Avoidance behavior and brain monoamines in fish

www.elsevier.com/locate/brainres

Brain Research 1032

Research report

Avoidance behavior and brain monoamines in fish

Erik Hfglunda, Finn-Arne Weltziena,b, Joachim Schjoldenc, Svante Winbergc,

Holger Ursind, Kjell B. Døvinga,*

aDepartment of Molecular Biosciences, University of Oslo, PO Box 1041 Blindern, 0316 Oslo, NorwaybUSM 0401, UMR 5178 CNRS, Biologie des Organismes Marins et Ecosystemes, Departement des Milieux et Peuplements Aquatiques,

Museum National d’Histoire Naturelle, 7 rue Cuvier, 75231 Paris Cedex 05, FrancecEvolutionary Biology Centre, Department of Comparative Physiology, Uppsala University, Norbyvagen 18 A, SE-75236 Uppsala, Sweden

dDepartment of Biological and Medical Psychology, University of Bergen, Jonas Lies vei 91, 5009 Bergen, Norway

Accepted 30 October 2004

Available online 9 December 2004

Abstract

The crucian carp performs a typical avoidance behavior when exposed to olfactory cues from injured skin of conspecifics. They swim

rapidly to the bottom and hide in available material. This work examines the effects of skin extract exposure and availability of hiding

material on this behavior, and concomitant changes in brain monoaminergic activity in crucian carp.

Individual fish were exposed to skin extract in aquaria with or without hiding material. Exposure to skin extract resulted in the expected

avoidance behavior consisting of rapid movement towards the bottom of the aquarium. This lasted for 1–2 min. Activity then decreased

below the level observed before exposure, suggesting a bfreezingQ type of avoidance behavior. This behavior was independent of availabilityof hiding material.

Brain dopaminergic activity increased in telencephalon and decreased in the brain stem following skin extract exposure, again

independent of availability of hiding material. However, fish kept in aquaria without hiding material showed an elevation of serotonergic

activity in the brain stem and the optic tectum compared to fish with available hiding material. Absence of hiding material increased

serotonergic activity also without exposure to skin extract. In aquaria with hiding material, the fish stirred up a cloud of fine sediments and

showed a more pronounced decrease in locomotor activity in agreement with this being a more efficient freezing or immobile avoidance

behavior. These results show that basic components of avoidance behavior and related brain changes are present in the fish brain, in

accordance with the common phylogenetic roots of avoidance behavior in all vertebrates.

D 2004 Elsevier B.V. All rights reserved.

Theme: Neural basis of behavior

Topic: Monoamines and behavior

Keywords: Fear behavior; Teleost; Monoamine; Predator; Refuge; Olfactory cues

1. Introduction

The crucian carp (Carassius carassius) perform stereo-

typic avoidance behavior when exposed to olfactory cues

from injured skin of conspecifics. They rapidly seek the

bottom (bflightQ), stir up a cloud of fine sediment and hide

0006-8993/$ - see front matter D 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.brainres.2004.10.050

* Corresponding author. Fax: +47 22 85 46 64.

E-mail address: kjell.doving@bio.uio.no (K.B. Døving).

(bfreezeQ) in or on the material available at the bottom [6].

Similar predator-avoidance behavior is elicited by exposure

to olfactory cues from injured skin of conspecifics (i.e. skin

extract) in other fish as well, for instance in the rainbow trout

(Oncorhynchus mykiss) [20]. Direct exposure to predators

involves changes in monoamine signaling as well as behavior

in teleost fish (e.g. Ref. [26]), as it does in mammals [2,14].

Availability of a refuge during predator exposure affects

both the behavior and physiological stress responses in the

long nose killifish (Fundulus majalis) [27]. This implies that

(2005) 104–110

E. Hoglund et al. / Brain Research 1032 (2005) 104–110 105

availability of hiding material interacts with stimuli signal-

ing predators and injury, suggesting an interaction between

stress stimuli and context that in mammals is taken as an

indication of cognitive factors [24]. In this paper, we

examine whether availability of hiding material affects the

behavior as well as the brain monoamine activity also in the

crucian carp. There are two components of the apparently

innate avoidance behavior. Following the nomenclature for

avoidance behavior and limbic mechanisms in mammals

(see Ref. [23]); the active movements (bflightQ) qualify as

active avoidance, and the immobility in the hiding material

(bfreezingQ) as passive avoidance. These innate behaviors

are goal directed behaviors, and, therefore, represent

potential binstrumentalQ behaviors [19]. When instrumental

behavior is well established, the vegetative, endocrine, and

immunological response to the situation diminishes in

mammals (bstress reductionQ or coping [13,23,24]). This is

also documented for by changes in brain monoamines [24].

