PLEASE SCROLL DOWN FOR ARTICLE

This article was downloaded by: [Universitetsbiblioteket i Bergen]On: 2 May 2011Access details: Access Details: [subscription number 919195656]Publisher Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Marine and Freshwater Behaviour and PhysiologyPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713644420

Effects of light intensity on visual prey detection by juvenile Atlantic cod(Gadus morhua L.)Justin J. Meagerab; Olav Moberga; Espen Strandac; Anne Christine Utne-Palma

a Department of Biology, University of Bergen, 5020 Bergen, Norway b Faculty of Science, Health andEducation, University of the Sunshine Coast, Maroochydore DC, Queensland 4558, Australia c Instituteof Marine Research, 5817 Bergen, Norway

First published on: 29 April 2010

To cite this Article Meager, Justin J. , Moberg, Olav , Strand, Espen and Utne-Palm, Anne Christine(2010) 'Effects of lightintensity on visual prey detection by juvenile Atlantic cod (Gadus morhua L.)', Marine and Freshwater Behaviour andPhysiology, 43: 2, 99 — 108, First published on: 29 April 2010 (iFirst)To link to this Article: DOI: 10.1080/10236241003798910URL: http://dx.doi.org/10.1080/10236241003798910

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

Marine and Freshwater Behaviour and PhysiologyVol. 43, No. 2, March 2010, 99–108

Effects of light intensity on visual prey detection by juvenile

Atlantic cod (Gadus morhua L.)

Justin J. Meagera,b*, Olav Moberga, Espen Stranda,c

and Anne Christine Utne-Palma

aDepartment of Biology, University of Bergen, 5020 Bergen, Norway; bFaculty ofScience, Health and Education, University of the Sunshine Coast, Maroochydore DC,Queensland 4558, Australia; cInstitute of Marine Research, 5817 Bergen, Norway

(Received 18 July 2009; final version received 9 March 2010)

Despite the research on this important fish species for more than a century,surprisingly little is known of some fundamental aspects of the biology ofAtlantic cod, such as how light affects foraging behaviour. We measuredthe reactive distances of juvenile cod (age 1) over light intensities from 0.01to 64 mmolm�2 s�1 in a controlled laboratory environment, and used theseresults to estimate the visual range and the parameters for a predictivevisual model. The reactive distance at 0.01mmolm�2 s�1 indicated highsensitivity to low light conditions. The reactive distance was less atintermediate light levels (1.5–6.5 mmolm�2 s�1) and increased thereafter.Only a model with a different set of parameters above and below5mmolm�2 s�1 fitted the data, but the validation of the model againstanother dataset indicated that the generality of the model was poor.We interpret these results as a change in foraging behaviour of juvenile codat light intensities occurring at twilight in natural habitats, and the resultsillustrate how behaviour can complicate the relationship between light andreactive distance in a marine teleost.

Keywords: foraging; light; vision; fish; Atlantic cod; Gadus morhua L.

Introduction

Vision is a critical sensory modality in many marine fishes (Blaxter 1970). Fish oftenencounter environments where light can limit vision, which can affect the growth,survival and fitness of fish, and hence the productivity of fisheries. The response offish to light also affects fisheries gear avoidance and stock assessment surveys (Aglenet al. 1999; Ryer and Olla 2000).

Predicting the visual range of fish requires a complex integration of factorsrelating to the optical environment, target visibility and visual physiology, and henceoften relies on mechanistic models (e.g. Aksnes and Giske 1993). Such models areroutinely used to estimate food intake and predation mortality in spatially explicitmodels of foraging and growth (e.g. Beauchamp et al. 1999; Fiksen and

*Corresponding author. Email: justin.meager@bio.uib.no

ISSN 1023–6244 print/ISSN 1029–0362 online

� 2010 Taylor & Francis

DOI: 10.1080/10236241003798910

http://www.informaworld.com

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

Folkvord 1999), ecosystem models (e.g. Salvanes et al. 1992), adaptive individual-

based models of fish behaviour (e.g. Strand and Huse 2007) and models that simulatethe effects of eutrophication on fish (e.g. Fiksen et al. 2002). These models are very

sensitive to species-specific measures of the visual response of fish to light, which arepoorly understood in most exploited marine fishes.

