Skip to main content
European Commission logo print header

The effect of turbidity and hypoxia on the behaviour of coastal marine fishes

Deliverables

Intertidal mudflats are important nursery grounds for juveniles of many fish species. However, they are being used increasingly to farm bivalve molluscs, which produce large amounts of organic biodeposit, rich in exo-polyssacharides (EPs), overlying the mud. Fish such as sole, Solea solea, hide in this fluff from potential predators, but the energy consumed by respiring the fluff may be high due to its biorheological properties (viscosity). To investigate the potential impact of EPS on the bioenergetics and ventilation of juvenile sole we developed an ichthyoviscometer. We developed this method to measure the rheological properties of fluids and suspensions, including fluff, at scales encompassing those in fish mouth and gill cavity. We found that the rheological behaviour of fluff is close to that of a gel with a yield stress strongly dependent on particulate organic matter concentration ([POM]). This allowed us to model fluff flow through the gill channels in living sole as a function of fish size and [POM], showing that in a 26g sole, fluff would halve flow at a [POM] value of 3.2g/L, and stop it at 3.4g/L.
Habitat selection of juvenile Atlantic cod, Gadus morhua and turbidity. We examined the effects of turbidity on habitat preference of juvenile Atlantic cod in the laboratory. In the first experiment, we compared three turbidity levels of kaolin (2, 7 and 15 beam attenuation m-1). In the second experiment, we looked at the effect of turbidity media (kaolin vs algae). Although cod preferred intermediate kaolin turbidity over clear water, most combinations of turbidity did not influence habitat preference. Similarly, algae did not influence habitat preference by cod. We suggest that factors such as food availability and predation risk have more important effects on habitat preference by juvenile cod than differences in turbidity per se.
The relationship between whole blood-oxygen affinity (P50) and pH-dependent binding (i.e. cooperativity and the Bohr (F) and Root effects) was examined statistically under standardized conditions (10.0C) in four unrelated cold-temperate marine fishes that differ widely in their swimming performance and their expected responses to hypoxia: Cod (Gadus morhua), herring (Clupea harengus), mackerel (Scomber scombrus) and plaice (Pleuronectes platessa). An unexpected difference in blood oxygen affinity was found (herring > plaice > mackerel > cod) and this was independent of both swimming performance and the predicted low O2 response of each species. The ecotype of the four marine species was also unrelated to pH dependent binding because no difference in the Bohr effect was apparent (F varied insignificantly from -0.90 to -1.06) and differences in the magnitude of the cooperative binding reaction was associated only with the presence of the Root effect. Although several reviews propose a generalized link between blood-oxygen affinity and pH-dependent binding, our results advise against overestimating the adaptive functional properties of haemoglobin across unrelated species.
Increasing turbidity in coastal waters in the north Atlantic and adjacent seas has raised concerns about impacts on Atlantic cod (Gadus morhua) using these areas as nurseries. A previous experiment (Meager et al. 2005) has shown that turbidity (up to 28 beam attenuation m-1) had little effect on the foraging rate of juvenile cod. Although this was attributed to cod using chemoreception in conjunction with vision to locate prey, foraging rates may also be maintained by increased activity. Higher activity, however, is energetically costly and may offset benefits from increased foraging return. We examined the effects of turbidity on prey searching and spontaneous activity of juvenile cod in the laboratory, by measuring activity with and without prey cues. Activity of juvenile cod was non-linearly affected by turbidity and was lower at intermediate turbidity, regardless of the presence of prey odour. Activity increased over time when prey odour was present and decreased when absent, but the effects of prey odour were similar across all turbidity levels. Position in the tank was unaffected by turbidity or prey odour. Reduced activity at intermediate turbidities is likely to offset longer prey-search times. At high turbidity (> 17m -1), both longer prey-search times and higher activity indicate that increased energetic costs are likely.
Constraints from bioenergetics have repeatedly been used to explain behaviour, for example how optimal foraging maximizes energy intake or how habitat selection balances the trade-off between survival and growth. In aquatic environments, however, bioenergetics is complicated by the fact that oxygen availability may vary and even reach harmfully low concentrations; in addition, the respiration cost is also higher than in air. In aquatic habitats, the energy budget of an animal is therefore paralleled by an oxygen budget, which may change behavioural trade-offs fundamentally. We developed a state-dependent model for fish bioenergetics that includes the flow of both energy and oxygen through the organism. The oxygen for metabolism comes from aerobic respiration or anaerobic lactate build-up, and increased lactate reduces swimming ability and thus the probability that the individual can escape predators. The environment�s oxygen concentration puts a constraint on maximum aerobic respiration; other constraints are imposed by encounters with predators and prey given the ambient