www.elsevier.com/locate/aqua-online

Aquaculture 224 (2003) 89–103

Effect of feeding method and protein source

on Sparus aurata feeding patterns

M.J. Sanchez-Murosa,*, V. Corcheteb, M.D. Suareza, G. Cardenetec,E. Gomez-Milana, M. de la Higuerac

a Department of Applied Biology, University of Almerıa, 04120 Almerıa, SpainbDepartment of Applied Physics, University of Almerıa, 04120 Almerıa, Spain

cDepartment of Animal Biology and Ecology, University of Granada, 18071 Granada, Spain

Received 5 February 2002; received in revised form 14 February 2003; accepted 18 February 2003

Abstract

The influence of dietary protein source (fishmeal, soy-protein concentrate and soy-protein

concentrate supplemented with methionine) on voluntary feed intake, daily feeding rhythm and

nutritive utilisation of diet was studied in the gilthead sea bream (Sparus aurata) fed by hand or

demand feeding. Fish weighing 21 g were maintained indoors under natural conditions of

temperature and photoperiod (transparent ceiling) and allowed ad libitum or self-feeding of

experimental diets for 26 days, with three replicates per treatment. In the first experiment, the

influence of hand or demand feeding on growth rate and feed utilisation of a fishmeal-based control

diet was studied. In a second trial, involving different protein sources, fish were maintained under the

experimental conditions for 6 days after 20 days training. The general composition of the

experimental diets was: 45% protein, 14% lipids and 20% carbohydrates. Results showed that: (1)

gilthead sea bream preferentially feed in the afternoon and evening; (2) demand feeding improved

both food conversion and protein efficiency; (3) the protein source appeared to induce changes in the

timing of feeding; and (4) supplements of methionine advanced the time of feeding and lengthened

ingestion phases.

D 2003 Elsevier Science B.V. All rights reserved.

Keywords: Feeding behaviour; Feed intake; Sparus aurata; Sea bream; Soy-protein; Methionine

0044-8486/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved.

doi:10.1016/S0044-8486(03)00211-4

* Corresponding author. Fax: +34-950-015476.

E-mail address: mjmuros@ual.es (M.J. Sanchez-Muros).

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–10390

1. Introduction

Maximum diet performance involves two important advantages: reduction of feeding

costs and decreased waste. The study of feeding behaviour in several fish species has

revealed that the adjustment of feeding times to match natural rhythm improves nutritional

efficiency, feeding frequency, food conversion efficiency and can even vary the utilisation

of certain nutrients (Bolliet et al., 2001).

Timing of feeding appears to influence the growth of fish. In Heterobranchus longifilis,

growth improved when fish fed at night rather than during the day (Kerdchuen and

Legendre, 1991). In Heteropneustes fossilis, growth was enhanced when feed was

supplied in the scotophase rather than at dawn (Sundaranaj et al., 1982). Thus, the same

food ingested at different times of the day is absorbed with differing efficiencies (Madrid,

1994). Likewise, voluntary daily feeding peaks have been recorded in species such as

European sea bass Dicentrarchus labrax (Anthouard et al., 1993; Sanchez-Vazquez et al.,

1995), rainbow trout Onchorhynchus mykiss (Alanara, 1996), siluridae (Boujard and

Luquet, 1996), turbot Psetta maxima (Burel et al., 1997), and gilthead sea bream Sparus

aurata (Anthouard et al., 1996).

An ability to regulate the intake of macronutrients has been demonstrated in some

species, such as sea bass (Rubio et al., 2001), rainbow trout (Sanchez-Vazquez et al., 1999)

or goldfish Carassius auratus, the latter even having an ingestion peak for each macro-

nutrient (Sanchez-Vazquez et al., 1998a). Hence, the dietary amino acid profile and

availability (as free or proteic amino acids) could affect feeding patterns.

Protein is a basic component of fish diets, both in terms of quantity and quality, protein

requirements being higher than those of other animals (Cowey, 1975). Fishmeal provides

an adequate balance of amino acids, but its increasing demand and price, as well as

uncertain supply, make it necessary to find alternative protein sources. Furthermore,

problems associated with animal proteins make vegetable proteins the most promising

candidates. The high quality of fishmeal proteins makes substitution difficult; however,

partial substitutions are being made (Shiau et al., 1987; Reigh and Ellis, 1992) and

fishmeal dietary levels could be reduced even further by adequate supplementation with

essential amino acids, providing that the availability of the supplementary amino acids

coincides with that of protein amino acids at the sites of protein synthesis. An adequate

postprandial pattern of amino acids is not only necessary to ensure protein synthesis and

growth, but might also determine feed acceptability. In the case of using soy-protein to

feed gilthead sea bream, the maximum substitution to maintain similar growth to that with

fishmeal is around 20% (Robaina, 1995).

