Water quality requirement and holding conditions of Arctic charr (Salvelinus alpinus L.) under intensive fish farming conditions.

Water quality requirement and holding conditions of Arctic charr (Salvelinus alpinus L.) under intensive fish farming conditions. Bjørn-Steinar Sæther og Sten Ivar Siikavuopio Nofima Marin Arctic charr (Salvelinus alpinus L.) have the most northern distribution of all freshwater fish and may be the most cold-adapted species within the salmonid family (Johnson, 1980). Due to its northern distribution Arctic charr experience considerable seasonal environmental changes throughout the year, but are well adapted to utilize the associated rapid seasonal changes in resource availability (Johnson, 1980). Arctic charr has a high degree of biological plasticity and life history patterns can vary considerably also within groups of siblings. Growth of Arctic charr populations are known to be highly variable under natural conditions (Johnson, 1980; Hammer, 1984; Klemetsen et al., 2003), typically varying with seasonal changes in availability of food and environmental conditions. Temperature How temperature affects fish depends to a large degree on what temperature the fish is adapted to. In its most extreme ranges, lethargic temperatures, it is independent of adaptation temperature, but for levels within the survival range, the tolerance temperatures are often referred to as the temperature tolerance polygon (e.g. Jobling, 1994; figure 1). The polygone is a model of understanding temperature tolerance and how tolerance is related to acclimation temperature;the recent temperature history of the fish. The ultimate upper and ultimate lower incipient lethal temperatures are independent of acclimation and as such reflect the limits at which survival does not go beyond. For Arctic charr these limits spans from 0 to 23-24°C (Lyytikainen & Jobling 1998). The area covered by this polygon (quantified in terms of degrees squared ((°C)2) reflects the tolerance zone. Eurythermal fish, which are tolerant of a wide range of temperatures, will have tolerance zones enclosing larger areas (1100 degrees squared), whereas stenothermal fish, like most salmonids are usually 650 degrees squared or less. Temperature preference and tolerance change with life stages. Maturing females usually have the narrowest tolerance limits and this is likely because low water temperatures (4 -7°C) are required during the final stages of reproductive cycle (Jobling et al., 1995) whilst the growth rates of juveniles peaks at 12-14C. Studies have shown relatively high growth rates even during winter periods with low water temperatures (Brännäs et al., 1992; Siikavuopio et al., 2009). 1

Figure 1. Graphical presentation of the thermal relations of fish (Jobling, 1994), and how preference (AP) and tolerance limits depends on acclimation temperature. AP: acute thermal preferendum, CTM: critical thermal maximum (survival time above CTM is virtually zero), LE: line of equality, where the response temperature is equal to the acclimation temperature. LILT: lower incipient lethal temperature, UILT: upper incipient lethal temperature, UUILT: Ultimate upper incipient lethal temperature, the maximum temperature that the fish can be acclimated. The incipient lethal limits mark the boundaries between the zone of tolerance and the zone of resistance, and also define the limits of the temperature tolerance polygon. The tolerance zone is marked with crosses, and the black square is the zone of final preferendum (from Jobling, 1994).

High-density intensive farming imposes practical limits on temperature, as both oxygen consumption and ammonia excretion increase dramatically with increasing temperature (Lyytikäinen & Jobling 1998). Although the optimal feed utilisation of juvenile Arctic charr occurs at approx. 9C (according to Uraiwan 1982), 12°C seems to be the most likely optimal temperature for aquaculture. This supports high growth rates, good feed utilisation and a reduced risk of contracting diseases as well as low fungus growth compared to higher rearing temperatures. If fish is to undergo large changes in temperature it is advised that these changes are made in 1-2°C steps per day over consecutive days to allow the fish to gradually acclimate. The take home message from the polygon-model is that temperature tolerance and preference is relative, and should be treated accordingly. Both the short term history (acclimation temperature)

2

and effects of life stages needs to be considered in temperature management, as what is optimal for growth is not optimal for feed utilisation and is not optimal for maturation etc.

Salinity The highly variable life strategies of Arctic charr affect the salinity tolerance of this species. Anadromous charr tolerate full strength seawater (33-35 ppt.) during the approximately 2 month seaward migration period in mid summer. During the rest of the year charr seem to cope well with brackish water (20 ppt), but this is very dependent on fish size and temperature (Johnston, 2002; Duston et al., 2007).

