• No results found

The relationship between behaviour and metabolic rate of juvenile Brown trout Salmo trutta

N/A
N/A
Protected

Academic year: 2021

Share "The relationship between behaviour and metabolic rate of juvenile Brown trout Salmo trutta"

Copied!
1
0
0

Loading.... (view fulltext now)

Full text

(1)

Faculty of Social and Life Sciences

Johanna Bengtson

The relationship between

behaviour and metabolic rate of

juvenile Brown trout

Salmo trutta

Länken mellan ämnesomsättning och

beteende hos bäcköring Salmo trutta

Biology

D-level thesis

Date/Term: 14-04-2009

(2)

Examiner: Monika Schmitz

(3)
(4)

The relationship between behaviour and metabolic rates of juvenile brown

trout (Salmo trutta)

ABSTRACT

In salmonids, the decision to migrate or remain resident is influenced by the status, and hence condition, of individuals. Status has been suggested to arise from the temperament of fish. In this study the links between standard metabolic rate and the levels of aggressiveness and shy/boldness were examined for 0+, hatchery-raised brown trout (Salmo trutta). I hypothesized, from the results of earlier studies (Cutts

et al., 1998; Yamamoto et al., 1998), that high metabolic rates (MR) would be

positively correlated to levels of aggression and boldness. The study was conducted in 200 L aquaria in which aggressiveness was measured by allowing each fish to interact with a mirror image of itself, and shy/boldness was tested by measuring the amount of time a fish used before exploring a new area. Standard metabolic rate was measured in a flow-through respirometer. In contrast to my expectations, there was no correlation between the different behavioural measures and the metabolic rate of fish. Also, no correlation between boldness and aggressiveness of fish was found. In additional testing aggressiveness correlated positively with the condition of fish (in coherence with Harwood et al., 2003) but, contrary to earlier studies (Överli et al., 2004; Schjolden & Winberg, 2007), not with the speed of acclimatization. The difference in results between this test and earlier studies, concerning the degree of correlation between MR and aggressiveness, suggests that the strength of this link differs between species of salmonids. Also, it may suggest changeability in the MR – behaviour link in different environments. Last, the status and condition of individuals cannot be unambiguously explained by temperament alone, but arise from a wider array of physiological and environmental factors.

SAMMANFATTNING

Salmoniders val att migrera eller stanna kvar i sitt födelsehabitat påverkas av individens status och den fysiska kondition som är kopplad till denna. Fiskens temperament har i sin tur föreslagits som en viktig faktor för dess status. I denna studie undersöks länken mellan ämnesomsättning och graden av aggressivitet och mod hos en årsgammal kull bäcköring (Salmo trutta) från odling. Hypotesen, baserad på tidigare studier

(Cutts et al., 1998; Yamamoto et al., 1998)

, var att aggressivitet och mod korrelerar positivt med ämnesomsättning. Försöksfiskarna testades individuellt i 200 l akvarier där aggressiviteten testades genom en simulerad konkurrenssituation, med hjälp av en spegel, och individernas mod bedömdes utifrån deras snabbhet att gå ut i, och exploatera födotillgångar i, ett främmande utrymme. Basal ämnesomsättning (standard metabolic rate) hos öringarna mättes med en ”flow-through” respirometer. I motsats till mina antaganden hittades inga signifikanta korrelationer mellan beteenden och ämnesomsättning. Inte heller kunde någon korrelation mellan aggressivitet och mod hittas. I vidare jämförelser kunde man se en positiv korrelation mellan aggressivitet och kondition, även konstaterat hos

Harwood et al. (2003)

, medan, i kontrast till tidigare studier av Schjolden & Winberg (2007), korrelationen mellan aggressivitet och acklimatisering saknades. Skillnaden i resultat mellan detta och tidigare tester, vilka gjorts på lax, antyder att länken mellan beteende och ämnesomsättning skiljer sig åt mellan arter. Graden av samband mellan ämnesomsättning och beteenden kan också variera mellan olika miljöer och förutsättningar. Individers status och kondition kan inte heller tillskrivas enstaka faktorer som inneboende temperament, utan kan snarare antas uppstå som en kombination av ett sort antal fysiologiska och miljömässiga förhållanden.

(5)

INTRODUCTION

BACKGROUND

Bridging the gap between understanding individual variation in behaviour and

physiology of individuals and the decision to migrate or remain resident is an area

of considerable interest in salmonid ecology. Previous studies of salmonid

migration have shown that fish in dense populations, and individuals displaying

social dominance and a high growth rate are the most prone to migrate (Metcalfe

et al., 1989; Morita

et al.