Vegetable proteins generally have an inadequate balance of amino acids, being deficient

in some essential amino acids (Murai, 1992). Soy-protein supplementation with methio-

nine allows significant replacement (50%) of fishmeal (Viola et al., 1982) but not complete

substitution. It has been suggested that the inefficiency of supplementation is due to the

faster uptake and subsequent catabolism of the supplemented amino acids with respect to

protein ones (Cowey and Walton, 1988). In fact, the utilisation of supplemented amino

acids for protein synthesis and growth is improved when they are coated, as shown in carp

(De la Higuera et al., 1998) and gilthead sea bream (Sierra, 1995). Another possibility of

improving dietary free amino acids for growth would be by increasing feeding frequency,

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103 91

with the aim of making all amino acids available at the same time and at the sites of protein

synthesis. However, since sea bass can discriminate between diets containing different

amounts of methionine (Hidalgo et al., 1988), it may be possible for gilthead sea bream to

adapt its feeding behaviour when facing either a methionine-deficient or a free-methio-

nine-supplemented diet.

The aim of the present study is to determine the effect of the feeding method and

protein source on S. aurata feeding patterns. In this sense, we seek to establish (1) whether

demand-feeding improves the notional utilisation of diet for growth; (2) whether gilthead

sea bream displays preferential feeding times throughout the day when fed on demand

with unrestricted access to feed; and (3) whether soy-protein-based diets, supplemented or

not with free methionine, affect voluntary-feeding patterns.

2. Material and methodology

2.1. Animal-rearing conditions

Gilthead sea bream of an initial weight of 21.0F 3.5 g, obtained from a local fish farm

(Carmar, Carboneras, Almerıa, Spain), were transported to the University of Almerıa and

placed in 250-l tanks with sea water at a flow of 10.4 l/h. The fish were kept under ambient

temperature (15.5F 0.6 jC) and photoperiod. Prior to the experiments, the fish were

divided into two groups, one fed by hand and the other by demand using Aquarium MA-

32 self-feeders (Procam Ingenieria), as described by Sanchez-Vazquez et al. (1994) that

delivered approximately 9–11 pellets (about 0.3 g feed/pellet) each time a fish activates

the rod located 3 cm below the water surface. Demand-fed fish were given a 20-day

training period. The experiments started when the fish appeared to have learned to feed

themselves on demand, following the criteria described below.

Feed-intake demands were recorded daily from the beginning of the training period.

Feeders were checked every day and the remaining food weighed. An approximate weight

of uneaten pellets at the bottom of the tanks was used to calculate the amount of food

ingested. The recording of early-day self-feeding was sporadic. Over the next few days,

the fish learned self-feeding, reacting to the food after hitting the bar. After 15–19 days,

the request for food approached food intake, and the self-feeding became more regular,

with feed demands concentrating within some hours of the day. The experiment started

when this behaviour remained constant for at least 3 days.

2.2. Diet

The mean general composition of the experimental diets was: 45% protein, 14% lipids

and 20% carbohydrates. Fishmeal (diet D-1), soy-protein concentrate (diet D-2) and soy-

protein concentrate supplemented with free methionine (diet D-3) were the protein sources

used as the only source of protein. The ingredients (Table 1) were blended with sodium

alginate and then thoroughly stirred with distilled water to obtain a moist, homogeneous

mixture. Pellets were made as described by Sanchez-Muros et al. (1998). Gross-energy

content of the experimental diets was calculated using metabolizable-energy values of

Table 1

Ingredients and proximate composition (% dry weight) of experimental diets

Control (D-1) Met-deficient (D-2) Met-supplemented (D-3)

Ingredient (g/100 g)

Fish meal 60.4 – –

Soy meal – 55.13 54.1

Methionine – – 0.77

Fish oil 7.28 13.69 13.69

Corn starch 11.90 6.78 6.78

Dextrin 6.65 5.79 5.79

Vitamin mixturea 2 2 2

Mineral mixturea 3 3 3

Sodium alginate 2 2 2

Cr2O3 0.5 0.5 0.5

Betaine 1 1 1

Micronized cellulose 5.43 10.11 10.11

Proximate composition (% dry matter)

Crude protein 47.71 47.30 47.23

Fat 13.21 13.42 13.41

Ash 10.23 11.45 11.02

Gross energy (MJ/kg) 22 22 22

aVitamin and mineral mixture according to Sierra (1995).

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–10392

19.6, 39.5 and 17.2 kJ g� 1 for protein, lipids and carbohydrates, respectively (Brett and

Groves, 1979). Crude protein, total lipids and moisture were analysed by standard methods

(AOAC, 1984). Methionine supplement for diet D-3 was calculated after determining the

amino acid content of the protein sources. Free methionine was added to soy-protein diet

to reach fishmeal-diet levels.

2.3. Experiment 1

Experiment 1 lasted from the 25th of February (sunrise 07:54 h; sunset 19:04 h) to the

21st of March (sunrise 07:10 h; sunset 19:30 h). The performance with hand feeding was

compared to that of demand feeding for 26 days. Groups of 15 fish were used: 3 groups

were hand-fed twice daily (at 10:00 h and 16:00 h) to apparent satiation and the other 3

were self-fed. All groups were fed a fishmeal-based diet (designated ‘‘Control’’ in Table

1). After about 30 min, non-ingested food was removed from hand-feeding tanks, those

having self-feeders were checked frequently. Uneaten food was quantified to calculate the

amount ingested.

