229. Stets, E.G. & Cotner, J.B. (2008). The influence of dissolved organic carbon on bacterial phosphorus uptake and bacteria-phytoplankton dynamics in two Minnesota lakes. Limnol. Oceanogr., 53, 137–147. Tarutani, K., Niimura, Y. & Uchida, T. (2004). Short-term uptake of dissolved organic nitrogen by an axenic strain of Ulva pertusa (Chlorophyceae) using 15N isotope measurements. Bot. Mar., 47, 248–250. Touratier, F., Legendre, L. & Vezina, A. (1999). Model of copepod growth influenced by the food carbon:nitrogen ratio and concentration, under the hypothesis of strict homeostasis. J. Plankton Res., 21, 1111–1132. Trautwein, K., Feenders, C., Hulsch, R., Ruppersberg, H.S., Strijkstra, A., Kant, M., et al. (2017). Non-Redfield, nutrient synergy and flexible internal elemental stoichiometry in a marine bacterium. FEMS Microbiol. Ecol., 93. Tupas, L.M., Koike, I., Karl, D.M. & Holm-Hansen, O. (1994). Nitrogen metabolism by heterotrophic bacterial assemblages in Antarctic coastal waters. Polar Biol., 14, 195–204. Urabe, J. (1993). N and P cycling coupled by grazers’ activities: Food quality and nutrient release by zooplankton. Ecology, 74, 2337–2350. Vanni, M.J. (2002). Nutrient cycling by animals in freshwater ecosystems. Annu. Rev. Ecol. Syst. Weber, T.S. & Deutsch, C. (2010). Ocean nutrient ratios governed by plankton biogeography. Nature, 467, 550–554. Yvon-Durocher, G., Schaum, C.E. & Trimmer, M. (2017). The temperature dependence of

phytoplankton stoichiometry: Investigating the roles of species sorting and local adaptation. Front. Microbiol., 8, 2003. Zeng, J., Zhao, D.-Y., Liu, P., Yu, Z.-B., Huang, R. & Wu, Q.L. (2014). Effects of benthic macrofauna bioturbation on the bacterial community composition in lake sediments. Can. J. Microbiol., 60, 517–524.

Thanks

Finally, after all those years of work for my PhD thesis, I would like to extend thanks to many people, in many countries, who contributed so generously to the work presented in this thesis. First and foremost, I would like to thank my family for all their support and love during all these years. Mom and Dad, thanks for raising me up and always supporting my choice of pursuing. Thanks to my mother in law, father in law and grandma in law, for their endless support and encouragement. Most of all to my loving daughter Yimo and my patient husband Renyuan. They are the most important people in my world and I dedicate this thesis to them.

Here special thanks to my supervisor Mehdi Cherif. It has been an honor to be his first Ph.D. student. I appreciate all his contribution in time and ideas to my Ph.D. My co-supervisors, Sebastian and Antonia, thanks for all the help though my PhD study.

The members of the Marine group have contributed immensely to my personal and professional time at Umeå University. The group has been a source of friendships as well as good advice and collaboration. Agneta, who gave me generous help in both funding and writing. These help contributed to half of my PhD studies. Without you, it would not have been possible for me to finish my PhD thesis. Sonia, my honest and selfless friend. Your help and company means a lot to my PhD and my personal life. You guide me all the way through the darkest period. You deserve half of my PhD!!! Owen, you did not only help me through my master but also through my PhD. You are always there and give good solution for any of my problems. Danny, you gave me many good suggestions on my thesis and always offered such good help even before I asked for it. Thanks to Jon, Anki, Ryan and Tomas for their support and suggestions for my PhD study, it meant a lot to me. Thanks to the UMF people, who helped me finishing my last two studies and it was such an honor to work with you.

Also, many thanks to many colleagues Anders, Joanna, Hélène, Xiaoru, Doreen, Marcus, Maria, Johans, Li, Wei, Yuqing, Jing, Biyue, Baosheng, Jin, Xi for good company and advice at EMG. Besides academic work, I spent such a good time on teaching allocation with Anne and Svetlana, thanks a lot for all the hard work we did together.

Special thanks to all my friends outside of EMG. Keni, Björn, Nongfei, Hanqing, Ning, Qun, Hongyu, Ying, Xiaoyan, you all contributed to enjoy my PhD life. Thank you!