light conditions, flow through the digestive tract, and the capacity to store lactate. Hypoxic water layers are often safer from predators but leave little oxygen for swimming (predator escapement and feeding) and digestion (growth and avoiding starvation). We investigated the emerging trade-offs as individuals have to balance oxygen flow, energy flow, predation risk, and starvation risk in a vertically structured water column. By using models, we show how hypoxic layers may i) decrease mortality by serving as a refuge from predators or ii) increase mortality and reduce growth by increasing exposure to predators. Aquatic environments are typically characterised by strong vertical gradients of light, oxygen content, temperature, prey, and predator densities. In the Oslo fjord, sprat (Sprattus sprattus) actively utilise vertical gradients of light and oxygen saturation to maximise survival during their over wintering period. The deep layers are hypoxic and provide safety because predators are scarce. By remaining in hypoxic habitats, however, sprat build up an oxygen debt that must be repaid at some point. In addition, oxygen-demanding activities such as foraging, digestion, growth, or escape-reactions come at the cost of increased predation risk, since sprat are forced to seek oxygen in water layers where predators are present. We developed predictions on optimal habitat use and foraging activity of fish faced with this trade-off using a state-dependent bioenergetics model accounting for the flow of both oxygen and energy in an organism. Sprat stay in hypoxic layers during the day, and seek oxygen-rich waters during night when encounters with predators are rarer. We also explore interactions between internal physiological constraints and external environmental factors such as prey- and predator density and vertical profiles of hypoxia, light, and temperature. Behavioural strategies depend strongly on the distribution of oxygen in the vertical habitat. This illustrates the need to integrate internal physiology with external ecology to fully understand both individual behavioural and population level responses of fish under environmental stress.
Atlantic cod, Gadus morhua, were exposed to a progressive stepwise decline in water oxygen pressure (19.9, 13.2, 10.5, 8.4, 6.2 and 4.3 kPa PO2). Fish swimming speed and indicators of primary and secondary stress (e.g. blood cortisol and lactate) were measured to assess whether a severe shift in physiological homeostasis (i.e. stress) preceded any change in behaviour or vice versa. Swimming speed increased by 18% when PO2 was reduced rapidly from 19.9 to 13.2kPa and was interpreted as an initial avoidance response. However, swimming speed was reduced by 21% at a moderate level of steady PO2 (8.4 kPa) and continued to drop by 41% under progressively deep hypoxia (4.3 kPa). Elevations in plasma cortisol and blood lactate indicated major physiological stress but only at 4.3kPa, which corresponds to the critical oxygen tension of this species. We propose that the drop in speed during hypoxia aids to offset major stress and is adaptive for the survival of cod in extensive areas of low oxygen.
In hypoxia, gray mullet surface to ventilate well-oxygenated water in contact with air, an adaptive response known as aquatic surface respiration (ASR). Reflex control of ASR and its behavioral modulation by perceived threat of aerial predation and turbid water were studied on mullet in a partly sheltered aquarium with free surface access. Injections of sodium cyanide (NaCN) into either the bloodstream (internal) or ventilatory water stream (external) revealed that ASR, hypoxic bradycardia, and branchial hyperventilation were stimulated by chemoreceptors sensitive to both systemic and water O2 levels. Sight of a model avian predator elicited bradycardia and hypoventilation, a fear response that inhibited reflex hyperventilation following external NaCN. The time lag to initiation of ASR following NaCN increased, but response intensity (number of events, time at the surface) was unchanged. Mullet, however, modified their behaviour to surface under shelter or near the aquarium edges. Turbid water abolished the fear response and effects of the predator on gill ventilation and timing of ASR following external NaCN, presumably because of reduced visibility. However, in turbidity, mullet consistently performed ASR under shelter or near the aquarium edges. These adaptive modulations of ASR behavior would allow mullet to retain advantages of the chemoreflex when threatened by avian predators or when unable to perceive potential threats in turbidity.
Until the end of the first year of life, juvenile plaice (Pleuronectes platessa) occupy shallow water sandy habitats, which are often turbid. The behavioural responses of juvenile plaice to a controlled turbid environment were examined and although a clear pattern of cyclical activity with a tidal periodicity was evident, turbidity had no effect on activity levels or timing of increased activity. Variable amounts of time were spent in both clear and turbid habitats but, given a choice, juvenile plaice were more likely to be found in turbid water. No effect of turbidity was identified on the depth of water occupied by the fish, although more pelagic swimming activity occurred during darkness in clear water, possibly as a means of predator avoidance. The presence of turbid waters in habitats occupied by juvenile plaice may be advantageous, as the reduced visibility decreases the likelihood of detection by visual predators.