2.4. Experiment 2

Fish trained for demand-feeding were divided into 9 groups (three replicates per

treatment) of 15 fish each. Each replicate per treatment was fed the corresponding test diet

for 6 days. Experimental diets had similar macronutrient compositions but varied in

protein source and methionine content (Table 1).

2.5. Spectral data analysis

Feed-demand results were subjected to spectral analysis based on the fast Fourier

transform. Spectral analysis was applied to the recordings of daily food delivered to each

tank, in order to identify any feeding rhythms of voluntary intake. Thus, the Fourier

spectrum H( f ) of the recording sample denoted by h(t) was obtained (Brigham, 1988)

Hðf Þ ¼Z l

�lhðtÞe�j2pftdt

where f is frequency, t is time and j is the imaginary unit ð j ¼ffiffiffiffiffiffiffi�1

pÞ. In general, H( f ) is a

complex quantity called Fourier transform of h(t). H( f ) can also be expressed as

Hð f Þ ¼ Rð f Þ þ jIð f Þ ¼ AHðf ÞAe jUðf Þ

where R( f ) is the real part of H( f ) and I( f ) is the imaginary part. The amplitude spectrum

of h(t) is denoted by AH( f )A and the phase spectrum is denoted by U( f ). These are

defined as

AHðf ÞA ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiR2ðf Þ þ I2ðf Þ

pUðf Þ ¼ tan�1 Iðf Þ

Rðf Þ

� �:

When studying the amplitude spectrum of h(t), it is possible to identify the principal

harmonic components of the sample because they always appear with major amplitude

(Bath, 1982). Frequencies corresponding to principal harmonic components are the

frequencies of the rhythms present in the sample (Aschoff, 1992). These rhythms are

identified visually and their period T (T= 1/f ) can be read directly from the amplitude

spectrum (see Fig. 4). By reading the period value at the highest peak points of the

amplitude spectrum, the phases of maximum feeding rhythms were derived for each group

studied. The spreading of these peaks reveals the intake phase or time during which the

maximum intake occurs (Papoulis, 1962; Corchete et al., 1989). The phase spectrum is

considered to determine the timing in which feed demands began to increase, because this

time is the phase delay of the principal harmonic component considered (Papoulis, 1962;

Bath, 1982). The harmonic component with major amplitude corresponds to the daily

rhythm. This harmonic component is easily identified in the amplitude spectrum as

described above. Thus, we can read the phase value for this period in the phase spectrum.

In this way, the phase delay, or timing in which demands began to increase, was derived.

As a result, spectral analysis can be used to identify and quantify any rhythm in feed

demand. Today, the spectral analysis is considered a standard tool for the analysis of time

series in several scientific fields. Particularly, in biomedical sciences, spectral analysis

proves to be a powerful analysis technique to reveal information which is hidden in the

original data and which is extremely difficult to discover with other analysis techniques

(Childers, 1989; Aschoff, 1992). This information hidden in the time domain may possibly

be easier to discover in the frequency domain (Corchete et al., 1989). This powerful

technique has been applied by Corchete et al. (1995) in another field of science, as

geophysics, with successful results.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103 93

Fig. 1. Daily feeding demand of gilthead sea bream fed a fishmeal-based feed recorded in 26 days (experiment 1).

Each graph corresponds to the record for 1 day and represents the average demand of three tanks. All graphs are

normalized to have the maximum of 1. Shadow zones show the hours of darkness.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–10394

Fig. 2. Average of the 26 records plotted in Fig. 1 (experiment 1). Y-axis units are number of demands in 10 min.

Shadow zones show the hours of darkness. Horizontal arrow shows the intake phase.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103 95

2.6. Biological indices and statistical data analysis

The biological indices determined were: feed conversion ratio (FCR), defined as dry

feed intake/biomass gain; and protein efficiency ratio (PER), defined as wet weight gain/

protein intake. Student’s t-test was used to compare the averages of the biological indices

determined.

3. Results

Over the first experiment, fish exhibited a certain feeding pattern throughout the

day; that is, feeding demands increased during the afternoon, starting at about 17:00 h

and ending at about 23:00 h, with a maximum of demands between 19:00 and 20.00 h

Table 2

Growth (W0 and Wf means initial and final weight, respectively), feed intake, feed efficiency ratio (FER) and

protein efficiency ratio (PER) of gilthead sea bream fed using different feeding systems (meanF standard

deviation)

Hand Demand

Wf�W0 (g) 19.95F 2.10 27.45F 2.50

(Wf�W0)/fish (g) 1.33F 0.15 1.83F 0.17

Feed intake (g) 52.50F 4.32 45.00F 5.06

Feed intake/fish/day (g) 0.14F 0.01 0.12F 0.02

FCR (feed intake/weight gain) 2.63F 0.78a 1.64F 0.03

PER (weight gain/protein intake) 0.80F 0.29 1.28F 0.54

FCR: feed conversion ratio = dry feed intake/wet weight gain.