Most of these models have relied on the experimental data obtained for larvae,

despite dramatic physiological changes to fish vision during larval development(Fuiman and Magurran 1994). Such has been the case with Atlantic cod (Gadus

morhua L.), historically one of the world’s most important fish species, an emergingaquaculture species and a focus of many simulation modelling studies. Compared to

many other fish species, the visual system of cod is comparatively well understood(e.g. Anthony 1981; Anthony and Hawkins 1983), but very little is known of the

effect of light intensity on prey detection and foraging behaviour in post-settlement

life-history stages.Cod use a range of senses to locate prey, but vision is particularly well developed

(Anthony 1981) and has an important role in detecting prey and avoiding predators

(Chinarina and Troshicheva 1975; Meager et al. 2005, 2006). Cod are found indiverse optical environments spanning summer at 42�N to polar winter (Drinkwater

2005). The variation in the light environment is particularly pronounced in shallowcoastal nurseries of juvenile cod (520m depth, e.g. Hawkins et al. 1985). Although

they have been shown to have a range of diel foraging patterns, varying betweenpopulations, ages and seasons (e.g. Clark and Green 1990; Mattson 1990; Grant

and Brown 1998a, 1998b), almost nothing is known on the role of light illuminationper se on foraging by juvenile cod.

This study examines the influence of light intensity on visually mediated foraging

of juvenile cod, and uses the relationship between light intensity and visual reactivedistance to develop a predictive visual model, based on Aksnes and Utne (1997).

This model includes criteria for optical conditions (light and turbidity) and targetvisibility (size and contrast), and hence can be generalised to other situations (Aksnes

and Giske 1993). Estimating parameters for this model for cod relies on determining

visual sensitivity as a function of photon irradiance (i.e. photon flux per unit area),rather than the radiance measured in earlier studies (Anthony 1981; Anthony and

Hawkins 1983).

Methods

Collection and maintenance of experimental animals

Experimental fish were the offspring of wild-caught parental Atlantic cod (60�160N;

4�580E), reared at the University of Bergen, Norway, and ranged in size from 17 to22 cm SL (standard length) (mean� SD¼ 20� 1.41 cm) and in total weight from 37

to 88 g (mean� SD¼ 60.7� 14.5 g). Fish were kept in 500L tanks under the naturalphotoperiod of Bergen and were fed frozen gobies, decapods and pellets twice daily.

Mysids (Praunus neglectus Sars) were used for the experiments as they are thecommon prey items of juvenile cod (e.g. Hussy et al. 1997). Praunus neglectus were

collected using dip and seine nets, kept in tanks (90L) for up to 2 months prior to theexperiments and fed live Artemia. Fish were fed live mysids ad libitum for 6 months

prior to the experiments.

100 J.J. Meager et al.

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

Experimental setup

Experiments were conducted in a large rectangular glass aquarium(70� 300� 50 cm3), filled to a depth of 20 cm with seawater (salinity 32–35)(Figure 1). A removable black partition separated the acclimation section of theaquarium (70� 100� 50 cm3) from the prey compartment (70� 200� 50 cm3).In each trial, fish were presented with visual signals from mysid prey inside twoprey cylinders (one mysid in each) (Figure 1, Meager et al. 2005). The cylinders weremade of clear polyvinyl chloride, had a diameter of 7 cm and were used because theydid not distort the visual image of the prey, and did not prevent repeated attacks byjuvenile cod (J.J. Meager, unpublished data). During the experiments, temperaturewas maintained at 9.5� 1�C and clear seawater (beam attenuation coefficient50.2m�1) was supplied at approximately 2 Lm�1. Diffuse light conditions wereprovided by indirectly illuminating the aquarium with different combinations of 150and 500W halogen floodlights. Light intensity was reduced by covering the lampswith metal plates of different degrees of perforation. In this way, it was possible togenerate the same spectral composition at different intensities. The halogen lightswere directed towards the white walls and ceiling of the room, and curtain aroundthe aquarium. The experimental aquarium was also illuminated from underneathwith disperse infrared light (Derwent, wavelength, �� 950 nm) to record fish silhou-ettes at low light intensities using an overhead video camera (Panasonic WV BP550)and video cassette recorder (JVC S-VHS ET).