Oxygen and Carbon dioxide In water, the dissolved gases that are of particular biological and ecological importance are oxygen and carbon dioxide. The solubility of dissolved gases in water depends on temperature, salinity and their individual partial pressure gradients across the surface (Timmons & Ebeling, 2007).

O2 Dissolved oxygen (DO) is often the critical (e.g. limiting) water quality parameter in flow-through and recirculation aquaculture systems. When the level of DO in the water is low, feed intake may be suppressed. This is thought to be due to the reduced oxygen availability being unable to support the energy demands of the animal. A reduction in feed intake at low levels of DO would obviously have consequences for growth, and attempts have been made to determine critical levels of DO at which feed intake and growth become affected (Jobling, 1994). It is difficult to specify critical dissolved concentrations because the response to low DO is not always fatal, but rather, a continuum of physiological effects (Harmon, 2009). These effects are also influenced by the exposure time, the size and health of the animal, water temperature, concentration of carbon dioxide and other environmental conditions (Jobling, 1994). The need for water flow is related to the need for oxygen in fish holding systems, oxygen is consumed at a rate dependent on the metabolic rate of the animals held in the system. The oxygen demand increase as the fish grow, but the metabolic rate relative to bodyweight decrease as body weight increase. Thus, 1 Kg of 10 gram fish needs more oxygen than 1 Kg of 100 gram fish. The oxygen consumption is also dependent on factors such as activity rate, temperature, metabolic state and growth rate. Usually, average consumption per 24 hours in charr is in the region of 100-200 mg O2 Kg-1 h-1. Fish with good growth rates have an average oxygen consumption above 150mg O2 Kg1

h-1. Following feeding the demand may peak at 40-50% above average consumption. 3

The availability of oxygen depends on the partial pressure of the gas. In water saturated under normal atmospheric composition and pressure (760 mg Hg), oxygen holds a pressure of 158 mmHg (approx. 21%). That is, when the water is fully saturated only 21% of the gas is oxygen, most of it is nitrogen (approx. 78%). The partial pressure of each gas is their contribution to the total gas saturation in the water. Oxygen uptake across the gills is passive, and depends on this partial pressure as a driving force. Under conditions with low oxygen content in the water (hypoxia), the fish can compensate by increasing the water flow over the gills, increase their gill surface or reduce the oxygen demand. The compensation eventually has a cost, leading to reduced feed intake and growth. Between 100 and 70% saturation, no negative effects on performance have been reported. Below 70% saturation both feed intake and growth is reduced (Christiansen et al. 1990), and should represent the lowest recommended gas saturation for the farming situation. In the case of fry and alevins, this lower threshold should be set to 80%. Some advocate the use of absolute oxygen content, as in mg or ml per litre of water, whereas other prefer to use a relative measure such as percentage of fully saturated water. The reasoning behind the use of relative measures is based on the reduced solubility of gases in water with temperature increase. Thus, 9.9mg oxygen per litre of water represents 100% saturation at 10C, but only 88% at 5C. It is important to realise that the partial pressure of a particular gas depends upon its fractional concentration in the overall gas mixture and the barometric pressure, but not on the absolute quantity of gas present. The following procedures should be followed under hypoxic conditions;increase the waterflow, stop feeding and reduce the seawater temperature. Under extreme hypoxic conditions, when it is a matter of keeping fish alive, it may help to reduce the water level to such an extent that the fish are kept moisturised, but can get oxygen from the air. Many bony fishes in fact shift to an aquatic surface respiration under severe hypoxia (Jobling 1994). Supersaturation of oxygen (hyperoxia) is usually not a problem in aquaculture. Under handling and transportation, it may be necessary to add oxygen to the water. Fish responds to hyperoxia by reducing the ventilation rate and volume. Increased oxygen content in the water as a result of adding pure oxygen, or using modified atmosphere under aeration of water, increases the partial pressure of oxygen, making more oxygen available per litre of water. This reduces the water needs, which may be favourable and even necessary under some conditions. By removing nitrogen before adding oxygen, the total gas saturation may be held under 100% even when the oxygen content is increased.

4

In general Arctic charr should be kept in water with sufficient oxygen concentration, usually 70 100 % of fully saturated water is advised to avoid reduced appetite and growth (Johnston, 2002).