PER: protein efficiency ratio =wet weight gain/protein intake.a Significant difference between both columns for p< 0.05 (Student’s t-test analysis).

Fig. 3. Daily feeding rhythms of gilthead sea bream fed on diets of: (a) soy protein supplemented in methionine;

(b) soy protein deficient in methionine; and (c) fishmeal as control diet (experiment 2). On the left side, each

graph corresponds to the record for 1 day and represent the average demand of three tanks for each treatment (all

graphs are normalized to have the maximum of 1). Shadow zones show the hours of darkness. On the right side,

average of the six records plotted in the left side for each treatment. Y-axis units are number of demands in 10 min.

Shadow zones show the hours of darkness. Horizontal arrows show the intake phases.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–10396

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103 97

(see Figs. 1 and 2). During the experiment, maximum intake was registered between March

3 and March 5, a period which coincided with a full moon. However, due to the duration of

the experiment, it cannot be statistically determined whether the moon influenced the

amount of feed ingested by the fish. On March 19, coinciding with new moon, there was

also an increase of food consumption associated with a rise in temperature of 1.5 jC. Fishfed on demand had a significantly lower FCR than those fed by hand, although the

improvement in protein utilisation (PER) did not significantly differ (Table 2).

In the second experiment, the data recorded showed that the general feeding pattern was

maintained, the maximum intake being concentrated in the afternoon/evening, irrespective

of the diet composition (see Fig. 3). However, the soy-protein-based diet supplemented

with free methionine appeared to lengthen the intake phase until 2:30 h (right part in Fig.

3). In this experiment, the intake phases for each diet were: 7:05 h for the control diet (D-

1), 7:05 h for soy-protein-methionine-deficient diet (D-2) and 9:30 h for the soy-protein

diet supplemented with methionine (D-3). The spreads of peaks in the amplitude spectrum

were considered to establish the intake phases (see Fig. 4 as example of spectral analysis).

Feed demands for the control diet (D-1) began to increase at 17:10 h (the spectrum phase

was considered to determine this time) and the timing for maximum demand-feeding

Fig. 4. An example of spectral analysis applied in this study: (a) Sample record with daily feeding demand for 6

days of gilthead sea bream fed on diets of soy protein deficient in methionine, as showed in Fig. 3b (left side). (b)

Normalized amplitude spectrum of this sample data. (c) A zoom view of the sample record spectrum shown in

panel b. Both spectrum graphs are normalized to have the maximum amplitude of 1.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–10398

activity was between 19:00 and 20:00 h (see Fig. 3). For diet D-2 (methionine-deficient),

feed demands began to increase at 17:10 h and the maximum occurred between 19:00 and

20:00 h. For the diet supplemented with methionine (D-3), demands started at 14:20 h,

reaching a maximum between 17:30 and 19:30 h. In the second experiment only, the

spectral analysis revealed a second rhythm with a period of 12 h for each experimental diet

(see Fig. 4). This secondary rhythm had a frequency of two peaks per day, the second peak

coinciding approximately with the peak of the main rhythm.

4. Discussion

Three behaviour patterns were identified in the experimental sea bream: daytime,

nighttime and evening (Aschoff, 1992). A dual behaviour has been described for the

European sea bass (Sanchez-Vazquez et al., 1995, 1998b), and a seasonal change in the

turbot (Muller, 1978); although, in larvae and juveniles of this latter species, daytime

behaviour was described by Burel et al. (1997) and Huse (1994). According to the results

of the present study, gilthead sea bream showed greater feed demand in the evening,

although a certain amount of demand remained throughout the day with a slight surge at

dawn. These results, showing two daily peaks, partly coincide with those of Anthouard et

al. (1996) for this species, who found another peak at midday, although this third peak may

have resulted from a higher experimental temperature (18–24 jC) than that of the present

study (14.5–16.5 jC). In fact, during the second experiment, a secondary rhythm with a

period of 12 h at a temperature 1.5 jC higher was found, while in the first experiment that

secondary peak was not detected, probably because the temperature was maintained

between 14 and 15 jC. One noticeable feature was the considerable increase in demand-

feeding in all the tanks during the 2 days of the full moon, although we cannot confirm

these results, since the experiment only lasted 1 month. Lunar cycles have previously been

shown to affect juvenile European sea bass growth rhythms (Planes, 1993). Growth in

length, body weight and food intake have also been reported to be influenced by lunar or

semi-lunar rhythms in several species, such as coho salmon Oncorrhynchus kisutch,

rainbow trout O. mykiss, and Arctic char Salvelinus alpinus (for a review, see Madrid et al.,

2001). However, no reference concerning the influence of lunar cycles for gilthead sea

bream or other Mediterranean species is available, nor for the relationship between

moonlight and feeding behaviour. Therefore, we cannot state whether this effect is a

lunar rhythm or an effect due to the greater clarity of the full-moon nights. This effect

could not be tested for full-moon nights with a cloudy sky. Juvenile silver barb (Haroon

and Pittman, 1997) reportedly feed more actively on full-moon nights. Since gilthead sea

bream were housed in tanks in the present study, the increased intake appears to be due

rather to moonlight than to the tides caused by the moon’s gravitational pull. At the end of

the experiment, demand feeding increased further, coinciding with the new moon,

although this may have been due to a water temperature increase of approximately 1.5

jC during the last 4 days of the assay. The effect of temperature on fish food intake is well

known (Kestemont and Baras, 2001). Nonetheless, a combined effect of the new moon

and temperature cannot be dismissed although no reference new moon effects on food

intake is available in the literature.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103 99

Nutritional utilisation of the diet appears to be influenced by the time of feeding, since

the parameters of nutritional efficiency improved when the fish were allowed to feed at

will. In this way, they regulate their intake to meet energy requirements (Kaushik and

Medale, 1994) according to their natural feeding rhythms (Boujard and Leatherland,

1992a).

