iNtroduCtioN

the welfare of farmed fish is increasingly coming to

the forefront of public concern (fSBi, 2002). Physi-

cal, chemical and other perceived stressors can all

evoke non-specific responses in fish, which are con-

sidered adaptive to enable the fish to cope with the

disturbance and maintain its homeostatic state. if the

stressor is overly severe or long-lasting to the point

that the fish is not able to regain homeostasis, then

the responses themselves may become maladaptive

and threaten the fish health and “well being” (Barton,

2002). in order to develop effective systems for pro-

tecting the welfare of farmed fishes, it is necessary to

know how individuals with different behavioral pro-

files live under different husbandry conditions [i.e. at

different stocking densities: huntingford (2004)].

Journal of Biological Research-Thessaloniki 19: 99 – 110, 2013

J. Biol. Res.-Thessalon. is available online at http://www.jbr.gr

indexed in: WoS (Web of Science, iSi thomson), SCoPuS, CaS (Chemical abstracts Service) and doaJ (directory of open access Journals)

99

Effect of temperature, stocking density, feeding conditions and

experimental day on the horizontal and vertical distribution

of sea bass fry Dicentrarchus labrax (Linnaeus, 1758)

Maria Neofytou1*, Stylianos SoMarakiS2, Vassilis PaPadakiS3,

Pascal diVaNaCh2, aspasia Sterioti4 and Maroudio keNtouri1

1 Biology Department, University of Crete, 71409 Heraklion, Crete, Greece2 Hellenic Centre for Marine Research, Institute of Aquaculture,

P.O. Box 2214, 71003 Heraklion, Crete, Greece3 Foundation for Research and Technology Hellas (FORTH),

Institute of Electronic Structure and Laser (IESL), P.O. Box 1385, 70013 Heraklion Crete, Greece4 Hellenic Centre for Marine Research, Cretaquarium, P.O. Box 2214, 71003 Heraklion, Crete, Greece

received: 29 July 2011 accepted after revision: 24 June 2012

the working hypothesis of the present study is to verify whether, the spatial distribution of seabass fry remains invariable under the pressure of certain rearing conditions, due to its extendedpla sticity or, certain situations lead to specific distributions. for this reason, we investigated andcom pared horizontal and vertical distribution of sea bass reared in aquaria, at three differentstocking densities (5 ind. l–1, 10 ind. l–1 and 20 ind. l–1), two different temperatures (16 and 23οC)and two different feeding conditions (feeding fish and starving fish). the analysis of the videorecords and General Linear Models revealed a strong effect of stocking density, feeding condi-tion, and day of experimentation on fish behavior interpreted by the vertical and horizontal dis-tribution of fish. temperature also affected distribution when it was combined with stocking den-sity and feeding condition. however, this effect seemed to be lessened by the effect of the othervariables (i.e. stocking density, feeding condition and day of experimentation). the exposure time(chronic stress) to the stocking density, temperature and feeding condition as well as to their ad-ditive cumulative effect (variables interactions) was the cause of the development of stereotypicbehaviors that increased in occurrence frequency and over an experimental time lapse and grad-ually led to the weakening and death of the fry (bad welfare). thus, it is assumed that stockingdensity, feeding condition, day of experimentation and temperature can have negative effects onsea bass “well being” and in order to maintain its welfare during rearing it is important to takeall necessary pre cau tions of handling when dealing with these variables in the fish environment.

Key words: distribution, Dicentrarchus labrax, fish, welfare.

* Corresponding author: tel.: +30 2810 394062/63, fax: +30

2810 394404, e-mail: neofytou@biology.uoc.gr

usually, intense farming implies that animals/fish

are kept in small volumes/spaces in densities much

higher than the ones encountered in natural popula-

tions. this indicates that rearing conditions are main-

ly manipulated so as to increase profit. these impos -

ed conditions may evoke reactions of impaired wel-

fare when the animal becomes unable to cope with

them. Behaviors such as food acquisition, predator

avoidance, migration, and habitat preference may

hamper welfare of an organism and thus of a popula-

tion, and were commonly used as indicators of the

suitability of the living conditions (Little, 2002). in

some respects, behavioral responses are an animal’s

first line of defense against adverse living conditions,

changes often being triggered by the same stimuli

that initiate the primary stress response. the expres -

s ed behavioral response depends on the kind and the

intensity of the stressor concerned. for example, when

aggressed or attacked by a predator, fish may respond

by shoaling (Pitcher & Parrish, 1993), “freezing” (Goo -

dey & Liley, 1985), taking shelter (Brown & Warbur-

ton, 1999) or changing its color (endler, 1986; o’Con -

nor et al., 2000), whereas feeding activity may be mo -

di fied, delayed or suppressed (hart, 1997) and fish

may avoid areas in which they have been attacked

(Li ma, 1998). Specific adaptive behavior patterns are

also observed in response to tissue damage (Verhei-

jen & Buwalda, 1988) or parasites (furevik et al.,

1993). in such cases, different behavioral patterns

may occur, ranging from reduced feed intake, to ag-

gression and death. the spatial distribution of a fish

as a part of its behavioral repertoire is therefore af-

fected and modified by the altered rearing conditions

applied and can serve as a proxy measure of potential

stressful conditions (Pihl et al., 1991; kristiansen et

al., 2004; Stien et al., 2007). thus, information about

the animal’s behavior and distribution in such condi-

tions may help us to evaluate its welfare condition.

european sea bass, Dicentrarchus labrax (Linna -

eus, 1758), is one of the two marine fish species along

with gilthead sea bream (Sparus aurata Linnaeus,

1758) which have characterized the development of

marine aquaculture in the Mediterranean region over

the last two decades. the substantial increase in pro-

duction levels of this high value species has been pos-

sible thanks to the progressive improvement in the

technologies involved in the production of fry in ha -

tc heries (Moretti et al., 1999). it is characterized as an

euryhaline and an eurythermal species since it can

survive from 0 to 38‰ salinity (Chervinsky, 1974)

and at a temperature range of 2°C to 32°C as an a -

dult although it will stop feeding below 7°C (Barna -

bé, 1990). it is considered to be a hypotropical species

that is mostly found in water temperatures of 8°C to

24°C (Moreira et al., 1992), while it can also be rear -

ed in a tropical ambient at temperatures between 25

and 29°C (Barnabé, 1990).