2000). Migration, in any species, generally occurs when

the risk of starvation and demise in the present habitat overweighs the danger and

energy expense of migration. The circumstance that dominant individuals, who

generally have an advantage in gaining food, are the first to reach this threshold

can be explained by a relationship between social status and metabolic rate

(Metcalfe

et al.

1995)

A high metabolic rate (hereafter called MR) is associated with high

maintenance costs, making the fish vulnerable to dwindling food supplies

(Metcalfe et al., 1995). MR has also been tied to traits crucial for the competitive

success and status of the fish: previous ownership (of territory), winning

experience and fighting ability (Yamamoto et al., 1998; Johnsson et al., 2000,

Harwood et al., 2003). A high metabolic rate forces the fry to rise earlier from the

gravel bed to feed, as the contents of its yolk sac is exhausted, giving it a head

start in occupying a territory and using the assets to increase growth rate (Cutts et

al., 1998; Harwood et al., 2003). Fighting ability at this age determines the

outcome of the first bouts, and the perceived status of the parr becomes a factor in

subsequent competition (Harwood et al., 2003). Metcalfe (1995) holds the MR as

the primary and immediate factor for status in salmon fry. According to Metcalfe

the increased rate of conversion of food to “body power”, in a high MR fish,

allows it to become involved in energy demanding activities such as fighting

(Metcalfe et al., 1995). The fighting ability, in turn, is considered to be strongly

influenced by aggressiveness (Huntingford et al., 1990).

It is argued that the energy demands of high MR fish create a need for a

reliable and uncontested food source, making the fish extremely motivated to

defend this resource. This motivation is thought to cause the aggressiveness of

high MR fish (Johnsson & Björnsson, 1994). However, in 2002, Cutts et al.

concluded that aggressive behaviour is more likely to be a consequence of the

fish’s inherent temperament than of its motivation, since high MR fish had a

lower feeding motivation than the low MR fish, in solitude. This suggests a more

direct link between aggressiveness and metabolic rate, not involving the step via

feeding motivation.

Given that metabolic rate and aggressiveness may be linked beckons the

question if other behavioural traits may be linked to metabolic rate and/or

aggression. One such trait is that of shyness/boldness, a concept used to describe

the basis of behavioural traits or “personality” of humans, other mammals and

fish. This concept describes differences in individual aggressiveness and also

entails the willingness to explore and take non-social risks (here, the “boldness” of

fish) (Sneddon, 2003). The relative expression of shyness to boldness balances the

risks of elevated energy expenditure and exposure to predation against possible

benefits of exploration, foraging and competition, which has consequences for the

physical condition, hence survival and reproduction, of wild trout. (Jakobsson et

(6)

al., 1995; Reinhardt, 1999) Generally, studies of animals (including humans)

other than fish have found that aggressiveness and boldness are positively

correlated with each other (Abrahams & Sutterlin, 1999; Frost et al., 2007). In

fish it is not clear that there is a relationship between aggressiveness and boldness

(Reinhardt, 1999; Wilson and Stevens, 2005). Reinhardt (1999), Wilson and

Stevens (2005) and Frost

et al.

(2007) all suggest there may be a relationship but

that this relationship is context dependent. To my knowledge, no studies have

tested for a relationship between MR and boldness in fish.

PURPOSE OF THE STUDY

The purpose of this study was to test if there is a relationship between the

individual behaviour and physiology of under-yearling (0+) brown trout (Salmo

trutta).

Specifically, I hypothesized that there is a positive correlation between

metabolic rate (MR) and aggressiveness (Hypothesis one). High MR fish were

therefore expected to display a higher frequency and intensity of aggressive acts

in a simulated competitive situation than the low MR fish. I also hypothesized that

there is a positive correlation between MR and exploratory boldness (Hypothesis

two). Thus the high MR fish were expected to more readily (faster) explore a

novel environment then the low MR fish. This test was conducted in a laboratory

environment, using hatchery-raised fish, comparing the boldness, aggression and

MR of individual fish.

MATERIALS AND METHODS

TEST STOCK AND FACILITY

The test stock consisted of 0+ brown trout, first generation in captivity,

originating from a mixed batch of wild-caught parent-fish. The parr were hatched

and reared in outdoor, flow-through tanks at the Fortum Energy fish hatchery in

Gammelkroppa, Sweden. The trout were transported to Karlstad University in

oxygenated, plastic 50 L containers. Upon arrival they were placed in opaque

plastic acclimatization-troughs, where water temperature was slowly raised from

5

o

C to 10

o

C, which was the temperature subsequently used during testing.

