CO2 in aquaculture

ARTICLE IN PRESS

CO2 IN AQUACULTURE PETER VILHELM SKOV1 Technical University of Denmark, Hirtshals, Denmark 1 Corresponding author: [email protected]

1. Introduction 2. CO2 Sources in Aquaculture 2.1. Fish Metabolism 2.2. Bacterial Metabolism 3. Safe Levels and Welfare Guidelines 4. Free CO2, pH and Alkalinity 4.1. Current Removal and Control Practices 5. Dynamics of CO2 in Aquaculture 6. Effects of Dissolved CO2 on Growth 6.1. Feed Intake and Appetite 6.2. Digestive Function 7. Seawater Transfer 8. Pathological Effects of CO2 8.1. Nephrocalcinosis 8.2. Swim Bladder Inflation 8.3. Cataract Formation 9. Confounding Water Quality Effects 9.1. Hypercapnia and Hypoxia 9.2. Hypercapnia and Metal Toxicity 9.3. Carbon Dioxide and Aluminum 10. Conclusions and Perspectives References

Aquaculture of fishes is increasingly using recirculation technology to conserve water and reduce the environmental footprint from production. Recirculating aquaculture systems (RAS) require an intensification of production, which gives rise to a number of water quality issues that must be considered, including the accumulation of carbon dioxide (CO2). This chapter provides an initial overview on the excretion of CO2 from fish and 1 Carbon Dioxide FISH PHYSIOLOGY

Copyright # 2019 Elsevier Inc. All rights reserved DOI: https://doi.org/10.1016/bs.fp.2019.07.004

ARTICLE IN PRESS 2

PETER VILHELM SKOV

bacterial metabolism, before discussing what is currently being considered as safe levels, or levels of no-effect, of dissolved CO2 for fish, highlighting some differences between aquaculture and ecological studies. The correlation between alkalinity and pH, and the chemical form of CO2 is briefly covered, before providing an overview of current removal practices and technologies. The chapter provides an overview of the physiological effects of CO2 in relation to aquaculture production, specifically feed intake, appetite, digestive function and growth, before considering the effects of CO2 exposure on the ability of fish to migrate from fresh to seawater, in light of structural and functional changes in the gills. Finally, some pathologies associated with CO2 exposure, such as nephrocalcinosis and eye cataracts are highlighted, before closing with an overview of interacting effects between CO2 and other water quality variables, such as hypoxia or dissolved metals.

1. INTRODUCTION Aquaculture production of fish, molluscs, and crustaceans has increased fivefold in volume over the past 3 decades, and has now reached 80 million tons. Of this, approximately 10 million tons represents finfish farmed in a marine environment, while 50 million tons is fish farmed in inland production systems (FAO, 2018). The exact contribution from different production systems is unknown, but it is fair to assume that inland production is dominated by pond culture in many parts of the world. In European and North American aquaculture, the implementation of recirculating aquaculture systems (RAS) has facilitated part of the growth in aquaculture production in these areas. Although net pen production has contributed significantly in the growth of the aquaculture industry, as has been seen in Norway, fish produced in the marine environment also spend a part of their life on land in RAS facilities until ready for transfer to sea. Past and ongoing development of recirculation technology stems from a multitude of requirements, including eliminating the risk of escapees that mix with wild populations, reducing nutrient discharge to the environment, reducing water consumption, and preventing the transmission of disease (from wild to farmed, and vice versa) (Martins et al., 2010). In addition to this, RAS presents a possibility to provide fish with a stable environment, in which physical and chemical water quality parameters can be controlled. The availability of dissolved oxygen is generally considered the first limiting factor that determines the biomass carrying capacity of an aquaculture production system (Wurts, 2000). However, oxygen as a limiting factor can be overcome in land-based aquaculture with the application of pure oxygen

ARTICLE IN PRESS CO2 IN AQUACULTURE

3

from either on-site production systems or liquid oxygen (Colt and Watten, 1988), as well as a variety of other aeration technologies. Logically, other water quality parameters then become limiting in production; which is determined by the design of the water treatment components in a facility, along with cultured species, rearing density and temperature. The water quality parameters of interest to RAS managers, apart from oxygen, include CO2, pH, dissolved nitrogen (total ammonia nitrogen and nitrite) (Dalsgaard et al., 2013), and more recently hydrogen sulfide in saltwater systems. Control of pH can be achieved by adding buffering capacity in the form of alkalinity, or directly by the addition of base, while a properly dimensioned nitrification system (biofilter) will efficiently convert ammonia and nitrite to the less toxic nitrate. Arguably, CO2 can also be removed with the use of degassing technology, but such devices are not always installed, and may not be sufficiently efficient to fully remove CO2. 2. CO2 SOURCES IN AQUACULTURE 2.1. Fish Metabolism In natural waters, the increase in dissolved levels of CO2 is considered to originate largely from the absorption from atmospheric air as well as increased productivity (and subsequent bacterial decomposition) of algae, due to fertilizer run-off (Sunda and Cai, 2012). In aquaculture production systems, dissolved CO2 originates from the metabolism of organisms in the system, predominantly fish, but also from bacterial populations. Although standard metabolic rates (SMR), the metabolic rate of an inactive individual in an unfed state, have been determined for a wide range of species, it is difficult to generalize about the oxygen consumption rates (Ṁ O2) to fish in aquaculture systems, due to contributions from routine activity (routine metabolic rate, RMR) and feeding (specific dynamic action, SDA). However, tank based measurements from a number of species are available in the literature. For tilapia (Oreochromis niloticus) reared at 28°C, routine metabolic rates (RMR) range from 5.3 to 7.0 mmol O2 kg
1 h
1, while peak O2 requirements during feeding range from 9.9 to 15.4 mmol O2 kg
1 h
1 (Skov et al., 2017). For rainbow trout (Oncorhynchus mykiss) reared at 15 °C, RMR may vary between 4.9 and 6.4 mmol O2 kg
1 h
1, with peak O2 requirements between 6.8 and 10.0 mmol O2 kg
1 h
1 (Skov et al., 2015). For Atlantic salmon (Salmo salar) reared at 12 °C, RMRs are in the range of 5.0–7.5 mmol O2 kg
1 h
1 (Atkins and Benfey, 2008). These values are likely to represent the range of ṀO2 for cultured fish, and can thus be used to estimate CO2 excretion rates. Since fish do not exclusively consume oxygen to satisfy their SMR and feeding metabolism, but also for swimming and other routine activity, the true oxygen

ARTICLE IN PRESS 4

PETER VILHELM SKOV

requirements for fish in aquaculture may be higher. For a more detailed insight into O2 consumption rates and correlation with culture conditions, Forsberg (1994) provided a detailed model taking into account growth trajectories, temperature, swimming speed and feeding rates. Assuming a mixed substrate utilization with equal contribution from fat and carbohydrate, fish would have a respiratory quotient of 0.85, meaning that fish excrete 0.85mol CO2 for every mole of O2 consumed. As such, the corresponding CO2 excretion for the oxygen consumption rates outlined above would be in the range between 4.2 and 13.1 mmol CO2 kg
1 fish h
1. As rearing densities in intensive aquaculture production may exceed 50 kg of fish per m3 of water, this means that the CO2 excretion of fish may add up to 0.7 mmol CO2 h
1 to every liter of water. 2.2. Bacterial Metabolism Feeding and activity of fish in aquaculture leads to the excretion of large amounts of nitrogenous waste to the water, predominantly in the form of ammonia, with smaller contributions from urea and other nitrogenous compounds (Wood, 2001). In RAS, unlike net pens, this necessitates nitrification of ammonia into nitrite and then nitrate, a process which is achieved by biofilters, in which media with a high specific surface area are used to culture biofilm, ideally comprised of nitrifying bacteria. As a rule of thumb, a biofilter should have a specific surface area of 100 m2 per kg feed used. Stoichiometry dictates that for every mole of ammonia nitrogen nitrified, 2 mols of oxygen are consumed. Actual O2 consumption rates may be higher, and we have shown that biofilters that are not substrate or oxygen limited may consume 0.85 mmol O2 m
2 h
1 (P.V. Skov et al., unpublished observation). At equimolar CO2 excretion rates, feeding fish a 2% ration under the conditions described for fish above, this would contribute an additional 0.085 mmol CO2 L
1 h
1. In addition to the oxygen consumption of fish and biofilter, comes a smaller contribution from the biological O2 demand (BOD) of bacteria in the water. This rarely exceeds 0.3 mmol O2 L
1 day
1 (Dalsgaard and Pedersen, 2011), and thus makes only a moderate contribution to dissolved CO2, although in some production systems, the breakdown of particulate organic matter in the system may contribute on a more significant level. 3. SAFE LEVELS AND WELFARE GUIDELINES Sensitivity to dissolved CO2 is likely to be highly species specific, and establishing generally applicable safety levels is not possible. As such, there are no unequivocally accepted guidelines for safe levels of dissolved CO2 in

ARTICLE IN PRESS CO2 IN AQUACULTURE

5

aquaculture systems. The majority of research on safe levels of CO2 in aquaculture systems has been conducted on salmonids, and even within this species group, it has been difficult to find consensus. Colt (1991) proposed a division between cold- and warm water fishes, with CO2 thresholds of 10–20 mg L
1 and 20–40 mg L
1, respectively, with no apparent distinction between species or life stage. Currently, there are few countries with regulations or guidelines on this issue. Norway has set a threshold for dissolved CO2 at 15 mg L
1 (FOR, 2004) while the UK has set a more conservative threshold at 10 mg L
1. The latter is in good agreement with no effect concentrations of dissolved CO2 found for Atlantic salmon in both sea- and freshwater (Fivelstad et al., 1998; Khan et al., 2018; Noble et al., 2012). The tolerance levels of salmonids to dissolved CO2 are largely based on data relating to production variables, such as specific growth rates, feed conversion efficiencies, and condition factors. The levels at which animal welfare may become compromised has received little attention, although it is known that already at PCO2 levels below 0.15 kPa a range of fishes show changes in their sensory ability, cognition, behavior, and learning abilities (Heuer and Grosell, 2014, and references therein, but see also Chapter 5, Vol 37: Heuer et al., 2019; Chapter 9, Vol 37: Munday et al., 2019). As highlighted by Ellis et al. (2017), tolerance studies in aquaculture are typically characterized by the use of control treatment groups that experience CO2 concentrations well above saturation levels, and this may represent a bias in data interpretation. In the absence of operational welfare indicators for fish, such studies are limited to investigating conventional stress response parameters, such as changes in plasma cortisol and glucose concentrations (Petochi et al., 2011), which are poor indicators of chronic stress in fishes (Aerts et al., 2015).