Sea bass tolerance and ability to adapt to temper-

ature variations (alliot et al., 1983), as well as to high

stocking densities (dalla Via et al., 1998; Papou tso -

glou et al., 1998; Paspatis et al., 2003) indicates a

species with high adaptability and plasticity in differ-

ent rearing conditions. the working hypothesis of the

present study is to verify whether, under the pressure

of certain rearing conditions, the spatial distribution

of the species remains invariable, because of its ex-

tended plasticity or, if extreme situations, lead to spe-

cific distributions. We also tried to ascertain whether

any change of the distribution behavior or an incre-

ment of certain behaviors (such as the emergence of

stereotypic behaviors) could serve as a non invasive

indicator of the good or bad welfare of the fish as a

result of the effect of the parameters studied. for this

reason, we investigated and compared fry sea bass

horizontal and vertical distribution reared in aquaria,

at three different stocking densities under two tempe -

ra tures and two different feeding conditions.

MateriaLS aNd MethodS

Experimental design

the experimental fish were provided from the hatch -

e ry facilities of the hellenic Center of Marine re-

search (hCMr) at the age of 76 days post hatch, 0.057

±0.010 g mean weight (±s.e.) and 1.810±0.040 cm

total Length (±s.e). they were transferred to the fa-

cilities of the Cretaquarium “thalassokosmos” of

hCMr where the experiments took place. the fish

were randomly assigned to 24 plastic transparent a -

qua riums of 6 l volume each (29.5 cm × 19 cm × 17

cm) in one replicate (a and b aquarium), two differ-

ent temperatures (16 and 23οC: cold and hot), two

different feeding conditions (feeding fish and starving

fish) and three different stocking densities (5, 10 and

20 ind. l–1: table 1). each aquarium was filled with

filtered seawater with the same salinity (S) as the

sour ce rearing tanks (35 g l–1), at a rate that assures a

renewal of 32% of its volume h-1 and was supplied

with a porous stone through which air bubbles were

diffused from the bottom of the aquarium to the sur-

100 Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry

face and in a way that the water was circularly mixed

gently. in order to avoid visual disturbance between

adjacent aquariums, the rear and lateral walls of each

aquarium were dyed with dark blue paint (fig. 1).

this also facilitated the video recording and analysis

due to the contrast between the fish and the back-

ground. Photoperiod was kept to 10L:14d. to avoid

stressing those fish being introduced to higher tem-

perature levels, the mean water temperature was gra -

dual ly increased from 17±2°C (temperature of the

source rearing tanks) to 23±1°C. after their intro-

duction to the aquariums fish at both temperatures

(16 and 23°C) were allowed to acclimatize, for at

least 30 min (estimated time during which fish occu-

pied the whole water volume of the tank) before any

experimental handling began, to allow them to recover

from transfer stress. the fish that were fed (feeding

fish) were provided with enriched instar i Artemia

nauplii (“Selco”, iNVe S.a., Belgium) equal to 80%

of their present biomass in wet weight and pellets [2/3

commercial crumbles (r1 100, Proton 2/3 particle

size 200-300 μm)] equal to 20% of their present bio-

mass in dry weight. food was provided 4 times during

the day, at 9:00, 12:00, 15:00 and 18:00. daily mortality

was measured and the dead fish were removed from

the aquaria.

Data collection

the observation was held with digital video cameras

(SoNy haNdyCaM, dCr-hC27e; SoNy diG-

Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry 101

taBLe 1. experimental setup. Stocking densities: low

(5 ind. l–1), medium (10 ind. l–1), high (20 ind. l–1)

aquarium feeding Stocking temperature

replicates condition density (°C)

1a 1b feeding fish low 16±1 (cold)

2a 2b feeding fish medium 16±1 (cold)

3a 3b feeding fish high 16±1 (cold)

1a 1b Starving fish low 16±1 (cold)

2a 2b Starving fish medium 16±1 (cold)

3a 3b Starving fish high 16±1 (cold)

1a 1b feeding fish low 23±1 (hot)

2a 2b feeding fish medium 23±1 (hot)

3a 3b feeding fish high 23±1 (hot)

1a 1b Starving fish low 23±1 (hot)

2a 2b Starving fish medium 23±1 (hot)

3a 3b Starving fish high 23±1 (hot)

fiG. 1. experimental set up where the fish groups were plac ed respectively per stocking density and per replicate. the same

set up was positioned in two sides (feeding and starving group) and in two different temperatures (table 1). the dark grey

color indicates the dark blue paint that was applied in the rear and adjacent walls. (i) front side of the experimental set up

showing feeding group, (ii) enlargement of an aquarium. a, water outlet (filter); B, water inlet; C, aeration (po rous stone).

itaL haNdyCaM, dCr-trV830e) placed in

front of the aquarium at an angle of 90o giving a who -

le view ability of the water column. recordings were

made before the first food supply of the day for all

the fish that were provided with food (feeding fish)

and at a corresponding time for all the fish that were

not provided with food (starving fish). the analysis of

the recordings stopped after 14 days of experiment

due to high mortality of the starving fish at 23οC. da-

ta was used for the analysis of the fish horizontal and

vertical distribution but also for the identification of

possible abnormal behavior patterns.