Lighting was kept on a diurnal 8h daylight / 14h darkness cycle, with gradual 1h

transitions, simulating dusk and dawn.

After acclimatization, all fish were marked with PIT tags (11*2 mm, Trovan).

This was accomplished after anaesthetizing the fish using Ethyl 3-aminobenzoate

methane sulfonate salt (MSS). In addition, I weighed them (to + 0.1 g), measured

(fork length (Ricker 1979) to + 0.5 mm), and calculated condition factor

(100*mass/(length

-3

)) (Sundström et al., 2003).

Trout were collected from the hatchery on two separate occasions. The first

group of fish (batch 1) was collected on 23 January 2008 and was kept in 600 L

glass aquaria at the facility until the onset of experimenting in March. The other

group (batch 2) was collected at the end of March and kept in storage tanks for

four to nine weeks, before used in experiments. These aquaria were fitted with a

gravel substrate and some shelter, in the form of clay pots and/or stones, also

creating some heterogeneity in the weak current generated from the pump. The

tank had three opaque sides and a semi-transparent, dark sheet of solar film

(7)

(commonly used on car windows) covered the fourth side, facing the observer.

The top of the aquarium was covered by a black opaque plastic sheet. Food, an

approximate maintenance ration of chironomidae larvae, was added once a day

through a fixed plastic tube in the outer corner of the aquarium, near the water

inlet.

The tanks used in the experiments were twelve 200 L glass aquaria with three

opaque sides. The fourth side (facing the observer) was partly covered by solar

film, allowing for observations. Black plastic sheets were used to screen off the

top of the aquaria from movements in the room. The tank was divided into two

compartments (a “home” section measuring 45*45*45 cm and a second “barren”

section, hereafter referred to as the outer section, measuring 45*45*55 cm) by an

opaque plastic wall (Fig. 1). There was an opening in the wall, measuring 30*24

cm (30*20 cm effective area, limited by water depth), covered by a sliding-door

that could

be opened without the observer being revealed to the fish.

In the home

section, where the water inlet was situated, the bottom of the tank was covered by

gravel substrate, a clay pot for the fish to hide in and a silk plant. The outer

section of the tank was bare, with a white floor. The front side of the “barren”

outer section was not covered by the solar film. Water depth was about 24 cm in

all aquaria, creating a water volume of 110 L in each tank. The water current in

the home section was circulated at a low speed.

(8)

EXPERIMENTAL PROCEDURE

Before the onset of the first experiment the fish were allowed to acclimatize in

the testing-tanks for seven days. The fish were fed four times daily at 10.00 a.m.,

12.00, 14.00 and 16.00 p.m., and their response to feeding was categorized on a

four grade scale: Motionless in shelter and no attempts to feed gave 0 points.

Feeding on nearby food particles without leaving the shelter received 1 point.

Feeding actively, yet returning to the shelter between each prey capture, gave 2

points and active feeding independent of shelter gave 3 points (Överli et al., 2004.

Each fish was observed for five minutes or until all larvae were eaten, whichever

came first. Notes were also taken of the fish’s position (resting on bottom or

swimming) between feedings. The daily rations were calculated to 2 % of body

weight, which is considered a maintenance ration (Abbot et al., 1988). On the

eighth day, the day before the trials were initiated, the fish were not fed.

When testing for exploratory boldness the water current in the tank was turned

off, and food was carefully placed at the bottom of the aquaria in the distant part

of the outer section (i.e. furthest away from the home section). A small amount of

food was presented in the home section, to activate the fish before opening the

sliding-door. Variables measured were: 1. the time spent in the home section after

opening of the sliding door, 2. the time spent in the outer section before first

feeding attempt and 3. the time from opening of the sliding door to the first

feeding attempt. Fish movements and behaviour during the test were recorded by

videotaping. The test was started at 10.00 a.m. Four tanks were filmed

simultaneously, and the cameras were subsequently moved to the other eight

aquaria as individual tests were completed. Individual tests lasted until the fish

attacked food in the outer compartment and the entire round of testing was

completed within the normal daily time-frame for feeding (10.00 – 16.15 h).

After the boldness experiment a new acclimatization period was initiated, with

a few modifications intended to reduce the stress of the following experimental

procedure. This time the water current was turned off 30 min before the first

feeding and the food was presented in the water column near the partition wall, to

make the fish recognize this as a good feeding spot. During the last two days I

rattled the partition wall slightly during feeding, to make the fish accustomed to

the disturbance created when the sliding-door is opened. The eighth day, the day

before the trials were initiated, the fish were not fed.