4. FREE CO2, pH AND ALKALINITY In aquaculture production systems, it is most common to refer to the amount of CO2 present in free form, as dissolved CO2 L
1, rather than expressing CO2 in partial pressures. Measurement of CO2 in water can be done using a CO2 electrode, which is an infrared gas analyzer situated behind a gas permeable membrane, or by coulometric titration. The former is a more practical approach for fish farmers, allowing for a direct assessment of the levels of CO2 in a system, without having to measure temperature, alkalinity and pH of the water. In this chapter, the originally reported CO2 concentrations or partial pressures are used throughout the text, but to facilitate comparison, Fig. 1 shows a correlation between partial pressures and dissolved CO2 levels for freshwater and saltwater at three different temperatures.

ARTICLE IN PRESS 6

PETER VILHELM SKOV

Fig. 1. The correlation between partial pressure of CO2 and dissolved CO2 concentrations, and the effects of temperature (black 5 °C, red 15 °C, green 25 °C). Solid lines represent fresh water, while dotted lines represent seawater.

As CO2 is introduced into the water from the metabolism of fish and bacteria, it becomes hydrated to form carbonic acid, H2CO3. In all practical senses, this is still considered part of the free CO2 pool, as it can rapidly be dehydrated and revert back to free CO2 gas. Once dissolved in water, H2CO3 can further enter into the carbonic acid-base equilibrium, where it may donate one or two protons, to become either bicarbonate (HCO
3 ) or carbonate (CO2
). The concentration of dissolved inorganic carbon (DIC) in 3 water constitutes the sum of carbonic acid, bicarbonate, and carbonate,   ½DIC ¼ ½H2 CO3  + ½HCO3
 + CO3 2
: Ultimately, the chemical form of inorganic carbon dissolved in water, is determined by the pH of the water. Alkalinity, the stoichiometric sum of bases in a solution, is the main determinant of the pH buffering capacity of water, of which carbonate alkalinity makes up the main part. Carbonate alkalinity is consumed by nitrifying bacteria, while, at the same time, protons are liberated from the conversion of ammonia to nitrite. Therefore, it is necessary to add alkalinity, typically achieved by the use of various hydroxide, carbonate or lime complexes. Due to the pH buffering capacity of alkalinity, this also means that in waters with different alkalinities, the absolute amount of dissolved

ARTICLE IN PRESS CO2 IN AQUACULTURE

7

inorganic carbon present as free CO2 at a given pH value will increase with increasing alkalinity. For most water sources, the amount of DIC present as free CO2 at pH 7–8, ranges from 1 to 10%, but below pH 7, CO2 quickly becomes the dominant carbon species, and at pH 5, nearly all inorganic carbon is present as CO2. 4.1. Current Removal and Control Practices The necessity of a dedicated CO2 stripper or degassing unit depends largely on the oxygenation method in a given aquaculture facility. In facilities that use air for the oxygenation of water, CO2 removal is achieved as a by-product, due to the large volumes of air delivered to the system. On the other hand, in facilities that use pure oxygen injection systems, there will be a build-up of CO2 that requires removal. In intensive systems, the concentration of CO2 in the water is much higher than in the ambient air. The air that is brought into contact with water during degassing quickly becomes laden with CO2. An example of this is given by Summerfelt et al. (2000), showing that the CO2 concentration in the outlet air from degassing may reach 6–8000 ppm. This rapid increase in CO2 reduces the transfer gradient and impairs removal efficiency. With the use of airlift pumps, Moran (2010b) showed that with an incoming CO2 concentration between 10 and 40 mg L
1, the maximum obtained removal efficiency did not exceed 40%, even at considerable flow rates. Efficient removal of CO2 from water therefore requires that degassers operate at gas to liquid (G:L) ratios as high as 10:1 in order to function efficiently (Summerfelt et al., 2000). Such high G:L ratios make it difficult to perform subsurface aeration, and instead, degassing facilities frequently use surface aeration, where water is brought into contact with air, rather than vice versa. This is achieved by the use of cascade columns or packed columns, where water is pumped to height and allowed to trickle down through perforated plates or a packed media, while air is delivered via the bottom to achieve a counter current flow. Using this approach enables removal efficiencies as high as 75–85% even at moderate CO2 concentrations (Moran, 2010b; Summerfelt et al., 2000). The efficiency of these degassing installations are influenced by the height of the degassing unit (contact time) and design of the packing media, and have been reviewed in a number of publications (Colt et al., 2012a,b; Moran, 2010a,b; Summerfelt et al., 2000). Reportedly, CO2 control is occasionally also achieved by chemical means by increasing alkalinity of the water. This has multiple effects, one being that a higher pH of the water drives a larger fraction of the inorganic carbon into bicarbonate or carbonate. The other is that this allows for an overall increase in the total amount of dissolved inorganic carbon that can accumulate in the system. This means that for any given decrease in water pH, the amount of

ARTICLE IN PRESS 8

PETER VILHELM SKOV

dissolved CO2 increases proportionally with alkalinity (Loyless and Malone, 2004), and furthermore introduces the risk that in the face of quite modest decreases in pH considerably more free CO2 can be mobilized. In the aquaculture of European eel (Anguilla anguilla) efficient CO2 removal is achieved by maintaining the majority of dissolved inorganic carbon in the form of CO2. RAS facilities for eel control water pH in a range between 5.8 and 6.5. This ensures that close to all DIC is present as free CO2 and facilitates efficient degassing. 5. DYNAMICS OF CO2 IN AQUACULTURE Daily variations in levels of dissolved CO2 are likely to be largely dependent on the type of aquaculture operation, the biomass of fish in the system, feeding levels and implementation of water treatment installations. Degassers will obviously play a large role in CO2 dynamics in a facility, but also the type of aeration used, if any, as well as sludge removal practices. Recirculating facilities, despite considerable variation in design and specifications, have a great deal of control over dissolved gases. While dissolved CO2 concentrations may be generally elevated above saturation levels, they tend to be stable. For example, in Danish trout farms which are constructed as outdoor concrete raceways with fish densities up to 50 kg m
3, CO2 levels rarely exceed 10 mg L
1, and daily fluctuations are typically <2–3 mg L
1 (personal observation). These facilities operate without dedicated degassers, but instead employ the airlift principle to achieve oxygenation of water and for water circulation throughout the farm. The airlift principle is based on delivering air at depth via diffusors located at the bottom of a 2–4 m deep U-tube interconnecting rearing units. The addition of air reduces the overall mass of water on one side of the tube, enabling gravity to push water through the tube. Although large volumes of air are delivered, G:L ratios are still modest, but a large number of airlifts are positioned throughout the farm, which ensures the removal of a large proportion of CO2. RAS facilities that are structurally enclosed typically use pure oxygen, or a combination of oxygen and air, for subsurface oxygenation. In this type of facility a dedicated degasser is required to maintain acceptable levels of dissolved CO2.

6. EFFECTS OF DISSOLVED CO2 ON GROWTH There is a great deal of evidence showing that high levels of dissolved CO2 have a negative impact on growth of fishes. This has been demonstrated for

ARTICLE IN PRESS CO2 IN AQUACULTURE

9

several species in aquaculture production, predominantly Atlantic salmon (Fivelstad et al., 2003a,b, 2007, 2018, 1998, 1999; Hosfeld et al., 2008; Khan et al., 2018; Gil Martens et al., 2006; Mota et al., 2019), but also in rainbow trout (Oncorhynchus mykiss) (Danley et al., 2005; Good et al., 2010; Smart et al., 1979), Atlantic cod (Gadus morhua) (Foss et al., 2006; Moran and Støttrup, 2011), turbot (Scopthalmus maximus) (Stiller et al., 2015), white sturgeon (Acipenser transmontanus) (Crocker and Cech, 1996) spotted wolfish (Anarhichas minor) (Foss et al., 2003) and pike perch (Sander lucioperca) (Steinberg et al., 2017). The effects of CO2 on Atlantic salmon performance were reviewed by Fivelstad (2013). The clear pattern that emerges is that as the severity of CO2 exposure increases, a reduction in feed intake becomes apparent. Numerous species experience progressive loss in metabolic scope (MS, defined as the difference between MMR and SMR) as CO2 levels increase (see also Chapter 6, Vol 37: Lefevre, 2019), but a clear correlation between loss of MS and reduced appetite is lacking. Recently, Khan et al. (2018) found that in Atlantic salmon, there is a 35% loss in MS at 20 mg CO2 L
1, and a 55% loss in MS at 30 mg L
1, compared to control levels, originating from an inability to uphold maximum metabolic rates (Fig. 2). Data on effects of CO2 on growth in nonsalmonids is given in Table 1. With a few exceptions in gilthead seabream (Ben-Asher et al., 2013), Atlantic cod (Foss et al., 2006), and rainbow trout (Danley et al., 2005), CO2 exposure is accompanied by a decrease in feed intake. Regardless of feed intake, feed conversion ratios are almost always negatively affected, although exceptions are found for turbot (Stiller et al., 2015) and spotted wolffish (Foss et al., 2003), while specific growth rates are consistently reduced. Based on data from Atlantic salmon, Fivelstad et al. (2015) established a model describing the relationship between dissolved CO2 concentrations and changes in specific growth rate (SGR) (Fig. 3A). From this it is apparent that there is a critical CO2 level close to 15 mg L
1, beyond which growth performance quickly becomes compromised. The model is specific to parr. Smolts do not appear to show a breakpoint in their SGR when exposed to increasing CO2 levels, but rather a linear response (Fig. 3B). A multitude of other factors are also likely to contribute to reduced growth, including digestive efficiency, overall cost of living, and increased protein oxidation. In attempts to elucidate the role of these mechanisms, a number of recent publications have approached the issue from a bioenergetic angle, examining how dissolved CO2 influences factors such as metabolic rate, metabolic scope, and instantaneous fuel use (Khan et al., 2018; Methling et al., 2013; Steinberg et al., 2018; Stiller et al., 2015). A summation of possible, although not exhaustive, mechanisms are highlighted in Fig. 4.

ARTICLE IN PRESS

Fig. 2. Standard metabolic rate (SMR) (A), maximum metabolic rate (MMR) (B) and the associated linear depressant effect of CO2 on the difference between MMR and SMR; the metabolic scope (MS) (C) of Atlantic salmon exposed acutely to different [CO2]. Values are measured as the rate of oxygen consumption (Ṁ O2) in mg O2 kg
1 h
1. Modified from Khan et al. (2018).