Data analysis

Software

the observation and processing of the recordings

were conducted in order to identify the location of

the individuals in the water column. the recorded vi -

deo clips were initially processed with the Windows

Mo vie Maker application software in order to extract

selected frames per minute. each frame was then a -

na lyzed by the use of a custom-made software (Xy-

quantifier) build in LabView (National instruments)

programming language. in every individual frame the

total length of the tank and the position of the center

of the mass of each fish were pointed by the user. all

these positions, which represent the real location of

each fish in the water column, were then calculated

automatically in both X (horizontal) and y (vertical)

dimensions. finally all measurements were exported

in text format for further processing and analysis.

Measurements were performed for three different

dates (i.e. 1st, 6th and 14th day) when noticeable chan -

ges in behavior were observed.

Statistical analysis

the effects of feeding condition, stocking density and

tem perature on the horizontal (X) and vertical (y)

position of fish in the tank (aquarium) were analyzed

using General Linear Models (GLM). day of experi-

mentation (d) was also included in the model. fur-

thermore, replicate (r) was considered as an aquari-

um effect (“cage effect”) nested within treatments

(dytham, 2011) to examine its significance for the

ver tical and horizontal position variables. to correct

the differences in the number of fish measured due to

the increasing cumulative mortality over the course of

experiments (e.g. very few remaining fish on day 14

in the starving treatments) position measurements

were weighted by a factor proportional to the number

of fish in the tank (survival rate) (faraway, 2005).

the general model used was of the form:

Position = μ + fC + Sd + t + d + interactions + ε

where Position is the vertical (y) or horizontal (X)

position, fC the feeding condition, Sd the stocking

density, t the temperature, d the day of experimen-

tation, interactions is any possible combination of two

effects interaction, μ the global mean and ε the error

term assumed to be distributed normally.

all variables and their first order interactions

were initially included in the model but only signifi-

cant (p<0.05) variables and interactions were finally

selected by using a stepwise backward removal me -

thod. finally, marginal means for each significant mo -

del term and their 95% confidence intervals were es-

timated and plotted.

reSuLtS

Trends and evolution of survival and behavior

the overall experiment lasted 19 days, but because

most individuals in the starving fish were dead by day

15, analysis of the recordings stopped at day 14. Gen-

erally, the fish kept at 23°C were those that died first,

although no starving fish survived, in any of the con-

ditions (i.e. high density, medium density or cold tem-

perature) (fig. 2).

the starving group exhibited abnormal swimming

patterns in different day of experimentation, stocking

density and temperature (table 2) like: swimming at

the upper levels of the water column, upstream posi-

tioning, “sitting” on the bottom, or laying in the water

column with their heads plump (a characteristic be-

havior of the autotrophic stages: kentouri, 1985) or

hiding behind the water filter and the air diffuser sto -

ne, expressing a stereotypic swimming behavior which

is characteristic of feeble fish during the first days of

exogenous feeding (kentouri, 1985).

on the contrary, the behavior of the feeding fish

did not show abnormal patterns besides swimming to

the more upper parts of the water column near the

surface. Some evidence of aggressive and cannibalis-

tic behavior was also observed at certain stocking

densities on the 6th day of experimentation and that

seemed to be the main cause of mortality within those

groups (fig. 2 and table 2).

102 Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry

Statistical data processing

results of the GLM analysis are shown in tables 3

and 4. With regard to vertical position (y), stocking

density, feeding condition and day of experimenta-

tion entered significantly into the model together

with three interaction terms (feeding condition × day

of experimentation, stocking density × temperature

and feeding condition × temperature) (table 3). hor-

izontal po si tion (X) was significantly related to day of

experimentation and the interaction day of experi-

mentation temperature (table 4). the replicate term

[(the interaction: r × (stocking density) × (feeding

condition) × (temperature)] was not statistically sig-

nificant neither for the vertical (f = 2.76, p = 0.10)

nor for the horizontal position (f=1.13, p=0.33).

Both models were highly significant (tables 3 and

4) with coefficients of determination (r2) 0.46 and

0.24 for vertical (y) and horizontal (X) position, re-

spectively. residual plots showed that the models

were satisfactory in terms of residual properties (ran-

domly distributed, homoscedastic) (fig. 3).

estimated standardized means for the main effects

entering into the two GLM models are shown in fig-

ure 4. Stocking density was a significant variable for

vertical position with the highest density treatment (20

ind. l–1) exhibiting the highest mean value (fig. 4a).

the starving fish had a significantly lower mean verti-

Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry 103

fiG. 2. range of survival in the experimental time for the fish populations of three different densities (low, medium, high)

for both replicates of different temperature (hot and cold) and feeding condition (feeding fish and starving fish).

cal distribution compared to the feeding fish (fig. 4b).

in the beginning of the experiment (day 1) the fish

were lower in the water column compared to day 6

104 Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry

taBLe 2. Summary of the behavioral observations for each group of fish, t: temperature

Behavioral element fish group

feeding condition/t day of Stocking

experimentation density

(1) Nervous swimming at the upper Starving fish/ from 2nd day all stocking densities

levels of the water column all temperatures

(2) Continuous struggle to retain Starving fish/16°C from 6th day Medium, Lowin the position in the water column

by positioning themselves opposite Starving fish/16°C from 8th day highto the gentle water current

(3) drifted away from the gentle

water current (moving passively Starving fish/23°C 6th day Mediumthrough the water)

(4) “Sitting” on the bottom

(5) Laying in the water column Starving fish/23°C from 8th day Lowwith their heads plump