CO2 Increased CO2 increases the concentration of hydrogen protons (H+) in the water, and thereby reduces pH. This initiates a response in the blood carrying protein (haemoglobin), known as the Bohr-effect, that reduces the affinity for O2 thereby reducing the fish’s ability to take up and transport O2 in the blood. Carbon dioxide occupies the same binding sites on the haemoglobin that carries O2, and the partial pressure of O2 has to be increased to load the same amount of O2 when CO2 is increased. Carbon dioxide is highly soluble in water, but the concentration in pure water is low (0.54 mg/L, at 20 °C) (Timmons et al., 2001). Most of the carbon dioxide in an aquaculture water volume is the product of animal respiration and the decomposition of organic matter, with a small percentage coming from atmospheric diffusion (Timmons et al., 2001). Recirculating water systems with inadequate degassing can also be a source of high CO2 levels (Timmons et al., 2001). The concentration of carbon dioxide in water is determined by a gas-liquid equilibrium and also by series of acid-base reactions (Summerfelt, 2000). In fluids, dissolved CO2 is hydrated to carbonic acid (H2CO3) and further to bicarbonate (HCO3-) and carbonate (CO32-). The proportion of each type of carbonate state (free carbon dioxide, bicarbonates and carbonate) depends on the H+ concentration of the water and, conversely, a change in these values will affect the pH.

In Arctic charr culture, recommended levels of CO2 in water are less than 10 mg /L at alkalinity lower than 100 mg/L and less than 15 mg/L at higher alkalinities (Johnston, 2002).

Ammonia, nitrite and nitrate In intensive closed or semi-closed rearing systems, even with frequent water changes, it is common to observe an accumulation of inorganic nitrogen, such as ammonia, nitrite and nitrate (Timmons et al., 2001). High levels of nitrite and ammonia in the water are potential factors triggering stress and can cause high mortality in aquatic organisms (U.S. EPA, 1989; 1999; Timmons & Ebeling, 2007). Nitrogen occurs in aquatic environments in several forms: dissolved gaseous N2 (from the atmosphere), dissolved gaseous NH3 (most as waste from animal metabolism), ammonium ions (NH4+), nitrite ions (NO2-) and nitrate ions (NO3-) (Timmons et al., 2001). The term “ammonia” is often used to include both the dissolved un-ionized gas (NH3) and the ionised NH4+, and as such is better expressed as total ammonia (TAN). The relative concentration of ammonia (NH4+ and NH3) 5

is a function of water pH, salinity and temperature. The un-ionized ammonia (NH3) is the most toxic form of ammonia because of its ability to move across cell membranes (Timmons & Ebeling, 2007). An increase in temperature, pH or salinity increases the proportion of the un-ionized ammonia (Timmons et al., 2001). Excreted ammonia-N is oxidized by bacteria in a process called nitrification. In this two-step process, ammonia is converted to nitrite (NO2-) and nitrite is converted to nitrate (NO3-) (Timmons et al., 2001). Nitrite can be toxic to salmonids at relatively low concentrations, resulting in reduced oxygen transportation capability in the blood due to the conversion of haemoglobin to methaemoglobin; a molecule which can not bind oxygen (Johnston, 2002).

Safe levels of ammonia, nitrite and nitrate for producing Arctic charr are still unknown. However, based on the experience from RAS farming of Arctic charr and recommendations from Atlantic salmon farming, ammonia levels in fresh water should not exceed 0.015 mg/L and ammonia-N levels should remain below 1.0 mg /L. If Arctic charr are held in sea water, the recommended safe levels will be somewhat higher (Johnston, 2002). Nitrite-N should always remain below 0.2 mg/L (Skybakmoen et al., 2009).

pH The pH balance is essential for fish metabolism (Oxygen uptake, salt-water balance and acid-base regulation). The pH is a measure of free hydrogen protons (H+), given in a logarithmic scale from 1 to 14, where 7 is neutral. Water with pH below 7 (high in H+) is acidic, whereas water above 7 are alkaline. Acceptable pH depends on a variety of water quality factors, e.g. the concentration of humic acids, aluminium, CO2, calcium, etc. Where possible, pH shall be kept as stable as possible because all changes in pH initiate complex water quality changes which may cause harm to the fish, especially their gills Although other salmonids can tolerate pH within the range of 5 to 9, maximum productivity occurs at pH between 6.5 and 8.5 (Jobling 1994). Arctic charr may be less sensitive to changes in water pH than other salmonids.