The experiment on aggressiveness was conducted in much the same way as the

boldness test. In this test, opening of the sliding door revealed a mirror, fitted just

behind the partition wall (Metcalfe, 1991). Food was distributed only in the home

section of the tank; a small ration (about 2 larvae) just before opening the

sliding-door, eight more just after opening and another ten larvae after ten minutes. The

fish were observed for five minutes following each feeding (after opening the

door), one tank at a time, and variables measured were: 1. portion of time spent on

aggressive acts towards the image in the mirror (Keenlyside & Yamamoto, 1961).

2. number of attacks per five minute period. Here a distinction was made between

two kinds of attacks, both resulting in nipping at the mirror; one fast and violent

“charge-attack” and one prolonged, seemingly low-intensity, “push-attack”. Fish

movements and behaviour during the test were also recorded by videotaping. All

experiments were completed within 10.00 – 16.15 h, when fish were normally fed

(9)

during the acclimatisation test. When the experiments were completed the fish

remained in the tanks for five days before being moved to the respirometry

chambers. The fish were not fed for 24 h prior to respirometry testing.

The standard metabolic rate (SMR) of each trout, i.e. the basal level of

metabolic rate (in inactive fish), was determined by measuring oxygen

consumption of each fish at ten-minute intervals over 24 h. Each fish was put in a

cylindrical glass chamber, 168 mm long and 32 mm in diameter, thereby limiting

its movements. The chamber was connected to a double circuit system, consisting

of one closed water circuit, containing the oxygen measuring probe of a LDAQ-4

respirometer, and one open circuit letting in fresh, oxygenated water between the

measuring periods. Two such systems were placed in one 100-litre aquarium,

using the same water for both systems. The water in the aquaria was circulated to

maintain oxygen levels and passed through a UV-light filter to eliminate

microorganisms, which may affect oxygen consumption. A computer program

controlled the pumping of water into each chamber, alternating between a

five-minute period of measuring the decline in oxygen saturation of closed-circuit

water and five minutes of reoxygenation, six times per hour. This method is often

used because of its accuracy and a low mortality of fish (Steffensen, 1989). A

graph of oxygen concentration over 24 h was produced. From the graph I could

identify stable stretches of the lowest readings (minimum levels of metabolic rate)

to estimate median SMR for each fish. 18 readings, from a stretch in the end of

each testing round (mean 17.1 h, SD = 0.27 h ), were chosen from each graph.

After testing the first twelve fish (originating from batch one) all testing tanks

were emptied and thoroughly cleaned. After refilling the tanks, the experiments

were repeated with twelve fish from the second batch. Thus, in all, 24 fish were

used in experiments.

SPSS and Microsoft Excel were used for subsequent statistical testing. The

scores of the acclimatization test, the boldness test and the aggressiveness test

were plotted against each other and against the weight, condition and SMR of

each fish, using multiple linear regression- and bivariate correlation tests in SPSS.

Outliers were identified with a box-plot and excluded. Normal distribution in the

variables was checked for using the Kolmogorov-Smirnov test and, when needed,

established using a logarithmic transformation of the dataset. All sets of data were

compared for equal variances (f-test) in Excel.

RESULTS

Three of the tested fish were immediately excluded from calculations due to

technical failures in the testing tanks and, in one case, to apparent illness. The

remaining test subjects are included in all or most of the following calculations.

The mean length of the remaining group was 98.1 mm (SE 2.88), mean weight 9.3

g (SE 0.80) and their condition factor was 939.2 (100*mass/(length

-3

)) (SE 11.51).

Mean SMR of all fish was 98.6 mg O

2

/kg/h (SE 2.98). SMR was compared with

weight (regression t=0.503, df=17, p=0.622 ns) and condition factor (regression

t=1.742, df=17, p=0.102 ns), revealing no significant relationships. To determine

if there were differences between batches, all test results and physiological factors

(10)

were compared between the two batches. No significant differences were found

(t-test, all p>0.05 ns). Adjustments concerning normal distribution and outliers were

made in some cases. Two outliers were excluded from the SMR dataset. In the

aggressiveness data, both sets of attack rate measures were log-transformed and

one outlier was excluded from the set of charge attacks. One outlier was excluded

from all the measurements of the boldness test. Adjustments of normal

distribution of the results from the acclimatization testing were unsuccessful,

leading to the use of non-parametric tests. Also, three outliers were excluded from

acclimatization results. All sets of data compared during testing had equal

variances (f-test, all p<0.01 ns)

In the aggression test, 82 % of the fish actively fed, and all fish that fed also

reacted to the mirror, attacking it to some degree. The mean time spent in

aggressiveness was 3 minutes, with a standard error of 0.19 min. Means of attack

rates were: Charge attacks: 4.2 (SE 0.58) Push attacks: 7.2 (SE 1.58). No

significant relationship between the different measures of aggression and SMR

could be found (Table 1).