Experimental protocol

The visual reactive distance of juvenile cod to mysid prey was examined at five lightintensity levels (0.01, 1.6, 6.5, 29 and 64 mmolm�2 s�1). A total of 32 fish were testedat each light level. Fish were tested in pairs, because earlier experiments have foundthat this minimises stress and is similar to the natural foraging situation for juvenilecod (e.g. Brawn 1969). Testing followed the randomised blocks experimental design,

Figure 1. Experimental setup. (A) Glass aquarium (300 cm� 70 cm� 50 cm).(B) Experimental arena (200 cm� 70 cm� 30 cm). (C) Polyvinyl chloride (PVC) prey cylinders(diameter, 7 cm) containing a mysid with vertical movement restricted by two clear disks.(D) Removable black partition. (E) Acclimation section. (F) Overhead video camera(Panasonic WV BP550). (G) Infrared lamps (2�Derwent 70W, 830 nm). (H) Reflectivewhite boards. (I) Water flow. Disperse lighting was obtained by directing the halogen lightstowards the white walls, ceiling and curtain around the aquarium.

Marine and Freshwater Behaviour and Physiology 101

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

with each of the 16 pairs of fish receiving the light treatments once in different orders.

We recorded data for each pair of fish and therefore did not treat observations

for each individual fish as an independent variable. Light intensity was measured

as photon irradiance with a photometer (Li-Cor Photometer, model LI-189,

�¼ 400–700 nm). Light levels were close to uniform over the experimental aquarium

(mean� SE¼ 0.01� 0.005, 1.6� 0.1, 6.5� 0.1, 29� 1 and 64� 1 mmolm�2 s�1).

These light levels represent a range from twilight to low-level diurnal irradiance in

habitats of age 1 cod in western Norway during late winter and early spring

(J.J. Meager, unpublished data).Mysids were introduced into the cylinders at the start of an experiment, when fish

were outside the experimental arena and behind a black screen (Figure 1). One

cylinder was placed randomly in each half of the experimental area. To verify that

vision rather than other sensory modalities was used to detect the mysids, six trials

were conducted in complete darkness by recording under infrared illumination

(�� 950 nm).Prior to the experiments, each fish was starved for 4–5 days (to increase feeding

motivation, e.g. Confer et al. 1978). Fish were acclimated overnight to the

experimental tank. Acclimation to a given light intensity treatment lasted for an

additional 90min. Trials commenced when the screen separating the acclimation

section from the prey section was removed by a pulley system from outside the room.

Trials lasted for 60min or until each fish reacted to the prey at least three times.

Behavioural measurements

The foraging behaviour of cod towards the visual prey cues over the range of light

intensities tested was determined by analysing the video recordings. The reactive

distance was measured every time a fish responded to the prey. A response towards

prey was easily distinguishable from other behaviour, because initiation of a strike

was accompanied by direct orientation of the long axis of the fish towards the prey

(Meager et al. 2005). Responsiveness was defined as the proportion of fish pairs that

responded to the prey at each light level.Visual reactive distance was defined as the distance between the position from

which the fish first oriented towards the prey to the centre of the prey cylinder being

attacked (Meager et al. 2005). The maximum reactive distance of the first three

observations for each fish pair was more likely to be at the limit of visual range and

was therefore used to estimate prey-sighting distance in a given trial (following