(6) Shelter searching (hiding Starving fish/23°C 14th day highbehind the water filter and the air

diffuser stone

(7) Stereotype (moving for a period Starving fish/16°C 14th day Mediumof time in the water column

in a repeated fixed way whilst

adhering to the aquarium walls

in up and down movements)

(8) aggressiveness (chasing or Starving fish/ 14th day Medium, high

biting a fish, or being chased or all temperatures

bitten by another fish)

(9) Cannibalism (biting or eating feeding fish/23°C 6th day Medium, highanother fish or being bitten or

eaten by another fish)

taBLe 3. aNoVa table for the final weighted general lin-ear model fitted to the vertical position variable (y). Sd:stocking density, fC: feeding condition, d: day of experi-mentation, t: temperature

type iii Sum Mean

Source of Squares df Square f p

Corrected 9207.28 14 657.66 11.75 0.000

Model

intercept 105803.36 1 105803.36 1890.10 0.000

Sd 635.00 2 317.50 5.67 0.003

fC 263.19 1 263.19 4.70 0.030

d 488.11 2 244.05 4.36 0.013

fC × d 1526.26 2 763.13 13.63 0.000

Sd × t 2062.62 2 1031.31 18.42 0.000

fC × t 743.59 1 743.59 13.28 0.000

error 113634.88 2030 55.98

total 766148.88 2045

Corrected

total 122842.16 2044

taBLe 4. aNoVa table for the final weighted general linear

model fitted to the horizontal position variable (X). fC: feed-

ing condition, d: day of experimentation, t: temperature

Source type iii Sum df Mean f p

of Squares Square

Corrected 11173.00 7 1596.14 8.33 0.000

Model

intercept 971472.52 1 971472.52 5067.79 0.000

d 2041.12 2 1020.56 5.32 0.005

fC × t 5798.11 2 2899.05 15.12 0.000

error 390483.46 2037 191.69

total 2503490.87 2045

Corrected

total 401656.46 2044

and day 14 (fig. 4c). With regard to horizontal distri-

bution, the only significant main effect was the day of

experimentation with fish exhibiting a gradual change

in their mean positioning towards the left part of the

aquaria (opposite to aeration) with time (fig. 4d).

estimated standardized means for the interaction

terms entering significantly into the two models are

shown in figure 5. the starving fish exhibited a de-

crease in their mean vertical distribution with time,

while an opposite trend was found for the feeding fish

(fig. 5a).

different stocking densities exhibited differences

in vertical positioning for the two temperature re gi -

mes. this was particularly evident for the medium

density (10 ind. l–1) in which the vertical positioning

was lower at the cold temperature (fig. 5b).

although at hot temperature no difference was ob-

served in the mean vertical distribution between the

feeding and starving fish, this was not the case for the

cold temperature treatment in which the feeding fish

were distributed higher than the starving fish (fig. 5c).

With regard to horizontal position, differences

were mainly observed for the feeding fish in which fish

were concentrated closer to the right part of the aquar-

ia (towards aeration) at cold temperature (fig. 5d).

Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry 105

fiG. 3. residual plots of the final general linear models fit-

ted to a) vertical (y) position, and b) horizontal (X) posi-

tion.

fiG. 4. Standardized estimates of vertical (y) or horizontal (X) position with 95% confidence intervals for the significant main

effects of the two models. (a) y: stocking density, (b) y: feeding condition, (c) y: day of experimentation and (d) X: day of

experimentation. 0: feeding fish, 1: starving fish.

diSCuSSioN

as the British farm animal Welfare Council (in kee -

l ing & Jensen, 2002) declared, there are five freedoms

that animals must have: 1) freedom from thirst, hun -

ger and malnutrition, 2) appropriate comfort and

shel ter, 3) prevention, or rapid diagnosis and treat-

ment of injury and disease, 4) freedom to display

most normal patterns of behavior, and 5) freedom

from fear. in the case of farmed fish, these five free-

doms can be translated into: sufficient quantity and

quality of food, suitable temperature, good oxygen

level, water free of pollution, and a stocking density

which takes into account normal swimming and social

interactions, good possibilities to avoid perceived

danger and good monitoring of the health status of

the fish (faWC, 1996).

fish behavior monitoring, especially the study of

spatial distribution, may be performed by means of

different methodologies, more or less complicated

and time consuming (kato et al., 1999, 2004; hader,

2001; Xu et al., 2005, 2006; Stien et al., 2007). the ob-

servation and analysis approach we used proved to be

a handy, non invasive methodology that could be

used as a very valuable tool for researchers that deal

with the behavior of fish. Nevertheless, it is suscepti-

ble to further improvements for more automation

and expansion to the three dimensional level.

the statistics applied, showed important effect of

the stocking density, feeding condition and day of ex-

perimentation at the vertical distribution and of the

day of experimentation for the horizontal distribution

of the sea bass fry. Specifically, if the effect of feeding

condition is standardized by the effect of all other sig-

nificant main and interaction terms, it was observed

that starving fish were swimming on average lower in

the water column than feeding fish. the interaction of

feeding condition × temperature indicates that the

starvation effect (lower positioning of the starving

fish) is particularly important at lower temperatures.

as far as the day of experimentation is concerned we

observed that on the first day of the experiment the

fish were swimming lower in the water column when

compared to the later experimental days. We also ob-

served in their horizontal distribution that on the first

day they appeared to shoal in the right area of the

aquarium while there-after they began to distribute

over the whole of the aquarium. the fact that the

starving fish tended to swim lower in the water column

106 Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry

fiG. 5. Standardized estimates of vertical (y) or horizontal (X) position with 95% confidence intervals for the significant in-

teraction terms of the two models. (a) y: feeding condition × day of experimentation, (b) y: stocking density × temperature,

(c) y: feeding condition × temperature, and (d) X: feeding condition × temperature. 0: feeding fish, 1: starving fish.

