Effects of temperature and terrestrial carbon on fish growth and pelagic food web efficiency Robert Lefébure

Department of Ecology and Environmental Science Umeå University 901 87 Umeå Umeå 2012

Copyright©Robert Lefébure ISBN: 978-91-7459-412-6 Printed by: Print & Media Umeå, Sweden 2012

2

To Annie

3

Men wanted for hazardous journey! Small wages, bitter cold, long months of complete darkness, constant danger, safe return doubtful. Honor and recognition in case of success. -

Sir Ernest Shackleton

4

LIST OF PAPERS This thesis is a summary and discussion of the following papers, which are referred to by their roman numerals.

I.

Lefébure R., S. Larsson and P. Byström, 2011. A temperature dependent growth model for the three-spined stickleback (Gasterosteus aculeatus), Journal of Fish Biology, 79: 1815-1827

II.

Lefébure R., P. Byström and S. Larsson, 2012. Temperature and size dependent attack rates of the three-spined stickleback (Gasterosteus aculeatus); are sticklebacks in the Baltic Sea food-limited? (Submitted Manuscript)

III.

Degerman R., R.Lefébure, P. Byström, U. Båmstedt , S. Larsson, L-O. Eriksson and A. Andersson, 2012. Bottom up and top down control of pelagic food web efficiencies. (Manuscript)

IV.

Lefébure R, R.Degerman, P. Byström, S. Larsson, A. Andersson , L-O Eriksson and U.Båmstedt, 2012. Impacts of elevated terrestrial nutrient loads and temperature on pelagic food web efficiency and fish production. (Submitted Manuscript).

Paper I has been reprinted with kind permission from the publisher

5

Table of Contents ABSTRACT

7

INTRODUCTION

8

Effects of terrestrial carbon on planktonic food web dynamics

8

Effects of temperature on planktonic food webs

10

Physiological changes in fish in response to temperature

10

THE BALTIC SEA AND PREDICTED IMPACTS OF CLIMATE CHANGE

12

THE THREE - SPINED STICKLEBACK, GASTEROSTEUS ACULEATUS

13

AIM

14

MATERIALS AND METHODS

15

MAIN RESULTS

18

CONCLUSIONS

20

The thermal response of the three-spined stickleback; implications of climate change for sticklebacks in the Baltic Sea 20 Impacts of climate change on pelagic food web structure and function; implications for fish production in the Baltic Sea 21 THANKS

23

REFERENCES

24

AUTHOR CONTRIBUTIONS

28

6

ABSTRACT Both temperature and terrestrial dissolved organic carbon (TDOC) have strong impacts on aquatic food web dynamics and production. Temperature affects vital rates of all organisms and terrestrial carbon has been shown to alter the dynamics of phytoplankton and bacterial production and affect the trophic structure of planktonic food webs. As climate change predictions for the Baltic Sea suggests future increases in both terrestrial carbon run-off and increases in temperature, the aim of thesis was to adopt a system-ecological approach and study effects of these abiotic variables, not only on interactions within planktonic food webs, but also on the growth and consumption rates of one of the most common zooplanktivorous fish in the Baltic Sea, the three-spined stickleback Gasterosteus aculeatus. Results showed that three-spined sticklebacks display a high degree of resilience against increasing temperatures, as both growth rates as well as consumption rates on zooplankton were high at temperatures well over 20 °C. Furthermore, it was shown that the minimal resource densities required to sustain individual and population growth, actually decreased with increasing temperatures, implying that sticklebacks around their optimum temperature for growth at 21 °C will actually have an increased scope for growth. As stickleback population densities have increased over the last decade in the Baltic Sea and are now suggested to out-compete other coastal fish species for shared zooplankton resources, the results presented in this thesis suggest that increased water temperatures would only serve to increase sticklebacks competitive advantage. As the structuring role of this small zooplanktivore on pelagic communities might be considerable, further studies investigating competitive interactions as well as patterns of population abundances are definitely warranted. TDOC was overall shown to stimulate bacterial production and the microbial food web. Because of the longer trophic pathways required to transport carbon from bacterial production to higher trophic levels, the addition of TDOC always reduced food web transfer efficiency. However, it became apparent that the full effect of TDOC additions on pelagic food webs was complex and depended heavily not only on the existing trophic structure to which the carbon was introduced, but also on ambient temperature levels. When three-spined sticklebacks were part of food webs with significant TDOC inputs, the presence of fish, indirectly, through predator release of lower trophic levels, amplified the magnitude of the effects of carbon addition on bacterial production, turning the base of the system significantly more heterotrophic, which ultimately, impacted negatively on their own production. However, when a pelagic food web containing sticklebacks was simultaneously subjected to realistic increases in temperature and TDOC concentrations, food web efficiency and fish production increased compared to present day conditions. These results were explained by a temperature dependent increased production potential of zooplankton, sustained by an increased production of heterotropic microzooplankton via TDOC additions, which lead to higher fish production. Although the increased number of trophic linkages in heterotrophic food webs should have reduced energy transfer efficiency, these negative effects seem here to have been overridden by the positive increases in zooplankton production as a result of increased temperature. These results show that heterotrophic carbon transfer can be a viable pathway to top-consumers, but also indicates that in order to understand the full effects of climate change on trophic dynamics and fish production, abiotic variables cannot be studied in isolation.

7

INTRODUCTION

Effects of terrestrial carbon on planktonic food web dynamics Due to technological advances that allowed for the identification of microbes in marine ecosystems, including viruses, bacteria, algae and heterotrophic protists (such as ciliates and flagellates), the image of the pelagic food web (See Fig 1) has undergone drastic change in recent decades. Initially, this compartment of the food web was called a “microbial loop” (Pomeroy, 1974) while it has now become more customary to call it a “microbial food web” to describe the combination of trophic interactions at this basal microbial level (Azam et al, 1983).

Fig. 1. Generalized view of the pelagic food web, including interactions on the microbial level and their role in recycling DOC (Modified from Sommer et al, 2002).

Simplified, the microbial food web describes lower level trophic pathways in aquatic environments and its main function is to recycle dissolved organic carbon (DOC) back into higher levels of the food web via the incorporation of DOC into bacterial biomass. It has been shown that bacteria are particularly efficient at assimilating this carbon, thanks chiefly to their small size and large surface-to-volume ratio, which allows them to absorb nutrients at very low concentrations (Vaccero et al, 1977; Larsson and Hagström, 1979). This gives them a competitive advantage over phytoplankton and as a result promotes their production (Azam et al, 1983). However, they are too small to be captured by mesozooplankton and planktonic larvae. Instead, these cells are consumed by heterotrophic protists, such as flagellates and