Aksnes and Utne 1997). When both fish responded to prey in a short succession,

only the first response was used to exclude the fish that may have reacted to the much

more conspicuous visual target of the other fish. Reactions of cod to prey were

digitised (Picolo capture board, Euresys) and distances were measured using Image

J (version 1.31). The depth of mysids (from 7.5 to 12.5 cm) was constant throughout

all trials, because two small, clear disks were used to standardise prey location and

movement to within a 5 cm vertical section in the middle of each cylinder (Figure 1).

As an earlier experiment using a side mirror indicated that most fish responded to

these prey cues from a similar depth (i.e. 87% of responses at 10–15 cm depth,

Meager et al. 2005), we ignored depth-related changes in fish position in our analysis

of reactive distance.

102 J.J. Meager et al.

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

Data analysis

Data were analysed using R (version 2.6.1, R Foundation for StatisticalComputing 2007) and SPSS (Release 9.0, SPSS Inc. 1998–1999). A log-likelihoodratio test was used to test whether responsiveness was associated with light intensity.Random-blocks analysis of variance (ANOVA) was used to test for significantdifferences in the mean reactive distance between the light intensities. Light level wasthe fixed effect, fish pairs were the random blocks, and the Type IV sum of squareswas used because some fish pairs did not respond to the prey at some light levels.

Test of parameters for the visual range model

To minimise the effect of outliers, the visual range of cod at each light level was firstestimated as the upper 90% confidence interval of maximum reactive distances(Aksnes and Utne 1997). An estimate of visual range at 9.5 mmolm�2 s�1 was takenfrom another study using the same general protocol and prey (Meager et al. 2005).Visual range was therefore estimated at six light levels from 0.01 to 64 mmolm�2 s�1.The following visual model was then fitted to the data (Aksnes and Utne 1997):

r 2 expðcrÞ ¼ C0j jApEb

Ke þ EbEmax � DS

�1

e ð1Þ

where visual range (r) is a function of c, the beam attenuation coefficient, C0 is theinherent prey contrast, Ap is the prey size (measured as the area, m2), Ke is thecomposite light saturation parameter, Eb is the background irradiance(mmolm�2 s�1), Emax is the maximal retinal irradiance that can be processed andDSe is a species-specific sensitivity parameter. The model was simplified before fittingit to the data by aggregating the parameters C0, Ap, Emax and DSe into anexperiment-specific parameter (T ) (Aksnes and Utne 1997):

r 2 expðcrÞ ¼ TEb

Ke þ Eb: ð2Þ

The values of Ke and T were estimated by non-linear regression (NLR)(R package, self-starting Michaelis–Menten equation). The predicted visual rangewas then plotted against the observed visual range, using a Visual Basic program toestimate r from Equation (2) by iteration. We then cross-validated the predictions ofthe model against the visual reactive distances measured across the four differentturbidity levels by Meager et al. (2005). The 95% BCa confidence intervals of theobserved visual ranges were determined by bootstrapping (10,000 iterations, LibraryBoot of the R package) and used to assess the fit of a given model.

Results

No prey were detected in complete darkness (0 reactions from six trials). Preywere detected across the full range of light intensities tested (0.01–64 mmolm�2 s�1).The responsiveness of fish towards prey was not significantly associated with lightintensity (G2

¼ 2.8, df¼ 4, p¼ 0.59), and at least 7 of the 16 fish pairs respondedwithin an hour at each light level. Mean reactive distance range differed significantlybetween light intensities (F4,31¼ 206, p¼ 0.008, Figure 2). Fish pairs had asimilar response to light, that is, there was no significant random-blocks effect

Marine and Freshwater Behaviour and Physiology 103

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

(F13,31¼ 1.64, p¼ 0.13). Mean reactive distance tended to decrease from 0.01 to6.5mmolm�2 s�1, and increased thereafter to 40.6 cm at 64 mmolm�2 s�1 (Figure 2).

