Sperm was collected from testes and from three or four parts of the deferent ducts of 4 to 6 winter-spawning male rainbow trout, about every two weeks over a three month period (January toMarch). Milt volume, spermatocrit, sperm motility, 11-ketotestosterone (11-KT) and 17α-hydroxy,20β dihydroprogesterone (17,20P) levels and ionic composition were determined. Steroids and ions were also measured in blood plasma. The K+ concentrations and steroid levels in the seminal fluid from each part of the ducts did not vary significantly over the period of study. In two out of the three parts of the ducts, Na+ (parts 1 and III) Ca2+ (parts I and III) and Mg2+ (parts I and II) levels were significantly higher at the end of March than two months earlier. Motility of testicular sperm peaked (42%) in mid-February while that of sperm from ducts usually remained stable (80 to 100%) throughout the three months. Seminal fluids and blood plasma were confirmed to be isotonic (300 mmol/kg). Levels of K+, Na+, Ca2+, 11-KT and 17.20P in the plasma and seminal fluid differed significantly (respectively 0.3 vs 34 mmol, 134 vs 116 mmol, 3.4 vs 2.0 mmol, 87 vs 17 to 39 ng/ml, 11 vs 3 to 13 ng/ml), with the Na+ and steroid levels being positively correlated in both liquids. Osmolality, Cl-, Na+, K+, Mg2+ and Ca2+ concentrationsremained constant along the genital tract. In contrast, spermatocrit and 11-KT and 17.20P levels decreased significantly and progressively from testis to the posterior or middle part of the ducts (respectively 97 to 49%, 39 to 17 ng/ml, 13 to 3 ng/ml) while sperm motility (percentage of motile cells and duration of motility) increased (74 to 96% and 41 to 51 s). No significant correlation was found between sperm motility and the studied physicochemical parameters of the seminal fluid.

Juveniles of the common dentex (Dentex dentex) (2.4 g mean weight), were grown over a 6 week period, fed exclusively on a commercial sea bream diet or an isocaloric moist pellet diet, which was produced on site. Ambient temperature (24.4-26.2 °C) and oxygen content (4.8-5.9 mg/L), as well as the relatively low stocking density (max. 2.62 kg/m3) provided favourable rearing conditions. Survival (63.8% vs 51.6%), specific growth rates (6.1% vs 4.5%) and food conversion efficiency (115.8% vs 76.3%) were considerably elevated in the group fed on moist pellets. The growth performance parameters recorded are considered to be favourable for aquaculture requirements. Agonistic behaviour (particular biting of the tail) was found to be the main cause of the continuous mortality that occurred in both feed groups (78% and 64.8% of total mortality in the dry- and moist-pellet group, respectively). These biting attacks targeted the smallest individuals in a tank, whose mean individual weight made up 49-62% (with dry pellet diet) and 28.9-46.6% (with moist pellet diet) of the respective mean weights. The results indicate a strong relationship between nutrition, size variation and agonistic behaviour in the rearing of cornmon dentex juveniles.

Two female barbel Rarbus barbus (L.) were tagged with implantable radio transmitters and tracked in the Ourthe River (Southern Belgium) from April 12th to July 7th, 1989 to study their movements and activity rhythms during the spring-summer period. Both barbels defined a 1.6 km home range and occupied four residence areas. Greatest average daily movements within the home range limits occurred during the prespawning and spawning periods and were associated with variations of mean daily water temperature and level. Barbel 2 made a 9.6 km excursion from its spring home range and homed a week later in its residence area. Barbels displayed dusk and dawn activity patterns, character- ized by the occupation of riffle areas. The activity timing was found to be dependent on water temperature while the size of daily activity area was more influenced by river flow conditions.

