Settling velocity and total ammonia nitrogen leaching from commercial feed and faecal pellets of gilthead seabream (Sparus aurata L. 1758) and seabass (Dicentrarchus labrax L. 1758)

The physico-chemical characteristics of particulate wastes of Sparus aurata and Dicentrarchus labrax were investigated. Changes in the dimensions, settling velocity and total ammonia nitrogen (TAN) leached from commercial feed pellets were investigated after soaking. Also, the settling velocity and TAN leached from faecal pellets of these fish were assessed at 15 and 25 °C. The settling velocity of feed pellets was influenced positively by pellet weight and negatively by immersion length as a result of changes in pellet dimensions after soaking. The settling velocity of faecal pellets was determined by pellet weight. The experimental design did not allow identification of any consistent effect of water temperature on settling velocity. Total ammonia nitrogen leaching over time from feed and faecal pellets was successfully explained by means of a first-order kinetic equation. For feed pellets, water temperature significantly affected the speed of the process and the time at which the maximum TAN leached was reached, but did not influence the maximum TAN leached. Leaching was related to feed pellet size, and so the smaller the pellet, the higher the leaching. Total ammonia nitrogen leaching from faecal pellets was greater per unit weight than in feed pellets. However, neither water temperature nor fish species influenced TAN leaching from faeces.     

Physical Characteristics of Commercial Pelleted Atlantic Salmon Feeds and Consideration of Implications for Modeling of Waste Dispersion through Sedimentation

Models used to employ waste dispersion from sea cages rarely take into account the physical characteristics of fish feed and faeces. In this study we determine the hardness, friability, water absorption properties and settling velocity of a range of commercial Atlantic salmon (Salmo salar L.) feeds, under defined laboratory conditions.Hardness of large, high energy pellets ( 8.5 mm) was generally greater than that of 6 mm high energy pellets. With the exception of 100 mm pellets, the friability of large pellets was also greater than that of the smaller pellets. Settling velocities were significantly greater at 20 psu than at 33 psu and significantly higher for most pellet types at 10 °C than at 20 °C. Settling velocities for unsoaked pellets were found to increase with pellet size, from a mean of 5.6 cm s^–1 for the smallest pellet (2 mm) to 13.9 cm s^–1 for the 10 mm standard (20 to 24% fat) pellets. The smallest feed pellets (2 mm) had the greatest percentage wet weight increase on immersion, irrespective of salinity. Settling velocities of pellets were not significantly affected by immersion time (0–15 min).Given the water depths at fish cage sites and the settling times involved, it is concluded that it is unnecessary to take account of changes in food pellet settling velocity as a result of immersion. Effects of pellet size and salinity on settling velocity are best incorporated into waste dispersion models through the use of 'look-up' tables.     

Determining spatial feed distribution in sea cage aquaculture using an aerial camera platform

Within this paper we describe an unmanned aerial vehicle (UAV) based method to estimate the spatial feed pellet distribution in salmon fish-cages and exploit it within a case study to determine the radial pellet distribution for different rotor spreaders and blower configurations. Compared to previously used methods, capturing pellets thrown from a rotary feed spreader in rows of Styrofoam boxes, the UAV based method is simpler and faster to setup and allows to cover a larger portion of the sea cage surface area. We compare results obtained with the Styrofoam box method with results we obtained by an automatic analysis of aerial videos taken by the UAV during feeding experiments. The employed method helps to gain insight into the spatial feed pellet distribution in full-scale salmon fish-cages where the feeding pipe and actual fixation of the spreader may influence the dynamic behaviour of the feeding system. This can also be seen as a step towards a continuous measuring of the feed pellet distribution, particularly in view of possible future feeding systems that may allow to adapt the feed pellet distribution to the actual fish distribution.     