Using the parameters estimated from Equation (2) by NLR, the predicted visualrange reached an asymptote at 1.5 mmolm�2 s�1 (T¼ 0.16� 0.00076;Ke¼ 0.0030� 0.00019 mmolm�2 s�1), and thus diverged from the observed valuesat higher light intensities (Figure 3a). The closest fit to the data was achieved byusing a two-stage approach (Figure 3a) with different parameters above andbelow 5 mmolm�2 s�1 (Eb5 5 mmolm�2 s�1: T¼ 0.26, Ke¼ 0.0001 mmolm�2 s�1;Eb� 5 mmolm�2 s�1: T¼ 0.11, Ke¼ 5 mmolm�2 s�1). This gave an estimate of11,000 for the visual capacity parameter, Emax �DS�1e (Equation 1) at low lightintensities (Eb5 5 mmolm�2 s�1), and an estimate of 25,000 at high light intensities(Eb� 5 mmolm�2 s�1) (Ap¼ 3.48� 10�1m; C0¼�0.3, Thetmeyer and Kils 1995).The two-stage model performed poorly against the data of Meager et al. (2005),underestimating visual ranges in turbidities exceeding 3m�1 (Figure 3b).

Discussion

Overall, the reactive distance increased with light intensity, approaching anasymptote that presumably represented the maximum amount of retinal light thatcould be processed (Aksnes and Utne 1997). Juvenile cod were tested over lightintensities that spanned over four orders of magnitude, but the change in the visualperformance of fish was comparatively slight, i.e. the mean reactive distance rangedfrom 29 to 41 cm. This is not surprising, given the range of optical environmentsencountered by a species that historically has had a nearly ubiquitous range incoastal and continental shelf areas of the North Atlantic and adjacent seas. Theseresults are also supported by earlier studies that noted that cod have a well-developedvisual system with high visual acuity and contrast perception (Chinarina andTroshicheva 1975; Anthony 1981).

The estimated visual range of juvenile cod at the lowest light intensity in ourexperiment (0.01 mmolm�2 s�1) suggests that the light threshold for foraging issubstantially lower. Although little is known of the light threshold of cod, asubstantial change in the herding effect of trawl gear on adult cod occurs between0.01 and 0.001 mmolm�2 s�1 (Blaxter 1970). An extremely low light threshold forfeeding is also characteristic of the few other gadoids that have been investigated(Theragra chalcogramma: 5� 10�7 mmolm�2 s�1, Ryer and Olla 1999; Molva molva:0.002mmolm�2 s�1, Blaxter 1970).

20

25

30

35

40

45

0 10 20 30 40 50 60

Irradiance (µmolm–2s–1)

Rea

ctiv

e di

stan

ce (

cm)

Figure 2. Reactive distance of juvenile cod with light intensity. Each point represents themean of maximum reactive distance (�1 SE).

104 J.J. Meager et al.

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

One possible explanation for the reduction in reactive distance at intermediatelight intensities in our experiment (Figure 2) was that foraging behaviour variedbetween light intensities. The relationship between the distance from which fish seethe prey (sighting distance) and the distance from which they react to the prey(reactive distance) is expected to vary with foraging mode, with reactive distances lessthan sighting distances when fish are pause–travel searching, but equal when fish arecruise searching (O’Brien et al. 1989; Browman and O’Brien 1992). Like larval cod(Hunt von Herbing and Gallager 2000), juvenile cod are flexible foragers that usesearch strategies that vary over the continuum from pause–travel to cruise searching(J.J. Meager, unpublished observations). Whether these search strategies co-varywith light intensity or the amount of visual information available is unknown.However, other studies have found that prey-search activity levels of juvenile codwere the lowest at intermediate turbidities (Meager and Batty 2007), and that latelarval stages of cod (428 days post-hatching) spent less time swimming and had ahigher predation success at intermediate light intensities (11.7 mmolm�2 s�1) than ata higher intensity (43 mmolm�2 s�1) (Monk et al. 2006).