A tracer (51Cr-EDTA) study was undertaken with juvenile (20 g) freshwater rainbow trout, Oncorhynchus mykiss, using an experimental design that minimized stress effects during feeding and drinking trials. A calculation procedure was developed, where feed intake (pellet number), tracer intake (mL), water in stomach contents (g) and drinking rate (mL·kg–1·h–1) are essential to discriminate between the major stomach water sources prandially and postprandially: water bound in food; initial water absorption of pellets; prandial water intake; postprandial water intake and endogenous stomach secretion. We put forward the hypothesis that intake of dry food with a minor water content (10 %) may impose a demand for water to moisture the feed up to the level in natural feed (75 %) as preparation for gastric emptying, whereafter food is ready to pass from the stomach through the pyloric sphincter. Moisture content of pellets increased from 9.4 to 24.9 % in the pre-meal period. The pellets were ingested with 4.0 to 19.3 μL water per pellet, reflecting high inter-individual variation. Prandially moisture content rose to 52 % and further increased to 56 % in the delay period. Moisture content was ca. 65 %, when pellets began to disintegrate and move through the sphincter in accordance with the hypothesis. Stomach secretion contributed 34–44 % of the stomach water and ingested water 25–35 %. The sampling and calculation procedure gave convincing evidence for the detailed stomach water budget and this individual approach can be very useful in comparisons of artificial and natural diets.

Levels of nickel and vanadium, trace metals found in high concentration in the oil spilled from the “Erika” tanker off Brittany (France) in December 1999 (41 ± 1 and 87 ± 3 mg kg−1 respectively), have been monitored over a one-year period following this accidental event in several benthic invertebrates. Ni and V in mussels (Mytilus edulis) and oysters (Crassostrea gigas) soft tissues were analysed twice a month by graphite furnace atomic absorption spectrometry and a scallop (Pecten maximus) shell daily growth bands were sampled every three bands by laser ablation and analysed by ICP-MS for these two contaminants. Survey data for mollusc tissues were compared with reference data arising from the national monitoring database. Ni concentrations in dried tissues, comparable with reference data (1.8 ± 0.9 µg g−1 in mussels and 1.2 ± 0.5 µg g−1 in oysters) do not show any additional input during the whole period. Conversely, a sharp increase in vanadium concentrations (up to 4.6 and 3.2 µg g−1 for mussels and oysters respectively) is observed around May 2000, i.e. 5 months after the wreck, compared with the monitoring data (1.4 ± 0.6 µg g−1 in mussels and 1.3 ± 0.6 µg g−1 in oysters). Furthermore, no Ni peak is detectable in the time profiles of scallop shell growth bands where a V peak is observed also in May 2000. This study shows that although a “mussel-watch-type” network, based on the monitoring of vanadium concentration in mollusc tissues, is able to identify contamination due to oil spills, similar information might be obtained a posteriori by analysing daily growth bands of scallop shells.

We describe daily allocation patterns of fishing effort (hookah diving) of the artisanal fleet in San Felipe, Yucatan (Mexico), using catch, fishing effort, catch per unit of effort, variable costs, quasi rent and distance from port to four fishing grounds as performance variables. Two vessel categories were defined by the presence/absence of a LOng RAnge Navigation (LORAN) system. Hookah divers caught a daily average of four species in 95% of the trips, thus precluding the quantification of effective fishing effort allocated by species. Highest catches and quasi rent were mainly obtained on Thursdays and Fridays, coinciding with the highest catches of Octopus maya and spiny lobster (making up 50% of the total economic revenue) on Fridays in 3 of the 5 weeks analysed. An upper ceiling of daily catch per diver occurred close to the weekend, suggesting a limited capacity in terms of handling time and diving hours, and also a catch level that fulfills daily economic expectations. Both LORAN and non-LORAN vessels preferred to work the nearest ground to port, in spite of higher yield and economic rent from more distant grounds. Nonlinear modelling of fishing effort allocation showed distance from port as the key decision factor. Generalized linear modelling (GLM) revealed significant effects of vessel type and fishing ground, with LORAN vessels having significantly higher catch rates and tending to allocate more effort to distant grounds than vessels without LORAN. GLM performed by species corroborated that Friday was the most productive day concerning spiny lobster and octopus. This day effect in the most valued species also suggests that pressure for higher economic benefits occurs before diminishing (Saturday) or ceasing (Sundays) fishing activities.