8

ciliates, which in turn can be preyed upon by larger organisms and thus, DOC is made available to higher levels of the food chain (Azam et al, 1983). DOC in aquatic environments either derives from within the system or through the transportation of organic matter from the terrestrial watershed, via rivers. Carbon of terrestrial origin is labeled terrigenous dissolved organic carbon (TDOC) or allochtonous dissolved organic carbon (ADOC), in contrast to DOC produced within the system; the autochthonous supply (Eby, 2004). This in-situ carbon production can come from several sources, such as the excretion of waste products by aquatic animals and microbes or by the breakdown and dissolution of organic particles (Cole et al, 1982; Strom et al, 1997). The main autochtnonus source of DOC production in the ocean is from photosynthetic algae (Cole et al, 1982). During carbon fixation by algae undertaking photosynthesis, a fraction of carbon is released as dissolved organic carbon (e.g. Larsson and Hagström, 1979; Wangersky, 2000). This release goes on to be used by bacteria in the microbial loop (Larsson and Hagström, 1979; Azam et al, 1983). As bacteria are efficient in assimilating carbon (Larsson and Hagström, 1979), the extracellular release by phytoplankton is considered to be one of the primary sources for bacterial growth, with bacterial production values usually averaging about 20 % of the total basal primary production (Cole et al, 1988). The ability of bacteria to use carbon of terrestrial origin to supplement their growth was initially overlooked, most likely because early concepts of aquatic microbial ecology arose from research conducted on open oceans and not coastal or inland waters (Tranvik, 1992). However, it is now accepted that humic riverine organic carbon contains readily available carbon and that when a pelagic system is influenced by sources of terrestrial dissolved organic carbon (TDOC), the microbial food web can play an increasingly larger role in determining overall carbon production at the base of the food web, through bacterial heterotrophic production (Tranvik 1988; Jansson et al, 2000; Jansson et al, 2007). In addition, it has been shown that it can support growth of bacteria to such a degree that it can even uncouple them from the need for autochthonous DOC from phytoplankton (Tranvik, 1992; Blomqvist et al, 2001; Sandberg et al, 2004; Berglund et al, 2007). If a pelagic system is driven primarily by bacterial production, the carbon produced at the base of the food web will have to pass through the entire microbial food web, chiefly mediated by heterotrophic protists such as flagellates and ciliates, before reaching top consumers (Azam et al, 1983; Breteler et al, 1999). The increase in food chain length ultimately means that less energy is available for higher trophic levels, a consequence of a loss of carbon of about 70 – 90 % at each additional step in the food web, due amongst other things to respiration and sloppy feeding (Straile, 1997; Strom et al, 1997; Sommer et al, 2002). These food web dynamics are characterized and measured as food web efficiency or ecological efficiency and is defined as the ratio between the productivity at the highest trophic level and the productivity at the base level (Rand and Stewart, 1998). In a study estimating food web efficiency in two contrasting food webs (one phytoplankton-based and one bacterial-based) in the Baltic Sea, it was shown that the FWE was 22% in the phytoplanktonbased food web and 2% in the bacterial-based food web. This discrepancy was explained by the 1–2 extra trophic levels in the bacterial-based food web through which carbon had to pass (via flagellates and ciliates) before reaching mesozooplankton (Berglund et al, 2007). In aquatic systems with high inputs of TDOC, other factors are of consequence for the dynamics between bacterial and phytoplankton productivity. In the photic zone, the relationship between phyto- and bacterioplankton is one of constant competition for inorganic nutrients (Blomqvist et al, 2001). Because of their high uptake capacity for nutrients, 9

bacterioplankton are believed to triumph, and if present in sufficient numbers, might out compete phytoplankton (Azam et al, 1983; Blomqvist et al, 2001). Furthermore, a major share of TDOC is made up of coloured substances, decreasing light availability (and quality) required for photosynthesis, which may lead to detrimental conditions for phytoplankton (Klug et al, 2002). Thus, the potential effects of ADOC on primary producers and on phytoplankton–bacterioplankton relationships are numerous, with increasing consensus being built that organic carbon can indeed play an important role in determining the structure and function of pelagic food webs (Blomqvist et al, 2001; Berglund et al, 2007; Jansson et al, 2007).

Effects of temperature on planktonic food webs Temperature affects all levels of the pelagic food web due to its intrinsic role in all vital rates of the food web organisms. At the base of the food web, temperature can influence ratios of bacterial and phytoplankton growth, higher temperature favoring bacteria, therefore altering the heterotrophic: autotrophic production ratio (Müren et al, 2005; Hoppe et al, 2008). With increasing temperatures, the species composition of phytoplankton communities have also been shown to change from a dominance of relatively large diatoms and dinoflagellates at low temperatures to a dominance of small algae (pico- and nano-plankton) at higher temperatures (Andersson et al, 1994). In addition, increased temperatures favour the productivity of heterotrophic protists, like flagellates and ciliates, and at higher trophic levels both development times and total biomass of zooplankton are affected (Heinle and Flemer 1975; Dippner et al, 2000). Therefore, with increasing temperatures, a pelagic system can move in increasing degrees towards heterotrophy, implying that moderately elevated seawater temperatures may affect the entire ecosystem functioning by favoring bacterial productivity and altering the balance between autotrophy and heterotrophy (Müren et al, 2005).

Physiological changes in fish in response to temperature

Of all the abitioc factors, changes in ambient water temperature has the largest effect on physiological properties in fish (Brett, 1979). Since fish in general are ectotherms, increases in ambient temperatures will lead to increases of their metabolic rates (Elliot, 1976) and these will translate to a need to increase their consumption rates to meet these demands (Jobling, 1995). Therefore, temperature will have a strong impact on predator prey interactions as it will affect all aspects of the predation cycle, including encounter rate, prey avoidance capacity, capture success and handling time (Persson, 1986; Persson et al, 1998; Wahlström et al, 2000; Englund et al, 2011). Temperature will also affect all aspects of fish physiology and dictate fundamental properties of the energy budget, metabolic demands, digestion rates and assimilation efficiencies (Werner, 1994; Jobling, 1995; Persson et al, 1998; Byström et al, 2006; Englund et al, 2011). Just as temperature affects consumption rates, growth rates of fish are intimately connected to ambient temperature levels. For most fish species increases in growth rates with increasing temperatures will be seen, up to a certain point, only to decline abruptly once the critical limit 10

of the species is reached (Brett 1979; Elliot, 1994; Jobling, 1995). The temperature peak at which growth is maximized is called the “optimum temperature for growth” and is usually a few degrees lower than the temperature at which food intake is the greatest. There are large variations in the optimum temperature for growth between species. Many salmonids, for instance, have their temperature optima in the range of 12 – 17 °C (e.g. Elliott and Hurley, 1995; Elliott and Hurley, 1997) in contrast to cyprinids which have optima of 20 °C or higher (Jobling, 1995). In addition, ontogenetic intra-specific differences also exist, with juveniles often having higher optima for growth than adults (Jobling, 1995). This shows that species are adapted to different temperature tolerance ranges and that typically energy allocation towards growth will decline once temperatures approach the range extremes (Brett, 1979). However, patterns of growth are strongly correlated to the available food supply and restricted feeding possibilities will have a marked influence on growth rates at any observed temperature.

Production E (P) = E( IN) – E (OUT)

The most detailed experimental studies on the effects of temperature and food ration on fish growth have been carried out on salmonids, e.g. Atlantic salmon Salmo salar (Forseth et al, 2001), brown trout Salmon trutta (reviewed in Elliot, 1994) and Arctic charr Salvelinus alpinus (Larsson and Berglund, 1998). These allow scientists to determine at which temperature food conversion is the most efficient, whilst simultaneously allowing for predictions of temperature related optimal growth at any given ratio level (Fig. 2).