Determination of physical behaviour of feed pellets in Mediterranean water

Settled uneaten feed causes the most intense impact under sea cages, and settling velocity of the feed pellets represents a key parameter for waste dispersion models. Even if some data about physical properties of feed pellets have been published in the framework of salmonid rearing, there is a complete lack of information related to the Mediterranean Sea, as regards typical values of temperature, salinity and feed composition for Gilthead Sea Bream (Sparus aurata L.) and Sea Bass (Dicentrarchus labrax L.). In this study we try to fill this lack, determining dimensions, water adsorption properties, floating times and settling velocities of a typical growing sequence of pellets for the species mentioned above, under defined laboratory conditions reproducing Mediterranean Sea water. The settling velocity increases with pellet size from 0.087, for the smallest pellet (3 mm), to 0.144 m s^?1, for the 5 mm pellet. The biggest extruded pellet (6 mm) falls slower (0.088 m s^?1). The floating time before pellet's fall is found to be a critical parameter in determining settling velocity. The latter depends on pellet's size, water temperature and salinity. The examined pellets reach a 42% of weight increase after 10 min of immersion, while no appreciable dimension change is observed. Our results are in part different from previous ones and could play a role in evaluating and modelling Mediterranean aquaculture environmental impact.     

Modelling and simulation of rotary feed spreaders with application to sea cage aquaculture – A study of common and alternative designs

Rotary feed spreaders are extensively used in large scale fish aquaculture and are tasked with distributing pelletized feed in a spatially uniform manner over the water surface. Questions have been raised with respect to the performance of classical spreader designs regarding the size and uniformity of the covered area as well as their inability to adapt to changing environmental conditions. This study presents a robotic model of rotary spreaders with experimental validations. Classic rotary spreaders are simulated as well as two alternative designs in the form of a spreader releasing pellets at an optimal initial ballistic angle and a motorized version to increase throw length and provide a more spatially homogeneous surface feed distribution. The alternative designs both yielded improved surface coverage without the need of higher conveying airspeed which may lead to increased pellet attrition. In addition, the motorized design may be used to position pellets at a given location within the sea cage, such as into the wind and current so that pellets can reside inside the sea cage for a prolonged period of time. The presented model may be of interest to researchers and equipment manufacturers who desire to explore performance of a given spreader design.     

Detection and target strength measurements of uneaten feed pellets with a single beam echosounder

The possibility of detecting uneaten food pellets in aquaculture floating cages using a scientific single beam echosounder is demonstrated. The applied methodology is based in a basic scheme with the ultrasonic beam facing upwards the sea surface from the cage bottom. The target strength of single pellets is measured resulting in a linear relationship between target strength and pellet size. These results are the basis to quantify the uneaten food falling and to develop an automated feeding system based on demand.     

Nocturnal feeding reduces sea bass (Dicentrarchus labrax, L.) pellet-catching ability

European sea bass exhibits a feeding rhythm with seasonal phase inversions: nocturnal feeding is predominant in winter, and diurnal feeding during the rest of the year. This explains the increased growth of sea bass fed at night during the winter months, but to fully assess such feeding practices, it is important to monitor food-catching efficiency in the dark. To evaluate the influence of nocturnal feeding on pellet-capture efficiency under laboratory conditions, we developed an infrared pellet detector linked to on-line recording with a microcomputer. Feed waste was recorded for eight groups of four sea bass (54.6±15.8 g body weight) fed either diurnal or nocturnal meals with long (>60 s) or short (<20 s) feed availability times. With short feed availability times, food-catching efficiency in complete darkness (0 lx) was reduced to 78.6%, compared to 93.5% in diurnally fed fish. This was confirmed in a second experiment, performed during winter, involving 4000 sea bass in four sea cages (1000 per cage) provided with self-feeders and pellets detectors. Feed waste amounted to less than 13% of the food distributed, but nocturnal demands (29% of total feed demands) were accompanied by higher levels of uneaten pellets (68% of the total feed waste) than diurnal demand. This suggests that if the aim is to reduce feed waste, nocturnal feeding should be restricted in sea cages.     

Survival and growth of western rock lobster Panulirus cygnus (George) fed formulated diets with and without fresh mussel supplement