could be attributed to their predictable seedy physio-

logical functions due to possible alterations in body

composition, e.g. reduction in the size of the gut and

of the liver, as was proven for various species (ehrlich

et al., 1976; rios et al., 2002; Pirhonen et al., 2003; tri-

pathi & Verma, 2003; Lemieux et al., 2004), and con-

sequently their passive sinking due to the water circu-

lation in the aquarium. this is a situation that was ev-

idenced also by time lapse since fish were getting

more and more stressed due to food deprivation.

day of experimentation was found to be an impor-

tant factor affecting the distribution behavior since it

represented the exposure time (chronic stress) to the

conditions of stocking density, temperature and food

availability. all of these factors over a period of time

lead to a sum of the effects (interaction effects) whe re -

by the intensity and frequency of the stereotypic behav-

iors observed led fish to gradual exhaustion and death.

food availability is one of several factors that may have

a direct effect on fish behavior and its variations during

the day. hecht & Pienaar (1993) show ed that in the

african catfish, Clarias gariepinus (Bur chell, 1822), a

gradual reduction in food availability resulted in an in-

crease in territorial behavior and also increased the

number of aggressive acts. the majority of studies on

the effects of starvation has focused on the effects on

growth, muscle protein and fat composition (rios et

al., 2002; Pirhonen et al., 2003; tripathi & Verma,

2003; Lemieux et al., 2004). food deprivation in

farmed fish for short periods under appropriate condi-

tions (e.g. temperature and season) may not cause wel-

fare problems (fSBi, 2002). however, it is important

to consider other effects of starvation and reduced nu-

trition, such as changes in metabolic acti vi ty (Martinez

et al., 2003) and behavior related to competition, as

well as the potential for increased aggression. for ex-

ample, reduced food abundance can cause changes in

territorial behavior strategies and a cti vity patterns in

brown trout, Salmo trutta Linnaeus, 1758 (alanara et

al., 2001; Brannas et al., 2003). in the present study,

food deprivation led to incidences of aggressive and

cannibalistic behavior. Nevertheless, such behaviors al-

so occurred in the feeding fish of the medium and high

stocking density and high tem perature reflecting the

impact of these two variables on the development of

this unfavorable behavior patterns.

important effects on the fry distribution in the

present study also appeared in the case of the stock-

ing density impact. this has been marked by many

authors (Paspatis et al., 2003; Gornati et al., 2004; di

Marco et al., 2008; Sammouth et al., 2009). Generally,

it is believed that increased stocking density has a

negative effect on fish performance (growth, feed in-

take, survival and feed conversion), because the stress

induced, has a metabolic cost, leading to the reduc-

tion in growth and food utilization (hengsawat et al.,

1997). in the present study the population with the

highest stocking density was found to swim higher in

the water column. this was also the case for halibut,

Hippoglossus hippoglossus Linnaeus, 1758 (kristian -

sen et al., 2004) and may occur because the fish are

hungry due to high mobility and high metabolism and

occupy the surface where they are used to expect/find

the food as usual. increasing stocking density was also

found to result in stress that leads to enhanced energy

requirements (Siddiqui et al., 1993; hengsawat et al.,

1997; hecht & uys, 1997). Nevertheless, studies in

both laboratory and aquaculture settings show that

the effect of stocking density on measures of welfare

varies between species. for example, sea bass and

gilthead sea bream, Sparus aurata, showed higher

stress levels at high densities (Montero et al., 1999;

Vazzana et al., 2002; Gornati et al., 2004). in contrast,

arctic charr, Salvelinus alpines (Lin naeus, 1758), feed

and grow well when stocked at high densities but show

depressed food intake and growth rates at low densi-

ties (Jorgensen et al., 1993). Moreover, from the stud-

ies of Greaves & tuene (2001) and kristiansen et al.

(2004) in halibut, tolerance for high stocking density

appears to be stage dependent.

When the combined effect of the stocking density

and temperature was examined the movement to-

wards the surface was also observed in the middle

stocking density populations and the high tempera-

ture conditions. this indicated that temperature may

have a potential impact on fry’s distribution. temper-

ature has been proven to affect the metabolism. in

particular, the high temperatures are more stressful

either because of the temperature itself or from the

combined effect of reduced oxygen content of the

war mer water and the increased metabolic oxygen de-

mand in the fish. the metabolism of the fish rises and

consequently its demand for food which is interpreted

by a behavior similar to the behavior described earlier

for the population of the higher stocking density

(Beamish, 1978; Schurmann & Steffensen, 1997; Clai -

reaux & Lagardère, 1999, Person-Le ruyet et al.,

2004; Sfakianakis & kentouri, 2010). temperature

exerts a significant effect on the swimming perfor -

mance of the fish with the effects mainly attributed to

the modification of the molecular kinetics and the

rates of the biochemical reactions that convert chemi -

Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry 107

cal energy into propulsive thrust (Beamish, 1978), as

well as to the alteration of the physical properties of

the water, which may affect the fish movement (fui -

man & Batty, 1997; Johnson et al., 1998).