Table 1. The regression coefficients (t) for the relationships between

SMR, weight and condition factor and the different measures of aggressiveness. Levels of significance are indicated by * = p <0.05 and **= p <0.01. Degrees of freedom excluded listwise (SPSS)

.

df SMR Weight Condition

Time aggressive 15 1.463 -1.112 1.1477

"Charge attacks" 14 1.341 1.736 a -0.430

"Push attacks" 15 0.362 -2.273 * -0.362

a: p=0.118

There was, however, a significant negative relationship between the number of

“push” attacks and the weight of the fish (Fig. 2). Note the difference in patterns

between push- and charge attacks.

There

were no other significant relationships

between weight or condition factor and the different measures of aggression

(Table 1). Comparing the measurements of aggressiveness amongst themselves,

the charge-attacks showed a strong tendency towards correlating with time in

aggressiveness

(regression test: t=2.059, df=16, p=0.059, near sig.), while the

push-attacks did not (regression test: t=0.660, df=16, p=0.520, not sig.).

(11)

There was much between-fish variation in the boldness experiment. The mean

finishing time of the group, after exclusion of one outlier, was 29.1 minutes, with

a standard error of 4.43. No significant correlations between SMR, weight and

condition factor and the different behavioural variables measured were found

(Table 2). In addition, the scores for time spent in the home section and the total

time were compared. A strong positive correlation was found (correlation test:

r=0.08, df=19, p<0.01 sig.), indicating that fish leaving the home section early

were generally the fastest to initiate feeding.

Table 2. The regression coefficients between SMR, weight and condition

factor and the different behavioural-based variables measured in the boldness trials. The “Total time” is measured from hatch-opening to initiation of feeding. Levels of significance (when present) are indicated by * p <0.05 and ** p <0.01.

df SMR Weight Condition

Time in home section 15 0.129 -0.038 -1.330

Time in outer section 15 1.314 -0.077 -1.126

Total time 15 0.701 -0.068 -1.681

Time in outer section/

Total time 15 1.709 a -0.585 0.003

a: p = 0.113

The scores of boldness and aggressiveness were also tested against each other.

No significant correlations could be found (Table 3). However, there were weak

tendencies for the time spent on aggressive acts to correlate with the time spent in

the outer section (correlation r=0.387, df=15, p=0.125) and the portion of total

testing time that was spent in the outer section (correlation test: r=0.405 df=18

p=108) in the boldness test.

Table 3. The correlation coefficients (t) between the results of the

boldness test and the aggression tests. Levels of significance (when present) are indicated by * p <0.05 and **

p <0.01.

Df Time in homesection Time in outersection Total time

Time in outer section / Total time Time aggressive 17 0.051 0.387a 0.234 0.404b Charge-attacks 16 -0.027 0.161 0.059 0.260 Push-attacks 17 0.185 0.031 0.168 -0.017 a: p=0.125 b: p=0.108

During the two acclimatization periods, most of the fish fed without seeking

shelter, receiving top scores on all feedings (see “execution of experiments”)

within four days. No significant correlations (multiple regression, SPSS) were

found between the variables of acclimatisation and the SMR, weight or condition

of fish. However, a strong tendency for correlation was found between the

condition of fish and the total score in period one (regression test: t=2.098, df=16,

p=0.056, near sig.)

(12)

The results from boldness and aggression tests were tested against the fish’s

speed of acclimatisation using the Spearman rank correlation test (non-parametr.).

A positive correlation between the timing of first top score in acclimatization (the

number of days until first daily top score) and the boldness variable Time in home

section was found (Table 4). Thus, fish that failed to feed maximally in

acclimatization were the slower to enter into a novel environment (Figure 3)

.

Figure. 3. The relationship between the “time in the home

section”-variable of the boldness test (the time between opening of the sliding-door and the fish’s passage into the outer compartment) and the number of days spent until first top score, in acclimatization.