In mice [1] and rats [14], limbic monoaminergic signal-

ing is affected by direct exposure to a predator (cat and

ferret). Direct predator exposure affects monoaminergic

signaling also in teleost fish like the bicolor damselfish

(Pomacentrus parties) [26], rainbow trout [12,16], and

Arctic charr (Salvelinus alpinus) [8,9].

If the presence or absence of hiding materials that offer

avoidance possibilities affects the behavior of the crucian

carp, it may also affect the brain monoamines. Our

hypothesis, therefore, is that availability of hiding material

may affect avoidance behavior as well as monoamine

signaling in the brain much in the same way as in other

vertebrates including mammals.

2. Materials and methods

2.1. Animals

Crucian carp (C. carassius) were caught in a small lake

outside of Oslo, Norway (608N), in June 2002. The fish

were kept indoors at the Department of Molecular Bio-

sciences, University of Oslo, in 700 L tanks continuously

supplied with aerated water at 7–9 8C, and with a photo

cycle of 12/12-h (light/dark). The fish were fed once a day

with TetraR pond sticks (Tetra, Germany). The experiment

was performed in February–March 2003 with fish (37.5%

males and 62.5% females) with a mean body weight of

29F7.5 g (meanFstandard deviation).

2.2. Preparation of skin extract

For preparation of the skin extract used in the experi-

ments, crucian carp were killed by decapitation and skin was

taken from the sides of the fish. Approximately 2 g of skin

was homogenized in 100 mL distilled water with pestle and

mortar. The homogenate was centrifuged at 2400 rpm for 5

min at 4 8C. The supernatant was frozen immediately and a

working concentration of 1:10 in pond water was made

fresh before use.

2.3. Experimental protocol

Two days before the behavioral observations, six fish

were transferred to six visually separated 40 L observation

aquaria. In three of the observation aquaria, about half of the

bottom floor was covered by an approximately 4 cm thick

layer of fine sediments taken from the lake where the fish

were caught. In the other three observation aquaria the

bottom floor consisted of uncovered glass.

After 2 days of acclimatization, each individual fish was

video filmed for 30 min, during which they received two 5

mL injections of either skin extract or aquarium water

(vehicle control). The first injection was given after 10 min,

followed by a second addition after 20 min to ensure

exposure of the fish to stable levels of alarm substance. To

minimize disturbance during the behavioral studies, the fish

were filmed through a hole in a black curtain covering the

front of the aquaria, and skin extract was delivered to the

aquarium water through a plastic tube from outside the

covering curtain. The video filmed fish were analyzed for

locomotor activity and position in the water tank. Locomo-

tor activity was quantified throughout the 30-min observa-

tional period as the distance the fish had moved during 15 s

periods. The position in the water tank was measured every

15th second, throughout the 30-min observational period, as

the distance from the bottom floor in the aquaria. A total of

14 fish were exposed to skin extract, 7 with hiding material

and 7 without. Ten fish, five with hiding material and five

without, received aquarium water instead of skin extract and

served as controls. All experiments were conducted at room

temperature in the light period of the 12/12-h photo cycle.

2.4. Brain tissue sampling and analysis of brain

monoaminergic activity

Directly after the 30-min observation period, the fish

were netted and rapidly killed by severing the spinal cord

just behind the head of the fish. The brain was then removed

and dissected into four regions: brain stem (including

medulla and part of the spinal cord but excluding

cerebellum and the vagal lobes), optic tectum, hypothal-

amus (di- and mesencephalon except optic tectum), and

telencephalon (excluding the olfactory bulbs). Each brain

part was immediately wrapped in aluminum foil, frozen in

liquid nitrogen, and stored at �80 8C.Upon HPLC analyses, the frozen brain parts were

homogenized in 4% (w/v) ice-cold perchloric acid contain-

ing 0.2% EDTA and 40 ng/mL epinine (deoxyadrenaline;

used as internal standard in the HPLC analyses). Homog-

enization was performed using either a Potter-Elvehjem

homogenizer (optic tectum and brain stem), or a MSE 100 W

ultrasonic disintegrator (telencephalon and hypothalamus).