Table 1 Summary of known effects of CO2 exposure and the corresponding experimental conditions on growth performance investigated for nonsalmon species Size

[CO2]

T

Sal (%o)

SGR

FCR

CF

FI

Duration

Atlantic cod (Gadus morhua)

80

2 8 18 3–4 4–5 7–9 10–12 5 15 21 5.2 20.5 32.3 56.3 7.2 22.2 39.3 0.8 4.6 23.8 4.9 25.8 41.6 4.4 14.7 29.7

10 10 10 10 10 10 10 7 7 7 22 22 22 22 22 22 22 26 26 26 18 18 18 23 23 23

35 35 35 35 35 35 35 0 0 0 35 35 35 35 35 35 35 35 35 35 20 20 20 0 0 0

1.7 1.35 0.75 1.13 1.15 0.98 0.93 0.75 0.65 0.5 2.96 2.60 2.14 1.99 0.40 0.36 0.28 6.05 5.77 4.52 1.5 1.3 1.0 0.89 0.86 0.78

– – – 1.08 0.97 1.11 1.18 1.08 1.39 1.86 1.15 1.52 2.14 2.25 2.77 2.87 4.32 0.98 0.98 1.26 0.7 0.7 0.7 1.04 1.14 1.13

1.07 1.01 0.97 0.87 0.87 0.85 0.84 1.04 0.99 0.97 – – – – – – – – – – 3.2 2.9 2.75 – – –

– – – 833 TFC 839 TFC 837 TFC 845 TFC 190 TFC 130 TFC 90 TFC 25.8 TFC 27.8 TFC 25.2 TFC 23.6 TFC 249 TFC 224 TFC 242 TFC 3080 TFC 2820 TFC 2120 TFC 1.3% d
1 1.2% d
1 0.9% d
1 0.92% d
1 0.98% d
1 0.88% d
1

40 40 40 63 63 63 63 60 60 60 64 64 64 64 71 71 +71 32 32 32 56 56 56 58 58 58

Arctic charr (Salvelinus alpinus)

Gilthead seabream (Sparus aurata)

Yellowtail kingfish (Seriola lalandi)

Turbot (Scopthalmus maximus)

Pikeperch (Sander lucioperca)

33.4 30.5 31.8 32.4 380 350 320 26.5 22.7 18.9 14.7 405 384 294 164 153 114 160 130 100 447 435 420

Comments

Ref 1

5

2

a

3

3

b,c

4

6

7

(Continued )

ARTICLE IN PRESS

Species

Table 1 (Continued) Species

Size

[CO2]

T

Sal (%o)

SGR

FCR

CF

FI

Duration

Spotted wolffish (Anarhichas minor)

33.1 30.0 30.7 21.4 8.99 5.57 817 849 200 200 135 650 550 500

1.1 18.1 33.5 59.4 0.5 42 8 24 12 24 55 22.1 34.5 48.7

6 6 6 6 19 19 15 15 9 9 9 13 13 13

33 33 33 33 0 0 0 0 0 0 0 0 0 0

1.0 0.9 0.9 0.4 2.86 1.17 1.65 1.67 0.84 0.84 0.69 1.05 0.85 0.73

0.69 0.65 0.62 0.71 – – 1.25 1.25 1.41 1.45 1.66 – – –

0.97 0.90 0.92 0.92 – – – – – – – – – –

235 TFC 213 TFC 174 TFC 62 TFC – – – – – – – 2.5% 2.5% 2.5%

70 70 70 70

White sturgeon (Acipenser transmontanus)

a

Ref 8

12 155 155 275 275 275 84 84 84

9 10

11

TFC was recalculated from FCR and weight gain. TFC recalculated from individuals feed intake, duration and number of fish. c CO2 calculated from pH with eq. 7 in Appendix F in National Academies of Sciences, Engineering, and Medicine (2017). Size refers to final body mass in the experiments (g); T is temperature in °C; SGR refers to specific growth rate in % BM increase per day; FCR is feed conversion ration (g feed per g weight gain); CF, Fulton’s condition factor; FI, feed intake as either % of body mass per day or total feed consumption (TFC) over the duration of the experiment, duration of the experiment is in days. 1 Moran and Støttrup (2011); 2 Musa and Thorensen (2013); 3 Ben-Asher et al. (2013); 4 Abbink et al. (2011) 5 Foss et al. (2006); 6 Stiller et al., (2015); 7 Steinberg et al. (2017); 8 Foss et al. (2003); 9 Good et al. (2010); 10 Smart et al. (1979); 11 Danley et al. (2005); 12 Crocker & Cech (1996). b

ARTICLE IN PRESS

Rainbow trout (Oncorhynchus mykiss)

Comments

ARTICLE IN PRESS CO2 IN AQUACULTURE

13

Fig. 3. Percentage reduction in mean specific growth rate in Atlantic salmon parr compared to normocapnic (2 mg CO2 L
1) values followed the equation SGR (% day
1) ¼1.32 + 0.0158 ∙ [CO2] – 0.00093 ∙ [CO2] (n ¼ 193, r2 ¼ 0.50; P < 0.05), which was indexed to 100 at a CO2 concentration of 2 mg L
1 (A), and SGR for smolts, which followed the equation SGR ¼
0.0095  [CO2] + 1.18 (r2 ¼ 0.99) (B). Panel A modified from Fivelstad et al. (2015) and Panel B from Khan et al. (2018).

ARTICLE IN PRESS 14

PETER VILHELM SKOV

Fig. 4. A schematic illustration of possible mechanisms influencing the growth of fish in response to CO2 exposure. See text for details.

6.1. Feed Intake and Appetite Appetite in fishes is regulated by a number of complex pathways under the influence of intrinsic and extrinsic factors (Volkoff et al., 2009). Factors such as dissolved oxygen, temperature, and photoperiod have received considerable attention. A rapid and common response to hypoxia is a decrease in feed intake (Buentello et al., 2000; Chabot and Dutil, 1999; Pedersen, 1987; Pichavant et al., 2001), which is considered a strategy to conserve or prioritize oxygen consumption, and which is typically accompanied by a reduction in voluntary activity (Chabot and Dutil, 1999; Pichavant et al., 2001). The responsible mechanisms for appetite suppression has received increasing interest, and Bernier et al. (2012) recently found that in common carp (Cyprinus carpio) an increased expression of the anorexic hormone leptin, appears to be the main driver of decreased feed intake during hypoxia. Surprisingly, comparable studies in which the effects of high levels of dissolved CO2 on appetite regulating hormones are absent, despite documented

ARTICLE IN PRESS CO2 IN AQUACULTURE

15

effects of CO2 on feed intake. It is plausible that changes in oxygen availability caused by elevated CO2 levels could potentially exert a hypoxia-like effect on hormones that regulate appetite, considering that high CO2 levels reduce both the oxygen affinity and oxygen carrying capacity of hemoglobin (Jensen et al., 1993). Feeding trials conducted under hypercapnic conditions do not show equivocal results, and in a number of experiments, determination of feed intake is not feasible (Khan et al., 2018; Moran and Støttrup, 2011). In turbot, fish exposed to 41.6 mg CO2 L
1 had a daily feed intake that was 40% lower than for fish reared at 4.9 mg CO2 L
1 (Stiller et al., 2015). While appetite hormone levels were not investigated, the authors found that routine metabolic rates decreased from 80 to 45 mg O2 kg
1 h
1, suggesting that at high concentrations of dissolved CO2, oxygen availability becomes compromised, despite normoxic rearing conditions. 6.2. Digestive Function A possible reason for the observed reduced feed efficiency and growth during exposure to unnaturally high CO2 concentrations may be an associated reduction in digestive performance, possibly due to a decreased ability to digest macronutrients or a prolongation of the digestive processes. This could be caused by stress induced impairment of digestive function, as has been observed in Arctic charr (Salvelinus alpinus) (Olsen & Ringø, 1999) or redistribution of blood flow from the gastrointestinal tract (Farrell et al., 2002). Acute exposure to hypercapnia (PCO2 of 20 mmHg) in unfed white sturgeon causes a short-lived near cessation of gastric blood flow (GBF), and although a partial recovery did occur, a 40% reduction in GBF persisted for at least 20 min (Farrell et al., 2002). During chronic exposure to CO2 fish show an ability to recover their GBF over time, and do show differences in comparison to normocapnic individuals. In these experiments animals were unfed, and while GBF can be recovered under these conditions, it remains unknown to what extent CO2 exposure might influence the ability of fish to upregulate GBF by 60–80% following feeding, which has been shown to be the normocapnic response in rainbow trout (Eliason et al., 2008). Evidence from feeding experiments in Atlantic cod and European eel show that the duration of digestion is prolonged during exposure to hypercapnic conditions (Methling et al., 2013; Tirsgaard et al., 2015). In Atlantic cod, the duration of the digestive processes (the specific dynamic action, SDA), in individuals fed a 5% ration of herring fillet increased from 81 h under normoxic conditions, to 110 h in fish exposed to a PCO2 of 7 mmHg (Tirsgaard et al., 2015). Similarly, European eel fed 0.5% of their body mass with commercial eel pellets, increased the duration of their SDA response from

ARTICLE IN PRESS 16

PETER VILHELM SKOV

36 to 44 h (Methling et al., 2013). In both instances, this effect appeared, at least in part, to be caused by a reduction in metabolic scope, which prevented fish from obtaining peak Ṁ O2 values comparable to the control groups. While nitrogen excretion rates during digestion were unaffected in eel, this topic remains to be determined for Atlantic cod and other species. Interestingly, the SDA coefficient, which represents the fraction of ingested energy that is used to fuel the digestive processes, was unaffected. This indicates, that although digestion is prolonged, it does not appear to become more costly. These observations are somewhat in contrast to the results of Heuer and Grosell (2016) who measured oxygen consumption rates in isolated intestinal tissues in Gulf toadfish (Opsanus beta). Here, an acclimation to moderate hypercapnia (PCO2 ¼ 1.45 mmHg) resulted in an 8% increase in the surface specific metabolic rate of the anterior intestine, probably associated with the metabolic cost of increased secretion rates of HCO
3 . It is relevant to point out that the anterior intestine of fishes plays a minor role, if any, in the absorption of nutrients, so that this increased energy expenditure is unlikely to compromise digestive function. Furthermore, as pointed out by the authors themselves, minor increases in the oxygen consumption rates of the anterior intestine may not be reflected in whole animal metabolic rates (Heuer and Grosell, 2016). The data provided by Methling et al. (2013) demonstrates the significance of acclimation to chronically and constant CO2 exposure, as is employed in most studies. Here, the authors included a treatment group with oscillating PCO2 levels, from which it was apparent that the magnitude of the SDA response fluctuated in response to changes in CO2 tension (Fig. 5). Although acute stress is associated with a decrease in lipid digestibility (Olsen & Ringø, 1999), this does not appear to be the case during chronic exposure in acclimated fish. Unpublished data from our lab shows that fish acclimated to 11, 21, and 36 mg CO2 L
1 do not show changes in the digestibility coefficients of lipid and protein. Despite significant changes in the ability of fish to convert feed to growth (FCR) (Fig. 6A), no significant changes in the digestion of feed energy were observed (Fig. 6B). The net result of CO2 exposure was a near proportional increase in the gross cost of growth, which increased from 15.5 kJ g
1 in the normocapnic group, to 20.7 kJ g
1 in the highest CO2 exposure group (Fig. 6C).