Signal crayfish (Pacifastacus leniusculus) originating from Lake Tahoe, California, were introduced into Karisjärvi, a small (11 ha) lake in central Finland (61° 58'N, 25° 32'E), in 1968 and 1969. Since then, the population has been monitored regularly by trap catches. The stocked signals (3–5 years old) were caught until 1973, by which time they had reached ages of 7–9 years. Catches have increased slowly since the early 1980s, the peak occurring in 1991 at 2.0 individuals per trap per night. The population size was estimated as 60 trappable specimens in 1974, 95 in 1981 and 420 in 1988. The mean density of adult population in suitable biotopes was low, 0.07 specimens per m–2, or 0.3 per shore metre, in 1988. The slow development of the P. leniusculus population has been attributed to environmental factors, mainly the limited area of good crayfish habitat. More than 90 % of all signal crayfish were caught in one-third of the shore area, i.e. in steeply sloping lake beds suitable for burrowing or rich in shelters such as rocks and submerged trees. The signals avoided flat, soft bottoms. Fifty per cent of the female P. leniusculus matured at 90 mm TL (smallest 64 mm), i.e. half of the females entered the breeding population in the autumn of their fourth year. The mean size of newly hatched (stage 2) juveniles was 9.7 mm TL and of one-summer olds 30.3 mm. The largest specimen trapped measured 159 mm. The signal crayfish imported into Finland and stocked were infected with crayfish plague (Aphanomyces astaci), but no mortality has been recorded. Two Branchiobdellidae (Annelida, oligochaeta) epibiont species new to Finland were imported from North America with P. leniusculus. The continuous occurrence of these commensals in the signal crayfish population indicates that they have adapted to Finnish conditions.

The size of pelagic fish schools depends on several parameters related to internal factors such as species, number of fish, fish swimming speed and physiological status and to external factors, such as hydrological factors and presence of predators. In order to better understand these relations, results coming from echosounder and multibeam sonar databases are analysed. Field data are collected during four acoustic surveys in the Mediterranean Sea in two different areas (Catalan and Adriatic Seas). The analysis shows differences between the two areas regarding size and position in the water column: schools are deeper and their mean size is lower in the Catalan Sea in comparison with Adriatic Sea. The differences in size of schools are mainly related to differences in school length. Moreover, the elongation of schools seen with the sonar is greater than one and half higher in the Adriatic Sea than in the Catalan Sea, whereas one would expect similar values for the two areas. The results are discussed in terms of environmental influence, avoidance reaction and acoustic capabilities of both tools. A hypothesis is proposed: the variation of school length and consecutively the variation of the correlated dimensions is first related to the strength of the avoidance reaction in front of the vessel and this effect can be reinforced depending on the environmental conditions. The model takes into account the effect of the boat, the vertical constraints undergone by the schools, and the internal requirements of the schools, such as the necessity for fish to keep visual contacts and the cohesion of the group.

An experimental small scale partial recirculating system for rainbow trout was assembled. The system components were two 1.3-m3 fish tanks with sloping bottoms, each connected to a sedimentation column and containing 50 kg rainbow trout biomass, an anaerobic up-flow digester (total volume 0.424 m3, available volume 0.382 m3) connected to the funnel shaped bottom of the sedimentation column by means of a peristaltic pump, an aerobic submerged plug-flow filter (total volume 1 m3; filled with 0.83 m3 plastic rings with a specific surface of 194 m2·m−3) and a submerged pump. Aeration was provided through porous stones. The anaerobic digester was filled with 35 mm cubes of expanded polyurethane foam (25 pores·cm−2, specific surface 1.375 m2·m−3, filtering volume 0.291 m3) and kept at a temperature of 24−25 °C using an electric heater. The gas chamber at the top of the anaerobic digester was connected to a gas meter and to an infrared continuous gas analyser. Measures on system performance with a recirculation rate of 60 % were done following three feeding levels (1, 1.5 and 2 % live weight). At the highest feeding rate, 2.8 L of faecal sludge collected from the trout tanks were pumped every four hours in the anaerobic digester. Slurry characteristic were: total N 0.197 g·L−1, TAN 0.014 g·L−1, volatile solids (VS) 16.91 g·L−1, suspended solids (SS) 21.39 g·L−1 and pH 6.9. Biogas production was 144 L·d−1 (mean value) with a methane content higher than 80 %. Methane volumetric production was 0.3 m3·m−3·d−1 and methane daily yield was 0.4 and 0.32 m3·kg−1 VS and SS respectively. After passing through the anaerobic digester, effluents were characterized by a total N content of 0.243 g·L−1, TAN 0.222 g·L−1, VS 1.1 g·L−1, SS 1.32 g·L−1 and pH 6.8. The anaerobic digester was able to significantly reduce VS and SS content of wastewater and the zeolite ion-exchange column significantly improved water quality of effluent produced by the digester. The aerobic biofilter significantly reduced the ammonia content of the water leaving the fish tanks.