Two trials were conducted to compare the growth and survival of western rock lobster Panulirus cygnus post-pueruli and year 1 juveniles fed different diets. Trial 1 compared lobsters fed a formulated pelletised feed developed for Panulirus ornatus supplemented with fresh mussels (Mytilus edulis) two days per week, with lobsters fed a daily diet of pellets. Trial 2 compared lobsters fed the pellet-mussel supplement diet with those fed a daily diet of fresh mussels.In the first trial, diet treatment did not significantly affect survival, although it was higher in tanks supplemented with mussels. Growth of post-pueruli and year 1 juveniles fed the pellet diet (0.03 g/day; 0.14 g/day, respectively) was significantly lower than growth of post-pueruli and year 1 juveniles fed pellet-mussel supplement diets (0.07 g/day; 0.2 g/day, respectively). Year 1 lobsters fed the pellet-mussel supplement diet consumed considerably less pellet (0.27 vs 0.40 g pellet dry matter/day (DM/day)) than those fed pellets indicating that they will preferentially favour mussel over pellet diets. Feed conversion ratio (FCR) was significantly better for lobsters fed the pellet-mussel supplement diet as a result of the poor growth of pellet-fed lobsters.In the second trial, survival was significantly higher for lobsters fed mussels only. Growth of post-pueruli fed a mussel diet (0.14 g/day) was significantly higher than those fed pellet-mussel diets (0.11 g/day), but there was no significant difference in growth for year 1 juveniles (0.33 g/day). Feed intake of mussel on a dry matter basis was higher than pellet intake for both size classes, with mussel intake being more than double pellet intake for post-pueruli (0.20 vs 0.49 g DM/day).Poor growth of P. cygnus on a pellet only diet indicates that the current formulation for P. ornatus may be inadequate for P. cygnus. However, as pellet intake declined dramatically for year 1 lobsters offered a mussel supplement, and pellet consumption was significantly lower than mussel consumption, it is more likely that poor growth may be attributed to inadequate palatability and intake. The similarly good survival of lobsters fed pellet and pellet-mussel diets indicates that the pellet diet is an acceptable maintenance diet for P. cygnus, but is not suitable for maximising growth for this species. Further research is needed to improve palatability, intake and potentially nutritional composition of this pellet diet before its suitability for P. cygnus is assured.     

Manufacturing Different Diet Sizes and Its Effect on Pellet Water Stability and Growth of Three Size Classes of Pacific White Shrimp,Litopenaeus vannamei

Abstract

A feed manufacturing method involving the production of large pellets followed by crumbling to desired diet sizes was used to produce six shrimp diets to determine its effect on growth of three size classes of Pacific white shrimp, Litopenaeus vannamei. The six diets containing 45% protein were 0.7 mm crumble, 1.2 mm crumble, 1.7 mm crumble, 2.2 mm crumble, 2.6 mm crumble, and 3.0 mm pellet. These diets were fed to three size classes of shrimp subjected to a four-week growth trial under indoor laboratory conditions. Stocking density was 12 shrimp/aquarium for size class 1 (1.13iö.01 g), 9 shrimp/aquarium for size class 2 (7.31iö.06 g), and 6 shrimp/aquarium for size class 3 (13.12iö. 10 g). Shrimp were fed three times daily and were evaluated for live weight, growth rate, feed conversion ratio (FCR), and survival. Results for shrimp growth appeared to be affected by diet size and pellet water stability. The effects of diet size, however, were more evident on smaller sized shrimp. Weekly growth rates for class 1 and class 2 shrimp indicated significant responses (P < 0.05) to the effects of diet size, while the class 3 shrimp did not show any significant result. The growth rates for both class 1 and class 2 shrimp were lowest at the smallest diet size (0.7 mm crumble) but differed with respect to highest growth rates. The largest diet (3.0 mm pellets) yielded the highest growth rates for class 1 shrimp, while medium sized feed (2.2 mm crumble) produced the best growth for class 2 shrimp. This could be partly attributed to pellet water stability, which had a direct correlation to diet size (r = 0.97). Smaller diet sizes resulting from crumbling larger pellets tend to have lower water stability. The water stability decreased from 82.5% to 76.3% as the 3.0 mm pellets were reduced to 0.7 mm crumbles. Stress cracks are created during crumbling, and surface area to volume ratio are increased as larger particles are reduced to crumbles. These would make smaller diet particles more susceptible to disintegration and nutrient leaching. For class 3 shrimp, the largest pellets produced the highest growth rate. This high growth value, however, was not significantly different (P > 0.05) from growth results produced by the crumbled feed sizes. This observation suggests that diet sizes ranging from 0.7 mm crumble to 3.0 mm pellet could be used for a 13 g shrimp, although preference would be given to larger pellets due to higher growth performance and better pellet water stability. Results for shrimp live weights were similar to results obtained for weekly growth rates. FCR and survival of all shrimp sizes did not show any significant response from the effects of various diet sizes. However, larger shrimp were less efficient in their diet utilization than smaller shrimp.     