deviations from the normal fish behavior (behav-

ioral variations) that usually accompany pre-lethal be-

haviors are rarely taken into consideration and/or are

interpreted in the wrong way leading to misconcep-

tions about the early assessment of stressed conditions

that may produce non-reversible results (kentouri &

divanach, 1983). Chronic stress has an effect on fish

behavior and sometimes can lead to atypical, abnor-

mal behavior like stereotypes. therefore the “quan-

tification” of the behavior in relation to chronic stress

could serve as a very useful tool for welfare assess-

ment. Bibliography about stereotypic behavior in fish

is scarce. Stereotypic behavior in animals is character-

ized by a repeated pattern (usually a phenotypically

non functional behavior). Stereotypes could be the in-

evitable result of a restricted environment (Jen sen,

2002) and is often correlated with the exposure to

chronic stress that is induced from unfavorable feed-

ing conditions (Mench & Mason, 1997; robert et al.,

1997). the development of stereotypes is usually the

response to frustration, thwarting or conflict (Wür bel

& Stauffacher, 1997). the results of the present stu dy

indicate that sea bass fry could develop stereotypic be-

havior like the countercurrent continuous swimming

or the periodic up and down movements in a repeated

fixed way adherent to the aquarium walls. this is also

a characteristic behavior pattern of feeble fish during

the first days of their exogenous feeding (kentouri,

1985), introducing in both cases a warning of a bad

non reversible health status. in the present study, the

exposure time (chronic stress) to the stocking density,

temperature and feeding condition as well as to their

additive cumulative effect (variables interactions) was

the cause of the development of stereotypic behaviors

that increased in occurrence frequency and over an

experimental time lapse and gradually led to the

weakening and death of the fry (bad welfare).

in conclusion, in this study, by the use of a handy,

data acquisition and analysis methodology, we evi-

denced a strong effect of the stocking density, feeding

condition, and day of experimentation on the fish be-

havior as interpreted in its vertical and horizontal dis-

tribution. also the temperature appeared to affect

the distribution when it was combined with the vari-

ables of stocking density and feeding condition but

this effect seemed to be lessened by stocking density,

feeding condition and day of experimentation. Con-

sequently, it is assumed that the previously men-

tioned variables (stock ing density, feeding condition,

day of experimentation and temperature) can have

negative effects on sea bass’s “well being” and in or-

der to maintain its welfare during rearing it is impor-

tant to take precautions of care when dealing with

these variables in the fish’s environment. the distri-

bution behavior could also be considered as a non in-

vasive indicator of the good or bad welfare of the fish.

aCkNoWLedGeMeNtS

We express our thanks to the Crete Marine aquari-

um “thalassokosmos” for the kind providing of their

facilities for the conducting of the experiment. Spe-

cial thanks must also be given to Petroutsos iasonas

for his assistance with the analysis of the video re -

cord ings and to Mrs. Pauline Chambers and Christos

Neofytou for their useful comments in english gram-

mar. financial support for this study has been provid-

ed by the eu project 022720 faStfiSh on farm as-

sessment of stress level in fish (2006-2009).

refereNCeS

alanara a, Burns Md, Metcalfe NB, 2001. intraspecific re -

sour ce partitioning in brown trout: the temporal distri-

bution of foraging is determined by social rank. Journal

of Animal Ecology, 70: 980-986.

alliot e, Pastoureaud a, thebault h, 1983. influence de la

tem pérature et de la salinite sur la croissance et la com -

po sition corporelle d’alevins de Dicentrarchus labrax.

Aquaculture, 31: 181-194.

Barnabé G, 1990. rearing bass and gilthead bream. in:

Bar nabé G, ed. Aquaculture. ellis horwood, New york:

647-686.

Barton Ba, 2002. Stress in fishes: a diversity of responses

with particular reference to changes in circulating cor-

ticosteroids. Integrative and Comparative Biology, 42:

517-525.

Beamish fWh, 1978. Swimming capacity. in: hoar WS,

ran dall dJ, eds. Fish Physiology, Vol. 7. academic

Press, inc., New york: 101-187.

Brannas e, Jonsson S, Lundqvist h, 2003. influence of food

abundance on individual behaviour strategy and growth

rate in juvenile brown trout (Salmo trutta). Canadian

Jour nal of Zoology, 81: 684-691.

Brown C, Warburton k, 1999. Social mechanisms enhance

escape responses in shoals of rainbowfish, Melanotaenia

duboulayi. Environmental Biology of Fishes, 56: 455-459.

Chervinsky J, 1974. Sea bass, Dicentrarchus labrax Linnaeus

(Pisces, Serranidae) a “police-fish” in fresh water ponds

and its adaptability to various saline conditions. Ba mi -

dgeh, 26: 110-113.

108 Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry

Claireaux G, Lagardère JP, 1999. influence of temperature,

oxy gen and salinity on the metabolism of european sea

bass. Journal of Sea Research, 42: 157-168.

dalla Via J, Villani P, Gasteiger e, Niederstatter h, 1998.

o xygen consumption in sea bass fingerling Dicentrar-

chus labrax exposed to acute salinity and temperature

changes: metabolic basis for maximum stocking density

estimations. Aquaculture, 169: 303-313.

di Marco P, Priori a, finoia MG, Massari a, Mandich a,

Marino G, 2008. Physiological responses of european

sea bass Dicentrarchus labrax to different stocking densi-

ties and acute stress challenge. Aquaculture, 275: 319-328.

dytham C, 2011. Choosing and Using Statistics: A Biologist’s

Guide. Blackwell Science, uk.

ehrlich kf, Blaxter JhS, Pemberton r, 1976. Morphologi cal

and histological changes during the growth and starvation

of herring and plaice larvae. Marine Biology, 35: 105-118.

endler Ja, 1986. defence against predators. in: feder Me,

Lau der GV, eds. Predator Prey Relationships: Perspecti -

ves and approaches from the study of lower vertebrates.

university of Chicago Press, Chicago: 109-134.

faraway JJ, 2005. Extending the Linear Model with R: Gene -

ra lized Linear, Mixed Effects and Nonparametric Regres-

sion Models (Texts in Statistical Science). Chapman &

hall/CrC.

faWC, 1996. Report on the welfare of farmed fish. Farm An-

imal Welfare Council (FAWC). Surbiton, Surrey.

fSBi, 2002. Fish Welfare (Briefing Paper 2). fisheries Socie -

ty of the British isles. Granta information Systems, Cam -

bridge, uk.

fuiman La, Batty rS, 1997. What a drag it is getting cold:

par titioning the physical and physiological effects of

tem perature on fish swimming. The Journal of Experi-

mental Biology, 200: 1745-1755.

furevik dM, Bjordal Å, huse i, fernö a, 1993. Surface ac-

tivity of atlantic salmon (Salmo salar L.) in net pens.