7. SEAWATER TRANSFER Anadromous species such as Atlantic salmon undergo a parr-smolt transformation in preparation to seaward migration. This entails a number of physiological and morphological adaptations, which have recently been reviewed

ARTICLE IN PRESS

Fig. 5. Postprandial oxygen consumption in A. anguilla exposed to hypercapnia and/or low pH. (A) SDA response in control fish under normocapnic conditions, (B) under hypercapnic conditions at a constant high PCO2 of 60 mmHg, (C) under oscillating PCO2 conditions from 20 to 60 mmHg, and (D) in a low pH group (6.5) under normocapnic conditions. Eels were acclimated to hypercapnia/low pH for at least 3 weeks. The meal consisted of 0.5% BW in commercial feed pellets. Data points are hourly averages  s.e.m. (N ¼ 6–8). Straight lines represent standard metabolic rates (SMR)  s.e.m. as determined before feeding. Dotted line in B represents the actual CO2 partial pressures during the postprandial phase. Note how fluctuating CO2 levels causes opposing fluctuations in Ṁ O2 values in panel B. Modified from Methling et al. (2013).

ARTICLE IN PRESS 18

PETER VILHELM SKOV

Fig. 6. Feed conversion, energy digestibility and cost of growth (kJ g
1) in rainbow trout (100 g) during a short term growth trial (21 days) in fish acclimated for 3 weeks at different CO2 concentrations. Feed intake and digestibility of nutrients were unaffected by CO2, but feed conversion ratios and cost of growth increase. P.V. Skov, unpublished data.

ARTICLE IN PRESS CO2 IN AQUACULTURE

19

extensively by McCormick (McCormick, 2012). The Na+/K+ ATPase (NKA), Na+/K+/2Cl
transporter (NKCC) and the cystic fibrosis transmembrane regulator (CFTR) constitute the three main transport mechanisms responsible for salt secretion in fishes, of which NKA activity is up-regulated in preparation for, and in response to, saltwater exposure. Although limited data is available, there is evidence to suggest that mitochondria rich cells (MRCs) and pavement cells of the gill undergo changes under hypercapnic conditions, to compensate for the associated respiratory acidosis. To maintain plasma pH, fishes elevate their plasma bicarbonate concentrations by reducing Cl
/HCO–3 exchange (Cameron and Iwama, 1987) (see also Chapter 3, Vol 37: Brauner et al., 2019). This results in a net decrease in plasma Cl
levels, which in salmonids appears to be proportional to the PCO2 levels (see also Chapter 3, Vol 37: Brauner et al., 2019). Goss et al. (1992a) observed that this was achieved by a large (95%) reduction in MRC surface area, caused by covering of the adjacent pavement cells (PVCs). Following 6 h of hypercapnic exposure in brown bullhead (Ictalurus nebulosus), MRCs recessed into the filament, and the influx rate of sodium
Jin Na + increased almost threefold. The increase in PVC surface area was also associated with an increase in surface microvilli, indicative of an increased activity, as well as ultrastructural changes within the PVCs, presumably in the excretion of acid and uptake of Na+ (Goss et al., 1992b). Similar results have been observed for Atlantic salmon, in which gill NKA activity decreased significantly during short-term (24 h) exposure of postsmolts (Seidelin et al., 2001) as well as long-term (37 days) exposure of smolts (Hosfeld et al., 2008) to hypercapnia. However, Seidelin et al. (2001) also investigated the effects of longer exposure times, and observed that after 4 days, gill vacuolar H+-ATPase mRNA expression, and NKA activity were comparable to control values. Thus, although the implications of the functional changes associated with hypercapnia are not entirely clear, and probably related to smolt status, it is plausible that the modifications to the acid-base regulation and osmoregulatory function of the gill associated with CO2 exposure may represent a challenge in the face other osmotic disturbances. Net pen production of salmonids is typically characterized by the freshwater life stage occurring on land in recirculated aquaculture systems, following by a transfer to sea for grow out when fish reach a mass of 200–800 g. In some instances, it is possible for the RAS facility to gradually increase salinity in the tanks prior to transfer, but most commonly, the transfer occurs directly from freshwater to saltwater. Brauner et al. (2000) investigated the effects of previous hypercapnic exposure in Atlantic salmon on their ability to hypoosmoregulate following transfer to seawater. Contrary to expectations, it appeared that CO2 exposure in fact facilitated successful transfer, and fish that

ARTICLE IN PRESS 20

PETER VILHELM SKOV

had been exposed to CO2 showed no mortalities, and an ability to regulate plasma Cl
and Na+ that was as good, or even better, than control fish. The authors propose that this observation may be facilitated by the decreased plasma Cl
levels associated with acclimation to high CO2 levels, and that this may leave fish with a bigger tolerance window for influx of Cl
during seawater exposure, or that the morphological changes outlined above simply decrease permeability of the gills, and thus exerts a protective effect (Brauner et al., 2000). Despite an ability to successfully adapt to SW following CO2 exposure in the FW stage, other carry over effects seem to occur. Mota et al. (2019) investigated the effects of long term (12 weeks) CO2 exposure on the growth performance of Atlantic salmon reared in freshwater, and subsequently followed their growth performance for 6 weeks following transfer to SW under normocapnic conditions. Exposure to high CO2 concentrations higher than 12 mg L
1 resulted in significantly reduced growth trajectories in the last 6 weeks of the trial. Interestingly, subsequent to transfer to normocapnic SW, fish did not show an ability to improve their growth, but in fact showed a continuous decline in the growth performance. Since feed intake was not quantified, changes in appetite or feed conversion could not be assessed, but no CO2 induced damage, such as eye cataracts or nephrocalcinosis (see below) were observed, that could provide an explanation for the reduced growth performance. 8. PATHOLOGICAL EFFECTS OF CO2 8.1. Nephrocalcinosis Nephrocalcinosis is a pathological condition of the kidney which can be recognized macroscopically, as white calcareous deposits in the ureters and tubules, consisting mainly of calcium, phosphorous and magnesium (Smart et al., 1979). The prevalence of nephrocalcinosis is thought to depend on a variety of factors, including dietary mineral content, as well as dissolved CO2 levels in the water. The functional consequences of nephrocalcinosis in fish have not been determined, but is an area for further research, as tissue necrosis may ultimately lead to kidney failure (Fivelstad et al., 2018). While it has been suggested that other tissues such as intestine and gills may compensate for a loss of ionoregulatory function in the kidney during progressive nephrocalcinosis (Harrison and Richards, 1979), the occurrence of nephrocalcinosis is certainly a concern for the welfare of fish in aquaculture. Studies on the prevalence of nephrocalcinosis in Atlantic salmon exposed to hypercapnia in fresh water have provided unequivocal results.

ARTICLE IN PRESS CO2 IN AQUACULTURE

21

Hosfeld et al. (2008) quantified the occurrence of nephrocalcinosis exposed to 18 mg L
1 CO2 and found that 8% of individuals displayed the pathology after 42 days of exposure. In a study investigating the effects of up to 24 mg L
1 CO2 exposure for a period of 57 days, all groups, including a 6 mg L
1 control group, displayed more than 50% occurrence of nephrocalcinosis after 3 months of rearing (Fivelstad et al., 2003b). Fivelstad et al. (2018), in a 3-month study, found that nephrocalcinosis did not occur at CO2 concentrations less than 21 mg L
1, while studies by Gil Martens et al. (2006), Good et al. (2010), Fivelstad et al. (2015), and Mota et al. (2018) did not observe pathologies in fish reared at up to 40 mg L
1 dissolved CO2 for periods of up to 135 days. In contrast, Fivelstad et al. (1999) observed an almost linear correlation between the frequency of occurrence of nephrocalcinosis and dissolved CO2 levels after 61 days of exposure, showing that in fish exposed to 8, 19 and 32 mg L
1, approximately 30, 40, and 80% of individuals were inflicted. The majority of research on nephrocalcinosis has been conducted in salmonids in fresh water, but a few studies on marine species have shown that high CO2 concentrations may also lead to calcium deposits in the kidney of sea bass (Dicentrarchus labrax) (Petochi et al., 2011) and wolfish (Anarhichas minor) (Foss et al., 2003). From the data available, it seems that there are other contributing factors in the formation of nephrocalcinosis, which are, as yet, unidentified. It has been proposed that elevated bicarbonate concentrations in the urine may be a contributing factor (Fivelstad et al., 2018), although presumably these would be comparable between groups exposed to similar CO2 levels, unless the chemical composition of water and body fluids differed between experiments. Water hardness could potentially be an explanatory factor, if calcium ions were contributing to increasing hardness levels, which may lead to a net calcium influx across the gill epithelium (Flik et al., 1995), or if branchial or opercular calcium transport mechanisms are affected by exposure to high CO2 levels. In the face of high Ca2+ and bicarbonate levels in the plasma, aqueous calcium bicarbonate could potentially react with CO2 and water to form precipitations of calcium carbonate. Exposure to high CO2 levels alone does not appear to cause an increase in the whole body (Cameron, 1985; Fivelstad et al., 2018; Gil Martens et al., 2006) or whole blood content of calcium (Good et al., 2018). Despite this, evidence suggests that an increased rate of bone remodeling does occur (Gil Martens et al., 2006), potentially elevating the flux of free Ca2+ levels in the plasma. Mechanisms by which increased calcium complexes are formed remains to be investigated further. Interestingly, if fish are simultaneously exposed to moderate (111%) or severe (127%) hyperoxia and hypercapnia (18 mgL
1), the prevalence of nephrocalcinosis increased in 33–50% of the individuals (Hosfeld et al., 2008).