4

3

2

1

Temperature (ºC)

Fig. 2. Influence of temperature on growth (production) of fish held under different conditions of food restriction. (1) unlimited food supply, (2) slight restriction (3) moderate restriction and (4) low level of food supply. The arrowheads indicate the temperatures at which growth rates will be greatest under the different regimes. (Modified from Jobling, 1995).

Under conditions of limited food availability there will be marked changes in the relationship between growth and temperature as opposed to under unrestricted food supply conditions. In the latter scenario, increased temperature would lead to increased growth up to the optimum and a subsequent decline towards a lethal point. However, if food is a limiting factor, a shift of the optimum growth towards increasingly lower temperatures will be seen. This shift 11

occurs because higher temperatures lead to higher metabolic rates in fish, thereby increasing their food demand as required to maintain a constant body weight, lessening their available scope for growth (Elliot, 1976; Brett, 1979; Jobling, 1995). Therefore, at low food allowances, it is more beneficial to be at low temperatures, where metabolism is low thus leaving a greater surplus of energy available for growth. Inversely, at high temperatures with low food availability, the high metabolic demands of the fish will lead to fish losing weight and eventually starving to death (Brett et al, 1969; Elliot, 1976; Jobling, 1995). As temperature influences rates of ingestion, metabolic rates and growth rates; increasing our understanding of these physiological properties in fish becomes paramount if we are to understand the effects of climate change, both at the individual (Elliot, 1994) and population level (Ohlberger et al, 2001) and also the impacts on food web dynamics such as predatorprey interactions (Hairston and Hairston, 1993).

THE BALTIC SEA AND PREDICTED IMPACTS OF CLIMATE CHANGE The Baltic Sea, with a surface area of 415 000 km2, is one of the largest and most studied brackish water bodies in the world. The catchment area of the Baltic Sea covers 1.74 million km2 and includes territory from a total of fourteen countries (HELCOM, 2007; Kautsky and Kautsky, 2000). About 80% of the river runoff and 85% of the net precipitation enters through the Gulf of Bothnia, northern Baltic Sea. Due to this gradient in fresh-water inflow the Baltic Sea is commonly divided into three distinct basins, going from the Bothinan Bay in the north, to the Bothnian Sea and finally the Baltic Proper in the south. This inflow controls the low salinity of the Baltic Sea waters, which range from about 2 PSU (Practical Salinity Units) in the northern Bay of Bothnia to about 10 PSU in the southern Baltic proper (Kautsky and Kautsky, 2000). Additionally, the high fresh water inflow, water runoff from various terrestrial environments, also explains between 71% and 97% of the variability in dissolved nutrients (nitrogen, phosphorus, silicon) input, as well as concentrations of total organic and inorganic carbon in the Baltic Sea (HELCOM, 2007). Current climate change predictions for the northern Hemisphere suggest both an increase in sea surface temperature of on average 4 °C and increased terrestrial nutrient runoff from rivers due to increased annual levels of precipitation (Meier, 2006; HELCOM, 2007; IPCC, 2007). Consequently, the future environment of the Baltic Sea is predicted to look very different from present conditions. These changes to ambient conditions will affect all levels of the pelagic food web, altering conditions for bacteria and phytoplankton, as well as affecting fish populations. In order to gain a better understanding of the effects of climate change on pelagic food webs in the Baltic Sea, a system-ecological approach has therefore to be adopted, where concurrent effects of temperature and DOC are studied on the planktonic food web, zooplanktivorus fish and interaction effects these parts of the food web exert on each other.

12

THE THREE - SPINED STICKLEBACK, GASTEROSTEUS ACULEATUS The typical stickleback is a small streamlined fish with a wide distribution in both marine and fresh waters of the northern hemisphere (Wootton, 1984) (Fig 3). The typical adult stickleback will reach about 11 cm in length and is distinguished by its bony armor with three prominent dorsal spines, the first two of which are large, serrated and isolated from the dorsal fin (Bell and Foster, 1994).

Fig. 3. Global distribution of the three-spined Stickleback (Gasterosteus aculeatus). Red areas indicate higher relative likelihood of occurrence (Froese and Pauly, 2007).

Generally sticklebacks take 2-3 years to mature and experience only one breeding season before dying (Wootton, 1984). As males come into reproductive condition, typically the eye and body take on a blue tinge and the ventral surface of the head and trunk become red. The stickleback males build a benthic nest where they care for the eggs by oxygenating them whilst simultaneously protecting them from rival males. After the eggs are laid, the females take no further part in rearing of the young (Bell and Foster, 1994). In the Baltic Sea, after the spawning period, G. aculeatus migrate from the littoral spawning grounds into the open ocean, this usually occurs around August-September (Lemmetyinen and Mankki, 1975). Although it is one of the main planktivorous species in the Baltic Sea (Jurvelius et al,, 1996), its ecological role has long been overlooked because of its lack of commercial importance. However, three-spined stickleback dynamics and their role in the ecosystem, have received recent attention since it has been suggested that population abundances have increased substantially in the last decade (Ljunggren et al, 2010; Eriksson et al, 2011). This increase in abundance has been attributed to a decrease in piscivours predators causing a predator release scenario for small fish species such as three-spined stickleback (Ljunggren et al, 2010; Eriksson et al, 2011). In addition, it has been suggested these increases have now led to sticklebacks out-competing other coastal fish species for shared zooplankton resources (Eriksson et al, 2009; Ljungren et al, 2010).

13

Even-though they are a well-studied fish species in both behavioural and evolutionary ecological terms (Bell and Foster, 1994; Huntingford and Gomez, 2009), their function and role in ecosystems is less well defined and important baseline parameters, such temperature dependent growth and attack rates, remain unknown. As three-spined sticklebacks occupy both in-shore and off-shore habitats of the entire Baltic Sea during their life cycle, coupled with increasing population numbers, the establishment of baseline parameters for growth and foraging capacity are clearly vital if we are to understand their full ecological role in the Baltic Sea. In addition, due to their ecological relevance, wide distribution, general small size and proven adaptability in laboratory environments they are ideal candidates for laboratory investigations into effects of climate change in pelagic food webs (Bell and Foster, 1994). AIM The aims of this thesis were two-fold. Firstly, we set out to establish baseline parameters of temperature and size dependent growth (Paper I) as well as foraging capacity for G. aculeatus (Paper II). Once these had been established, a better understanding of the role of the stickleback in the Baltic Sea, currently and in a future climate change scenario, would be gained. Secondly, we aimed to investigate the impacts of allochtonous dissolved organic carbon (ADOC) on food web dynamics in a marine system to better elucidate patterns of carbon transfer from the microbial food web to zooplanktivorous fish (Paper III). Finally, by building on results from previous papers, we aimed to test the impact of terrestrial dissolved organic carbon (TDOC) and temperature alteration, as predicted in relevant climate change scenarios, on pelagic food web efficiency and fish production (Paper IV). More specifically for each paper we wanted: Paper I - to parameterize a growth model describing the temperature dependent growth capacity and range of G. aculeatus under unlimited food conditions. In addition, we wanted to test whether growth rates measured using commercial pellets could be applied to more natural conditions by investigating differences in growth potential between fish fed invertebrate prey or pellets and finally, to investigate whether maximum growth potential was dependent on seasonality. Paper II -

to first estimate the size and temperature dependent attack rates of the three-spined stickleback through a series of laboratory experiments using the calanoid copepod Eurytemora affinis as prey. Secondly, by using these estimates, literature data on size and temperature dependent consumption capacities, metabolic rates and a time series of zooplankton biomass for all three basins in the Baltic Sea, we wanted to investigate past and present consumption rates and resource limitation thresholds which may have supported the suggested increase in population abundances, as well as provide an estimate of the critical resource density (CRD), i.e. the minimal resource density required to sustain metabolic needs for this species.