Water currents Water currents are necessary for self-cleaning in most tank systems, and circular tanks usually require a water flow of 4-6 cm sec-1 (but dependent on tank design). Water currents affect the behaviour of the fish. At low water speeds, fish swim in a less organised manner in the tank and seems to engage in more aggressive interactions. Dominant fish attempt to hold territories, 6

preferably nearby feeding stations. In contrast, at higher water speeds, fish become more organised and start to school in the tank. The fish are then evenly distributed and are less occupied with social interactions, which are reflected in reduced level of fin damage (Jørgensen and Jobling 1993). The water current also helps distributing the feed in the water column. As the fish prefer to hold the same position in the tank when schooling, water current also determine the swimming speed of the fish. The increase in exercise caused by increased swimming speed leads to improved growth and feed utilisation. The threshold speed for schooling is probably somewhere between 0.5 and 1 body length per second. Growth rate increase until a speed of 1.5-1.75 body lengths per second are reached. Water currents in a holding tank therefore serves several purposes; self-cleaning, distribution of feed, reducing aggression and fin damage to the fish (schooling) and providing exercise for the fish.

The water current speed should therefore always be kept high enough to support schooling, usually approximately 1 body length per second, but this is best judged by observing the behaviour of the fish. This sustains self cleaning, distribution of feed and reduced aggression.

Water depth Low water depth may cause swim bladder stress syndrome, as a certain depth seem necessary for proper buoyancy adjustment. It is advised that juvenile charr has a minimum water depth of 15 cm (Kolbeinshavn and Wallace, 1985).

Density Arctic charr tolerate high densities without negative effects on feed intake and growth. In fact, low densities should be avoided as it increases social interactions within groups (Jørgensen et al., 1993). At 15 kg m-3 charr showed poorer growth rate, and larger variability in growth rate when compared to charr held at 60 kg m-3. An increase to 120 kgs m-3 did not lead to reduced growth or negative effects on individual growth variability. In another study, a density of 30 kg m-3 showed higher degree of fin damage, and a higher degree of weight loss, as compared to charr held at 90 or 150 kg m-3. This is likely due to the organisation of fish into schooling groups, as charr tend to shift to schooling when density increases. This leads to a more organised swimming behaviour with less social interactions and aggression. Density will act in concert with water currents to reduce aggression, hence at low swimming speeds impact of rearing density becomes a more important modulator of behaviour. Sixty kilos per 1000 litres of water should be considered minimum density 7

under growing conditions. The upper limit is uncertain, but it seems clear that at least 150 kg m-3 is not a problem (Jobling et al.,1993, Siikvuopio & Jobling, 1995, Johnston, 2002).

Light Arctic charr are affected by photoperiod (day length) which is the most fixed environmental cue that can be used due to the regular changes in day length over the calendar year. Salmonids use changes in day length to time important seasonal events, such as parr-smolt transformation and maturation. Changes like these are complex, involving morphological, behavioural and physiological changes and exact timing of these is of immense importance. If the fish is expected to follow the seasonal changes seen in nature, they need to get accurate information regarding the change of season through appropriate changes in day length. The fish do have the ability to generate an internal calendar, as is it equipped with one or more biological clocks. However, these continuously need adjustment, as they tend to lag behind or move to fast in the absence of environmental cues. This means that if fish are held under constant conditions, they will still go through the seasonal changes, but these changes will drift in relation to natural seasons. This can also be utilised by farmers, as time of smolting or spawning can be shifted in a controlled manner by manipulating the changes in day length. Simulated changes in seasons, does not have to include daily changes in day length, but can be achieved by adjusting the day length the fish are exposed to at weekly intervals. The fish’s response to shifts in photoperiod information depends on where its internal biological clock is set, e.g the fish will respond differently to a change in its perceived winter as compared to summer. Thus, great care should be taken before applying photoperiod manipulations on fish groups, and successful results depends on long time planning.