ARTICLE IN PRESS 22

PETER VILHELM SKOV

The authors did not discuss the possible mechanisms by which this occurred, but suggested that hyperoxia induced hypercapnia (Wood, 1991) may have resulted in a more severe CO2 exposure than indicated by the dissolved CO2 levels in the water. Such observations lend further evidence to the notion that supersaturated oxygen conditions exacerbate the effects of hypercapnia (Brauner et al., 2000). 8.2. Swim Bladder Inflation The initial gaseous inflation of the swim bladder of larval fishes is achieved by gulping air at the water surface, followed by transfer of air to the swim bladder via the pneumatic duct (Woolley and Qin, 2010). Although this is considered the mechanism for swim bladder filling in all species, Elsadin et al. (2018) suggested that initial filling of the swim bladder in white grouper (Epinephelsu aeneus) is achieved, at least in part, by means of oxygen unloading from hemoglobin to the swim bladder via the rete mirabile, and that dissolved CO2 may impair this crucial step in development. A limited amount of research has been conducted on the effects of dissolved CO2 on larval fishes in culture. Arguably, this is because biomass loads are much lower than in grow out production systems, and it is assumed that good water quality, including dissolved gases, are easier to maintain. However, aquaculture facilities for larval rearing are still based on recirculation technology with limited water exchange, and while biomass may be low, feeding rates are high, and the possibility of CO2 accumulation cannot be ignored. Elsadin et al. (2018) investigated the effects of three levels of dissolved CO2; 0.8, 5.6, and 28.6 mg L
1, on swim bladder inflation rate and volume in white grouper. In the normocapnic treatment, 79% of all fish had normal swimbladder inflation at 105 days post hatch, only 57% of fish in the moderate hypercapnic successfully inflated their swim bladder, and only 42% were capable in the high CO2 group. How CO2 interferes with swim bladder filling if the process is physostomous, (i.e. by filling the swimbladder with engulfed atmospheric air via a pneumatic duct in the alimentary canal) is unknown, but improper swim bladder inflation is a common problem in larval culture, and the potential role of dissolved CO2 warrants further attention. Following swim bladder filling, gas regulation is achieved by unloading oxygen to the swim bladder via the rete mirabile, and in this process, dissolved CO2 appears to have an effect, for reasons as yet unknown. White grouper that had been exposed to medium and high concentrations of CO2, but had successfully inflated swim bladders, showed over a 50% reduction in swim bladder volume (Elsadin et al., 2018). In the absence of successful swim bladder inflation, fish are unable to properly orient themselves in the water column, tend to display erratic swimming and elevated metabolic rates, and have much higher

ARTICLE IN PRESS CO2 IN AQUACULTURE

23

mortality rates (Lund et al., 2014). It is generally assumed that larval fishes are more sensitive to dissolved CO2, therefore future work on hypercapnia in aquaculture should also include aspects of larval husbandry, particularly at more moderate levels relevant for such rearing systems. 8.3. Cataract Formation Cataracts in the eyes of farmed fish is a major concerns from an economical and ethical standpoint. Cataracts are a clouding of the lens of the eye that may initially be small and nucleated, but which can progress into a cortical cataract affecting the entire lens. The cause of cataracts is not clear, and a multitude of different possible mechanisms have been suggested, including diet composition, growth rates, parasite infection, and water quality parameters (Bjerkås et al., 2004). The transfer of salmon from fresh- to seawater has also been shown to cause significant increases in cataract formation, indicating that osmotic disturbance may be a causative agent (Breck and Sveier, 2001). Only recently have correlations between CO2 exposure and cataract formation been documented, although it does not appear to apply to all species. In juvenile Atlantic cod, a positive correlation between not only the prevalence of cataracts, but also their severity, and dissolved CO2 concentrations, was documented by Moran et al. (2012). In fish exposed to 18 mg L
1 CO2, more than 80% of individuals displayed some sort of eye damage after 55 days of rearing, and 73% had some degree of cataract formation, with nearly half being severe cortical cataracts affecting more than 75% of the lens (Fig. 7). In contrast, fish exposed to 8 and 2 mg CO2 L
1 had only 11–14% incidence of cataracts, which were generally less severe. A study by Neves and Brown (2015) confirmed these results in a study where cod were exposed to slightly higher CO2 levels of 7, 12, and 20 mg L
1 for a much longer period of 5 months. Inspection of fish on a monthly basis revealed that cataract formation continued to develop, such that in the high CO2 group at the end of the trial all had cataracts in the most severe form, and that even the lowest CO2 group had a 30% incidence rate affecting 51–75% of the lens. Moran et al. (2012) suggested a number of possible mechanisms by which dissolved CO2 might induce cataract formation, including CO2 disruption of the pH levels in the eye, ionic balance of the lens, or alterations to the oxygen delivery to the eye. In contrast to the results from cod, studies on Atlantic salmon have not documented any correlation between dissolved CO2 levels as high as 40 mg L
1 and the prevalence of cataracts (Mota et al., 2019; Waagbø et al., 2008). Instead, Waagbø et al. (2008) found a correlation between hyperoxygenated water and cataract formation, once more highlighting the possible interactive roles of oxygen and carbon dioxide.

ARTICLE IN PRESS 24

PETER VILHELM SKOV

Fig. 7. Prevalence of different eye lesions in Atlantic cod subjected to different CO2 treatments over a period of 55 days (A), and cataract intensity observed in each treatment. Modified from Moran et al. (2012).

ARTICLE IN PRESS CO2 IN AQUACULTURE

25

9. CONFOUNDING WATER QUALITY EFFECTS For the large majority of research conducted, practical and logistical conditions often dictate that experimental work in aquaculture is largely limited to studies where CO2 levels fluctuate as the sole variable. More often than not, multiple water quality parameters change in concert in aquaculture systems. Given that there is continued pressure to increase the degree of recirculation and loading density within RAS, new water quality parameters are becoming relevant, such as the accumulation of metals of dietary origin, levels of dissolved atmospheric gases, and nitrogenous waste compounds.

9.1. Hypercapnia and Hypoxia In a modern RAS facility the occurrence of hypoxic events are unlikely, due to the option of supplemental oxygenation. However, more than 60% of the global aquaculture production volume still comes from extensive or semiintensive pond culture (Bostock et al., 2010) which rely on limited or absent aeration technology. While some semi-intensive pond systems may employ aeration devices, such as paddle wheel aerators or diffusors, extensively cultured ponds do not. Instead, these ponds rely on photosynthesis for the addition of dissolved oxygen to the water during daylight hours, and maintain low fish densities in the ponds. Following sunset, oxygen becomes severely depleted from the respiration of fish, plankton and bacteria (Lefevre et al., 2011), while dissolved CO2 levels progressively increase. Despite this undisputable correlation, surprisingly little attention has been given to the effects of combined hypoxia and hypercapnia in aquaculture. Exposure to elevated CO2 levels induces a Bohr shift and Root effect, causing both a reduced oxygen binding affinity of hemoglobin as well as a reduced oxygen carrying capacity of the blood, respectively (Jensen et al., 1993; see also Chapter 3, Vol 37: Brauner et al., 2019). When hypoxia is concurrent, fish must increase their ventilatory work to compensate for the reduction in oxygen availability (see also Chapter 3, Vol 37: Brauner et al., 2019). The response of fish to combinations of hypoxia and hypercapnia are dependent on the dissolved levels of both gases. In many fish species, moderate levels of dissolved CO2 lead to either no change (Khan et al., 2018; Thomas et al., 1983) or only modest increases in Ṁ O2 (Crocker and Cech, 2002). Presumably this is driven in part by an increase in the cost of ventilation (Gilmour and Perry, 1994; Perry and Gilmour, 1996). As hypoxia progresses, venous PO2 decreases, which increases the solubility of CO2 (Permentier et al., 2017), the so-called Haldane effect, which further reduces the affinity of hemoglobin for O2 (Jensen et al., 1993).

ARTICLE IN PRESS 26

PETER VILHELM SKOV

The effects of combined hypoxia and hypercapnia on hypoxia tolerance has been demonstrated in the European eel (Cruz-Neto and Steffensen, 1997). Here it was shown that the critical oxygen tension, below which fish are no longer able to maintain their SMR, increases from 3.4 kPa under normocapnic conditions, to 6.4 kPa O2 at a PCO2 of 4 kPa. With a simultaneous decrease in Ṁ O2 with increasing CO2 levels, this shows that both hypoxia tolerance and metabolic scope are compromised when hypoxia and hypercapnia cooccur. This may not necessarily be the universal response of fishes; in white sturgeon, Ṁ O2 increases almost twofold when hypoxia (PO2 ¼ 12 kPa) and hypercapnia (PCO2 ¼ 2.7 kPa) occur simultaneously (Crocker and Cech, 2002). Presumably, the experimental conditions and acclimation history of the animals have a large influence on the outcome of experiments examining the effects of hypoxia–hypercapnia. The different responses observed in the studies above may also be a result of the different PCO2 and PO2 levels investigated. While fish exposed to long-term hypercapnia show an ability to compensate for respiratory acidosis by accumulation of plasma bicarbonate (Larsen and Jensen, 1997), this is, in all likelihood, not the case for fish that experience large fluctuations in PCO2 on a daily basis. The effects of concurrent hypoxia and hypercapnia under such conditions warrant further attention, specifically directed towards how it affects fish health, welfare and productivity under extensive culture. Concurrently, CO2 may also influence the tolerance of fish to hypoxia, which is the topic of Lefevre (2019: Chapter 6, Vol 37). 9.2. Hypercapnia and Metal Toxicity All fish have a nutritional requirement for a number of trace metals, including magnesium, zinc, selenium, manganese, iron and copper (Antony Jesu Prabhu et al., 2016), and as in other animal feed, vitamin and mineral premixes are added to satisfy nutritional requirements. Because fish do not completely digest their feed, a fraction of dietary metals are lost via fecal material (Bury, 2002). As water exchange rates in recirculating aquaculture systems (RAS) decrease, there is growing awareness that a number of organic and inorganic compounds may start to accumulate in the system. In a study by Davidson et al., (2009) the accumulation of metal ions was examined in RAS rearing rainbow trout (Oncorhynchus mykiss) with high or low hydraulic retention times (6.7 vs 0.67 days). Out of the 8 metals investigated, copper increased 10-fold compared to intake water, while zinc increased 2-fold. Similar results have been found in RAS with turbot; as water exchange rates decrease, significant accumulation of zinc, copper, cobalt and manganese was observed (van Bussel et al., 2014). Copper toxicity is influenced by several environmental variables, only a few of which are discussed here. Cu exposure in fish leads to ionoregulatory disturbances, affecting sodium homeostasis (Lauren and McDonald, 1987).