Seaweed farming is often depicted as a sustainable form of aquaculture, contributing to poverty reduction and financial revenues in producer countries. However, farms may negatively affect seagrasses and associated organisms (e.g. invertebrate macrofauna) with possible effects on the flow of ecosystem goods and services to coastal societies. The present study investigates the influence of a seaweed farm, and the farmed seaweed Eucheuma denticulatum in particular, on fishery catches using a traditional fishing method (“madema” basket traps) in Chwaka bay (Zanzibar, Tanzania). The results suggest that a seaweed farm, compared to a seagrass bed, had no influence on catch per unit effort (no. of individuals per catch, or catch weight) or no. of species per catch, but significantly affected catch composition (i.e. how much that was caught of which species). The two species contributing most to differences between the sites were two economically important species; the herbivorous seagrass rabbit fish Siganus sutor, which was more common in the seaweed site and is known to graze on the farmed algae; and the benthic invertebrate feeder chloral wrasse Cheilinus chlorourus, more common in the seagrass site. Compared to vegetation-free bottoms, however, the catches were 3−7 times higher, and consisted of a different set of species (ANOSIM global R > 0.4). As traps placed close to the seaweeds fished three times more fish than traps placed on sand patches within the seaweed farm, the overall pattern is attributed to the presence of submerged vegetation, whether seagrass or seaweed, probably as shelter and/or food for fish. However, qualitative differences in terms of spatial and temporal dynamics between seagrass beds with and without seaweed farms, in combination with other factors such as institutional arrangements, indicate that seaweed farms cannot substitute seagrass beds as fishing grounds.

The need for a better understanding of the role of biodiversity in the functioning of aquatic ecosystems has been raised recently by several authors. If the question is not entirely new, its formulation is. It is therefore the opportunity to re-examine the pool of data already available in order to discuss some results that may be relevant for that purpose, while some major ecological questions have been identified. The top-down view argues that the effects of fish predation cascade down the trophic chain and are responsible for controlling the state of the entire ecosystem. The influence of fish predation on prey assemblages (prey selection, ontogenic shift) has been fairly well documented by various studies of African fish. They are also challenging subjects for studies of resource partitioning, a well documented example being given by zooplanktivores in Lake Turkana. The principle of cascading trophic interactions has been documented in Lake Victoria, after the introduction of a top predator, Lates niloticus. Another striking example is thatof Lake Nakuru where the introduction of a tilapiine resulted in a substantial increase of bird diversity, by the extension of the food chain to fish-eating birds. This introduced tilapiine may be considered as a keystone species, but this concept is questioned. A possible origin of fish biodiversity is adaptive radiation, that is to say the differentiation of species in morphology and resource use through competition. The cichlid flocks of the African Great Lakes where hundreds of rather similar species coexist, provide one of the best known examples of adaptive radiation. The question of niche overlap and functional redundancy has been raised and various studies have been conducted in the East African Lakes. The role of rare species has also been questioned. They may provide an insurance of ecosystem stability, which means that the more stable ecosystems in terms of key functions are those richest in species. In fact, in many large Nilo-Sudan river basins, there are apparently several groups of species which are able to replace the current assemblage when it experiences new ecological situations. Finally, the question has been raised as to what extent patterns of biodiversity are important for ecosystem productivity. The cornparison of data for four African lakes shows no relationship between fish diversity and fishery production, and speciesrichness does not appear to be a major determinant of basic production trends. At present we have only anecdotal, sometimes controversial, evidence of the role of fish diversity on the functioning of African freshwater ecosystems. We need to improve this knowledge on the importance of biodiversity in order to convince decision-makers that fish biodiversity needs protection.