Standardized method for the Holmen pellet tester (NHP100) as a tool for evaluating the physical quality of fish feed

The fish feed production industry requires reliable methods measuring the physical quality of feed before it is being released to the customer. The Holmen pellet tester (NHP 100) could be adapted for this purpose, but the special characteristics of fish feed require particular care during testing. The aim of the study was to determine the standardized conditions for using the Holmen tester as a tool measuring the durability of fish feed. Screening tests were performed on commercial fish feed samples to assess the effects of temperature, pressure and cleaning the instrument between runs. The operating pressure was found to be the most important factor that can influence the results. The temperature of the operating air can also alter the results; increased temperatures can cause oil to melt and leak from the pellets. Cleaning the instrument between analyses can also affect the results. This study shows that when these factors are maintained at a standardized level, the Holmen pellet tester is a valuable tool for assessing the physical quality of fish feed giving additional information on the quality of feed pellets compared to other durability tests like DORIS.     

Increased ability to compete for food by growth hormone-transgenic coho salmon Oncorhynchus kisutch (Walbaum)

In salmonids, growth hormone (GH) stimulates growth, appetite and the ability to compete for food. This study tested the hypothesis that increased GH levels in GH-transgenic coho salmon Oncorhynchus kisutch (Walbaum) increase competitive ability through higher feeding motivation. The transgenic strain of salmon used contained a gene construct consisting of the sockeye metallothionein-B promoter fused to the type 1 growth gene coding region. The transgenic animals (mean size = 250 g) were F₁ individuals. In six consecutive feeding trials, the intake of contested food pellets by size-matched pairs consisting of one control (1 year older non-transgenic coho salmon) and one GH-transgenic coho salmon was compared. Pellets were provided sequentially until neither fish took three consecutive pellets; the identity of the fish taking each pellet was noted. Calculated on the three first pellets offered at each feeding trial, the transgenic coho salmon consumed 2.5 times more contested pellets than the controls, supporting the hypothesis that GH transgenesis increases the ability to compete for food. Overall, the transgenic fish consumed 2.9 times more pellets that the non-transgenic controls, indicating a high feeding motivation of the transgenic fish throughout the feeding trials. It appears that GH transgenesis and GH treatments can induce similar changes in the feeding behaviour of salmonids. Depending on how transgenic and wild individuals differ in other fitness-related characters, escaped GH transgenic fish may compete successfully with native fish in the wild.     

Disaggregation rates of extruded salmon feed pellets: influence of physical and biological variables

The culture of salmonids in the marine environment has led to concern regarding eutrophication of the water column and sediment from uneaten feed. This study used measurement of mass loss and change in carbon and nitrogen content of feed to quantify material lost from waste feed as a function of physical and biological variables. A laboratory flume was used to measure changes in cooked–extruded salmon feed pellets as a function of pellet size (6.5 and 12.0 mm in diameter), residence time in seawater (0, 24, 48, 72 and 120 h), and flow speed (0, 9.3 and 22.2 cm s^?1). The influence of sediment on pellet degradation was also determined. Loss of pellet mass increased with water flow velocity. Smaller pellets eroded faster than larger pellets, losing 32% (6.5 mm) and 21% (12.0 mm) of their mass after 120 h in seawater. The presence of sediment during preconditioning increased the percentage mass loss of the small pellets only. The presence of sediment also resulted in a greater percentage loss of carbon, but not nitrogen, relative to non‐sediment samples. These results allow quantitative assessment of the dynamics of feed pellets in the marine environment with respect to prediction of impact from salmon culture.     

Agar‐based pellets as feed for sea urchins (Paracentrotus lividus): rheological behaviour,digestive enzymes and gonad growth

The algal polysaccharide agar has long been used as a food binder due to its structure, rheological behaviour, stability and interactions – properties that help to generate firm, round, disk‐shaped pellets that may be used in recirculating sea urchin‐rearing systems. Three algae‐based diets (Ulva lactuca, Gracilaria gracilis, Cystoseira sp.) containing 3% and 6% agar were tested on the sea urchin Paracentrotus lividus in order to examine the effect of varying percentages of agar on pellet stability in water and sea urchin gonad growth. The kinetics of water absorption and solute leaching of pellets were measured by immersing quadruplicate samples of the pellets in water for 1, 2, 3, 4, 5 and 6 days. Our results show that the pellets had good water stability, were readily consumed by sea urchins and the presence of agar did not hamper sea urchin gonad growth. Animals fed Ulva‐containing pellets reached a more advanced gametogenic stage with respect to animals fed Cystoseira‐ and Gracilaria‐containing pellets. Moreover, the presence of agarase activity in the digestive system indicated that agar may be an energy source. Pellets are relatively low cost and easy to prepare and store. They may represent a useful resource for rearing sea urchins under intensive conditions.     