It is unlikely that differences in feeding motivation explain the results, asresponsiveness to prey did not differ between light intensities. A physiologicalexplanation for the anomaly in reactive distance at 1.5 mmolm�2 s�1 is also unlikely,

0

10

20

30

40

50

5 15 25 35 45 55 65

Irradiance (µmolm–2s–1)

Predicted visual range: two-stage modelObserved visual range (±95% CI)Predicted visual range: NLR estimateV

isua

l ran

ge (

cm)

10

20

30

40

50

60

0 4 8 12 16Turbidity (beam attenuation m–1)

Predicted visual range: two-stage modelObserved visual range (±95% CI)

*

Vis

ual r

ange

(cm

)

(b)

(a)

Figure 3. (a) Fit of Aksnes and Utne’s (1997) visual model to the experimental data (squares).Predicted values are based on a two-stage model with an inflection point at 5mmolm�2 s�1

(diamonds) and parameters determined from a non-linear regression (triangles). The bar at thebottom of the figure indicates the corresponding diel changes in irradiance encountered innatural habitats (black, night; grey, twilight; white, day). (b) Validation of the two-stage visualmodel against the data set from Meager et al. (2005). The point marked with an asterisk isincluded in both figures, and the grey-shaded areas indicate the 95% confidence intervals ofthe observed visual range.

Marine and Freshwater Behaviour and Physiology 105

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

because retinal light–dark adaptation in cod is thought to occur at a much lowerirradiance (2� 10�4 to 2� 10�3 mEm�2 s�1, Protasov 1964, as cited in Blaxter 1970).Our interpretation of the data as a change in foraging behaviour is further supportedby the fact that it occurred at an irradiance corresponding to dawn–dusk light levelsin natural habitats of juvenile cod in western Norway (Figure 3a). This is whenthey are thought to forage in shallower waters (Hofgaard 1995, as cited inLekve et al. 2006). Although foraging periodicity tends to vary between regions andseasons, most studies have noted that foraging by juvenile cod is under diel con-trol and crepuscular changes are often evident. For example, juvenile cod arecrepuscular in the Western Baltic (Arntz 1973), crepuscular/nocturnal onthe Swedish west coast (Pihl 1982), and age 1 cod are nocturnal in summer anddiurnal in winter in coastal Canadian waters (Clark and Green 1990; Grant andBrown 1998a, 1998b).

The poor fit of the visual model with a single set of parameters to our data(Figure 3a) was because a model fitted to the low light intensities would not fit thedata at high light intensities and vice versa. Hence, it was not the anomaly in sightingdistance at intermediate light intensities, but rather the relatively small differencebetween the sighting distances at the lowest light level compared to the highest lightlevel that made the model difficult to fit. The best fit was achieved when a differentset of visual sensitivity parameters were used above and below a light level of5 mmolm�2 s�1. A similar observation was made by Aksnes and Utne (1997),where Equation (1) using a single set of parameters would not fit contrast thresholdmeasurements made for cod over a large span of light intensities. Nevertheless,even the two-stage model underestimated sighting distances considerably whentested against independent data (Figure 3b), suggesting that the generality of themodel is poor.