ARTICLE IN PRESS CO2 IN AQUACULTURE

27

However, Cu toxicity appears to be higher for fish that live in an anisoosmotic environment presumably due to a higher requirement for sodium regulation (Blanchard and Grosell, 2006). Furthermore, the toxicity of Cu is influenced by water pH. Acute toxicity studies on the streaked prochilod (Prochilodus lineatus) showed that at pH 4.5 the LC50 increased 6-fold, from 0.016 mg Cu L
1 to 0.1 mg Cu L
1 (Carvalho and Fernandes, 2006). In their study on the interacting effects of Cu and hypercapnia exposure, Larsen et al. (1997) showed that Atlantic cod exposed to 0.4 mg L
1 copper responded with extracellular acidosis due to a decrease in plasma bicarbonate concentrations with no change in plasma PCO2. Simultaneous exposure to hypercapnia (1 kPa) resulted in an aggravation of this acidosis as well as a large increase in PCO2, and prolonged acid-base recovery compared to fish that had only been exposed to hypercapnia. These results show that Cu exposure exerts a negative effect on the ability of fish to recover from acute CO2 exposure. On the other hand, hypercapnia exerted a protective effect against the toxicity of Cu, by reducing the osmotic and ionic (Na+, K+, Cl
) disequilibria caused by copper alone. The exact mechanisms are not detailed, but the authors suggest that hypercapnia may have caused reduced gill permeability. Alternatively, the PCO2 used (1 kPa) may have reduced water pH sufficiently to reduce the toxicity of Cu (Carvalho and Fernandes, 2006). A somewhat similar response was observed for rainbow trout in freshwater (Wang et al., 1998). Here, the simultaneous exposure to 0.6 mg L
1 Cu and 0.8 kPa CO2 also aggravated the extracellular drop in pH and prolonged recovery, but contrary to Larsen et al. (1997) no protective effects of hypercapnia on ionic disturbance from Cu exposure were observed. Although the accumulation of Cu in RAS with low water exchange does not reach levels as high as those reported in the studies above, observed levels of 0.045–0.09 mg L
1 (Davidson et al., 2009; van Bussel et al., 2014) may be cause for concern, and potential toxic effects from long term exposure should be investigated. 9.3. Carbon Dioxide and Aluminum Aluminum toxicity in fishes has been extensively reviewed by (Exley et al., 1991). Briefly, aluminum (Al) is most toxic under acidic conditions, where Al speciation leads to high solubility in water (as Al(OH)2+). For this reason, toxicity of Al has been of interest to aquaculture for the past several decades, particularly in combination with elevated CO2 levels which reduces water pH. This applies in particular for areas where Al is abundant in the natural water supply due to either run-off from land or naturally high deposits of aluminum ore in the earth. Al causes irritation of the gills, and can precipitates on the gills of fish as aluminum hydroxide, leading to ionoregulatory disturbance and an increase in gill membrane permeability leading to uptake of protons and Al, respiratory

ARTICLE IN PRESS 28

PETER VILHELM SKOV

failure and loss of osmoregulatory capacity (Exley et al., 1991). Stunted fish growth has been shown to occur for brown trout (Salmo trutta) at Al concentrations as low as 20 μg L
1 at pH levels of 4.4–5.2, and increased mortality was observed at 50 μg L
1 (Fairman and Sanz-Medel, 1995). Tench (Tinca tinca) experiencing short-term Al exposure (2 mg L
1) at pH 5, display a rapid drop in arterial PO2, blood oxygen content and saturation, in addition to a doubling of plasma PCO2 and lactate, as well as a drop in plasma chloride levels (Jensen and Weber, 1987). This suggests that Al toxicity is a relevant concern for aquaculture, especially when exposure coincides with hypercapnia induced reductions in pH. Furthermore, in the streaked prochilod (Prochilus lineatus) exposed to much lower Al concentrations (0.2 mg L
1) significant and persistent reductions in plasma osmolality were observed after 24 h from decreases in Na+ and Cl
, demonstrating the osmoregulatory disturbance of Al. After 96 h exposure there was a significant reduction in the number of both filamentous and lamellar MRC cells. A concomitant elevation in plasma glucose illustrates that fish exposed to Al suffer from metabolic disturbances, as was also indicated by the increase in blood hemoglobin concentrations (Camargo et al., 2009). The studies above are more ecotoxicological in nature, and fish in aquaculture may not experience Al concentrations and pH levels in this range. However, Al can be toxic at higher pH levels, and the water in the microenvironment of the gill may be as much as 0.5 pH unit lower than the ambient water (Wilson, 2011). See also Wilson (2011) in this series, for a more detailed description of Al toxicity and mechanisms. Fivelstad (2013) provided a theoretical relationship between dissolved CO2 levels in culture water and the pH at which Al becomes toxic, showing that at least in low alkalinity water, Al becomes toxic at levels as low as 7 mg L
1 dissolved CO2, where pH levels drop to 6.2 (Fivelstad et al., 2003a). Fivelstad et al. (2003a) showed that when Atlantic salmon were reared in soft water containing 2, 9, or 19 mg L
1 CO2 the corresponding pH levels dropped from 6.6 to 6.0 and 5.7 respectively. This lead to significant increases in the accumulation of Al in the gills, from 20 μg g
1 in the low CO2 treatment to 100 and >140 μg g
1 in the medium and high CO2 group. These findings illustrate the importance of dissolved Al levels in aquaculture, even at levels considerably lower than conventional toxicology studies, and perhaps particularly for fish undergoing the parr-smolt transformation. 10. CONCLUSIONS AND PERSPECTIVES A considerable amount of work has demonstrated the effects of CO2 on growth in a variety of fish species, although with an overrepresentation of the salmonids, in the grow-out phase. For the majority of these studies,

ARTICLE IN PRESS 29

CO2 IN AQUACULTURE

experimental conditions have favored exposure to constant levels of dissolved CO2, and with CO2 as the sole experimental variable. Dissolved CO2 levels fluctuate on an hourly or daily basis in all aquaculture production systems, but less so in recirculating systems, than in extensive systems with no water treatment technology, such as earthen ponds. Fish show remarkable capacity to acclimate to different environmental conditions. However, acclimation comes at a metabolic cost, but once achieved may not exert a large expenditure for the individual, depending on the severity of conditions. Multiple water quality parameters, when not stable, tend to fluctuate in concert. In pond culture, dissolved CO2 levels and dissolved oxygen levels fluctuate in a reciprocative manner, accompanied by changes in pH, ammonia ionization, and more. These events happen on a diurnal rhythm, probably never allowing fish to achieve homeostasis, and induce a large allostatic load. Further research is required to understand how fish respond to daily fluctuations in multiple water quality parameters, relative to long-term acclimated individuals. From our current knowledge on hypercapnia, it is apparent that there are interacting effects of hyperoxia and hypercapnia. It may be hypothesized that hyperoxia either aggravates the effects of CO2 by additional respiratory acidosis, or causes oxidative damage to the gills and their adaptive capacity. In the light of the use of pure oxygen injection in RAS, this is a highly relevant are for further research. There is a remarkable scarcity of work on larval stages of fishes. Aquaculturists may perceive larviculture as an aspect of fish husbandry that is unproblematic in relation to CO2. However, husbandry practices dominated by low water exchange rates and high input of organic matter (feed), in combination with the increased sensitivity of fish larvae to hypercapnia, may prove this wrong. Although the mechanisms are not entirely clear, it appears that even quite low CO2 levels may potentially lead to development issues. In this respect, aquaculture researchers should join forces with ecologists, who already have a detailed understanding on the effects of increasing CO2 levels in relation to climate change science. Such research should aim to include carry-over effects from hypercapnic exposure during the larval stage to juvenile and adult fishes. REFERENCES Aerts, J., Metz, J.R., Ampe, B., Decostere, A., Flik, G., De Saeger, S., 2015. Scales tell a story on the stress history of fish. PLoS One 10, 1–17. Antony Jesu Prabhu, P., Schrama, J.W., Kaushik, S.J., 2016. Mineral requirements of fish: a systematic review. Rev. Aquac. 8, 172–219. Atkins, M.E., Benfey, T.J., 2008. Effect of acclimation temperature on routine metabolic rate in triploid salmonids. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 149, 157–161.

ARTICLE IN PRESS 30

PETER VILHELM SKOV

Ben-Asher, R., Seginer, I., Mozes, N., Nir, O., Lahav, O., 2013. Effects of sub-lethal CO2(aq) concentrations on the performance of intensively reared gilthead seabream (Sparus aurata) in brackish water: flow-through experiments and full-scale RAS results. Aquacult. Eng. 56, 18–25. Bernier, N.J., Gorissen, M., Flik, G., 2012. Differential effects of chronic hypoxia and feed restriction on the expression of leptin and its receptor, food intake regulation and the endocrine stress response in common carp. J. Exp. Biol. 215, 2273–2282. Bjerkås, E., Holst, J.C., Bjerkås, I., 2004. Cataract in farmed and wild Atlantic salmon (Salmo salar L.). Anim. Eye Res. 23, 3–13. Blanchard, J., Grosell, M., 2006. Copper toxicity across salinities from freshwater to seawater in the euryhaline fish Fundulus heteroclitus: is copper an ionoregulatory toxicant in high salinities? Aquat. Toxicol. 80, 131–139. Bostock, J., McAndrew, B., Richards, R., Jauncey, K., Telfer, T., Lorenzen, K., Little, D., Ross, L., Handisyde, N., Gatward, I., et al., 2010. Aquaculture: global status and trends. Philos. Trans. R. Soc., B 365, 2897–2912. Brauner, C.J., Seidelin, M., Madsen, S.S., Jensen, F.B., 2000. Effects of freshwater hyperoxia and hypercapnia and their influences on subsequent seawater transfer in Atlantic salmon (Salmo salar) smolts. Can. J. Fish. Aquat. Sci. 57, 2054–2064. Brauner, C.J., Shartau, R.B., Damsgaard, C., Esbaugh, A.J., Wilson, R.W., Grosell, M., 2019. Acid-base physiology and CO2 homeostasis: regulation and compensation in response to elevated environmental CO2. In: Grosell, M., Munday, P.L., Farrell, A.P., Brauner, C.J. (Eds.), Carbon Dioxide. Fish Physiology. vol. 37. Academic Press, San Diego. Breck, O., Sveier, H., 2001. Growth and cataract development in two groups of Atlantic salmon (Salmo salar L) post smolts transferred to sea with a four-week interval. Bull. Eur. Assoc. Fish Pathol. 21, 91–103. Buentello, J.A., Gatlin, D.M., Neill, W.H., 2000. Effects of water temperature and dissolved oxygen on daily feed consumption, feed utilization and growth of channel catfish (Ictalurus punctatus). Aquaculture 182, 339–352. Bury, N.R., 2002. Nutritive metal uptake in teleost fish. J. Exp. Biol. 206, 11–23. Camargo, M.M.P., Fernandes, M.N., Martinez, C.B.R., 2009. How aluminium exposure promotes osmoregulatory disturbances in the neotropical freshwater fish Prochilus lineatus. Aquat. Toxicol. 94, 40–46. Cameron, J.N., 1985. The bone compartment in a teleost fish, Ictalurus punctatus: size, composition and acid-base response to hypercapnia. J. Exp. Biol. 117, 307–318. Cameron, J.N., Iwama, G.K., 1987. Compensation of progressive hypercapnia in channel catfish and blue crabs. J. Exp. Biol. 133, 183–197. Carvalho, C.S., Fernandes, M.N., 2006. Effect of temperature on copper toxicity and hematological responses in the neotropical fish Prochilodus scrofa at low and high pH. Aquaculture 251, 109–117. Chabot, D., Dutil, J.D., 1999. Reduced growth of Atlantic cod in non-lethal hypoxic conditions. J. Fish Biol. 55, 472–491. Colt, J., 1991. Aquacultural production systems. J. Anim. Sci. 69, 4183–4192. Colt, J., Watten, B., 1988. Applications of pure oxygen in fish culture. Aquacult. Eng. 7, 397–441. Colt, J., Watten, B., Pfeiffer, T., 2012a. Carbon dioxide stripping in aquaculture. Part 1: terminology and reporting. Aquacult. Eng. 47, 27–37. Colt, J., Watten, B., Pfeiffer, T., 2012b. Carbon dioxide stripping in aquaculture—Part II: development of gas transfer models. Aquacult. Eng. 47, 38–46. Crocker, C.E., Cech, J.J., 1996. The effects of hypercapnia on the growth of juvenile white sturgeon, Acipenser transmontanus. Aquaculture 147, 293–299.