14

Paper III -

to investigate the role of allochtonous dissolved organic carbon (ADOC) on pelagic food web structure and efficiency. Currently, there is very little information available on how ADOC influences aquatic systems beyond the planktonic food web, and the interactive effects of an external carbon source and/or the addition of another trophic level (top consumer) are poorly established. To this end, a full-factorial experiment was set up in a large-scale mesocosm facility, with four treatments where food web transfer efficiencies and production potentials were measured in pelagic systems dominated by either bacteria or phytoplankton at the basal level and zooplankton or fish (three-spined stickleback) as the top consumer.

Paper IV -

to investigate the effects of natural TDOC, increased temperature and their combined effect on ratios of heterotrophy/autotrophy and marine food web efficiency, in a factorial large-scale mesocosm experiment. The aim was to measure pelagic food web function and structure under predicted climate change scenarios and to contrast these findings to present day conditions. More specifically we investigated the effects of these environmental drivers on (i) ratios of bacterial and phytoplankton productivity (ii) changes in species compositions and abundances in the food web (iii) impacts on growth of zooplanktivorous fish and (iv) the effects on food web efficiency in the four contrasting experimental treatments.

MATERIALS AND METHODS For all experiments, G. aculeatus juveniles were collected with a beach seine from a coastal area outside of Umea Marine Science Centre, in the Gulf of Bothnia, Sweden. Fish were then kept in 400 l holding tanks at 10 °C with continuous water flow-through and fed daily with commercial pellets or with a culture of Eurytemora affinis, a calanoid copepod common to the area, depending on the experiment. For papers I and II an indoor aquarium system was used (Fig. 4. A), built specifically to allow for experiments on growth and consumption rates to be carried out at up seven different experimental temperatures simultaneously. Fourteen aquaria (120 l) were arranged in temperature control systems (2 aquaria per temperature) and supplied with partly re-circulated 20µm filtered sea water. In each aquarium, 5 smaller plastic tanks (5 l) or 2 larger tanks (10 l) could be submerged. This allowed for either 10 fish per temperature to be reared singly during the growth experiments (Paper I) or capture rates to be measured on 4 fish per temperature (Paper II).

15

A

B

Fig. 4. A) Picture of the aquarium experimental facility. B) Picture of the indoor mesocosm facility at Umeå Marine Science Centre

Fish were always initially placed at a neutral temperature common to all aquaria (15 °C). The experimental temperatures were then obtained by increasing or decreasing the temperature by approximately 1 °C h-1. During this acclimatisation phase, fish were always fed ad lib rations of food. Once set to the desired temperature, electronic thermostats (Nobelektronik, Älmhult, Sweden) were used to regulate water temperatures. Temperatures never fluctuated beyond ±0.5 °C from the set value. The plastic tanks were perforated to allow for constant water exchange with the surrounding, temperature-controlled water in the aquaria. Light levels were adjusted to a 12 h light/12 h dark photo period. To counteract potential effects of increasing ammonium concentration and to replenish water lost by evaporation, 20 l of filtered (20 μm) sea water was added daily to each temperature system. Aquarium air pumps supplied oxygen to each 120 l aquaria tank. In paper I, growth was measured on juvenile sticklebacks with a continuous supply of commercial pellets as food. In addition, a growth study was conducted in the winter to test for seasonal differences in growth patterns and also a study was done testing for differences in growth rates between commercial pellets and invertebrate prey (chironomids). The different growth trials lasted between 10-14 days. In paper II, individual fish were placed singly in the 10 l tanks, and held at 15 °C. In order to measure capture rates on zooplankton, fish were first kept at the experimental temperature for 24 hours and fed ad lib rations of zooplankton. Before the start of each feeding trial, fish were then deprived of food for 24 h before experiments to standardize hunger level. At the start of each trial, the desired zooplankton density was introduced from above and the water was carefully stirred so that the zooplankton were evenly distributed. During this time fish were kept in a small net to prevent them from feeding whilst zooplankton was being distributed in the aquaria. Fish were then released from their enclosures and the trial started once the fish had captured one E. affinis, and lasted until it had successfully captured three subsequent prey. All experiments were carried out with one size class of E. affinis (0.7 mm ± 0.04, mean ± 1 SD), at 6 different prey densities (1, 2, 4, 8, 16 and 32 prey/l) and at 3 different temperatures (15, 21 and 24 °C). 16

In papers III and IV, experiments were conducted at the Umeå Marine Research Centre indoor mesocosm facility (Fig. 4. B). This facility consists of twelve polyethylen mesocosms, each with a volume of 2000 l, a diameter of 73 cm and a depth of 5 m. These can be filled simultaneously with unfiltered coastal water (5 PSU) from the northern Baltic Sea (63°34´N, 19°54´E) using a peristaltic pump system. By using unfiltered water, a natural assemblage of all planktonic organisms is evenly distributed to the mesocoms, which can thereafter be manipulated, either by altering ambient conditions, such as nutrients and temperature, or by changing the food web structure, by adding zooplanktivorous fish. In both experiments, treatment allocation was randomized between the 12 mesocosms. Air was gently bubbled (~20 ml/s) into the first 4 m of water in each mesocosm to create a well-mixed water column. Light was provided by 150 W metal halogen lamps (MASTERColour CDM-T 150W/942 G12 1CT) suspended over each individual tank and set to 12 h light/12 h dark regime. Temperature was regulated by a heating-cooling system containing circulating glycol and temperatures did not fluctuate more than ± 0.5 °C/day. For both experiments, biological variables at all levels of the food web were measured continuously, which includes production of bacteria and phytoplankton at the base of the food web, biomass measurements of all planktonic organisms, as well as fish growth. In addition, environmental variables, such as nutrient concentrations, water colour and ambient light were monitored throughout experiments. In paper III four experimental treatments, each triplicate replicated, were established in order to test for differences in food web efficiency between systems with contrasting structure at the basal production level and with food webs of differing length. Basal production was manipulated by adding either nitrogen and phosphorous to promote a phytoplankton based food web, or by adding carbon in the form of glucose in addition to nitrogen and phosphorus, to promote a food web mainly based on bacteria. Trophic positions at the top of the food web was manipulated by having either naturally occurring zooplankton as top predator or by the addition of zooplanktivorous fish, juvenile three-spined sticklebacks. In paper IV we wanted to simulate the effects of river-bound carbon input, decreased light conditions, as well as increased temperature on the pelagic food web. To achieve a carbon input, which as closely as possible represented natural conditions, we created a natural, colored humic soil solution. This terrestrial dissolved matter (TDM) would add not only TDOC to the system, but also organic and inorganic N and P, along with humic acids and other colored substances. Using the mesocosm facility we then created four experimental treatments, each with three replicates, with TDM additions and temperature increase as treatment variables and natural pelagic food webs containing all trophic levels from bacteria to zooplanktivorous fish. The treatments were chosen to reflect current, relatively low (TDML) and predicted high (TDMH) inputs of terrestrial organic and inorganic nutrients to the marine environment via rivers and/or an elevation in water temperature (TDML T; TDMH T). DOC concentrations were raised by 30% in our climate change scenario, based on simulations predicting future, river-bound nitrogen export to the Baltic Sea (Eriksson-Hägg, 2010, Wikner and Andersson in press). Temperature was raised by 4 °C based on current climate change predictions (Meier 2006: Helcom, 2007: IPCC, 2007).