The FCR and PER results showed that self-fed fish performed significantly better than

animals hand fed at set hours. This could be due to the feeding methods or feeding time. In

fact, it has been demonstrated that the same diet offered at different times of the day is

assimilated differently (Madrid, 1994). Different growth rates have been recorded in

H. longifilis (Kerdchuen and Legendre, 1991) and H. fossilis (Sundaranaj et al., 1982)

when fed at different periods of the day. In the present study, FCR and PER improved by

some 60% when fish were under a self-feeding regimen, this increase becoming significant

when FCR is considered. In rainbow trout, a higher feed conversion efficiency for growth

has been found when the fish are fed in phase with their natural feeding activity, a result

apparently related to increased protein synthesis and retention (Bolliet et al., 2001).

Better diet utilisation when fed at certain times of the day could be explained by

coincidence with natural rhythms of secretion, activation or synthesis of digestive and/or

metabolic enzymes. The intestinal protease activity in European sea bass has been

described as having a nychthemeral rhythm with increased activity at night, regardless

of the frequency of feeding or deprivation (Martınez-Bebia et al., 1995). Boujard and

Leatherland (1992b) have detected changes in the concentration of several hormonal

metabolites in trout during the day, this being associated with feeding and/or photoperiod.

The adaptation of the feeding schedule to these metabolic, digestive and other rhythmic

processes related to the utilisation of nutrients (i.e. hormone release) appears to improve

FCR and PER indices when the fish are fed according to their natural rhythm of ingestion.

Diet chemical properties, including nutrient availability, are factors expected to affect

voluntary feed intake. It has been demonstrated that fish can discriminate between diets

differing in their macro- and micronutrient composition (Aranda et al., 2001). Sea bass can

also discriminate between diets with different methionine concentrations (Hidalgo et al.,

1988). Therefore, it is possible for fish to discriminate and adjust feed intake when facing

unbalanced amino acid diets.

According to the results, both soy-protein experimental diets (either methionine-

deficient or free-methionine-supplemented) resulted in a similar behaviour to that for

controls, concentrating feeding demands during the evening (although feeding frequency

changed with respect to controls). In the case of the diet supplemented with free

methionine, the feeding period was longer, though the total daily number of requests

resembled that of the control diet and lower than the demand recorded for the methionine-

deficient diet; however, in no case were the differences significant. The lower feed intake

of diets containing protein sources other than fishmeal might be attributed to essential

amino acid deficiencies, as shown for arginine (Kim et al., 1992a; Walton et al., 1986),

leucine (Choo et al., 1991) and methionine (Kim et al., 1992b). Essential amino acid

deficiency has been widely demonstrated to reduce feed intake, where normal intake

values are reached only when the amino acid concentration in the diet meets the

requirements of the fish, as reviewed by De la Higuera (2001). On the contrary, when

the European eel, Anguilla anguilla, was fed on diets containing sunflower protein

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103100

supplemented with needed amino acids, feed intake and utilisation reached values similar

to those for fish fed on a fishmeal diet (Garcıa-Gallego et al., 1998; De la Higuera et al.,

1999).

Soy meal is the vegetable-protein source most frequently used in practical diets to

replace fishmeal. Nevertheless, its methionine deficiency for most fish species limits

fishmeal-protein substitution. Attempts to correct this deficiency by adding free methio-

nine to the diet have improved growth rates when compared to non-supplemented diets,

though growth rates remained lower than those for a control diet (Drabrowski and Kozak,

1979; Ketola, 1982; Murai et al., 1987). Gilthead sea bream responds to free methionine

supplements in the same way as other fish species: improving the nutritional parameters

with respect to unsupplemented soy meal but failing to reach fishmeal values (Sierra et al.,

1993).

Inefficiency of essential amino acid supplementation might be attributed to an

imbalance of postprandial amino acid availability due to the early absorption of dietary

free-supplemented amino acids. In this sense, in the European sea bass (Thebault, 1985),

trout (Cowey and Walton, 1988; Garzon, 1995), carp (Garzon, 1995) and gilthead sea

bream (Sierra, 1995) a postprandial imbalance was observed between protein and free-

supplemented amino acids. In the case of gilthead sea bream fed on a soy-protein-based

diet supplemented with free methionine, the maximum concentration of plasma methio-

nine was detected 2 h after ingestion, while protein amino acids reached maximum

concentrations 8 h after feed intake (Sierra, 1995). Nevertheless, when methionine

assimilation is delayed by microencapsulation with a protein coat, protein synthesis and

growth improve significantly, reaching values close to those for a fishmeal-control diet

(Sierra, 1995).