Serotonin (5-HT), 5-hydroxyindoleacetic acid (5-HIAA, the

Table 1

Effects of skin extract exposure and availability of hiding material on locomotor activity in crucian carp

Locomotor activity (m/min)

Control After 1st addition After 2nd addition

Vehicle no material 0.25F0.016a 0.27F0.072a 0.21F0.052a

hiding material 0.22F0.007a 0.16F0.020a 0.19F0.060a

Skin extract no material 0.16F0.048ac 0.10F0.012ac 0.040F0.015cb

hiding material 0.26F0.060a 0.011F0.001b 0.009F0.008b

Skin extract or vehicle (aquaria water) was injected twice to the aquaria (1st and 2nd addition after 10 and 20 min, respectively). Locomotor activity was

quantified as the swimming distance during 10 min before 1st addition of skin extract (controls) and during 8.5 min after 1st and 2nd addition of skin extract. A

shorter analysis period in the periods after 1st and after 2nd addition was chosen to exclude bias induced by the first 1.5 min of high swimming activity caused

by the alarm substance. Values are meanFS.E.M. for five individuals in the vehicle-treated groups, and for seven individuals in the skin extract-treated groups

Means with no common letters are significant at the level of Pb0.05. For MANOVA values, see text.

E. Hoglund et al. / Brain Research 1032 (2005) 104–110106

major 5-HT metabolite), dopamine (DA), 3,4-dihydroxy-

phenylacetic acid (DOPAC, the major DA metabolite), and

norepinephrine (NE) were quantified by HPLC, as described

by Ref. [8]. Because of interfering unidentified peaks in the

chromatogram, we were unable to quantify 3-metoxy-4-

hydroxyphenylglycol (MHPG, the major NE metabolite).

Samples were quantified by comparison with standard

solutions of known concentrations and corrected for

recovery of the internal standard with HPLC standard

solutions using HPLC software (CSW, Data Apex, the

Czech Republic). The ratio of [metabolite]/[parent mono-

amine] was used as an index for monoaminergic activity.

2.5. Statistical analyses

All values are presented as meanFstandard error of mean

(S.E.M.). To investigate the effect of skin extract and hiding

material in the aquaria on locomotor activity and position in

the water tank, a repeated two-way analysis of variance

(ANOVA) was performed with skin extract or vehicle

treatment, and hiding material or no hiding material as

independent variables. Furthermore, data on brain levels of

monoamines, monoamine metabolites and ratios of mono-

amine metabolites to parent monoamine concentrations (i.e.

[5-HIAA/5-HT], [DOPAC]/[DA]) were subjected to a two-

way ANOVA, with skin extract or vehicle treatment, and

hiding material or no hiding material as independent

variables. The Tukey post hoc test (Tukey honest significant

difference test for unequal N) was applied to compare means

Table 2

Effects of skin extract exposure and availability of hiding material on crucian carp position in the water tank

Position in the water tank (mm)

Control After 1st addition After 2nd addition

Vehicle no material 59F4.3a 60F3.0a 68F2.8a

hiding material 76F4.5a 84F4.1a 81F4.0a

Skin extract no material 58F12a 30F4.2b 29F4.8b

hiding material 78F8.4a 22F0.87b 22F2.0b

Skin extract or vehicle (aquaria water) was injected twice to the aquaria (1st and 2nd addition after 10 and 20 min, respectively). Position in the water tank was

quantified as mean distance from the bottom floor of the aquarium during 10 min before 1st addition of skin extract (controls) and during 8.5 min after 1st and

2nd addition of skin extract. A shorter analysis period in the periods after 1st and after 2nd addition was chosen to exclude bias induced by the first 1.5 min o

high swimming activity caused by the alarm substance. Values are meanFS.E.M. for five individuals in the vehicle-treated groups, and for seven individuals in

the skin extract-treated groups. Means with no common letters are significant at a level of Pb0.05. For MANOVA values, see text.

.

between different groups. To obtain normal distribution, data

on locomotor activity and distance from the bottom floor of

the aquaria were log transformed, and brain monoamine

ratios were arc-sin transformed. All statistical analyses

were performed using Statistica 5.1 (StatSoft, Tulsa, OK,

USA) software.