Table 4. Correlation coefficients (rs) of Spearman test between scores of

boldness and aggressiveness and the behavioural-based variables of acclimatisation period 1. Acclimatisation variables are listed on the top row. N=17 Levels of significance (when present)

are indicated by * p

<0.05 ** p <0.01.

First top score

period 1 Total score period 1

Aggressiveness Time aggressive 0.094 -0.323

"Charge attacks" -0.083 -0.094

"Push attacks" 0.319 -0.201

Boldness Time in home section 0.498 * -0.335

Time in outer section 0.113 -0.289

Total time 0.435 -0.376

Time in outer section

(13)

DISCUSSION

In the present study, no correlation between metabolic rate and behaviour could be found. Hypothesis one, that there is a positive correlation between standard metabolic rate (SMR) and aggressiveness, and hypothesis two, that there is a positive correlation between SMR and boldness, were both rejected. Thus, the results of the present study does not corroborate the results of the earlier, similar studies, by Cutts et al. (1998, 2002) and Yamamoto et al. (1998), from which my hypothesis’ were derived.

This discrepancy between results may be a reflection of the variation in physiology, behaviour and life-history choices between salmonid species. The studies referred to; Cutts et al. (1998, 2002) and Yamamoto et al. (1998), were conducted on Atlantic salmon (Salmo salar) and Masu salmon (Onchorhynchus masou), respectively, while the present study was conducted on Brown trout (Salmo trutta). Other studies have testified to the differences in aggressiveness (Jonsson & Jonsson, 2005) and in allocation of energy (Hutchinson & Iwata, 1997) between species of trout and salmon. Considering the present results, these differences seem to be accompanied by a variation in strength of correlation between metabolic rate and behaviour. In trout, being the least aggressive of the three species and having a relatively large variation in habitat choice and life-history decisions (Jonsson & Jonsson, 2005; Hutchinson & Iwata, 1997), the physiological and behavioural factors may be less important for individual success, compared to environmental clues and random events. When a trait is having a peripheral and largely random role for the success of individuals it is likely to be less obvious and inferior to other factors. Since environmental factors were not included as a variable in experiments of the present study, their effects could not be controlled for.

Another obvious difference between the present study and earlier studies is that I tested single, isolated individuals. Earlier studies of aggressiveness or dominance and metabolic rate have involved interactions in groups of fish (Cutts et al., 1998; Yamamoto

et al., 1998). I excluded testing two or more fish together to remove the confounding

secondary effects of social hierarchies. However, excluding social interactions may have caused other, unintended, effects on the fish’s responsiveness. As shown in the study by Huntingford et al. (1993) the sight of conspecifics triggers higher intensity and risk-taking in attacks on food. The isolation from other (real) fish during testing may have decreased or disrupted behavioural responses that can be triggered by social clues, such as boldness and feeding (and, possibly, aggressiveness). Such effects could cause inconsistencies in the responsiveness of fish, lowering the degree of correlation. Another potential unintended effect that I may have introduced is through the use of the mirror in the aggression tests, disrupting the normal course of events in territorial bouts. Fights are normally settled as soon as one fish is chased off or an imbalance in fighting ability between combatants is recognized (Huntingford et al., 1993). When battling a mirror, none of this will take place, possibly triggering the fish to excessive aggression or other unexpected reactions that may not be related to its metabolic rate. No clues to the relative influence of such effects were discovered when studying and processing the data.

An additional aspect of this study that may have been responsible for my lack of significant effects between behaviour and metabolic rate was my use of hatchery fish. Effects of social interaction and hatchery selection, so called “hatchery effects”, creating a skew in the order of physiological status in the population, has been shown in several studies (Cutts, 1998; Brännäs, 2002). The potential expression of such effects, in this study, encompasses symptoms of growth depensation and stress (Cutts, 1998; Brännäs, 2002) as a consequence of aggressive interaction and food monopolization of dominants. It has been shown that aggression is most intense between fish of high absolute status (Cutts, 1998). In accordance with this reasoning, symptoms will be most prominent in high MR fish loosing their battles (Metcalfe et al., 1995), causing bimodality in size and,

(14)

most likely, in MR. No such bimodality has been seen in the test subjects, suggesting that effects of the hatchery environment are neglectable in this study.