Results of the abovementioned works could be related to the lengthening described

above for the intake phase, to be interpreted as an adjustment of feeding behaviour to

correct postprandial imbalances after dietary amino acid uptake. The fish respond to an

earlier uptake of methionine by increasing their feeding-demand period, starting to feed 2

h earlier than the control. This change of feeding pattern would allow methionine from

most recent ingestion to coincide with protein amino acids from previous ingestion, so that

a sufficiently homogeneous amino acid pool is obtained at the protein-synthesis sites. Due

to the short duration of the second experiment (Table 1), we were unable to gather basic

data for growth rate and feed utilisation. Even though further long-lasting experiments are

needed, our preliminary results show that an appropriate feeding strategy might improve

dietary nutrient utilisation, self-feeding constituting an option for improving the utilisation

of alternative protein sources, especially when supplemented with the essential amino

acid(s) in which these sources are deficient.

Acknowledgements

The Direccion General de Investigacion, Ministerio de Ciencia y Tecnologıa, Spain,

Projects NMAR96-1881 and REN2000-1740-C05-03 RIES, supported this research. The

authors are grateful to Professor M. Jobling for his kindly help on the reading of this paper

and discussion of the results.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103 101

References

Alanara, A., 1996. The use of self-feeders in rainbow trout (Oncorhynchus mykiss) production. Aquaculture 145,

1–20.

Anthouard, M., Kentouri, M., Divanach, P., 1993. An analysis of feeding activities of sea bass (Dicentrarchus

labrax, Moronidae), raised under different light conditions. Ichtyophysiologica acta 16, 59A–63A.

Anthouard, M., Dermoncourt, E., Divanach, P., Pastatis, M., Kentouri, M., 1996. Les rythmes d’activite trophique

chez la dourade (Sparus aurata) en situation de libre acces alimentaire total ou temporellement limite.

Ichtyophysiologica acta 19, 91–113.

AOAC, 1984. Official Methods of Analysis. AOAC (Association of Official Analytical Chemists), Arlington,

VA (1141).

Aranda, A., Sanchez-Vazquez, F.J., Madrid, J.A., 2001. Effect of short-term fasting on macronutrient self-

selection in sea bass. Physiology and Behavior 73, 105–109.

Aschoff, J., 1992. Handbook of Behavioural Neurobiology. Biological Rhythms, vol. 4. Plenum, New York.

Bath, M., 1982. Spectral Analysis in Geophysics. Elsevier, Amsterdam.

Bolliet, V., Azzaydi, M., Boujard, T., 2001. Effect of feeding time on feed intake and growth. In: Houlihan, D.,

Boujard, T., Jobling, M. (Eds.), Food Intake in Fish. Blackwell Science—COST Action 827, Oxford, pp.

232–249.

Boujard, T., Leatherland, F.J., 1992a. Circadian rhythms and feeding time in fishes. Environmental Biology of

Fishes 35, 109–131.

Boujard, T., Leatherland, J.F., 1992b. Circadian pattern of hepatosomatic index, liver glycogen and lipids content,

plasma non-esterified fatty acid, glucose, T3, T4, growth hormone and cortisol concentrations in Oncorhyn-

chus mykiss, held under different photoperiod regimes and using demand-feeders. Fish Physiology and

Biochemistry 10, 111–122.

Boujard, T., Luquet, J., 1996. Rythmes alimentaires et alimentation chez les Siluriodei. In: Lengedre, M., Proteau,

J.P. (Eds.), Aquatic Living Resources, vol. 9, pp. 113–120.

Brett, J.R., Groves, T.D.D., 1979. Physiological energetics. In: Hoar, W.S., Randall, D.J., Brett, J.R. (Eds.), Fish

Physiology, vol. 8. Academic Press, New York, pp. 280–351.

Brigham, E.O., 1988. The Fast Fourier Transform and Its Applications. Prentice Hall, New Jersey.

Burel, C., Robin, J., Boujard, T., 1997. Can turbot, Psetta maxima, be fed with self-feeders? Aquatic Living

Resources 10, 381–384.

Childers, D.G., 1989. Biomedical Signal Processing. Prentice Hall, New Jersey.

Choo, P., Smith, T.K., Cho, C.Y., Ferguson, H.W., 1991. Dietary excesses of leucyne influence growth and body

composition of rainbow trout. Journal of Nutrition 121, 1932–1939.

Corchete, V., Badal, J., Payo, G., Seron, F.J., 1989. Filtrado de ondas sısmicas dispersadas. Revista de Geo-

fisica 45, 39–58.

Corchete, V., Badal, J., Seron, F.J., Soria, A., 1995. Tomographic images of the Iberian subcrustal lithosphere and

asthenosphere. Journal of Geophysical Research 100, 24133–24146.