3. Results

3.1. Effects of skin extract and hiding material availability

on locomotor activity and position in the water tank

Both skin extract and availability of hiding material

affected the behavior in the crucian carp. Skin extract

exposure resulted in rapid movements towards the bottom of

the aquarium. This behavior lasted 1–2 min.

Then, when hiding material was available, the locomotor

activity of the fish decreased below the level observed

before exposure (ANOVA: F2,38=7.62; P=0.0016). This

decrease in locomotor activity was significant after both the

first (Pb0.001) and the second addition of skin extract

(Pb0.001).

When there was no hiding material available, there was

no significant decrease in locomotor activity following the

first addition of skin extract, but after the second addition

there was a significant decrease compared to addition of

aquarium water (P=0.024). However, there was less

decrease in activity compared to the fish with hiding

f

Fig. 1. Effects of skin extract exposure on [DOPAC]/[DA] in the telencephalon (A), brain stem (B), hypothalamus (C) and optic tectum (D) of crucian carp. The

values are presented as meanFS.E.M. * indicates significant difference at a level of Pb0.05. See Table 3 for ANOVA values.

E. Hoglund et al. / Brain Research 1032 (2005) 104–110 107

material available (first addition, P=0.0015; Table 1).

Availability of hiding material did not affect locomotor

activity before addition of skin extract (or water) in any of

the two groups (Table 1).

The two-way repeated measurement MANOVA indi-

cated a combined effect of hiding material and skin extract

on position in the water tank (ANOVA: F2,38=3.40,

Fig. 2. Effects of hiding material availability on [5-HIAA]/[5-HT] in the telenceph

carp. The values are presented as meanFS.E.M. * indicates significant difference

P=0.043). The post hoc test, however, failed to show any

significant effects. Skin extract exposure resulted in a

position lower in the water tank, both after first (hiding

material, Pb0.001; no hiding material, P=0.009) and second

addition of skin extract (hiding material, Pb0.001; no hiding

material, P=0.004), with no significant difference in water

tank position between fish with or without available hiding

alon (A), brain stem (B), hypothalamus (C) and optic tectum (D) of crucian

at a level of Pb0.05. See Table 3 for ANOVA values.

Table 3

Effects of skin extract exposure and availability of hiding material on the concentration of monoamines and monoamine metabolites, and on the ratio bet een monoamine metabolite to parent monoamine in

crucian carp

Effect Skin extract Substrate Skin extract and substrate

ANOVA ANOVA ANOVA No substrate Substrate

Skin extract ehicle Skin extract Vehicle

Hypothalamus

[DOPAC] F(1,20)=0.174, P=0.681 F(1,20)=1.68, P=0.209 F(1,20)=4.09, P=0.057 8.80F2.10 5.43F2.58 7.28F1.68 12.3F2.27

[DA] F(1,20)=0.077, P=0.783 F(1,20)=0.026, P=0.871 F(1,20)=0.181, P=0.675 2390F244 165F469 2203F286 2260F460

10�3�[DOPAC]/[DA] F(1,20)=0.001, P=0.969 F(1,20)=2.18, P=0.155 F(1,20)=1.46, P=0.240 3.79F0.96 2.27F0.74 4.12F1.51 5.73F1.19

[5-HIAA] F(1,20)=0.587, P=0.452 F(1,20)=1.13, P=0.300 F(1,20)=0.230, P=0.636 166F34.9 206F59.9 166F6.77 156F25.3

[5-HT] F(1,20)=0.021, P=0.884 F(1,20)=0.0003, P=0.986 F(1,20)=0.046, P=0.831 2910F244 790F552 2840F260 2859F238

10�3�[5-HIAA]/[5-HT] F(1,20)=0.738, P=0.400 F(1,20)=0.848, P=0.368 F(1,20)=1.03, P=0.321 54.7F5.92 70.9F14.3 55.6F3.36 54.2F6.16

[NA] F(1,18)=1.40, P=0.250 F(1,20)=0.658, P=0.658 F(1,20)=0.0126, P=0.912 3530F412 090F547 3390F547 2860F534

Optic tectum

[DOPAC] F(1,19)=0.191, P=0.667 F(1,19)=0.0578, P=0.813 F(1,19)=2.53, P=0.128 4.26F0.681 .41F0.91 3.60F0.83 5.22F0.735

[DA] F(1,19)=0.833, P=0.372 F(1,19)=1.13, P=0.813 F(1,19)=0.366, P=0.850 124F33.8 154F46.0 155F46.0 105F8.5

10�3�[DOPAC]/[DA] F(1,19)=0.0001, P=0.989 F(1,19)=1.06, P=0.316 F(1,19)=2.47, P=0.133 42.5F7.72 0.4F9.61 38.3F6.21 50.2F6.70