17

MAIN RESULTS Paper I The results in this study showed that the generic model for temperature dependent maximum fish growth by Elliott et al, (1995) could be applied to describe experimentally derived growth data for G. aculeatus. This is the first time a base-line model over the whole temperature range suitable for growth in this species has been established. Optimal temperature for growth was estimated to be 21.7 °C and based on the estimates of lower and upper temperature limits for growth three-spined sticklebacks seem to be able to grow in conditions as low as 3.5 °C and up to 30 °C. The comparisons between commercial pellets and natural prey suggested no difference in growth potential between these two prey items and therefore our model predictions can be applied to natural systems. Still, we observed a strong seasonal effect on growth rates, with an average of 60% higher growth rates achieved at the optimum temperature in summer compared to the winter. This finding reduces the applicability of the model derived in this study to spring and summer seasons, but highlights that caution should be exerted when extrapolating results from experimentally derived conditions to natural populations and that the potential presence of seasonal effects warrants further investigation.

Paper II The results presented in this paper provided estimates of both the size and temperature dependent attack rate of the three-spined stickleback. We showed that three-spined sticklebacks appear to be tolerant of increases in ambient temperatures and, when approaching their optimum temperature of growth, just over 21 °C, their tolerance for decreasing resource levels will actually increase their potential scope for growth, due to a decrease in CRD at these temperatures. As higher summer temperatures should benefit stickleback populations, the impacts of climate change on this species and its competitive interactions with other species, definitely warrants further investigation. Furthermore, our estimates of consumption capacities showed that despite increasing densities in the Baltic proper (BP), sticklebacks in offshore areas have experienced relatively low levels of resource limitation over the investigated time period. However, in the coastal zones of the BP, resource levels have in recent years declined and approached CRD for a 1g stickleback, indicating that sticklebacks under these conditions would not be able to sustain their metabolic rates during the summer months. This may suggest that, at least in the coastal area where sticklebacks reproduce, populations are starting to reach levels were their densities will not increase further in the BP. In contrast, sticklebacks in the coastal zones of both the Bothnian Bay (BB) and Bothinan Sea (BS) have not been markedly resource limited during this last decade and have been able to feed close to maximum capacity, in addition to fish in these corresponding offshore areas having experienced a significant decrease in resource limitation over recent years. These results suggest that the off-shore habitats in BB and BS, devoid of pelagic predators, have during recent years been areas with good growth potential for migrating stickleback populations, but also that the suggested recent rise in stickleback densities in the BS might have been facilitated by this lack of resource limitation.

18

Paper III Results showed that ADOC additions significantly increased bacterial production (BP), yielding a higher total basal production than treatments without carbon addition. As a result of the increased heterotrophic production, food web efficiencies were lowered in the systems, regardless of the top consumer. However, in terms of absolute production at the top of the web, a top-consumer dependent response of ADOC additions was seen. In systems with zooplankton as top-consumer, ADOC seems to have acted as a subsidy to the production of heterotrophic protists, giving copepods a food source in addition to phytoplankton, resulting in a higher total zooplankton production compared to NP treatments. However, when fish was top-consumer, their production was instead lowered. The high energy loss at the top trophic level in fish dominated food webs is suggested to be due to top-down effects, which, via predator release of the lowest trophic levels, indirectly increased BP in the presence of ADOC without stimulating primary production (PP) and stimulated PP in absence of ADOC. Since the majority of energy in systems where fish and ADOC additions interacted was localized to the microbial food web, higher energy losses occurred as carbon had to pass through several additional trophic levels before reaching fish. Our results therefore suggests, that depending on ADOC inputs into the system, cascading effects of planktivorous fish will dictate both the magnitude of BP and PP production and the BP:PP ratios, causing feedbacks in the food web, which ultimately influence their own production. Moreover, a climate change increase of ADOC inputs may induce a shift towards heterotrophy in aquatic systems, which would according to this study, cause decreased fish production in areas expressing high BP: PP ratios.

Paper IV It was shown that both increases in temperature and increased concentrations of TDOC will independently affect the pelagic food web by stimulating bacterial production, and to varying extents, the trophic levels of the planktonic food web. Neither factor on their own produced effects that cascaded to the top of the food web, thus fish production or food web efficiency were not influenced. However, when these two treatment effects were combined a significantly higher fish production and food web efficiency was achieved in our simulated climate change scenario as compared to present day conditions. The results indicate that the combined effect lead to a temperature dependent increased production potential of zooplankton, which was sustained by the increased production of heterotropic microzooplankton stimulated by TDOC additions. The increased zooplankton biomass in turn lead to higher fish production and increased food web efficiency. Although the increased number of trophic linkages in the heterotrophic food webs should have reduced the energy transfer efficiency, these negative effects seem here to have been overridden by the positive increases in zooplankton production as a result of increased temperature. In terms of climate change, predicted increases in temperature and TDOC input might therefore yield positive effects on FWE and total production in higher trophic levels in present low productive systems, but further studies need to be conducted to determine the generality of such a statement. Furthermore, the results presented here strongly suggest that in order to fully understand the effects of climate change on pelagic food web transfer efficiency, abiotic variables cannot be studied in isolation.