ARTICLE IN PRESS CO2 IN AQUACULTURE

31

Crocker, C.E., Cech, J.J., 2002. The effects of dissolved gases on oxygen consumption rate and ventilation frequency in white sturgeon, Acipenser transmontanus. J. Appl. Ichthyol. 18, 338–340. Cruz-Neto, A.P., Steffensen, J.F., 1997. The effects of acute hypoxia and hypercapnia on oxygen. Fish Biol. 50, 759–769. Dalsgaard, J., Pedersen, P.B., 2011. Solid and suspended/dissolved waste (N, P, O) from rainbow trout (Oncorynchus mykiss). Aquaculture 313, 92–99. Dalsgaard, J., Lund, I., Thorarinsdottir, R., Drengstig, A., Arvonen, K., Pedersen, P.B., 2013. Farming different species in RAS in Nordic countries: current status and future perspectives. Aquacult. Eng. 53, 2–13. Danley, L., Kenney, P.B., Hankins, A., 2005. Effects of carbon dioxide exposure on intensively cultured rainbow trout Oncorhynchus mykiss: physiological responses and fillet attributes. J. World Aquacult. Soc. 36, 249–261. Davidson, J., Good, C., Welsh, C., Brazil, B., Summerfelt, S., 2009. Heavy metal and waste metabolite accumulation and their potential effect on rainbow trout performance in a replicated water reuse system operated at low or high system flushing rates. Aquacult. Eng. 41, 136–145. Eliason, E.J., Higgs, D.A., Farrell, A.P., 2008. Postprandial gastrointestinal blood flow, oxygen consumption and heart rate in rainbow trout (Oncorhynchus mykiss). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 149, 380–388. Ellis, R.P., Urbina, M.A., Wilson, R.W., 2017. Lessons from two high CO2worlds—future oceans and intensive aquaculture. Glob. Chang. Biol. 23, 2141–2148. Elsadin, S., Nixon, O., Mozes, N., Allon, G., Gaon, A., Kiflawi, M., Tandler, A., Koven, W., 2018. The effect of dissolved carbon dioxide (CO2) on white grouper (Epinephelus aeneus) performance, swimbladder inflation and skeletal deformities. Aquaculture 486, 81–89. Exley, C., Chappell, J.S., Birchall, J.D., 1991. A mechanism for acute aluminium toxicity in fish. J. Theor. Biol. 151, 417–428. Fairman, B., Sanz-Medel, A., 1995. Determination of aluminium species in natural waters. Tech. Instrum. Anal. Chem. 17, 215–233. FAO, 2018. The State of World Fisheries and Aquaculture 2018—Meeting the Sustainable Development Goals. (Rome). Farrell, A.P., Thorarensen, H., Axelsson, M., Crocker, C.E., Gamperl, A.K., Cech, J.J., 2002. Gut blood flow in fish during exercise and severe hypercapnia. Comp. Biochem. Physiol. A. Mol. Integr. Physiol. 128, 549–561. Fivelstad, S., 2013. Long-term carbon dioxide experiments with salmonids. Aquacult. Eng. 53, 40–48. Fivelstad, S., Haavik, H., Løvik, G., Olsen, A.B., 1998. Sublethal effects and safe levels of carbon dioxide in seawater for Atlantic salmon postsmolts (Salmo salar L.): ion regulation and growth. Aquaculture 160, 305–316. Fivelstad, S., Olsen, A.B., Kløften, H., Ski, H., Stefansson, S., 1999. Effects of carbon dioxide on Atlantic salmon (Salmo salar L.) smolts at constant pH in bicarbonate rich freshwater. Aquaculture 178, 171–187. Fivelstad, S., Waagbø, R., Zeitz, S.F., Diesen Hosfeld, A.C., Olsen, A.B., Stefansson, S., 2003a. A major water quality problem in smolt farms: combined effects of carbon dioxide, reduced pH and aluminium on Atlantic salmon (Salmo salar L.) smolts: physiology and growth. Aquaculture 215, 339–357. Fivelstad, S., Berit, A., Torbjørn, A., Baeverfjord, G., Rasmussen, T., Vindheim, T., Stefansson, S., 2003b. Long-term sublethal effects of carbon dioxide on Atlantic salmon smolts (Salmo salar L.): ion regulation, haematology, element composition, nephrocalcinosis and growth parameters Sveinung. Aquaculture 215, 301–319.

ARTICLE IN PRESS 32

PETER VILHELM SKOV

Fivelstad, S., Waagbø, R., Stefansson, S., Olsen, A.B., 2007. Impacts of elevated water carbon dioxide partial pressure at two temperatures on Atlantic salmon (Salmo salar L.) parr growth and haematology. Aquaculture 269, 241–249. Fivelstad, S., Kvamme, K., Handeland, S., Fivelstad, M., Olsen, A.B., Hosfeld, C.D., 2015. Growth and physiological models for Atlantic salmon (Salmo salar L.) parr exposed to elevated carbon dioxide concentrations at high temperature. Aquaculture 436, 90–94. Fivelstad, S., Hosfeld, C.D., Medhus, R.A., Olsen, A.B., Kvamme, K., 2018. Growth and nephrocalcinosis for Atlantic salmon (Salmo salar L.) post-smolt exposed to elevated carbon dioxide partial pressures. Aquaculture 482, 83–89. Flik, G., Verbost, P.M., Bonga, S.E.W., 1995. Calcium transport processes in fishes. Fish Physiol. 14, 317–342. FOR, Regulations relating to Operation of Aquaculture Establishments (Aquaculture Operation Regulations). FOR 2004-12-22 no. 1785 § 19. Norwegian Ministry of Fisheries and Coastal Affairs. Forsberg, O.I., 1994. Modelling oxygen consumption rates of post-smolt Atlantic salmon in commercial-scale, land-based farms. Aquac. Int. 2, 180–196. Foss, A., Røsnes, B.A., Øiestad, V., 2003. Graded environmental hypercapnia in juvenile spotted wolffish (Anarhichas minor Olafsen): effects on growth, food conversion efficiency and nephrocalcinosis. Aquaculture 220, 607–617. Foss, A., Kristensen, T., Åtland, Å., Hustveit, H., Hovland, H., Øfsti, A., Imsland, A.K., 2006. Effects of water reuse and stocking density on water quality, blood physiology and growth rate of juvenile cod (Gadus morhua). Aquaculture 256, 255–263. Gil Martens, L., Witten, P.E., Fivelstad, S., Huysseune, A., Sævareid, B., Vikeså, V., Obach, A., 2006. Impact of high water carbon dioxide levels on Atlantic salmon smolts (Salmo salar L.): effects on fish performance, vertebrae composition and structure. Aquaculture 261, 80–88. Gilmour, K.M., Perry, S.F., 1994. The effects of hypoxia, hyperoxia or hypercapnia on the acidbase disequilibrium in the arterial blood of rainbow trout. J. Exp. Biol. 192, 269–284. Good, C., Davidson, J., Welsh, C., Snekvik, K., Summerfelt, S., 2010. The effects of carbon dioxide on performance and histopathology of rainbow trout Oncorhynchus mykiss in water recirculation aquaculture systems. Aquacult. Eng. 42, 51–56. Good, C., Davidson, J., Terjesen, B.F., Takle, H., Kolarevic, J., Bæverfjord, G., Summerfelt, S., 2018. The effects of long-term 20 mg/L carbon dioxide exposure on the health and performance of Atlantic salmon Salmo salar post-smolts in water recirculation aquaculture systems. Aquacult. Eng. 81, 1–9. Goss, G.G., Laurent, P., Perry, S.F., 1992a. Evidence for a morphological component in acid-base regulation during environmental hypercapnia in the brown bullhead (Ictalurus nebulosus). Cell Tissue Res. 268, 539–552. Goss, G.G., Laurent, P., Perry, S.F., 1992b. Gill morphology during hypercapnia in brown bullhead (Ictalarus nebulosus): role of chloride cells and pavement cells in acid-base regulation. J. Fish Biol. 45, 705–718. Harrison, J.G., Richards, R.H., 1979. The pathology and histopathology of nephrocalcinosis in rainbow trout Salmo gairdneri Richardson in fresh water. J. Fish Dis. 2, 1–12. Heuer, R.M., Grosell, M., 2014. Physiological impacts of elevated carbon dioxide and ocean acidification on fish. Am. J. Physiol. Regul. Integr. Comp. Physiol. 307, R1061–R1084. Heuer, R.M., Grosell, M., 2016. Elevated CO2 increases energetic cost and ion movement in the marine fish intestine. Sci. Rep. 6, 1–8. Heuer, R.M., Hamilton, T.J., Nilsson, G.E., 2019. The physiology of behavioral impacts of high CO2. In: Grosell, M., Munday, P.L., Farrell, A.P., Brauner, C.J. (Eds.), Carbon Dioxide. Fish Physiology. vol. 37. Academic Press, San Diego.