Cowey, C.B., 1975. Aspect of protein utilisation by fish. Proceedings of the Nutrition Society 34, 57–63.

Cowey, C.B., Walton, M.J., 1988. Studies on the uptake of (14C) amino acids derived from both dietary (14C)

protein and dietary (14C) amino acids by rainbow trout, Salmo gairdneri Richardson. Journal of Fish Biology

33, 293–305.

De la Higuera, M., 2001. Effects of nutritional factors and feed characteristics on feed intake. In: Houlihan,

D., Boujard, T., Jobling, M. (Eds.), Food Intake in Fish. Blackwell Science—COST Action 827, Oxford,

pp. 250–268.

De la Higuera, M., Garzon, A., Hidalgo, M.C., Peragon, J., Cardenete, G., Lupianez, J.A., 1998. Influence of

temperature and dietary-protein supplementation either with free or coated lysine on the fractional protein-

turnover rates in the white muscle of carp. Fish Physiology and Biochemistry 18, 85–95.

De la Higuera, M., Akharbach, H., Hidalgo, M.C., Lupianez, J.A., Garcıa-Gallego, M., 1999. Liver and white

muscle protein turnover rates in the European eel (Anguilla rostrata): effects of dietary protein quality.

Aquaculture 179, 203–216.

Drabrowski, K., Kozak, B., 1979. The use of fish meal and soybean meal as a protein source in the diet of grass

carp fry. Aquaculture 18, 107–114.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103102

Garcıa-Gallego, M., Akharbach, H., De la Higuera, M., 1998. Use of protein sources alternative to fish meal

in diets with amino acids supplementation for the European eel (Anguilla rostrata). Animal Science 66,

285–292.

Garzon, A., 1995. Influencia de la encapsulacion de la lisina sobre la utilizacion nutritiva de la proteına de gluten

y el recambio proteico tisular en la trucha (O. mykiss) y en la carpa (C. carpio). Doctoral thesis. Universidad

de Granada, Spain.

Haroon, A.K.Y., Pittman, K.A., 1997. Diel feeding pattern and ratio of two sizes of silver barb, Puntius

gonionotus Bleeker, in a nursery pond and ricefield. Aquaculture Research 28, 847–858.

Hidalgo, F., Kentouri, M., Divanach, P., 1988. Sur l’utilization du self-feeder comme outil d’reuve nutritionnelle

du loup, Dicentrarchus labrax. Resultats preliminaires avec la methionine. Aquaculture 68, 177–190.

Huse, I., 1994. Feeding at different illumination levels in larvae of three marine teleost species: Cod (Gadus

morhua L.), plaice (Pleuronectes platessa L.) and turbot (Scophthalmus maximus L.). Aquaculture Fish

Management 25, 687–695.

Kaushik, S.J., Medale, F., 1994. Energy requirements, utilization and dietary supply to salmonids. Aquaculture

124, 81–87.

Kerdchuen, N., Legendre, M., 1991. Influence de la frequence et de la periode de nourrissage sur la croissance et

efficacite alimentaire d’un Silure africain, Heterobranchus longifilis (Teleostei Clariidae). Aquatic Living

Resources 40, 241–248.

Kestemont, P., Baras, E., 2001. Environmental factors and feed intake: mechanisms and interactions. In: Hou-

lihan, D., Boujard, T., Jobling, M. (Eds.), Food Intake in Fish. Blackwell Science—COST Action 827,

Oxford, pp. 131–156.

Ketola, H.G., 1982. Amino acid nutrition of fish: requirements and supplementation of diets. Comparative

Biochemistry and Physiology 73b, 17–24.

Kim, K., Kayes, T.B., Amundson, C.H., 1992a. Requirements for lysine and arginine by rainbow trout (Onco-

rhynchus mykiss). Aquaculture 106, 333–344.

Kim, K., Kayes, T.B., Amundson, C.H., 1992b. Requirements for sulfuraminoacid and utilization of D-methio-

nine by rainbow trout (Oncorhynchus mykiss). Aquaculture 101, 95–103.

Madrid, J.A., 1994. L’alimentation a la demande et les ritmes endogenes. Aqua Revue, 52–53.

Madrid, J.A., Boujard, T., Sanchez-Vazquez, F.J., 2001. Feeding rhythms. In: Houlihan, D., Boujard, T., Jobling,

M. (Eds.), Food Intake in Fish. Blackwell Science—COST Action 827, Oxford, pp. 189–216.

Martınez-Bebia, M., Suarez, M.D., Sanchez, M.J., Sanchez, F.J., Zamora, S., Madrid, J.A., 1995. Efecto de la

alimentacion sobre los ritmos de enzimas digestivos de lubinaDicentrarchus labrax. In: Castello y, F., Calderer,

A. (Eds.), Actas V Cong. Nac. Acuicultura, 10–13 de Mayo de 1995. Pub. Univ. Barcelona, pp. 558–563.

Muller, K., 1978. The flexibility of the circadian system of fish at different latitudes. In: Thorpe, J.E. (Ed.),

Rhythmic Activity of Fishes. Academic Press, London, pp. 91–104.