Aquaculture, 110: 119-128.

Goodey W, Liley Nr, 1985. Grouping fails to influence the

escape behaviour of the guppy (Poecilia reticulata). Ani -

mal Behaviour, 33: 1032-1034.

Gornati r, Papis e, rimoldi S, terova G, Saroglia M, Ber -

nar dini G, 2004. rearing density influences the expres-

sion of stress-related genes in sea bass (Dicentrarchus

labrax, L.). Gene, 341: 111-118.

Greaves k, tuene S, 2001. the form and context of aggres-

sive behaviour in farmed atlantic halibut (Hippoglossus

hip poglossus L.). Aquaculture, 193: 139-147.

hader dP, 2001. Image Analysis: Methods and Applications.

CrC Press, Boca raton, London.

hart PJB, 1997. foraging tactics. in: Godin J-GJ, ed. Be ha -

vioural Ecology of Teleost Fishes. oxford university

Press, oxford: 104-133.

hecht t, Pienaar aG, 1993. a review of cannibalism and

its implications in fish larviculture. Journal of the World

Aquaculture Society, 24: 246-261.

hecht t, uys W, 1997. effect of density on the feeding and

aggressive behaviour in juvenile african catfish, Clarias

gariepinus. South African Journal of Science, 93: 537-541.

hengsawat k, Ward fJ, Jaruratjamorn P, 1997. the effect

of stocking density on yield, growth and mortality of a -

fri can catfish (Clarias gariepinus Burchell, 1822) cultur -

ed in cages. Aquaculture, 152: 67-76.

huntingford fa, 2004. implications of domestication and

rearing conditions for the behaviour of cultivated fishes.

Journal of Fish Biology, 65: 122-142.

Jensen P, 2002. the study of animal behaviour and its ap -

pli cations. in: Jensen P, ed. The ethology of domestic a -

ni mals, an introductory text. CaBi Publishing, uk: 3-11.

Johnson rL, Geist dr, Mueller rP, Moursund ra, he d -

ge peth J, fuhriman d, Wirtz ar, 1998. Behavioral A -

cou stic Tracking System (BATS). report prepared for

u.S. army Corps of engineers Walla Walla district,

Walla Walla, Wa 99362, Battelle Pacific Northwest di-

vision P.o. Box 999 richland, Wa 99352.

Jorgensen eh, Christiansen JS, Jobling M, 1993. effects of

stocking density on food intake, growth performance

and oxygen consumption in arctic charr (Salvelinus al -

pin us). Aquaculture, 110: 191-204.

kato S, devadas M, okada k, Shimada y, ohkawa M,

Mu ramoto k, takizawa N, Matsukawa t, 1999. fast

and slow recovery phases of goldfish behavior after

tran section of the optic nerve revealed by a computer

image processing system. Neuroscience, 93: 907-914.

kato S, Nakagawa t, ohkawa M, Muramoto k, oyama o,

Watanabe a, Nakashima h, Nemoto t, Sugitani S, 2004.

a computer image processing system for quantification

of zebrafish behavior. Journal of Neuroscience Methods,

134: 1-7.

keeling L, Jensen P, 2002. Behavioural disturbances, stress

and welfare. in: Jensen P, ed. The ethology of domestic

animals, an introductory text. CaBi Publishing, uk: 79-

98.

kentouri M, 1985. Comportement larvaire de 4 Sparides

méditerranéens en élevage: Sparus aurata, Diplodus sar-

gus, Lithognathus mormyrus, Puntazzo puntazzo (Pois-

sons teleosteens). thèse de doctorat ès Sciences, uni-

versité de Sciences et techniques du Languedoc, Mont-

pellier.

kentouri M, divanach P, 1983. Sur l’utilisation des critères

pour déterminer l’état de sante et l’ évolution probable

des élevages de poissons marins.1. Cas des prelarves et

des larves de Diplodus sargus, Sparus aurata, Puntazzo

puntazzo, Lithognathus mormyrus. Bases biologiques de

l’aquaculture, Montpellier, ifremer. actes de Colloques

1: 525-538.

kristiansen tS, ferno a, holm JC, Privitera L, Bakke S,

fosseidengen Je, 2004. Swimming behaviour as an in-

dicator of low growth rate and impaired welfare in at-

lantic halibut (Hippoglossus hippoglossus L.) reared at

three stocking densities. Aquaculture, 230: 137-151.

Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry 109

Lemieux h, dutil Jd, Guderley h, Larocque r, 2004.

Growth, starvation and enzyme activity in white muscle

of atlantic cod: at what point do muscle metabolic ca-

pacities change? Marine and Freshwater Behaviour and

Physiology, 37: 287-294.

Lima SL, 1998. Stress and decision making under the risk

of predation: recent developments from behavioral,

reproductive, and ecological perspectives. Advances in

the Study of Behavior, 27: 215-290.

Little ee, 2002. Behavioral measures of environmental

stressors in fish. in: adams SM, ed. Biological Indicators

of Aquatic Ecosystem Stress. american fisheries Socie-

ty, Bethesda, Md: 431-472.

Martinez M, Guderley h, dutil Jd, Winger Pd, he P, Walsh

SJ, 2003. Condition, prolonged swimming perfo r mance

and muscle metabolic capacities of cod Gadus morhua.

