Reaction distance (RD) studies EXP SET-UP

Visual feeding in fish larvae EXP SET-UP

Prey Choice EXP SET-UP

Feeding at risk of predation VIDEO

Feeding and competitive interactions: VIDEO

Reproductive behaviour of Gobusculus flavescens : VIDEO

My work on visual feeding can be divided in to three categories: 1) look at the physical and optical properties of prey and the surroundings water, and how this affect the predator’s prey detection; 2) how a larvae’s visual abilities develop in relation to these factors (physical, optical and prey properties); 3) and to quantify and express the mentioned relations theoretically in models of visual feeding.

1) studies done on the planktivore goby Gobiusculus flavescens

2) studies done on herring (Clupea harengus) and cod (Gadus morhua) larvae.

3) a general model of visual feeding and a model of visual feeding in cod and herring larvae.

Study (Utne 1997) showed that increasing light had a positive effect (log-linear) on the gobies reaction distance (RD) until a saturation level, where a further increase had no positive effect on RD. This light saturation response was included in revised model of visual feeding (Aksnes & Utne 1997). Accordingly did an increasing turbidity give a log-linear decrease in the gobies RD (Utne 1997; Utne-Palm 1999). However, these studies showed a discontinuity in the log-linear relation, as an intermediate increase in RD or attack rate (Utne 1997; Utne-Palm 1999) was found at low turbidity levels. Turbidity has both positive and negative effects on prey detection, through its ability to increase contrast between prey and background (by increasing backlighting) concurrently with scattering light and subsequent loss of detection. An intermediate increase in prey detection rates with increasing turbidity can therefore be expected among fish that feed visually. However, the positive effect of turbidity on prey contrast depends on the optical properties of the target and the suspended particles as well as the visual sensitivity of the predator fish to colour and brightness contrast (Utne-Palm 2002). The positive effect of prey contrast on RD, as well as the positive effect of prey mobility, prey size and number is expressed in Utne (1997). (EXP SET-UP)

The positive effect of prey size, - mobility and -number is taken in to an account in or visual model (Aksnes & Utne 1997) as increasing prey size, -mobility or - number all increase the stimulated area on the fishes retina which again gives an increased RD.

A previous study on (Gobiusculus flavescens) indicated that visual pigments of a predator, as well as the turbidity, influence contrast threshold and, therefore, the predators reaction distance to prey. This study indicated a positive relationship between turbidity and contrast threshold. In the models of visual range, contrast threshold of predators is expressed as independent of turbidity. However, turbidity influences inherant contrast by increasing the background irradiance, functionally by increasing contrast between prey and background. This finding suggest that exsiting models largely ignore one of the primary effects of turbidity on prey contrast. Further, contrast threshold was found to decrease with decreasing wavelength (Utne-Palm 1999). This relationship has not been expressed in any visual model.

I have in collaboration with Erling Kåre Stenevik (student at our department) done a behavioural study on visual acuity in cod and herring larvae. We used the method optomotoric response to determine the development of visual angle (Optomotoric response).

Parallel to our behavioural study was a histological study performed on the same larvae group, this to be able to compare both histological and behavioural development of vision in these two species. The microbiologist doing the histological study was my colleague Jon Vidar Helvik, Dept. of microbiology, University of Bergen, and Bo Holmqvist, Johan Forsell and Peter Ekström all from the University of Lund, Sweden. A paper presenting the results of the histological and optomotoric study is under preparation (Helvik et al. in prep)

Optomotoric reflexes are used for position maintenance. (EXP SET-UP)

The optemotoric response apparatus consisted of a horisontal white Plexiglas dish that is rotated at variable speed, by a reversible electric motor. The visual stimulus consisted of cylinders with equal width black and white stripes, which is placed at the top of the rotating white Plexiglas dish. A glass tube, used as animal chamber, is mounted in a stationary position in the centre of the rotating stimulus cylinder. By decreasing the stripes with, one will get a subtended decrease in visual angle when seen from the animal chamber. One larva is introduced to the chamber at the time.

Individual larva is observed from above by camera while the stimulus cylinder is rotated. Each tested larvae is exposed to stimuli of progressively narrower black and white stripes, where minimum separable angle (MSA) is calculated from the smallest stripes the larva respond positively to. A positive response consisted of pursuit of the moving stripes, which is consistent across both swimming speed and direction.

The last two years I have been engaged in a strategic program called: Environmental influence on fish stocks. The main goal of this program is to derive quantitative relationships of how environmental variables affect important processes underlying fish stock dynamics. Within this program my job is to establish the relationships on how feeding in cod (Gadus morhua) and herring (Clupea harengus) larvae is affected by light and wind conditions by the use of experimental laboratory studies. The mentioned effects have been approached through modelling work (Fiksen et al. 1998), and it is therefore important to get the model prediction evaluated through experimental work and field investigations. Within this work I have collaborated with the physical oceanography Jan Erik Stiansen. Stiansen is a PhD student at the Geophysical Institute, University of Bergen. His thesis is on the effect of small-scale turbulence in relation to plankton ecology.

The experimental set-up, which we are using in our study, is almost a copy of the experimental set-up previously used by Brian MacKenzie and Thomas Kiørboe (MacKenzie & Kiørboe 1995. Limnol Oceanogr 40; 1278-89) in their study of effect of turbulence on herring and cod larva’s swimming behaviour (experimental set-up).