The aim of this study was to quantify the efficiency of different lengths of beach seine nets for each diel period in freshwater reservoirs. Only fish older than young-of-the-year were considered. Nets of 10, 20 and 50 m length (“in-nets”) were tested in an enclosed area framed by a 200 m long net (block net). The net efficiency estimate was calculated as the ratio of fish catches with the in-net and block net divided by the ratio of their respective areas. The net efficiency estimate was significantly different between day and night catches. At night, the efficiency estimate of nets depended on the size of the fish. The efficiency estimate of a 10 m long net decreased significantly in reverse correlation with fish size. A similar trend was found when using a 20 m long net. The efficiency estimate of a 50 m long net was independent of fish size. The variance in efficiency estimate between samples with a given net length was high, but decreased with longer nets. Of five species tested, only the efficiency estimate for catching bream (Abramis brama) increased significantly with the length of net. The biomass and abundance of larger fish was generally higher at night, although especially short nets exhibited a spuriously high efficiency estimate during the day, probably due to the concentrating (chasing) effect of the hauling ropes. We therefore recommend the use of a 50 m long net, since its nighttime efficiency estimate was about 0.9 in terms of both sampling abundance and biomass. A model relating the efficiency estimate and net length was developed with the data acquired.

This paper compares the species composition of free swimming schools and schools associated with fish aggregating devices (or FADs) observed by scientists in the eastern equatorial Atlantic during the early eighties and late nineties. This comparison shows that in free swimming schools, big changes in the species composition have occurred. The main change is a rarefaction of mixed species free schools (skipjack and small yellowfin or bigeye). This change is probably a real biological one, and possibly a consequence of the large numbers of FADs seeded in the area since 1990. Nowadays, most small tuna living in the equatorial area appear to be concentrated under these drifting FADs instead of in free schools. Further study is recommended in order to evaluate the validity and interpretation of this result and to examine its implications.

The diurnal cycle in density measured during acoustic surveys is difficult to investigate quantitatively from the data set of a single survey because the signal depends mainly on the spatial variability and/or horizontal fish migrations. In order to detect the diurnal cycle in spite of the noise due to spatial variability, several acoustic surveys from different tropical countries were combined, after standardization, into a single data base. A diurnal cycle is observed, with the highest values during the night and the lowest during the day. The transition periods are very short (less than one hour). Different hypotheses relative to fish behaviour or to acoustic biases are discussed. Among these hypotheses the lateral avoidance of schools by day and the diurnal change in patchiness seem predominant. Moreover, owing to the extreme skewness of the density distribution the applicability of the Central Limit Theorem is discussed.

Incidental catch or bycatch represents a significant threat for the conservation of seabird populations. The western Mediterranean is an important fishing area where the Spanish pelagic and semi-pelagic longline fleet targeting swordfish (Xiphias gladius), bluefin tuna (Thunnus thynnus) and albacore (Thunnus alalunga) operates. Bycatch of these fisheries includes several seabird species. Given the importance of conservation of the bycatch species (marine mammals, turtles, sharks and seabirds), an on-board observer program was implemented by the Spanish Oceanographic Institute (IEO); this included collecting data on effort and catch, as well as weight and number of individuals of the main bycatch species. The aim of the present study is to report data on seabird bycatch collected by the on-board observer program of the IEO in the Western Mediterranean. Data on seabird bycatch were collected for the period 2000-2008, throughout the year. Six longline gears targeting large pelagic fish were identified operating in the area of study, but only three had an effect on seabird species. Differences in catch per unit effort (CPUE, birds per 1000 hooks) for each gear, as well as their effect on particular seabird species, are reported in this study. A total of 4 786 466 hooks were monitored, which yielded 182 seabirds belonging to 7 different species. The average CPUE for the studied period was 0.038 birds per 1000 hooks. Cory’s shearwater (Calonectris diomedea) and yellow-legged gull (Larus michahellis) were the species the most highly represented in the catch. In contrast, Balearic shearwater (Puffinus mauretanicus) was only present in the longline fishery targeting albacore (CPUE = 0.005 birds per 1000 hooks). Our results suggest that Cory’s shearwater is the species the most affected by the longline fishery in the Western Mediterranean, probably due to its biological characteristics, and corroborate the well-established downward trend in its population.

In this paper, biochemical and physiological analyses were used to characterize changes associated with mortality event occurred during veliger development of American oyster, Crassostrea virginica. Biochemical analyses included the evaluation of lipid classes, fatty acid composition and total protein content. Larval physiology was evaluated by studying feeding activity, enzymes related to energy metabolism, oxidative stress levels and enzymatic antioxidant defenses. These analyses were complemented by bacterial community analyses as well as by measuring larval oyster performance. We observed that mortality events coincided with (1) strong changes in the surrounding bacterial community; (2) a progressive decrease in feeding activity; (3) higher levels of some lipid classes (free fatty acids, diglycerides, and acetone mobile phospholipids); (4) lower levels of phospholipids and protein; (5) higher contents of non-methylene interrupted dienoic fatty acids (22:2 NMI); (6) a decrease in energy metabolism activity (citrate synthase and cytochrome oxidase activities); (7) a higher oxidative stress (lipid peroxidation level); and (8) an activation of antioxidant defences before mortality (glutathione peroxidase and superoxide dismutase). We hypothesized that mortality emergence was related to higher energy consumption coupled with the progressive decline in feeding activity, lowered energy reserves and a decrease in energy metabolism activity. Thus, the low energy availability limited the efficiency of antioxidant defenses, resulting in a higher oxidative stress.