The Journal of Experimental Biology, 206: 503-511.

Mench Ja, Mason GJ, 1997. Behaviour. in: appleby MC,

hughes Bo, eds. Animal welfare. CaBi, uk: 127-141.

Montero d, izquierdo MS, tort L, robaina L, Vergara

JM, 1999. high stocking density produces crowding

stress altering some physiological and biochemical pa-

rameters in gilthead seabream, Sparus aurata, juveniles.

Fish Physiology and Biochemistry, 20: 53-60.

Moreira f, assis Ca, almeida Pr, Costa JL, Costa MJ,

1992. trophic relationships in the community of the

up per tagus estuary (Portugal): a preliminary appro -

ach. Estuarine, Coastal and Shelf Science, 34: 617-623.

Moretti a, Pedini fernandez-Criado M, Cittolin G, Gui da -

stri r, 1999. Manual on hatchery production of seabass

and gilthead seabream, Volume 1. fao, rome.

o’Connor ki, taylor aC, Metcalfe, NB, 2000. the stability

of standard metabolic rate during a period of food dep-

rivation in juvenile atlantic salmon. Journal of Fish Bi-

ology, 57: 41-51.

Papoutsoglou Se, tziha G, Vrettos X, athanasiou a, 1998.

effects of stocking density on behavior and growth rate

of european sea bass (Dicentrarchus labrax) juveniles

reared in a closed circulated system. Aquacultural Engi-

neering, 18: 135-144.

Paspatis M, Boujard t, Maragoudaki d, Blanchard G,

ken touri M, 2003. do stocking density and feed reward

level affect growth and feeding of self-fed juvenile eu-

ropean sea bass? Aquaculture, 216: 103-113.

Person-Le ruyet J, Mahé k, Le Bayon N, Le delliou h,

2004. effects of temperature on growth and metabo-

lism in a Mediterranean population of european sea

bass, Dicentrarchus labrax. Aquaculture, 237: 269-280.

Pihl L, Baden SP, diaz rJ, 1991. effects of periodic hypoxia

on distribution of demersal fish and crustaceans. Mari -

ne Biology, 108: 349-360.

Pirhonen J, Schreck CB, reno PW, ogut h, 2003. effect of

fasting on feed intake, growth and mortality of Chinook

salmon, Oncorhynchus tshawytscha, during an induced

Aeromonas salmonicida epizootic. Aquaculture, 216: 31-38.

Pitcher tJ, Parrish Jk, 1993. functions of shoaling behavi -

our in teleosts. in: Pitcher tJ, ed. Behaviour of Teleost

Fishes. Chapman & hall, London: 363-439.

rios fS, kalinin aL, rantin ft, 2002. the effects of long-

term food deprivation on respiration and haematology

of the neotropical fish Hoplias malabaricus. Journal of

Fish Biology, 61: 85-95.

robert S, rushen J, farmer C, 1997. Both energy content

and bulk of food affect stereotypic behaviour, heart

rate and feeding motivation of female pigs. Applied An-

imal Behavior Science, 54: 161-171.

Sammouth S, roque d’orbcastel e, Gasset e, Lemarié G,

Breuil G, Marino G, Coeurdacier JL, fivelstad S, Blan -

che ton JP, 2009. the effect of stocking density on sea -

bass (Dicentrarchus labrax) performance in a tank-bas -

ed recirculating system. Aquacultural Engineering, 40:

72-78.

Schurmann h, Steffensen Jf, 1997. effects of temperature,

hypoxia and activity on the metabolism of juvenile at-

lantic cod. Journal of Fish Biology, 50: 1166-1180.

Sfakianakis dG, kentouri M, 2010. effect of temperature

on muscle lactate metabolic recovery in sea bass (Di -

cen trarchus labrax, L.) juveniles exposed to exhaustive

exercise. Fish Physiology and Biochemistry, 36: 387-390.

Siddiqui aQ, alnajada ar, alhinty hM, 1993. effects of

density on growth, survival and yield of the african cat-

fish, Clarias gariepinus (Burchell) in Saudi arabia. Jour-

nal of Aquaculture in the Tropics, 8: 213-219.

Stien Lh, Bratland S, austevoll i, oppedal f, kristiansen

tS, 2007. a video analysis procedure for assessing ver-

tical fish distribution in aquaculture tanks. Aquacultural

Engineering, 37: 115-124.

tripathi G, Verma P, 2003. Starvation-induced impairment

of metabolism in a freshwater catfish. Verlag der Zeit -

schrift für Naturforschung, Tübingen, 58: 446-451.

Vazzana M, Cammarata M, Cooper eL, Parrinello N,

2002. Confinement stress in sea bass (Dicentrarchus la -

brax) depresses peritoneal leukocyte cytotoxicity. Aqua-

culture, 210: 231-243.

Verheijen fJ, Buwalda rJa, 1988. Do pain and fear make

a hooked carp in play suffer? CiP – GeGeVeNS. u -

trecht.

Würbel h, Stauffacher M, 1997. age and weight at weaning

affect corticosterone level and development in iCr-mi -

ce. Animal Behaviour, 53: 891-900.

Xu Jy, Miao XW, Liu y, Cui Sr, 2005. Behavioral respon -

se of tilapia (Oreochromis niloticus) to acute ammonia

stress monitored by computer vision. Journal of Zhe-

jiang University Science B, 6: 812-816.

Xu Jy, Liu y, Cui Sr, Miao XW, 2006. Behavioral respon -

ses of tilapia (Oreochromis niloticus) to acute fluctua-

tions in dissolved oxygen levels as monitored by compu -

ter vision. Aquacultural Engineering, 35: 207-217.

110 Maria Neofytou et al. — Variables affecting horizontal and vertical distribution of sea bass fry