[5-HIAA] F(1,19)=0.120, P=0.733 F(1,19)=3.46, P=0.0714 F(1,19)=0.305, P=0.862 36.1F3.90 7.0F4.54 27.4F4.52 26.3F2.09

[5-HT] F(1,19)=0.0279, P=0.869 F(1,19)=0.551, P=0.467 F(1,19)=0.120, P=0.733 234F51.1 220F51.2 203F43.6 190F12.1

10�3�[5-HIAA]/[5-HT] F(1,19)=0.0876, P=0.770 F(1,19)=6.77, P=0.0175 F(1,19)=0.398, P=0.535 167F13.4 178F15.4 142F10.8 138F7.10

[NA] F(1,19)=0.224, P=0.642 F(1,19)=0.215, P=0.647 F(1,19)=0.0836, P=0.775 791F62.7 856F189 778F130 793F79.3

Brain stem

[DOPAC] F(1,18)=6.13, P=0.0234 F(1,18)=2.22, P=0.15 F(1,18)=1.01, P=0.326 8.20F2.79 0.0F2.27 3.92F0.54 14.5F5.52

[DA] F(1,18)=1.65, P=0.214 F(1,18)=0.922, P=0.349 F(1,18)=1.08, P=0.310 270F46.5 178F35.0 181F17.0 279F75.5

10�3�[DOPAC]/[DA] F(1,18)=5.07, P=0.0370 F(1,18)=1.65, P=0.214 F(1,18)=2.07, P=0.167 27.7F4.25 3.4F26.5 21.4F2.14 40.6F9.83

[5-HIAA] F(1,19)=0.0937, P=0.763 F(1,19)=3.65, P=0.0711 F(1,19)=0.0150, P=0.904 73.9F12.9 1.1F11.5 46.4F4.38 50.5F15.5

[5-HT] F(1,19)=0.103, P=0.75 F(1,19)=0.819, P=0.379 F(1,19)=4.12, P=0.0567 437F68.2 455F72.9 318F33.0 404F92.0

10�3�[5-HIAA]/[5-HT] F(1,19)=2.44, P=0.134 F(1,19)=4.47, P=0.0479 F(1,19)=0.593, P=0.464 168F7.86 147F6.40 155F20.2 95.7F22.9

[NA] nd nd nd nd

Telencephalon

[DOPAC] F(1,19)=13.3, P=0.0017 F(1,19)=0.0604, P=0.808 F(1,19)=0.0663, P=0.799 23.4F4.61 1.74F7.65 24.5F6.34 6.77F1.98

[DA] F(1,19)=1.67, P=0.211 F(1,19)=0.172, P=0.682 F(1,19)=0.0198, P=0.682 674F120 569F113 637F100 459F70.0

10�3�[DOPAC]/[DA] F(1,19)=4.48, P=0.0402 F(1,19)=00009, P=0.975 F(1,19)=0114, P=0.916 41.0F11.5 18.9F12.1 41.8F9.66 17.4F7.40

[5-HIAA] F(1,19)=1.06, P=0.314 F(1,19)=0.691, P=0.416 F(1,19)=0.0229, P=0.881 342F110 272F96.2 213F25.2 152F24.7

[5-HT] F(1,19)=0.936, P=0.345 F(1,19)=0.0375, P=0.848 F(1,19)=0.772, P=0.129 2540F848 2430F800 2210F284 1460F146

10�3�[5-HIAA]/[5-HT] F(1,19)=2.19, P=0.154 F(1,19)=0.201, P=0.658 F(1,19)=0.382, P=0.554 136F28.8 118F13.2 99.3F8.09 102F10.3

[NA] F(1,19)=1.87, P=0.187 F(1,19)=0.0238, P=0.879 F(1,19)=0.115, P=0.738 6200F1010 5630F1320 6375F942 4650F587

Five milliliters of either skin extract or vehicle (aquaria water) was injected twice to the aquaria (after 10 and 20 min, respectively; see Materials and methods r details). Fourteen fish were exposed to skin extract,

seven with and seven without hiding material available. Ten fish, five with and five without hiding material available, received aquaria water and served as v icle treated controls. Values are meanFS.E.M. F, df

and P values are results of ANOVA (significant results are indicated in bold). DA, Dopamine; DOPAC, 3,4-dihydroxyphenylacetic acid; NE, norepinephrine; -HT, serotonin; 5-HIAA, 5-hydroxyindoleacetic acid.