The development and application of existing and novel methods for quantitative recording of fish behaviour was an important part of the project. New computer software was developed for recording and quantifying fish activity. The program has been designed as a motion detector that can score movements in a number of user-defined areas of a video image (allows the program to be used to log activity simultaneously in four tanks, or to record time spent by fish in up to four regions or habitats) and record all those frames of a video stream where movement is detected in an AVI file for later verification and/or analysis. In addition, it produces a text file containing a date- and time-stamped log of activity, with one record for every recorded video frame. Areas of the video image can be masked by defining up to 100 mask zones, in order to exclude reflections from the water surface that may falsely trigger the detection software. The application can also be used as a time lapse digital video recorder. MotionGrab is used in conjunction with another program MotionRead to replay recorded AVI files at variable speed, forward or reverse, as well as frame by frame, with the detection zones overlaid. The detection log can also be displayed. Other simple utilities have been added to analyse the data generated by MotionGrab and recorded in log files. MotionRead is also an event logger for analysing video files recorded by MotionGrab and other video files, enabling easier scoring and logging of behaviour data.
SUMMARY Result No. 1: The effect of hypoxia on the anti-predator response of grey mullets. Escape performance was investigated in the golden grey mullet Liza aurata exposed to various levels of oxygen: >85 (i.e. normoxia), 50, 20 and 10 % air saturation. Since the golden grey mullet performed aquatic surface respiration when air saturation approached 15-10 %, escape performance was tested at 10 % air saturation with and without access to the surface (10 % S and 10 % C, respectively). Various locomotor and behavioural variables were measured, such as cumulative distance, maximum swimming speed, acceleration, responsiveness (per cent of responding fish), response latency and directionality. Golden grey mullet showed a decrease in responsiveness when the oxygen level was reduced to 10 % air saturation, whether the surface access was obstructed or not. Hypoxia did not have any effect on the response latency. Cumulative distance and maximum swimming speed over a fixed time were significantly different between normoxic conditions and 10 % C, while no differences were found in maximum acceleration. While the fish escaped away from the stimulus in normoxia, the proportion of away and towards responses was random when the oxygen was <20 % air saturation. This suggested an impairment of the left-right discrimination at the initiation of the response. Hypoxia affected golden grey mullet escape performance mainly through an impairment of responsiveness and directionality, while locomotor performance was affected only in severe hypoxia when the surface was obstructed. The study showed that, in addition to forcing the fish to the surface as shown by previous studies, hypoxia may also increase the vulnerability of grey mullets facing a predator by directly impairing its escape performance.
The swimming speed and physiological (stress) status of individual Atlantic cod, Gadus morhua L., was monitored in response to an acute vs. chronic progressive decline in water oxygen pressure (PO2 = 20.9 to 4.3kPa in 1h at 10oC). The behavioural activity response of cod in acute hypoxia appears to augment rather than offset major physiological stress and, therefore, has minimal adaptive value for this species under these conditions. The behavioural response of cod in acute hypoxia appears to augment (vs. offset) major physiological stress. Therefore, with a noticeable imbalance between O2 supply and demand, the behavioural strategy of cod should be considered suboptimal in acute hypoxia and probably has minimal value for the survival of this species under such conditions. The Atlantic cod is only moderately tolerant of low O2 (Plante et al., 1998; Herbert & Steffensen, In press) and, compared with more hypoxia tolerant species such as Solea solea (Dalla Via et al., 1998), is not likely to survive long in an environment experiencing large and rapid reductions in PO2.
We investigated the effects of turbidity on the foraging behaviour of juvenile Atlantic cod (Gadus morhua) on mysid prey (Praunus neglectus) in the laboratory. The influence of turbidity on vision and chemoreception was examined by measuring reactive distances and search times to visual, chemical, and visual-chemical prey cues over turbidity levels ranging from 0.4 to 17 center dot m(-1) (beam attenuation center dot m(-1)). We also compared foraging rates of juvenile cod on mysids in highly turbid water and clear water under well lit and totally dark conditions. Juvenile cod using chemical cues were able to locate mysids from significantly longer distances than when only visual cues were available. Turbidity did not affect reactive distance to chemical cues, and had only a weak negative effect on reactive distance to visual and visual-chemical cues. Search time was variable, but tended to increase with turbidity. Turbidity did not affect predation rates on free-ranging mysids, but predation rates were significantly lower in dark conditions than in well lit conditions. We suggest that juvenile cod use chemoreception in conjunction with vision (at close ranges) to locate prey in highly turbid water.

Searching for OpenAIRE data...

There was an error trying to search data from OpenAIRE

No results available