ARTICLE IN PRESS CO2 IN AQUACULTURE

33

Hosfeld, C.D., Engevik, A., Mollan, T., Lunde, T.M., Waagbø, R., Olsen, A.B., Breck, O., Stefansson, S., Fivelstad, S., 2008. Long-term separate and combined effects of environmental hypercapnia and hyperoxia in Atlantic salmon (Salmo salar L.) smolts. Aquaculture 280, 146–153. Jensen, F.B., Weber, R.E., 1987. Internal hypoxia-hypercapnia in Tench exposed to Aluminium in acid water: effects on blood gas transport, acid-base status and electrolyte composition in arterial blood. J. Exp. Biol. 127, 427–442. Jensen, F.B., Nikinmaa, M., Weber, R.E., 1993. Environmental perturbations of oxygen transport in teleost fishes: causes, consequences and compensations. In: Rankin, J.C., Jensen, F.B. (Eds.), Fish Ecophysiology. Springer Netherlands, Dordrecht, pp. 161–179. Khan, J.R., Johansen, D., Skov, P.V., 2018. The effects of acute and long-term exposure to CO2 on the respiratory physiology and production performance of Atlantic salmon (Salmo salar) in freshwater. Aquaculture 491, 20–27. Larsen, B.K., Jensen, F.B., 1997. Influence of ionic composition on acid-base regulation in rainbow trout (Oncorhynchus mykiss) exposed to environmental hypercapnia. Fish Physiol. Biochem. 16, 157–170. Larsen, B.K., P€ ortner, H.O., Jensen, F.B., 1997. Extra- and intracellular acid-base balance and ionic regulation in cod (Gadus morhua) during combined and isolated exposures to hypercapnia and copper. Mar. Biol. 128, 337–346. Lauren, D.J., McDonald, D.G., 1987. Acclimation to copper by rainbow trout, Salmo gairdneri: biochemistry. Can. J. Fish. Aquat. Sci. 44 (1), 105–111. Lefevre, S., 2019. Effects of high CO2 on oxygen consumption rates, aerobic scope and swimming performance. In: Grosell, M., Munday, P.L., Farrell, A.P., Brauner, C.J. (Eds.), Carbon Dioxide. Fish Physiology. vol. 37. Academic Press, San Diego. Lefevre, S., Huong, D.T.T., Ha, N.T.K., Wang, T., Phuong, N.T., Bayley, M., 2011. A telemetry study of swimming depth and oxygen level in a Pangasius pond in the Mekong Delta. Aquaculture 315, 410–413. Loyless, J.C., Malone, R.F., 2004. A sodium bicarbonate dosing methodology for pH management in freshwater-recirculating aquaculture systems. Prog. Fish Cult. 59, 198–205. Lund, I., H€ oglund, E., Ebbesson, L.O.E., Skov, P.V., 2014. Dietary LC-PUFA deficiency early in ontogeny induces behavioural changes in pike perch (Sander lucioperca) larvae and fry. Aquaculture 432, 453–461. Martins, C.I.M., Eding, E.H., Verdegem, M.C.J., Heinsbroek, L.T.N., Schneider, O., Blancheton, J.P., d’Orbcastel, E.R., Verreth, J.A.J., 2010. New developments in recirculating aquaculture systems in Europe: a perspective on environmental sustainability. Aquacult. Eng. 43, 83–93. McCormick, S.D., 2012. Smolt Physiology and Endocrinology. Elsevier Inc. Methling, C., Pedersen, P.B., Steffensen, J.F., Skov, P.V., 2013. Hypercapnia adversely affects postprandial metabolism in the European eel (Anguilla anguilla). Aquaculture 416–417, 166–172. Moran, D., 2010a. Carbon dioxide degassing in fresh and saline water. II: degassing performance of an air-lift. Aquacult. Eng. 43, 120–127. Moran, D., 2010b. Carbon dioxide degassing in fresh and saline water. I: degassing performance of a cascade column. Aquacult. Eng. 43, 29–36. Moran, D., Støttrup, J.G., 2011. The effect of carbon dioxide on growth of juvenile Atlantic cod Gadus morhua L. Aquat. Toxicol. 102, 24–30. Moran, D., Tubbs, L., Støttrup, J.G., 2012. Chronic CO2 exposure markedly increases the incidence of cataracts in juvenile Atlantic cod Gadus morhua L. Aquaculture 364–365, 212–216.

ARTICLE IN PRESS 34

PETER VILHELM SKOV

Mota, V.C., Hop, J., Sampaio, L.A., Heinsbroek, L.T.N., Verdegem, M.C.J., Eding, E.H., Verreth, J.A.J., 2018. The effect of low pH on physiology, stress status and growth performance of turbot (Psetta maxima L.) cultured in recirculating aquaculture systems. Aquacult. Res. 49, 3456–3467. Mota, V.C., Nilsen, T.O., Gerwins, J., Gallo, M., Ytteborg, E., Baeverfjord, G., Kolarevic, J., Summerfelt, S.T., Terjesen, B.F., 2019. The effects of carbon dioxide on growth performance, welfare, and health of Atlantic salmon post-smolt (Salmo salar) in recirculating aquaculture systems. Aquaculture 498, 578–586. Munday, P.L., Jarrold, M.D., Nagelkerken, I., 2019. Ecological effects of elevated CO2 on marine and freshwater fishes: from individual to community effects. In: Grosell, M., Munday, P.L., Farrell, A.P., Brauner, C.J. (Eds.), Carbon Dioxide. Fish Physiology. vol. 37. Academic Press, San Diego. Neves, K.J., Brown, N.P., 2015. Effects of dissolved carbon dioxide on cataract formation and progression in juvenile Atlantic cod, Gadus Morhua L. J. World Aquacult. Soc. 46, 33–44. Noble, C., Kankainen, M., Set€al€a, J., Berrill, I.K., Ruohonen, K., Damsgård, B., Toften, H., 2012. The bio-economic costs and benefits of improving productivity and fish welfare in aquaculture: utilizing CO2 stripping technology in Norwegian Atlantic Salmon smolt production. Aquac. Econ. Manag. 16, 414–428. Olsen, R.E., Ringø, E., 1999. Dominance hierarchy formation in Arctic charr Salvelinus alpinus(L.): nutrient digestibility of subordinate and dominant fish. Aquacult. Res. 30, 667–671. Pedersen, C.L., 1987. Energy budgets for juvenile rainbow trout at various oxygen concentrations. Aquaculture 62, 289–298. Permentier, K., Vercammen, S., Soetaert, S., Schellemans, C., 2017. Carbon dioxide poisoning: a literature review of an often forgotten cause of intoxication in the emergency department. Int. J. Emerg. Med. 10, 17–20. Perry, S.F., Gilmour, K.M., 1996. Consequences of catecholamine release on ventilation and blood oxygen transport during hypoxia and hypercapnia in an elasmobranch Squalus acanthias and a teleost Oncorhynchus mykiss. J. Exp. Biol. 199, 2105–2118. Petochi, T., Di Marco, P., Priori, A., Finoia, M.G., Mercatali, I., Marino, G., 2011. Coping strategy and stress response of European sea bass Dicentrarchus labrax to acute and chronic environmental hypercapnia under hyperoxic conditions. Aquaculture 315, 312–320. Pichavant, K., Person-Le-Ruyet, J., Le Bayon, N., Severe, A., Le Roux, A., Boeuf, G., 2001. Comparative effects of long-term hypoxia on growth, feeding and oxygen consumption in juvenile turbot and European sea bass. J. Fish Biol. 59, 875–883. Seidelin, M., Brauner, C.J., Jensen, F.B., Madsen, S.S., 2001. Vacuolar-type H + -ATPase and Na+, K+-ATPase expression in gills of Atlantic Salmon (Salmo salar) during isolated and combined exposure to hyperoxia and hypercapnia in fresh water. Zoolog. Sci. 18, 1199–1205. Skov, P.V., Lund, I., Pargana, A.M., 2015. No evidence for a bioenergetic advantage from forced swimming in rainbow trout under a restrictive feeding regime. Front. Physiol. 6, 1–9. Skov, P.V., Duodu, C.P., Adjei-Boateng, D., 2017. The influence of ration size on energetics and nitrogen retention in tilapia (Oreochromis niloticus). Aquaculture 473, 121–127. Smart, G.R., Knox, D., Harrison, J.G., Ralph, J.A., Richard, R.H., Cowey, C.B., 1979. Nephrocalcinosis in rainbow trout Salmo gairdneri Richardson; the effect of exposure to elevated CO2 concentrations. J. Fish Dis. 2, 279–289. Steinberg, K., Zimmermann, J., Stiller, K.T., Meyer, S., Schulz, C., 2017. The effect of carbon dioxide on growth and energy metabolism in pikeperch (Sander lucioperca). Aquaculture 481, 162–168.

ARTICLE IN PRESS CO2 IN AQUACULTURE

35

Steinberg, K., Zimmermann, J., Meyer, S., Schulz, C., 2018. Start-up of recirculating aquaculture systems: how do water exchange rates influence pikeperch (Sander lucioperca) and water composition? Aquacult. Eng. 83, 151–159. Stiller, K.T., Vanselow, K.H., Moran, D., Bojens, G., Voigt, W., Meyer, S., Schulz, C., 2015. The effect of carbon dioxide on growth and metabolism in juvenile turbot Scophthalmus maximus L. Aquaculture 444, 143–150. Summerfelt, S.T., Vinci, B.J., Piedrahita, R.H., 2000. Oxygenation and carbon dioxide control in water reuse systems. Aquacult. Eng. 22, 87–108. Sunda, W.G., Cai, W.J., 2012. Eutrophication induced CO2-acidification of subsurface coastal waters: interactive effects of temperature, salinity, and atmospheric PCO2. Environ. Sci. Technol. 46, 10651–10659. Thomas, S., Barthelemy, L., Peyraud, C., 1983. Comparison of the effects of exogenous and endogenous hypercapnia on ventilation and acid-base balance in fish. J. Comp. Physiol. B 151, 51–52. Tirsgaard, B., Moran, D., Steffensen, J.F., 2015. Prolonged SDA and reduced digestive efficiency under elevated CO2 may explain reduced growth in Atlantic cod (Gadus morhua). Aquat. Toxicol. 158, 171–180. van Bussel, C.G.J., Schroeder, J.P., Mahlmann, L., Schulz, C., 2014. Aquatic accumulation of dietary metals (Fe, Zn, Cu, Co, Mn) in recirculating aquaculture systems (RAS) changes body composition but not performance and health of juvenile turbot (Psetta maxima). Aquacult. Eng. 61, 35–42. Volkoff, H., Unniappan, S., Kelly, S.P., 2009. Chapter 9 The endocrine regulation of food intake. Fish Physiol. 28, 421–465. Waagbø, R., Hosfeld, C.D., Fivelstad, S., Olsvik, P.A., Breck, O., 2008. The impact of different water gas levels on cataract formation, muscle and lens free amino acids, and lens antioxidant enzymes and heat shock protein mRNA abundance in smolting Atlantic salmon, Salmo salar L. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 149, 396–404. Wang, T., Knudsen, P.K., Brauner, C.J., Busk, M., Vijayan, M.M., Jensen, F.B., 1998. Copper exposure impairs intra- and extracellular acid-base regulation during hypercapnia in the fresh water rainbow trout [Oncorhynchus mykiss]. J. Comp. Physiol. B Biochem. Syst. Environ. Physiol. 168, 591–599. Wilson, R.W., 2011. Aluminum. In: Wood, C.M., Farrell, A.P., Brauner, C.J. (Eds.), Fish Physiology. Elsevier, pp. 67–123. Wood, C.M., 1991. Branchial ion and acid-base transfer in freshwater teleost fish: environmental hyperoxia as a probe. Physiol. Zool. 64, 68–102. Wood, C.M., 2001. Influence of feeding, exercise, and temperature on nitrogen metabolism and excretion. In: Wright, P.A., Anderson, P.M. (Eds.), Fish Physiology. Academic Press, pp. 201–238. Woolley, L.D., Qin, J.G., 2010. Swimbladder inflation and its implication to the culture of marine finfish larvae. Rev. Aquac. 2, 181–190. Wurts, W.A., 2000. Sustainable aquaculture in the twenty-first century. Rev. Fish. Sci. 8, 141–150.