Murai, T., 1992. Protein nutrition of rainbow trout. Aquaculture 100, 191–207.

Murai, T., Ogata, P., Kosutarak, P., Arai, S., 1987. Effects of methanol treatment and amino acid supplemen-

tation on utilization of soy flour by chum salmon fingerlings. Bulletin National Resources Inst. Aquaculture

12, 37–41.

Papoulis, A., 1962. The Fourier Integral and Its Applications. McGraw-Hill, New York.

Planes, S., 1993. Analysis of growth rhythm on the larvae and juvenile of sea bass Dicentrarchus labrax. Life

Sciences 316, 197–203.

Reigh, R.C., Ellis, S.C., 1992. Effects of dietary soybean meal and fish protein ratios on growth and body

composition of red drum (Scianops ocellatus) fed isonitrogenous diets. Aquaculture 104, 279–292.

Robaina, L.E., 1995. Utilizacion digestiva de fuentes proteicas alternativas a la harina de pescado en dietas de

engorde de dorada (Sparus aurata). Doctoral thesis. Universidad de Las Palmas de Gran Canaria. Spain.

Rubio, V.C., Sanchez-Vazquez, F.J., Madrid, J.A., 2001. Macronutrient self-selections in European sea bass using

microencapsulated nutrients. Fourth and Final Workshop of COST 827 Action on Voluntary Food Intake in

Fish, Reykjavik Iceland. http://www.st-pee.inra.fr/stpee/collab/cost/cost.htm.

Sanchez-Muros, M.J., Garcıa Rejon, L., Garcıa Salguero, L., de la Higuera, M., Lupianez, J.A., 1998. Long-term

nutritional effects on the primary liver and kidney metabolism in rainbow trout. Adaptive response to

starvation and a high-protein, carbohydrate-free diets on glutamate dehydrogenase and alanine amino trans-

ferase kinetics. Biochemistry and Cell Biology 30, 55–63.

M.J. Sanchez-Muros et al. / Aquaculture 224 (2003) 89–103 103

Sanchez-Vazquez, F.J., Martinez, M., Zamora, S., Madrid, J.A., 1994. Desing and performance of an accurate

demand feeder for the study of feeding behaviour in sea bass Dicentrarchus labrax L. Physiology and

Behavior 56, 4789–4794.

Sanchez-Vazquez, F.J., Madrid, J.A., Zamora, S., 1995. Circadian rhythms of feeding activity in sea bass,

Dicentrarchus labrax L.: dual phasing capacity of diel demand-feeding pattern. Journal of Biological

Rhythms 10, 256–266.

Sanchez-Vazquez, F.J., Yamamoto, T., Akiyama, J.A., Madrid, J.A., Tabata, M., 1998a. Selection of macro-

nutrients by goldfish operating self-feeders. Physiology and Behavior 65, 211–218.

Sanchez-Vazquez, F.J., Azzaydi, M., Martinez, F.J., Zamora, S., Madrid, J.A., 1998b. Annual rhythms of de-

mand-feeding activity in sea bass: evidence of a seasonal phases inversion of the dial feeding pattern.

Chronobiology International 15 (6), 607–622.

Sanchez-Vazquez, F.J., Yamamoto, T., Akiyama, J.A., Madrid, J.A., Tabata, M., 1999. Macronutrients self-

selection through demand-feeders in rainbow trout. Physiology and Behavior 66, 45–51.

Shiau, S.Y., Chuang, J.L., Sun, C.L., 1987. Inclusion of soybean meal in tilapia (Oreochromis niloticus�O.

aureus) reared in seawater. Aquaculture 81, 119–127.

Sierra, M.A., 1995. La encapsulacion como estrategia para establecer las necesidades de metionina y la suple-

mentacion de proteına de soja. Consecuencias sobre el recambio proteico tisular y el crecimiento de la dorada

Sparus aurata. Doctoral thesis. Universidad de Granada. Spain.

Sierra, M.A., Hidalgo, F., De la Higuera, M., 1993. Methionine microencapsulation as a way of supplementating

dietary protein for gilthead sea bream. Int. Conf. World Aquaculture’93, Torremolinos, Espana. European

Aquaculture Society, Special Publication, Torremolinos, Spain, vol. 19, p. 468.

Sundaranaj, B.I., Nath, P., Halberg, F., 1982. Circadian meal timing in relation to lighting schedule optimizes

catfish. Journal of Nutrition 112, 1085–1097.

Thebault, H., 1985. Plasma essential amino acid changes in sea bass (Dicentrarchus labrax) after feeding diets

deficient and supplemented in L-methionine. Comparative Biochemistry and Physiology 82A, 233–237.

Viola, S., Mokady, Y., Rappaport, U., Arieli, Y., 1982. Partial and complete replacement of fishmeal by soybean

meal in feed for intensive culture of carp. Aquaculture 26, 223–236.

Walton, M.J., Cowey, C.B., Coloso, R.M., Adron, J.W., 1986. Dietary requirements of rainbow trout for trypto-

pha, lysine and arginine determined by growth and biochemical measurement. Fish Physiology and Bio-

chemistry 2, 161–169.