The negative correlation between the number of “push attacks” and weight, does not concur with the reported relationship between aggressiveness and size (Huntingford et

al., 1990; Harwood et al., 2003). Meanwhile, “charge attacks” and weight showed no

negative, but rather a slight positive, tendency to correlate. This suggests that the attack strategies, “pushing” and “charging”, are two separate, size-dependent behaviours. The strategies may therefore be based on different causes and stimuli. The “charge” tactic may, for example, be a sign of higher aggression levels compared to the “push” tactic. The “push” attacks were, as it seemed to observers, somewhat resembling the escape attempts sometimes occurring between tests, only more ferocious. This suggests that various behaviours traditionally associated with aggressiveness (Keenlyside & Yamamoto, 1961; Hutchinson & Iwata, 1997; Johnson & Åkerman, 1998) can be interpreted in a number of different ways. This may cause contradictions between the results of aggressiveness tests based on different behavioural variables.

Fish that failed to feed maximally in acclimatization were the slower to enter into a novel environment in the boldness test. This is in concurrence with the conclusions of Överli et al. (2004) and Schjolden & Winberg (2007) stating that aggressive and bold fish are the fastest to resume feeding after transfer to a new environment. This was the single significant correlation in the entire set of comparisons between aggressiveness and boldness, and acclimatization. However, several other relations in the comparison reached p-values around 0.1-0.3, suggesting that significant results may have been reached using an extended sample.

In comparisons involving acclimatization scores, only the results from acclimatisation period one was used. Acclimatisation procedures were adapted to resemble the procedures of the following tests, causing differences in experimental procedure between the first and second period. Therefore, periods were not comparable. The first period was chosen for calculations as it was assumed to be less influenced by effects of learning and/or stress reactions and was the period most resembling “natural” conditions. Also, the usefulness of adding a small portion of food prior to the boldness test (to alert the fish), can be questioned. Rather, it could influence the test by lowering the feeding motivation of the fish. However, fish in acclimatization rarely decreased its interest in food after eating one or two larvae, and since all fish did enter the novel compartment sooner or later, no distinct effect of this feeding could be identified.

In comparisons between aggressiveness and boldness, the time in the outer section showed a weak correlation to the time spent in aggressiveness. No other correlations in the comparison were found significant, suggesting differences between the variables within the behavioural tests. Evaluating the variables of the significant correlation one note from the observation protocol is worth mentioning: To observers, the main activity of fish in the outer section seemed to be monitoring of the compartment (and possibly its surroundings, since this section lacked solar film). This may indicate that other, unidentified behaviours overrode the urge to feed in this case. The individuals’ skill in identifying the food as food was not measured precisely, but was interpreted as adequate.

No significant correlations between aggressiveness and boldness were found in this study. Earlier studies, too, have failed to find a strong and coherent correlation of this sort (Wilson & Stevens, 2005). A lack of correlation between aggressiveness and boldness in the basic temperament of fish may allow higher flexibility in the trade-off between holding territory (aggressiveness promotes access to food) and staying safe (boldness threatens survival). Reinhardt (1999) showed how aggression hierarchies were disrupted during predation threat, where the relative investment in aggressiveness, compared to

(15)

boldness, may change with the perceived level of threats. Thus, the present results corroborate the notion of Wilson & Stevens (2004) and Reinhardt (1999) that there is no strong, genetically determined correlation between aggressiveness and boldness, in fish.

The lack of correlation between metabolic rate and behaviour, shown in this study, is contradictive to the patterns found in earlier studies. The difference can be, in large, ascribed to variation in physiology and life-history choices between the species used in experimenting. Social and environmental clues of the testing environment are also likely to have contributed to the result. Thus, the link between MR and behaviour is not consistently present in all species and environments. However, the question as to what stimulates this relationship between traits remains.

Likewise, the life history choices of fish can not be unambiguously attributed to a single cause as the temperament or even the metabolic rate of individuals. Individual condition and energy demand are important factors in life history choices of salmonids and such traits may arise from a number of environmental and physiological factors. Nonetheless, the arguments for MR as a prominent basis of these traits are strong and several other theorized pathways between MR and migration remain to be studied.

ACKNOWLEDGEMENTS

I thank Linnea Lans, who initiated the project, for welcoming me into the stream-water laboratory and the experiment, and my supervisor, Larry Greenberg, for not losing patience with my constant struggle to write a coherent report. Thanks to Björn Arvidsson, Sven-Åke Bood and Jonas Andreen, also, for guiding me through the statistical processing in SPSS.