Prey detection in this model is based on the number of photons entering theretina, and it is therefore the only model for fish, that we are aware of, that takes intoaccount the relationship between prey size and distance. It does, however, haveseveral important limitations, as it was designed for planktonic fish sighting alonghorizontal paths at depths where colour vision is less important (Aksnes and Giske1993). Juvenile cod inhabit shallow depths where visual distances can also dependstrongly on the viewing angle and the spectral composition of light (Gagnon et al.2007), and cod are known to have colour vision (Anthony and Hawkins 1983).Colour, viewing angle and the angular distribution of light are factors that should bebuilt into future visual models. Whether or not the poor performance of the visualmodel over the light range tested was because it failed to capture the visual ability ofcod, or because the foraging behaviour of cod varied over the light range testedshould be the focus of further work. Regardless, the results call to question the valueof using a visual model to predict foraging rates or energy intake of fish withoutconsidering some of the behavioural effects of light, such as changes in foragingmotivation or searching behaviour.

Acknowledgements

The research was approved by the Norwegian Animal Research Committee(Forsøksdyrutvalget). Financial support was provided by the Norwegian Research Councilproject ‘Foraging behaviour of planktivorous fish and fish larvae.’

106 J.J. Meager et al.

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

References

Aglen A, Engas A, Huse I, Michalsen K, Stensholt BK. 1999. How vertical fish distribution

may affect survey results. ICES J Mar Sci. 56:345–360.

Aksnes DL, Giske J. 1993. A theoretical model of aquatic visual feeding. Ecol Modell.

67:233–250.Aksnes DL, Utne ACW. 1997. A revised model of visual range in fish. Sarsia. 82:137–147.Anthony PD. 1981. Visual contrast thresholds in the cod Gadus morhua L. J Fish Biol.

19:87–103.

Anthony PD, Hawkins AD. 1983. Spectral sensitivity of the cod, Gadus morhua L. Mar Behav

Physiol. 10:145–165.Arntz WE. 1973. Periodicity of diel food-intake of cod Gadus morhua in Kiel Bay. Oikos.

15:138–145.

Beauchamp DA, Baldwin CM, Vogel JL, Gubala CP. 1999. Estimating diel, depth-specific

foraging opportunities with a visual encounter rate model for pelagic piscivores. Can

J Fish Aquat Sci. 56:128–139.Blaxter JHS. 1970. Light-fishes. In: Kinne O, editor. Marine ecology: environmental factors.

London: Wiley-Interscience. p. 213–320.

Brawn VM. 1969. Feeding behaviour of cod (Gadus morhua). J Fish Res Board Can.

26:583–596.Browman HI, O’Brien WJ. 1992. The ontogeny of search behaviour in the white crappie,

Pomoxis annularis. Environ Biol Fishes. 34:181–195.

Chinarina AD, Troshicheva NV. 1975. Food-procuring reactions of the Atlantic cod,

Gadus morhua, and the short-horn sculpin, Myoxocephalus scorpius, in the absence of

visual recognition. J Ichthyol. 15:990–997.Clark DS, Green JM. 1990. Activity and movement patterns of juvenile Atlantic cod,

Gadus morhua, in Conception Bay, Newfoundland, as determined by sonic telemetry.

Can J Zool. 68:1434–1442.Confer JL, Howick GL, Corzette MH, Kramer SL, Fitzgibbon S, Landesberg R. 1978. Visual

predation by planktivores. Oikos. 31:27–37.Drinkwater KF. 2005. The response of Atlantic cod (Gadus morhua) to future climate change.

ICES J Mar Sci. 62:1327–1337.

Fiksen Ø, Aksnes DL, Flyum MH, Giske J. 2002. The influence of turbidity on growth and

survival of fish larvae: a numerical analysis. Hydrobiologia. 484:49–59.Fiksen Ø, Folkvord A. 1999. Modelling growth and ingestion processes in herring Clupea

harengus larvae. Mar Ecol Prog Ser. 184:273–289.

Fuiman LA, Magurran AE. 1994. Development of predator defences in fishes. Rev Fish Biol

Fish. 4:145–183.Gagnon YL, Shashar N, Warrant EJ, Johnsen SJ. 2007. Light scattering by selected

zooplankton from the Gulf of Aqaba. J Exp Biol. 210:3728–3735.