19

CONCLUSIONS

The thermal response of the three-spined stickleback; implications of climate change for sticklebacks in the Baltic Sea The results presented in this thesis have highlighted several aspects of stickleback behavior influenced by varying temperature and provided crucial baseline estimates of thermal response to both consumption rates and growth, estimates which until now were missing. The results in Paper I underscore the eurythermal nature of this species and clearly shows that being adapted to a wide range of temperatures most likely underlies the sticklebacks documented high capacity to adapt to new environments, which has allowed it to colonize various habitats including shallows, warm tidal pools and the open oceans (Wootton, 1984; Bell and Foster, 1994). Furthermore it shows a strong temperature dependent growth response up to over 21 °C. This sharp peak in optimum response is actually higher than temperatures usually experienced by the stickleback even in the midst of summer. In terms of growth, if climate change increases average water temperatures, the three-spined stickleback might therefore not be as adversely affected as other species. Conclusions from paper II further highlight the tolerance of sticklebacks to increasing temperatures. Even at the relatively elevated temperature of 24 °C stickleback foraging capacity does not only function normally, but appears to be significantly faster compared to lower temperatures. Furthermore, by measuring the CRD, i.e. the resource densities a species needs to maintain basic metabolic needs, it was shown that these levels decreased with increasing temperatures. This result seems at first glance to be counter-intuitive, as metabolic demands increase with increasing temperature, necessitating an increased consumption to meet those demands, which in turn is facilitated by and often dependent on higher resource levels (Elliott 1977; Jobling 1995). However, it became evident that attack rates for the threespined stickleback had a much steeper increase with temperature than routine metabolism, in turn decreasing CRD at higher temperatures. Therefore, based on our results, increased temperatures may, in accordance with increased metabolic demands and digestion capacity, lead to an overall increased consumption, but the levels actually needed to sustain individuals will be lowered, thus increasing their scope for growth at higher temperatures. It has been suggested that three-spined stickleback population densities in the Baltic Sea have increased substantially in the last decade (Ljunggren et al, 2010; Eriksson et al, 2011) and that competition from sticklebacks for shared zooplankton resources may affect recruitment of other fish species e.g. perch Perca fluvitialis in the Baltic proper (Eriksson et al, 2009; Ljungren et al, 2010). Although the results from this thesis cannot conclusively state that this is occurring based on our attack rate estimates, it is clear that their competitive advantage will only increase with increasing temperatures, as a combination of lower CRD and increased growth potential will lead to substantially higher scope for growth of stickleback populations. However, as the attack rate of a consumer is the net outcome of foraging related capacities of both the consumer and its prey, the species specific temperature scaling of attack rate with temperature, could likely vary substantially between consumers and for different prey types and sizes (Englund et al, 2011 and references therein). Our results also provide estimates on the levels of resource limitations which sticklebacks have experienced over the last decade, which may have facilitated the reported increased 20

population densities. We showed that sticklebacks in the Baltic proper, where the suggested population increase is supposed to be the strongest, have in recent years started to experience higher degrees of resource limitation. This means population levels here might not increase any further as they have started to reach their CRD. However, in both the coastal zones of the Bothnian Bay and Bothnian Sea, sticklebacks have not been markedly resource limited during the last decade and have been able to feed close to maximum capacity. Furthermore, degrees of resource limitation seem to have decreased in these offshore habitats, indicating that these areas, devoid of pelagic predators, would have provided areas of good growth potential for migrating stickleback populations. A recent assessment of coastal test fishing programs has indicated that spawning population of sticklebacks have increased over the last 5 years in the Bothnian Sea (Byström unpublished compilation). Since both in-shore and off-shore areas of the Bothnian Bay and Bothinan Sea seem to have provided good habitats for growth in recent years, it plausible that we will see further increases in stickleback population numbers in these basins and consequently patterns of negative relationships between sticklebacks and coastal fish species seen today in the Baltic proper might soon be replicated in the more northern basins. As the impacts of climate change on fish populations are hard to deduce, having established the baseline parameters of growth and foraging capacity in this species will at least provide a starting point from which future patterns of competitive interactions and population fluctuations can be deduced. Overall, since the physiological prowess of the three-spined stickleback seems to provide it with a high degree of resilience against increasing temperatures, the ecological role of this species in the Baltic Sea, today and in the future, is a topic which definitely warrants further investigation. Impacts of climate change on pelagic food web structure and function; implications for fish production in the Baltic Sea The results of paper III clearly show that allochotnous carbon (ADOC), i.e. originating from outside the pelagic system, does have a structuring role on the pelagic food web. This will mainly be expressed in decreased reliance of marine bacteria on autochthonous carbon (originating from within the system) produced by phytoplankton. This will lead to increased bacterial production and shift the ratio of bacterial to phytoplankton produced carbon at the base of the food web. However, in the presence of planktivorous fish, this relationship is strengthened and we saw ratios shifting even further in favor of bacterial production, leading to a severe reduction in fish growth, compared to when fish were living in food webs dominated by phytoplankton. The increased reliance on the microbial food web in these systems suggested that depending on ADOC inputs into the system, cascading effects of planktivorous fish will dictate both the magnitude of bacterial and primary production along with their ratio, causing feedbacks in the food webs, which ultimately influenced their own production. The results of this study thus clearly demonstrate that introducing ADOC to a food web dominated by zooplanktivourous fish will impact fish production negatively, due to the inefficiency of the microbial food web in providing energy to higher trophic levels. However, the results in paper IV showed that a reduction of FWE induced by ADOC can be offset by increases in temperature and that the combined effect of both factors can actually lead to increased production of fish with a consequent higher transfer efficiency. This was due to an increased temperature dependent production potential of zooplankton, leading to higher food availability for fish, which in turn also had a higher production potential with the 21

temperature increase. Since temperature and ADOC interacted in a way that we did not foresee, in addition to top-down effects profoundly altering the structure of the food web, it becomes obvious that the full effects of climate change will be hard to predict. If we are to extrapolate on these results beyond our study systems, we have to look at several factors which may have influenced our findings. The Baltic Sea is firstly a unique system in that its three main basins in a north-south gradient are very different from each other in terms of abiotic variables, such as nutrient levels and salinity but also in terms of species compositions, which differ markedly from the southern Baltic Proper to the northern Bothnian Bay (Kautsky and Kautsky, 2000). The study area where we have been working in the Bothnian Sea is an oligotrophic system, compared to the Baltic Proper, which is highly eutrophic. Previous work has indicated that the effects of TDOC on pelagic systems would be highly dependent on the existing nutrient regimes in the water and that its effects might be stronger under oligotriohic conditions (Carpenter et al, 2005). As the effects of TDOC have not been tested here under different stoichiometric conditions, we do not currently know how the effects of increased river-bound carbon input might affect pelagic food web function and structure in the Baltic proper or other marine areas with similar conditions. However, it could be postulated that, in systems with high phytoplankton production, the reliance of bacteria on ADOC will be lessened, as the carbon produced by phytoplankton would provide a much better substrate for growth (Carpenter et al, 2005). Therefore, the amount of ADOC which needs to be introduced to a eutrophic system before the ratio of BP: PP significantly shifts in favor of bacteria, might be a lot more than the 30% increase we simulated in paper IV. Furthermore, the choice of using the three-spined stickleback as a zooplanktivorous fish in these study systems will also have a bearing on the presented results. As shown here, the high resilience to increasing temperatures displayed by sticklebacks definitely implies that effects of food availability and increased temperature will uniquely affect this species and effects seen for sticklebacks might not translate directly to other species. Had our experimental treatments instead been tested on performance of e.g. herring (Clupea harengus), which has been shown to potentially be more sensitive to higher temperatures (Fey, 2001), it is likely that the outcome of our experiments would have been very different.

In conclusion, as both terrestrial carbon and temperature have profound effects on all levels of the pelagic food web, from influencing the competition for inorganic nutrients and production potentials of phytoplankton and bacteria, to determining development times of zooplankton and finally dictating production levels of fish, the full effects of climate change on the pelagic food web and fish production are hard to untangle. As the body of evidence on the effects of climate change accumulates it becomes apparent that only through increased knowledge will we be able to prepare for how climate change will alter our common ocean environments and hopefully the results put forward in this thesis will go some small way towards accomplishing that goal.