This work examines the effects of constant long and short days on the hormonal profile of plasma17β-oestradiol (E2) levels and the ovarian vitellogenic events, as well as on the fecundity, time of spawning and egg quality of sea bass. Constant short days first administered in April advanced spawning (by 45 days on average) whereas exposure to constant long days from the same date induced a delay (averaging 38 days) compared to controls reared on an ambient light regime (Lat. 40°N). These light regimes produced shifts in the profiles of plasma 17β-oestradiol and in the percentage of the vitellogenic oocytes according to the spawning time of the respective groups. Relative fecundity of the short day group was similar to controls (257,000 versus 230,000 eggs/kg spawned female, respectively). However, in the long day group, relative fecundity was reduced to one half that of control fish (124,000). Plasma oestradiol showed a bimodal pattern under long days and vitellogenetic and atretic oocytes were present for a longer period (6-7 months compared to 4-5 months in controls). In summary, in this paper we describe, for the first time in the sea bass, the effect of constant photoperiod regimes on the 17β-oestradiol plasma profiles, the timing of spawning and the quality of the eggs.

Beaked redfish inhabits North Atlantic waters in the depth range 100–950 m, over thecontinental shelf, slope and the open ocean. Individuals can live demersal or pelagic, atvarious stages of their life cycle. The geographical distribution of the species extendsto most of the Atlantic waters from Newfoundland and the Labrador basin in the west to theBarents Sea in the east. Monitoring beaked redfish is challenging because of the specieswide geographical distribution and large scale migrations; deep distribution, whichcomplicates trawling and hydroacoustic surveys; difficulties with tagging; and persistentdifficulties in taxonomic identification. These challenges make it a particularlyproblematic species to observe with conventional research methods. We review these keychallenges and provide recommendations for the coordinated observation of Sebastesmentella in the North Atlantic that would best contribute to the assessment andecological research on this species.

The effects of toxic and non-toxic strains of Alexandrium sp. on the initial feeding responses of the Pacific oyster Crassostrea gigas were determined by continuous monitoring of the clearance rate of individual oysters. This was determined fluorometrically by loss of fluorescence of live microalgae introduced as food. Feeding oysters with toxic or non-toxic A. tamarense and toxic A. fundyense resulted in stop/start clearance behaviour. In contrast, control oysters fed a reference microalga, Isochrysis sp., known to support their growth, did not exhibit this behavior. Pacific oysters acclimated to Isochrysis sp., fed mixtures of Alexandrium/Isochrysis, showed further evidence of stop/start clearance. Exposure for periods up to 48 hours with non-toxic A. tamarense did not enhance clearance of either Alexandrium or Isochrysis.

Sibling juvenile vundu Heterobranchus longifilis reared at 27 ± 1 °C under 12L:12D were fed during the day, at night or over the entire 24-h cycle, for periods of 14-15 d. Nighttime feeding gave the highest growth rates, best feed conversion, lowest mortality, fewest losses due to cannibalism and least growth heterogeneity. The trends were seen in fish of all ages and weights (31−101 d and 0.3−30.0 g), but differences were significant only for the smallest fish that were reared at the lowest stocking biomass (< 3.5 g·L−1). The effects of daytime and nighttime feeding were examined in 151-d old (85 g) juveniles stocked at different biomasses (3.4 and 13.6 g·L−1). The growth of vundu was density-dependent, being impaired at low stocking density, but this was substantially mitigated under nighttime feeding. Rearing at high stocking density, feeding at night or in darkness, may act through similar mechanisms that involve changes in the level of aggressive behaviour. It is suggested that nighttime feeding in intensive catfish culture could be replaced by 24-h feeding under conditions of permanent darkness.