The only difference between our set-up and the one used by MacKenzie & Kiørboe is that we don’t use plankton mesh to prevent the larvae from entering the area of the moving grid, as plankton mesh would affects the growth of the typical eddies downwards from the grid. Further, by designing a sluiced opening in the bottom of the experimental tank, where thorough the turbulence measuring equipment could enter the experimental tank, are we able to obtain better spatial and temporal resolution of the turbulence intensity within our tank.

In this set-up we are studying the effect of light, turbulence and turbidity on the feeding behaviour of herring (Clupea harengus) and cod larvae (Gadus morhua) at different larval stages (Utne-Palm & Stiansen, 2002) (Utne-Palm, submitted).

A prey choice study, also performed on the planktivore goby Gobiusculus flavescens (same species used in the RD studies), showed that of the preys characters: visibility, activity and catchability, is visibility (prey size, - mobility or - contrast) what determines prey detection. While activity (and not size or catchability) is what triggers an attack when detected (Utne-Palm 2000).

Habitat choice in a marine goby (Gobiusculus flavescens) and perch (Perca fluviatilis) has been investigated as a function of internal and external factors. Groups of fish differing in size, age and sex (internal factors) were tested for their activity and distribution between habitats differing in food, predation risk and shelter (external factors).

Gobies distributed proportionally to food availability when 10 individuals were offered equal amounts of prey items at to sites (habitats). However, with an increasing difference between food amounts offered at the sites or when the number of gobies and perch were respectively 20 and 18, their distribution no longer mirrored the food distribution. Input match, which is predicted by the Ideal Free Distribution (IFD) was violated (Utne et al 1993). This deviation was probably due to perceptual constraints, given that both increased difference in amount of food offered at the sites and increasing number of foragers may reduce the perceptual threshold below which differences in site profitability can be perceived by the foragers (Utne 1995). In the perch experiment could increased switching or sampling activity also have led to a more uniform distribution (Utne et al 1997). In an ever changing environment, where site profitability and competition are changing constantly, an improvement in perceptual limits is of less importance and increased searching activity is important.

When a predator was introduced in the experiments, riskbalancing behaviour was observed. This means a reduced feeding activity when little food is offered, but that greater predation hazard is accepted as the food level increases (Utne et al 1993; Utne & Aklsnes 1994; Utne et al 1997). This relation was, however, less clear in larger groups (Utne & Aksnes 1994; Utne et al 1997).

Age and size: How much the presence of a predator will influence the spatial distribution of foragers, seems to depend on age and size. My results indicate that juveniles (gobies and perch) expose themselves more to a predator, presumably to achieve a higher food reward, than adults do (Utne & Aksnes 1994).

Furthermore, were larger perch more willing to risk exposure to a predator while foraging than smaller perch of the same age (Utne et al 1997). This indicates size and age differences in the trade off between energetic gain and mortality risk, which lead to different growth rates. The complexity of natural groups of fish would therefore probably prohibit their distribution to mirror the distribution of their prey, as predicted by IFD.

Further have the response of wild caught and laboratory reared (naïve) gobies (Gobiusculus flavescens) to the visual image and or chemical cues of their natural predator, cod (Gadus morhua), been investigated. Although image had the most pronounced effect at the time of introduction, did odour have a stronger effect on the wild caught gobies persistence in avoiding the affected habitat. Introduction of an injured conspecific (alarm substance) had, however, no significant effect on the wild gobies distribution (Utne & Bacchi 1997).

In order to determine whether the anti-predator response of gobies to cod (Gadus morhua) is inherited, naive two-spotted gobies were exposed to visual or chemical cod stimuli, in order to determine whether such stimuli would trigger an innate anti-predator response. When a visual stimulus of a cod was presented at one of the feding sites, they avoided this habitat. The introduction of cod odour at one of the sites did not influence the distribution of the gobies. However, after having experienced a live cod in one of the feeding sites on three consecutive feeding sessions they responded to cod odour with avoidance. The response to cod odour was still significant six months after the cod experience treatment (Utne–Palm 2001).

In this study, pairs of juvenile sticklebacks (Gasterosteus aculeatus) either familiar with each other or pairs unfamiliar with each other met to share a food source. The study showed that sticklebacks were less aggressive towards a partner when sharing a common food source with a familiar conspecific, compared to when sharing a food source with an unfamiliar conspecific.

The results showed that the aggressive behaviour was built up and broken down gradually depending on how long the two competitors had been together or apart (Utne-Palm & Hart 2000).

My Can. Scient student (Mph) Rune Skolbekken investigated the male goby’s investment in its offspring in form of egg guarding and tending (Skolbekken & Utne-Palm, 2002).

The results showed that guarding males feeding activity decreased while cleaning and fanning of eggs increased, as the eggs developed. The males fanning and cleaning activity was significantly higher at a water temperature of 13.0 ° C than at 8.5 ° C, presumably due to higher oxygen demands by the eggs, and a greater risk of egg-infections. The condition factor of the males guarding eggs decreased over time, in contrast to the non-guarding males’ condition factor. The exploitation of the nest (percentage of total nest area covered by eggs) seemed to determine the amount of fanning and cleaning and loss of condition, while brood size (area of eggs) had no effect.

Fanning activity: a swimming movements caused by caudal, dorsal and pelvic fin beats, which send a water current past the clutch, aerating the eggs (for a more thorough description see Barlow, 1964).

Cleaning: with his mouth he is removing debris, dead eggs, unfertilised eggs and infected eggs.

Fertilising: upside down undulating movements while rubbing the genital papilla onto the ceiling of the nest (for more thorough description see Ota et al., 1996).