Grant SM, Brown JA. 1998a. Diel foraging cycles and interactions among juvenile Atlantic

cod (Gadus morhua) at a nearshore site in Newfoundland. Can J Fish Aquat Sci.

55:1307–1316.Grant SM, Brown JA. 1998b. Nearshore settlement and localisation populations of age 0

Atlantic cod (Gadus morhua). Can J Fish Aquat Sci. 55:1317–1327.

Hawkins AD, Soofiani NM, Smith GW. 1985. Growth and feeding of juvenile cod (Gadus

morhua L.). J Cons Int Explor Mer. 42:11–32.Hunt von Herbing I, Gallager SM. 2000. Foraging behaviour in early Atlantic cod larvae

(Gadus morhua) feeding on a protozoan (Balanion sp.) and a copepod nauplius

(Pseudodiaptomus sp.). Mar Biol. 136:591–602.

Hussy K, St. John MA, Bottcher U. 1997. Food resource utilization by juvenile Baltic cod

Gadus morhua: a mechanism potentially influencing recruitment success at the demersal

juvenile stage? Mar Ecol Prog Ser. 155:199–208.

Marine and Freshwater Behaviour and Physiology 107

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011

Lekve K, Enersen K, Enersen SE, Gjosaeter J, Stenseth NC. 2006. Interannual variability inabundance and length of young coastal cod in the subtidal zone. J Fish Biol. 68:734–746.

Mattson S. 1990. Food and feeding habits of fish species over a soft sublittoral bottom in thenortheast Atlantic. 1. Cod (Gadus morhua L.) (Gadidae). Sarsia. 75:247–260.

Meager JJ, Batty RS. 2007. Effects of turbidity on the spontaneous and prey-searching activityof juvenile Atlantic cod (Gadus morhua). Philos Trans R Soc. B 362:2123–2130.

Meager JJ, Domenici P, Shingles A, Utne-Palm AC. 2006. Escape responses in juvenile

Atlantic cod Gadus morhua L.: the effects of turbidity and predator speed. J Exp Biol.209:4174–4184.

Meager JJ, Solbakken T, Utne-Palm AC, Oen T. 2005. Effects of turbidity on the reaction

distance, search time, and foraging success of juvenile Atlantic cod (Gadus morhua). CanJ Fish Aquat Sci. 62:1978–1984.

Monk J, Puvanendran V, Brown JA. 2006. Do different light regimes affect the foraging

behaviour, growth and survival of larval cod (Gadus morhua L.)? Aquaculture.257:287–293.

O’Brien WJ, Evans BI, Browman HI. 1989. Flexible search tactics and efficient foraging insaltatory searching animals. Oecologia. 80:100–110.

Pihl L. 1982. Food intake of young cod and flounder in a shallow bay on the Swedish Westcoast. Neth J Sea Res. 15:419–432.

Ryer CH, Olla BL. 1999. Light-induced changes in the prey consumption and behaviour of

two juvenile planktivorous fish. Mar Ecol Prog Ser. 181:41–51.Ryer CH, Olla BL. 2000. Avoidance of an approaching net by juvenile walleye pollock

Theragra chalcogramma in the laboratory: the influence of light intensity. Fish Res

(Amst). 45:195–199.Salvanes AGV, Aksnes DL, Giske J. 1992. Ecosystem model for evaluating potential cod

production in a West Norwegian Fjord. Mar Ecol Prog Ser. 90:9–22.Strand E, Huse G. 2007. Vertical migration in adult Atlantic cod (Gadus morhua). Can J Fish

Aquat Sci. 64:1747–1760.Thetmeyer H, Kils U. 1995. To see and not be seen: the visibility of predator and prey with

respect to feeding behaviour. Mar Ecol Prog Ser. 126:1–8.

108 J.J. Meager et al.

Downloaded By: [Universitetsbiblioteket i Bergen] At: 07:48 2 May 2011