E.Hoglundet

al./Brain

Resea

rch1032(2005)104–110

108

w

V

2

2

3

3

3

3

1

7

7

1

fo

eh

5

E. Hoglund et al. / Brain Research 1032 (2005) 104–110 109

material after the first (P=0.78) or the second addition

(P=0.88; Table 2). There were no significant effects of

available hiding material on position in the water tank in

vehicle treated individuals, or before the experimental

additions in any of the groups, see Table 2.

3.2. Effects of skin extract and availability of hiding

material on brain monoaminergic activity

Skin extract exposure significantly affected the ratio of

[DOPAC]/[DA] and resulted in an increased ratio in the

telencephalon (ANOVA: F1,19=4.48, P=0.040) and a

decreased ratio in the brain stem (ANOVA: F1,19=5.07,

P=0.037), compared to vehicle-treated controls (Fig. 1).

Availability of hiding material affected the [5-HIAA]/[5-

HT] ratio. Fish with no available hiding material had

significantly elevated ratios in the brain stem (ANOVA:

F1,19=4.47, P=0.048) and the optic tectum (ANOVA:

F1,19=6.77, P=0.023). A trend towards elevated ratios was

observed also in the hypothalamus and the telencephalon of

fish with no available hiding material, although they were

not significantly different from control (Fig. 2).

However, there were no combined effects of hiding

material and skin extract exposure on brain monoaminergic

activity, see Table 3 for ANOVA values.

4. Discussion

Exposures to skin extract from conspecifics resulted in a

typical predator avoidance behavior in crucian carp. This

avoidance behavior consists of an initial bactive avoidanceQphase (1–2 min of intense swimming against the bottom

floor), followed by a bpassive avoidanceQ phase (decreased

locomotor activity and a position lower in the water tank).

Active and passive avoidance of predators and of other

types of threat have been described in a wide range of

species. Even if both active and passive avoidance are

regarded as fear behaviors, the brain mechanisms seem

distinct and separate, in rats [25], cats [22], and possibly

also in primates [7].

Exposure to skin extract affected the brain DAergic

activity, it increased in the telencephalon and decreased in

the brain stem. This was independent of availability of

hiding material. Exposure to skin extract did not change the

brain 5-HT activity.

Lack of hiding material led to an elevation of 5-HTergic

activity in the brain stem and the optic tectum compared to

levels in fish with available hiding material. Major 5-

HTergic systems in hypothalamic and pretectal areas were

not significantly affected [11]. The most remarkable finding

was that this was independent of exposure to skin extract.

Absence of hiding material, therefore, influenced brain stem

5-HT activity even when no threat was present. The fish

were captured from their natural habitat (a lake) and

introduced to a totally new environment. Availability of

hiding material was enough to reduce the brain stem 5-HT

activity. This is a response antedating any avoidance

behavior, and, therefore, may qualify for bexpectancyQ,referred to as bcognitiveQ activity in mammals [24]. An

alternative explanation may be that the presence of hiding

material makes the environment less novel. Under any

circumstance the brain biochemistry of carps that register

familiarity and /or potential escape routes differs from that

of carps that do not register this.

To what extent do the innate avoidance responses in

crucian carp represent instrumental behavior with the same

neurochemical concomitants as learned behavior in mam-

mals, including humans? Is the crucian carp brain capable of

registering the potential escape possibilities to the extent

that this reduces the 5-HT activation induced by a novel or

threatening environment?