Light intensity Salmonids are able to conceive light at very low densities (below 1 lux, Zachmann 1992). When comparing growth and mortality in juvenile Arctic charr exposed to light intensities from 0 to 700 lux, Wallace (1988) found highest growth rate at 50 lux, with survival rate only being higher in fish held in complete darkness (0 lux). Charr at 10 and 200 lux had similar growth with the latter having slightly lower survival. Poorest growth and highest mortality were recorded in the group held at 700 lux. In terms of optimising growth and survival it seems that the light intensity should be relatively low, approximately 50 lux. The difference between day and night is more important than the absolute intensity for the fish’s perception of day and night (Zachmann 1992). When using low intensity light, it is therefore important to keep the night as dark as possible. 8

Feeding practice The gastrointestinal system in charr is relatively short, with limited storage capacity in the stomach. It is therefore recommended to feed the fish on a daily basis. Charr are able to feed from the bottom of the tank, and also feed during darkness. However, best growth rates are obtained under long day length, although this is probably more closely connected to season. The time of day at which feeding occurs seems to be of less importance, as feeding at dusk did not result in differences in feed intake and growth when compared to fish fed at dawn (Dalen 1998). Competition for feed is often seen in groups of Arctic charr. If allowed, the most competitive fish (often referred to as social dominants) gain access to the feed at the expense of others. By forming such feeding hierarchies, these individuals grow at a higher rate, eventually leading to large size variability in the group. Size differences escalate, as large fish seems more likely to gain access to feed. Under extreme conditions and even if the feed resource is not limited the dominant fish controls the feed to such an extent that sub-ordinate fish cease feeding. Predictability of where and when feed is delivered increases competition and is likely to result in large size variation. Ensuring that the feed is presented throughout the entire water column during feeding reduces the opportunity of individuals to monopolise the feed, and thereby decrease the chances of dominant fish establishing feeding hierarchies. This reduces the competition and aggression in the fish. Aggression is stressful, and reduces growth potential, immunocompetence, and compromises animal welfare. Control of the formation of aggression based social hierarchies can be achieved by regular measurement of individual fish size (every month during fast growth periods, and every second month in periods with slow growth), and calculating variability. It is advisable to use a size independent measure of variability, such as coefficient of variation, to enable direct comparisons of variability over time as the fish grow (coefficient of variation (CV) = standard deviation divided by average times 100). Several feeding rate tables exist for Arctic charr. These estimate the growth reasonably well over a long period of time (e.g. 1 year) but can lead to extensive over or under feeding during periods within a year due to seasonality on growth. The growth potential changes with external factors, such as temperature, but is also controlled by endogenous factors. The growth appears to be connected to season (Sæther et al. 1996), possibly an adaptation to varying availability of food and large differences in opportunities for growth between seasons. Most feeding and growth models do not consider this, and should therefore be used with great care. The best feeding practice seems to be the ones based on the fish’s appetite. Ultimately the fish should be fed to satiation every day, and the feed should be evenly distributed evenly throughout in water volume during feeding. 9