22

THANKS First and foremost I would like to thank my advisors and all the staff at the Umeå Marine Science Centre for all the help and support they have given me. A big thanks goes of course to my thesis collaborator Richard Degerman. I have really enjoyed working with you and I will miss our discussions of crazy experimental mesocosm designs. Agneta and Ulf, without your help, my comprehension of marine plankton life would have been today what it was when I started this Ph.D, basically non-existent. Thank you for interesting discussions and also for all the financial support over the years. Stefan, thanks for agreeing to give me a job in the first place and although I am sorry you could not stick with me to the end, I am glad that I could finish what we started together and the overall scientific design of this thesis would not have been possible without you. Finally, I wish to especially thank Pär Byström for all the enormous support and hard hours you have put in to make sure this thesis reached completion. If you had not agreed to supervise me this last year, this thesis would not have been possible to finish. I cannot express enough thanks for this and working together with you and exchanging ideas has truly been a rewarding experience. Thanks to my family for all the support they have given me throughout the years and who always encouraged me to travel and never restricted my freedom to explore. Also, by allowing me to spend most of our early summer holidays under-water with a snorkel, played a big part in instilling my love for the ocean environment. I also wish to thank granddad for taking me out on numerous fishing trips as a kid which started my interest in fish and fishing. Last I want to thank my wife Annie for always supporting me, being a great partner in travels and in life and who always makes me want to do better.

23

REFERENCES Azam, F., T. Fenchel, J.G. Field, J. S.Gray, L. A.Meyer-Reil and F. Thingstad (1983). The ecological role of water-column microbes in the sea. Marine Ecological Progress Series 10: 257–263. Andersson, A., Haecky P., and Hagström Å. (1994). Effect of temperature and light on the growth of micro-, nano- and pico-plankton: impact on algal succession Marine Biology 120: 511–520 Bell, M. A. and Foster, S. A. (1994). The Evolutionary Biology of the ThreespineStickleback. Oxford University Press, Oxford Berglund J, Müren U, Båmstedt U and Andersson A .(2007). Efficiency of a phytoplanktonbased and a bacteria-based food web in a pelagic marine system. Limnology and Oceanography, 52: 121–131 Blomqvist, P., Jansson M., Drakare. S, Bergström A.-K. and L. Brydsten ( 2001). Effects of additions of DOC on pelagic biota in a clearwater system: Results from a whole lake experiment in northern Sweden. Microbial Ecology 42: 383–394 Breteler, W.C.M., N. Schogt, M. Baas S. Schouten and G. W. Kraay (1999). Trophic upgrading of food quality by protozoans enhancing copepod growth: role of essential lipids. Marine Biology.135: 191-198 Brett, J. R. (1979). Environmental factors and growth. In Fish Physiology, Vol. 8 (Hoar, W. S. Randall, D. J. and Brett, J. R., eds), pp. 599–675. New York: Academic Press. Brett,J . R., Shelbourn,J. E. and Shoop,C . T. (1969).Growth rate and body composition of fingerling sockeye salmon,O ncorhynchus nerka, in relation to temperature and ration size. Journal of Fisheries Research Board of Canada, 26: 2363-2394. Byström P., J. Anderson, A. Kiessling and L-O Eriksson (2006). Size and temperature dependent foraging capacities and metabolism: consequences for winter starvation mortality in fish. OIKOS 115: 43-52, Carpenter, Sr, J.J. Cole, Ml. Pace, M. Van De Bogert, Dl. Bade, D. Bastviken, C. Gille, J.R. Hodgson, J.F. Kitchell and E.S. Kritzberg (2005). Ecosystem subsidies: terrestrial support of aquatic food webs from 13C addition to contrasting lakes. Ecology. 86:2737–2750. Cole, J, J (1982). Interactions between bacteria and algae in aquatic ecosystems. Annual Review of Ecology and Systematics 13: 291-314 Cole, J.J Stuart Findlay S and Pac M.L (1988). Bacterial production in fresh and saltwater ecosystems: a cross-system overview, Marine ecology progress series 43:1-10 Dippner, J.W, G. Kornilovs and L. Sidrevics. (2000). Long-term variability of mesozooplankton in the central Baltic Sea. J. Mar Syst. 25:23–31 Eby, G.N. (2004). Principles of Environmental Geochemistry: Thomson Brooks/Cole, Pacific Grove, CA. 24

Elliott, J.M (1976). The energetics of feeding, metabolism and growth of brown trout (Salmo trutta) in relation to body weight, water temperature and ration size. Journal of Animal Ecology, 45: 923–948. Elliot J.M (1994). Quantitative ecology and the Brown Trout Oxford University Press, Oxford Elliott J.M. and Hurley M.A. (1997). A functional model for maximum growth of Atlantic salmon parr, Salmo salar from two populations in north-west England. Functional Ecology, 11: 592–603. Englund G., G. Öhlund, C.L. Hein and S. Diehl (2011). Temperature dependence of the functional response. Ecology Letters 14: 914-921 Eriksson, B.K., L. Ljunggren, A. Sandström, G. Johansson, J. Mattila, A. Rubach, S. Råberg and M. Snickars (2009). Declines in predatory fish promote bloom-forming macroalgae. Ecol. Appl. 19: 1975–1988. Eriksson B.K., K. Sieben, J. Eklöf, L. Ljunggren, J. Olsson, M. Casini and U. Bergström (2011). Effects of Altered Offshore Food Webs on Coastal Ecosystems Emphasize the Need for Cross-Ecosystem Management. AMBIO 40:786-797. Fey, D. P. (2001). Differences in temperature conditions and somaticgrowth rate of larval and early juvenile spring-spawned herring from the Vistula Lagoon, Baltic Sea, manifested in the otolith to fish size relationship. Journal of Fish Biology, 58: 1257–1273. Forseth, T., Hurley, M. A., Jensen, A. J. and Elliott, J. M. (2001). Functional models for growth and food consumption of Atlantic salmon parr, Salmo salar, from a Norwegian river. Freshwater Biology 46, 173–186. Froese, R. and D. Pauly. Editors ( 2007). FishBase. World Wide Web electronic publication. www.fishbase.org, version (10/2007) Gasol J.M, del Giorgio P. and Duarte C.M. (1997). Biomass distribution in marine planktonic communities. Limnology and Oceanography 42:1353-1363 Hairston, N.G and N.G Hairston (1993). Cause-Effect relationships in energy flow, trophic structure and interspecific interactions. The American Naturalist 142: 379-411 Heinle, D.R. and D.A. Flemer. (1975). Carbon requirements of a population of the estuarine copepod Eurytemora affinis. Mar Biol 31:235–247 HELCOM (2007). Climate Change in the Baltic Sea Area – HELCOM Thematic Assessment in 2007 Balt. Sea Environ. Proc. No. 111 Hoppe, Hg, P. Breithaupt , K. Walther , R. Koppe,S. Bleck, U. Sommer and K. Jürgens (2008). Climate warming in winter affects the coupling between phytoplankton and bacteria during the spring bloom: a mesocosm study. Aquat Microb Ecol 51:105–115 Huntingford, F. A. and M. L. Ruiz-Gomez, (2009). Three-spined sticklebacks Gasterosteus aculeatus as a model for exploring behavioural biology. Journal of Fish Biology 75: 1943– 1976 25