Presence or absence of hiding material also affected the

avoidance behavior, even if the 5-HT change was inde-

pendent of whether there was any adequate stimulus for

avoidance. When hiding material was available, the fish

stirred up a cloud of fine sediments and showed a more

pronounced decrease in locomotor activity in agreement

with this being a more efficient freezing or immobile

avoidance behavior. Lack of hiding material led to less

bfreezingQ (passive avoidance) behavior. In mammals,

established avoidance behavior is accompanied by low

levels of arousal, measured by behavior or plasma levels of

corticosterone [4,13,23]. If, however, the behavior is made

difficult, like in forced extinction of avoidance behavior,

arousal reappears. In our case, lack of hiding material may

be postulated to lead to more arousal and less avoidance

behavior. In mammals including humans, it has been

postulated that what the brain stores when avoidance is

acquired is a positive response outcome expectancy [3],

which leads to a low arousal level (the cognitive activation

theory of stress, CATS [24]). Similar expectancy functions

may be present in fish.

Such mechanisms appear to be present also in other

species of fish. In the long nose killifish submerged aquatic

vegetation as a refuge affected behavior and decreased

plasma cortisol content during visual predator exposure

[27]. Arctic charr exposed to a white background showed an

elevated NE activity in the optic tectum and the brainstem,

and a tendency for elevated 5-HT activity in the optic

tectum, compared to fish that were exposed to a black

background [10]. Absence of hiding material or avoidance

possibility elicits a high arousal also in other species, for

instance the rat. In rats, a typical response to threat by a

shock prod is to bury the shock prod, if burying material is

available. This leads to a low level of arousal. However, if

there is no burying material, the rats show a high level of

arousal [5].

DAergic systems in the fish brain are located in

telencephalon, in hypothalamic and pretectal areas, and in

the medulla oblongata [11]. The alarm response in our fish

involves increased DAergic activity in the telencephalon

E. Hoglund et al. / Brain Research 1032 (2005) 104–110110

and decreased DAergic activity in the brain stem, while

major DAergic systems in hypothalamic areas remained

unchanged. It is well established that the teleost tel-

encephalon is involved in the regulation of fear behavior

and the learning of avoidance responses, in particular with

the production and reinforcement mechanisms for avoidance

responses [18]. The telencephalon is involved in, but not

necessary for, both passive and active avoidance [17],

possibly through inhibitory mechanisms necessary for

freezing and passive avoidance. In particular, it seems to

be the telencephalic DAergic systems that are important for

fear behavior [15]. Our results confirm this position. The

active and the passive avoidance responses (flight and

freezing) observed in our experiment appear to be related to

the increase in telencephalic DAergic activity. However, our

study does not permit conclusions as to the possible

differential role of different structures in the telencephalon,

as has been demonstrated in other vertebrates [21].

Acknowledgments

Supported by the Norwegian Research Council. We

thank Trude Haug and ayvind averli for valuable com-

ments on earlier versions of this paper.

References

[1] C. Belzung, W. El Hage, N. Moindrot, G. Griebel, Behavioral and

neurochemical changes following predatory stress in mice, Neuro-

pharmacology 41 (2001) 400–408.

[2] D.C. Blanchard, G. Griebel, R.J. Rodgers, R.J. Blanchard, Benzodia-

zepine and serotonergic modulation of antipredator and conspecific

defense, Neurosci. Biobehav. Rev. 22 (1998) 597–612.

[3] R.C. Bolles, Reinforcement, expectancy and learning, Psychol. Rev.

79 (1972) 394–409.

[4] G.D. Coover, H. Ursin, S. Levine, Plasma-corticosterone levels during

active-avoidance learning in rats, J. Comp. Physiol. Psychol. 82

(1973) 170–174.

[5] S.F. De Boer, J.L Slangen, J. Van der Gugten, Plasma catecholamine

and corticosterone levels during active and passive shock-prod

avoidance behavior in rats: effects of chlordiazepoxide, Physiol.

Behav. 47 (1990) 1089–1098.

[6] K.B. Dbving, E.H. Hamdani, E. Hfglund, A. Kasumyan, A.

Tuvikene, A review on the chemical and physiological basis of alarm

reactions in cyprinids, in: A.S. Kapoor, K. Reuter (Eds.), Fish

Chemosenses, Science Publishers, Plymouth, 2004, ISBN: 1-57808-

319-2, (in press).

[7] J.A. Gray, The Neuropsychology of Anxiety: an Enquiry into the

Functions of the Septo-Hippocampal System, Oxford University

Press, Oxford, 1981.