Modelling of surface and 3D pellet distribution in Atlantic salmon (Salmo salar L.) cages

Feed represents the greatest single cost factor in the production of Atlantic salmon (Salmo salar L.). Focusing on the problem of maximising the available feed for the fish while minimising the feed waste, a mathematical model of the feeding process has been developed. The model covers the feed spreader delivering the feed, water currents, pellet sinking speed and turbulent diffusion, fish appetite as a function of temperature, gut fullness and population structure, and is intended as a tool both for optimising general feeding strategies, and to support the daily handling of the feeding process. The process of horizontal and vertical diffusion of feed pellets in the model has been parametrised and validated through two low level validation experiments. Furthermore, global distribution patterns simulated by the model were verified by comparisons with experimental data from the literature.     

Weaning of greater amberjack (<Emphasis Type="Italic">Seriola dumerilii</Emphasis> Risso 1810) juveniles from moist to dry pellet

The weaning of greater amberjack (Seriola dumerilii Risso 1810) to a dry diet was studied in two consecutive experiments. The experimental fish were hatchery-produced and grown to their initial size on a soft pellet containing 50% trash fish and 50% commercial dry pellet. In the first experiment, three homogenous groups of greater amberjacks (n = 153, mean weight: 374 ± 43 g), initially adapted for a period of 30 days to a diet with a 40% moisture content, were fed on commercial pellets containing three different levels of dietary moisture (7, 20 and 40%, respectively) for 126 days. We found that fish needed a month to adapt to the 7%- and 20%-moisture diets and that moisture content of the diet had a significant effect on fish growth in terms of increasing feed consumption and lowering feed conversion ratio. The group fed on the 20%-moisture diet not only compensated for the initial adaptation period but presented the higher growth rate and best feed conversion ratio. In the second experiment, all groups of experiment 1 were fed the same commercial diet with a 7% moisture content for 72 days. The group that had been previously fed on the 20%-moisture diet showed the best growth performance. The implication of the above results for the weaning of greater amberjack to dry diets is discussed in terms of the digestion process and physiology.     

Performance of Post-juvenile Rainbow Trout Oncorhynchus mykiss Fed Diets Manufactured by Different Processing Techniques

Four different methods of pelleting—extruded (EXT), steam pelleted (STM), using a Universal Pellet Cooker (UPC), and expanded (EXP)—were performed on subsamples of a single batch of mash to determine the effects on feed pellet characteristics, in vivo feed digestibility, and performance of rainbow trout Oncorhynchus mykiss fed the diets for 16 wk. Results indicated that the STM feed was the most (851 g/L) and UPC feed the least (672 g/L) dense, and that UPC feed had the highest % float (8.3%) and, of the pellets that sank, the fastest sink rate (15.1 cm/sec). While some differences occurred among pelleted diets in loss of dry matter (% dry loss) after incubation in a water bath for up to 10 min, the overall % dry loss was only approximately 1–2%. The method of pelleting did not affect fish weight gain, feed conversion ratio, or specific growth rate. Measurement of the apparent digestibility coefficient of dry matter of the test diets using Cr₂O₃ as the marker ranged from 71–89%. While it appeared that the method of pelleting can affect feed characteristics and the amount of feed offered to the fish, fish performance was not affected.     

Optimal feed pellet size for culturing juvenile spiny lobster Jasus edwardsii (Hutton, 1875) in New Zealand

There is intense international interest in the development of spiny lobster aquaculture and of particular interest is the development of artificial diets for spiny lobsters. However, there is little information on the feeding behaviour of spiny lobsters and their ability to handle and consume artificial feeds of different dimensions. Observations of the feeding behaviour of juveniles of the spiny lobster Jasus edwardsii in culture found that they waste significant amounts of food through externally fragmenting the feed pellets before sweeping particles into the mouth. The consumption efficiency of different sized pellets was found to change markedly with increasing size of the juvenile lobsters. Up to 50% of artificial food was wasted by the inefficient feeding behaviour of the lobsters. However, consumption of the feed could be reduced by as much as 19% by altering pellet size. These results indicate that the size of feed pellets, and their consistency and fragmentation characteristics, will be critical aspects of cost‐effective feeding of cultured spiny lobsters.     