References: Brännäs, E. and Wiklund, B.-S. (1992) Low temperature growth potential of Arctic charr and rainbow trout. Nordic Journal of Freshwater Research 67, 77-81. Christiansen, J.S., Jobling, M., and Jørgensen, E.H. (1990). Oksygen- og vannbehov hos laksefisk. Nye reviderte tabeller. Norsk Fiskeoppdrett 10: 28-29. (In Norwegian) Dalen, H. (1998). Effekt av morgen- kontra kveldsfôring på vekst, fôrinntak og kroppsammensetning hos røye (Salvelinus alpines). Cand. Scient. Oppgave i Havbruk, NOrges Fiskerihøgskole, Universitetet i Tromsø (in Norwegian) Duston, J., Astatkie, T. and Murray, S.B. (2007) Effect of salinity at constant 10 oC on grow-out of anadromous Arctic charr from Labrador. Aquaculture 273:679-686. Hammer, J. (1984) Ecological characters of different combinations of sympatric populations of Arctic charr in Sweden. In: Biology of the Arctic charr. In: Johnson, L., Burns B.L. (Eds). Proceedings of the international Symposium of Arctic charr, Winnipeq, Manitoba, May 1981. University of Manitoba Press: Winnipeg. 35-63 pp. Harmon, T.S: (2009). Methods for reducing stressors and maintaining water quality associated with live fish transport in tanks: a review of the basics. Reviews in Aquaculture, 1: 58-66. Jobling, M., Jørgensen, E.H., Arnesen, A.M. and Ringø, E. (1993) Feeding, growth, and environmental requirements of Arctic charr: a review of aquaculture potential. Aquaculture International 1, 20-46. Jobling, M.,(1994) Fish Bioenergetics. Chapman & Hall, Fish and Fisheries Series 13, pp. 155-201 Jobling, M., Johnsen, H.K., Pettersen, G.W. and Henderson, R.J. (1995) Effect of temperature on reproductive development in Arctic charr, Salvelinus alpinus (L.). Journal of Thermal Biology 20, 157-165. Johnson, L. (1980) The Arctic charr, Salvelinus alpinus. In Charrs, Salmonid Fishes of the Genus Salvelinus. In: Balon, E.K. (Ed). W. Junk, The Hague, Netherlands. 87 pp. Johnston, G. (2002) Arctic Charr Aquaculture. Fishing News Books, Oxford. Jørgensen, E.H., Christiansen, J.S., and Jobling, M. (1993). Effects of stocking density on food intake, growth performance and oxygen consumption in Arctic charr (Salvelinus alpinus) Aquaculture, 110: 191-204. Klemetsen, A., Amundsen, P-A., Dempson, J.B., Jonsson, B., Jonsson, N., O’Connell, M.F. and Mortensen, E. (2003) Atlantic salmon Salmo salar L., brown trout Salmo trutta L. and Arctic charr Salvelinus alpinus (L.): a review of aspects of their life histories. Ecology of Freshwater Fish 12, 1-59. Kolbeinshavn, A., Wallace, J.,1985. Observations on swim bladder stress syndrome in Arctic charr (Salvelinus alpinus), induced by inadequate water depth. Aquaculture, 46, 259-261. Lyytikäinen, T., Jobling, M., 1998. The effect of temperature fluctuations on oxygen consumption and ammonia excretion in underyearling lake Inari Arctic charr. J. Fish. Biol. 52: 1186-1198 Skybakmoen, S., Siikavuopio, S.I. and Sæther B-S. (2009). Coldwater RAS in an Arctic charr farm in Northern Norway . Aquacultural Engineering, 41: 114-121. Siikavuopio, S.I. and Jobling, M. (1995) The effects of stocking density on survival and growth of wild-caught Arctic charr. Nordic Journal of Freshwater Research 71, 419-423. Siikavuopio, S.I., Skybakmoen, S. and Sæther B-S. (2009). Comparative growth study of wild- and hatchery produced Arctic charr (Salvelinus alpinus L.) in a coldwater recirculation system. Aquacultural Engineering, 41: 122-126. Summerfelt, S. T., Vinci, B. J., and Piedrahita, R. H. (2000) Oxygenation and carbon dioxide control in water reuse systems. Aquacultural Engineering 22, 87-108. Sæther, B-S., Johnsen H, K. and Jobling, M (1996). Seasonal changes in food consumption and growth of Arctic charr exposed to either simulated natural or a 12:12 LD Photoperiod at constant water temperatures. Journal of Fish Biology 48: 1113-1122. Timmons, M. B., Ebeling, J. M., Wheaton, F. W., Summerfelt, S. T. and Vinci, B. J. (2001). Recirculating Aquaculture Systems. NRAC Publication NO.01-002. ISBN 0-9712646-0-0. 650 pp. Timmons, M. B and Ebeling, J. M. (2007) Recirculating Aquaculture. NRAC Publication NO.01-007. ISBN 0-978-09712646-2-5. 975 pp. Uraiwan, S., 1982. Effect of genotype, environment, and genotype-environment interactions on growth of rainbow trout (Salmo gairdneri Richardson). MSc thesis, University of Manitoba, Winnipeg, Canada. U.S. Environmental Protection Agency (1989). Ambient Water Quality Criteria for Ammonia (Saltwater), Office of Research and Development, Environmental Research Laboratory, Narragansett, Rhode Island, EPA 440/5-88-004. 59 pp. U.S. Environmental Protection Agency (1999). Update of ambient water quality criteria for Ammonia. Technical version-1999, USE-PA, Washington DC, USA, EPA, 823-F99024. 147 pp. Wallace, J.C., 1988. Observations on the effect of light inytensity on the growth of Arctic charr fingerlings (Salvelinus alpinus) and salmon fry (Salmo salar). Aquaculture, 72: 81-84. Zachmann, A., Knijff., S.C.M., Ali, M.A. & Anctil, M. (1992). Effects of photoperiod and different intensities of light exposure on melatonin levels in the blood, pineal organ, and retina of the brook trout (Salvelinus fontinalis Mitchill). Can. J. Zool., 70: 25-29.

10