[8] E. Hfglund, P.H. Balm, S. Winberg, Skin darkening, a potential social

signal in subordinate Arctic charr (Salvelinus alpinus): the regulatory

role of brain monoamines and pro-opiomelanocortin-derived peptides,

J. Exp. Biol. 203 (2000) 1711–1721.

[9] E. Hfglund, N. Kolm, S. Winberg, Stress-induced changes in brain

serotonergic activity, plasma cortisol and aggressive behavior in

Arctic charr (Salvelinus alpinus) is counteracted by l-DOPA, Physiol.

Behav. 74 (2001) 381–389.

[10] E. Hfglund, P.H. Balm, S. Winberg, Behavioural and neuroendocrine

effects of environmental background colour and social interaction in

Arctic charr (Salvelinus alpinus), J. Exp. Biol. 205 (2002) 2535–2543.

[11] J. Kaslin, P. Panula, Comparative anatomy of the histaminergic and

other aminergic systems in zebrafish (Danio rerio), J. Comp. Neurol.

440 (2001) 342–377.

[12] O. Lepage, a. averli, E. Petersson, T. Jarvi, S. Winberg, Differential

stress coping in wild and domesticated sea trout, Brain Behav. Evol.

56 (2000) 259–268.

[13] S. Levine, H. Ursin, What is stress? in: M.R. Brown, G.F. Koob, C.

Rivier (Eds.), Stress: Neurobiology and Neuroendocrinology, Marcel

Decker, New York, 1991, pp. 3–21.

[14] D.C. McIntyre, P. Kent, S. Hayley, Z. Merali, H. Anisman, Influence

of psychogenic and neurogenic stressors on neuroendocrine and

central monoamine activity in fast and slow kindling rats, Brain Res.

840 (1999) 65–74.

[15] E.Y. Mok, A.D. Munro, Effects of dopaminergic drugs on locomotor

activity in teleost fish of the genus Oreochromis (Cichlidae): involve-

ment of the telencephalon, Physiol. Behav. 64 (1998) 227–234.

[16] a. averli, C.A. Harris, S. Winberg, Short-term effects of fights for

social dominance and the establishment of dominant–subordinate

relationships on brain monoamines and cortisol in rainbow trout,

Brain Behav. Evol. 54 (1999) 263–275.

[17] J.B. Overmier, N.B. Flood, Passive avoidance in forebrain ablated

teleost fish, Carassius auratus, Physiol. Behav. 4 (1969) 791–794.

[18] J.B. Overmier, M.R. Papini, Factors modulating the effects of

teleost telencephalon ablation on retention, relearning, and

extinction of instrumental avoidance behavior, Behav. Neurosci.

100 (1986) 190–199.

[19] R.A. Rescorla, R.L. Solomon, Two process learning theory: relation-

ships between pavlovian conditioning and instrumental learning,

Psychol. Rev. 74 (1967) 151–182.

[20] G.R. Scott, K.A. Sloman, C. Rouleau, C.M. Wood, Cadmium disrupts

behavioural and physiological responses to alarm substance in

juvenile rainbow trout (Oncorhynchus mykiss), J. Exp. Biol. 206

(2003) 1779–1790.

[21] C.H. Summers, T.R. Summers, M.C. Moore, W.J. Korzan, S.K.

Woodley, P.J. Ronan, E. Hfglund, M.J. Watt, N. Greenberg, Temporal

patterns of limbic monoamine and plasma corticosterone response

during social stress, Neuroscience 116 (2003) 553–563.

[22] H. Ursin, The effect of amygdaloid lesions on flight and defense

behavior in cats, Exp. Neurol. 11 (1965) 61–79.

[23] H. Ursin, The psychology in psychoneuroendocrinology, Psychoneur-

oendocrinology 23 (1998) 555–570.

[24] H. Ursin, H.R. Eriksen, The cognitive activation theory of stress,

Psychoneuroendocrinology 290 (2004) 567–592.

[25] H. Ursin, P. Linck, R.A. McCleary, Spatial differentiation of

avoidance deficit following septal and cingulated lesions, J. Comp.

Physiol. Psychol. 68 (1969) 74–79.

[26] S. Winberg, A.A. Myrberg, G.E. Nilsson, Predator exposure alters

brain serotonin metabolism in bicolour damselfish, NeuroReport 4

(1993) 399–402.

[27] C.M. Woodley, M.S. Peterson, Measuring responses to simulated

predation threat using behavioral and physiological metrics: the role of

aquatic vegetation, Oecologia 136 (2003) 155–160.