Effect of Feeding Frequency, Water Temperature, and Stocking Density on the Growth of Tiger Puffer, Takifugu rubripes

Effects of daily feeding frequency, water temperature, and stocking density on the growth of tiger puffer, Takifugu rubripes, fry were examined to develop effective techniques to produce tiger puffer in a closed recirculation system. Fish of 4, 14, and 180 g in initial body weight were fed commercial pellet diets once to five times a day to apparent satiation each by hand for 8 or 12 wk at 20 C. Daily feeding frequency did not affect the growth of 14‐ and 180‐g‐size fish. However, the daily feed consumption and weight gain of the 4‐g‐size fish fed three and five times daily were significantly higher than those of fish fed once daily (P < 0.05). Fish of 4 and 50 g in initial body weight were reared with the pellet diet at 15–30 C for 8 wk. The weight gain of fish increased with increasing water temperature up to 25 C and decreased drastically at 30 C for both sizes. Similar trends were observed for feed efficiency, although 4‐g fish had highest efficiency at 20 C. Effects of stocking density on growth were examined with fish of 8, 13, and 100 g in initial body weight. Fish were reared with the pellet diet for 8 or 16 wk at 20 C. Fish were placed in floating net cages in the culture tank, and the stocking density was determined based on the total weight of fish and volume of the net cage. Fish of 8 g in body weight grew up to 35–36 g during the 8‐wk rearing period independent of the stocking density of 8, 15, and 31 kg/m³ at the end of rearing. Final biomass per cage reached 32, 60, and 115 kg/m³ for 13‐g‐size fish, and 10, 18, and 35 kg/m³ for 100‐g‐size fish, and the growth of the fish tended to decrease with increasing stocking density for both sizes.     

Effects of Particle Size and Processing Variables on the Properties of an Extruded Catfish Feed

Abstract

A catfish feed was extruded to determine how particle size, moisture content and screw speed affected the pellet durability, water stability index, and buoyancy. Increasing moisture content from 250 to 300 g/kg required less die pressure, specific mechanical energy and product temperature to produce durable pellets that were stable in water. Increasing screw speed from 200 to 400 rpm also lowered die pressure, due to a lower filled length of the extruder screw channels, but increased specific mechanical energy and product temperature; the resulting pellets had a lower apparent bulk density, but a higher degree of starch gelatinization and water stability index. Reducing particle size from 1200 to 700 m decreased die pressure and apparent bulk density, but increased product temperature, degree of starch gelatinization, pellet durability index and water stability index.     

Feeding behaviour, energy expenditure and growth of rainbow trout in relation to stocking density and food accessibility

The effect of stocking density on feed intake, feed utilization and feeding behaviour was studied in groups of rainbow trout submitted to different levels of food accessibility. In the first experiment, 18 groups of 100, 300 or 500 fish (initial individual weight 26 ± 0.5 g) were held in 1 m³ tanks where they were fed: in excess using belt feeders; or by means of self‐feeders with a fixed reward level (4 g per trigger actuation); or a reward level proportional to fish density (2 g/100 fish/trigger actuation). Final biomass after 125 days was c. 25, 70 and 100 kg/m³ in groups of 100, 300 and 500 fish respectively. Growth and feed intake was affected by fish density, feeding technique and reward level. However, feed efficiency, body composition, as well as nutrient retention and loss were unaffected by the treatment. In the self‐fed groups, daily pattern of feeding activity was affected by fish density when the reward level was fixed but not when the reward level was proportional to fish density. In a second experiment, groups of 100 or 500 fish (same initial individual weight) had access to one or three self‐feeders. Access to multiple self‐feeders had no positive effect on any of the studied parameters. The fish used only one of the three feeders at a time, and the three feeders where not used equally. We conclude that feed intake but not feed utilization was impaired by increased density, indicating that density in itself is not stressful to the fish. Food accessibility is the main factor responsible for the observed density – related decrease in growth.