IPCC (2007): Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 996 pp. Jansson, M, A-K. Bergström, P. Blomqvist and S. Drakare (2000). Allochthonous organic carbon and phytoplankton/bacterioplankton production relationships in clearwater and humic lakes. Ecology 81:3250–3255 Jansson, M., L.Persson, A.M. De Roos, R.I. Jones and L.J. Tranvik (2007). Terrestial carbon and intraspecific size-variation shape lake ecosystems. Trends in ecology and evolution 22: 316-322 Jobling, M. (1995). Fish bioenergetics, Chapman and Hall, London Jurvelius J, Leinikki J, Mamylov V and Pushkin S. (1996). Stock assessment of pelagic threespined stickleback (Gasterosteus acuelatus): A simultaneous up and down looking echo sounding study. Fisheries Research 27: 227-241 Kautsky L and Kautsky N. (2000). The Baltic Sea, including Bothnian Sea and Bothnian Bay Seas at the millennium, an environmental evaluation Vol 1, Regional chapters: Europe, the Americas and West Africa Pergamon, an imprint of Elsevier Science Klug J.L. (2002). Positive and negative effects of allochthonous dissolved organic matter and inorganic nutrients on phytoplankton growth. Published on the NRC Research Press Web site at http://cjfas.nrc.ca’ Larsson U, and Hagström A (1979). Phytoplankton exudate relaease as an energy source for the growth of pelagic bacteria. Marine biology 52: 199-206 Larsson S. And Berglund I. (1998) Growth and food consumption of 0+ Arctic charr fed pelleted or natural food at six different temperatures Journal of Fish Biology, 52: 230-242 Leinikki, J. (1995). The food of three-spined stickleback (Gasterosteus aculeatus) in the Bothian Bay during its pelagic phase in August. Aqua Fennica, 25: 71-75. Lemmetyinen, R. and Mankki, J. (1975). The three-spined stickleback (Gasterosteus aculeatus) in the food chains of the northern Baltic. Memntutkimuslaitoksen Julk. Havsforskningsinst. Skr., 239: 155- 161. Ljunggren, L., A. Sandström , U. Bergström, J. Mattila, A. Lappalainen, G. Johansson, G. Sundblad, M. Casini , O. Kaljuste and B.K. Eriksson (2010). Recruitment failure of coastal predatory fish in the Baltic Sea coincident with an offshore ecosystem regime shift. ICES J. Mar. Sci. 67:1587–1595 Meier, H.E.M.(2006). Baltic Sea climate in the late twenty-first century: a dynamical downscaling approach using two global models and two emission scenarios. Climate Dynamics 27: 39- 68 Müren U., Berglund J., Scheinersson K and Andersson A. (2005). Potential Effects of Elevated Sea-Water Temperature on Pelagic Food Webs Hydrobiologia 545:153–166 26

Ohlberger, J., E.Edeline, L. A. Völlestad, N.C. Stenseth and D. Claessen. (2011).Temperature-driven regime shifts in the dynamics of size-structured populations. Am.nat. 177: 211-223 Persson L. (1986). Temperature-Induced Shift in Foraging Ability in Two Fish Species, Roach (Rutilus rutilus) and Perch (Perca fluviatilis): Implications for Coexistence between Poikilotherms. The Journal of Animal Ecology. 55: 829-839 Persson, L., K. Leonardsson, A. M. De Roos, M. Gyllenberg, and B. Christensen (1998). Ontogenetic scaling of foraging rates and the dynamics of a size-structured consumer– resource model. Theoretical Population Biology 54:270–293. Pomeroy, L. R. (1974). The ocean's food web, a changing paradigm. BioScience 24: 499-504 Rand, P. S. and Stewart D. J (1998). Prey fish exploitation, salmonine production, and pelagic food web efficiency in Lake Ontario. Can. J. Fish. Aquat. Sci. 55: 318–327 Sandberg, J., A. Andersson, S. Johansson and J. Wikner (2004). Pelagic food web structure and carbon budget in three brackish water nvironments: Potential importance of terrigenous carbon. Marine Ecoogical Progress Series 268: 13–29 Sommer, U.H. Stibor, A Katechakis, F. Sommer, and T. Hansen (2002). Pelagic food web configurations at different levels of nutrient richness and their implications for the ratio fish production: Primary production. Hydrobiologia 484: 11–20 Straile, D. (1997). Gross growth efficiencies of protozoan and metazoan zooplankton and their dependence on food concentration, predator–prey weight ratio, and taxonomic group. Limnology and Oceanography 42: 1375–1385 Strom S,Benner R,Ziegler S and Dagg M (1997). Planktonic grazers are a potentially important sourceof marine dissolved organic carbon Limnology and oceanography 42:13641374 Tranvik J (1992). Allochthonous dissolved organic matter as an energy source for pelagic bacteria and the concept of the microbial loop Tranvik, L. J. and Höfle, M. G. (1988). Bacterial growth in mixed cultures on dissolved organic carbon from humic and clear waters. Applied Environmental Microbiology. 53: 482488 Vaccaro, R. F., Azam, F. And Hodson, R. E. (1977). Response of natural marine bacterial populations to copper: controlled ecosystem pollution experiment. Bulletin of marine science 27: 17-22 Wahlström, E., Persson, L., Diehl, S. and Byström, P. (2000). Size-dependent foraging efficiency, cannibalism and zooplankton community structure. Oecologia 123: 138–148. Wangersky P.J. (2000). Marine Chemistry, Springer Werner, E.E. (1994). Ontogenetic scaling of competitive relations: size-dependent effects and responses in two anuranlarvae. Ecology. 75: 197–213. Wootton, R. J. (1984) A Functional Biology of Sticklebacks. Croom Helm: London 27

AUTHOR CONTRIBUTIONS

Contribution/Paper

I

II

Original concept

RL,SL, PB

RL,SL, PB

Experimental design

RL, SL

RL, SL, PB

Data collection

RL

RL

Data analysis

RL, SL, PB

RL, PB

Interpretation of results and manuscript preparation

RL, PB, SL

RL, PB

III

IV

RD* and RL*, AA, SL, UB, LOE RD, RL, SL, AA, UB, LOE RD*, RL *

RL and RD*, AA, SL, UB, LOE

RD*, RL*, AA, PB, UB

RL*, RD*, PB, AA, UB

RD, AA, RL, PB, UB

RL, PB, AA, RD, UB

RL, RD, AA, SL, UB RL*, RD *

AA, Agneta Andersson, LOE Lars-Ove Eriksson PB, Pär Byström, RD, Rickard Degerman, RL, Robert Lefébure, SL, Stefan Larsson, UB, Ulf Båmstedt * Both authors contributed equally to the original concept, data collection and data analysis.

28