(16)

REFERENCES

Abbot, J. C & Dill, L. M. 1988 The relative growth of dominant and subordinate juvenile steelhead trout (Salmo gairdneri) fed equal rations Department of Biological Sciences, Simon Fraser University 1988

Abrahams & Sutterlin 1999 The foraging and anti-predatr behaviour of growth-enhanced transgenic Atlantic salmon Animal Behaviour 58: 933-942

Brännäs , E et al. 2002 Aggression and growth as an effect of size composition in groups of Arctic charr (short communication) Journal of Fish Biology 60: 1331-1334

Cutts, C. J., Metcalfe, N. B., and Taylor, A. C. 1998 Aggression and growth depression in juvenile Atlantic salmon: the consequences of individual variation in standard metabolic rate. Journal of fish biology 52: 1026-1037

Cutts, C. J., Metcalfe, N. B. & Taylor, A. C. 2002 Fish may fight rather that flee in a novel environment: metabolic rate and feeding motivation in juvenile Atlantic salmon Journal

of fish biology 61: 1540-1548

Frost et al. 2007 Plasticity in animal personality traits: does prior experience alter the degree of boldness? Proceedings of the Royal Society B 274: 33-339

Harwood, A. J., Griffiths, S. W., Metcalfe, N. B. & Armstrong, J. D. 2003 The relative influence of prior residency on the feeding behaviour of juvenile Atlantic salmon. Animal

Behaviour 65: 1141-1149

Huntingford, F. A., Metcalfe, N. B., Thorpe, J. E., Graham, W. D. and Adams, C. E. 1990 Social dominance and body size in Atlantic salmon parr, Salmo salar L. Journal of Fish Biology 36: 877-881

Huntingford, F. A., Metcalfe, N. B. & Thorpe, J. E. 1993 Social status and feeding in Atlantic Salmon Salmo salar Parr: The effect of Visual Exposure to a Dominant Ethology 94: 201-206

Hutchinson, M. J. & Iwata, M. 1997 A comparative analysis of aggression in migratory and non-migratory salmonids. Environmental Biology of Fishes 50: 209-215

Jakobsson, S., Brick, O. & Kullberg, C. 1995 Escalated fighting incurs increased predation risk

Animal Behaviour 49: 235-239

Johnsson, J. I. et al. 2000 Fitness-related effects of growth investment in brown trout under field and hatchery conditions Journal of Fish Biology 57: 326-336

Johnsson, J. I. & Björnsson B. T. 1994 Growth hormone increases growth rate, appetite and dominance in juvenile rainbow trout, Oncorhynchus mykiss Animal behaviour 48: 177-186

Jonsson, B. & Jonsson, N. 2005 Lipid energy reserves influence life-history decision of Atlantic salmon (Salmo salar) and brown trout (Salmo trutta) in fresh water Ecology of freshwater

fish 14: 296-301

Johnsson, J. I. & Åkerman, A. 1998 Intersexual differences in aggression in juvenile rainbow trout Journal of Fish Biology 52: 1292-1294

(17)

Keenlyside, Miles H. A. & Yamamoto, Fred T. 1961 Territorial behaviour of juvenile atlantic salmon (Salmo Salar L.) Fisheries Research Board of Canada, Biological Station, St.

Andrews, N.B. 1961

Metcalfe, N. B. 1991 Competitive ability influences seaward migration age in Atlantic salmon

Canadian Journal of Fish Biology 69: 815-817

Metcalfe, N. B. et al. 1989 Early status and the development of life-history strategies in Atlantic salmon. Proceedings of the Royal Society of London 236: 7-19

Metcalfe, N. B., Taylor, A, C. & Thorpe, J. E. 1995 Metabolic rate, social status and life-history strategies in Atlantic salmon Animal behaviour 49: 431-436

Morita et al. 2000 Extreme life history change of white-spotted char (Salvelinus leuchomaenis) after damming Canadian Journal of Fish Biology 57: 1300-1306

Reinhardt, U. G. 1999 Predation risk breaks size-dependent dominance in juvenile coho salmon (Oncorhynchus kisutch) and provides growth opportunities for risk-prone individuals

Canadian Journal of Fish Aquatic Science 56: 1206-1212

Ricker W. E. 1979 Growth rates and models Fish Physiology 8: 677-743

Schjolden, J & Winberg, S 2007 Genetically Determined Variation in Stres responsiveness in Rainbow trout: Behaviour and Neurobiology Brain, Behaviour and Evolution 70: 227-238

Sneddon, L. U. 2003 The bold and the shy: individual differences in rainbow trout. Journal of Fish Biology 62: 971-975

Steffensen, John Fleng 1989 Some errors in respirometry of aquatic breathers: how to avoid and correct for them. Fish Pysiology and biochemistry 1: 49-59

Sundström, L. F. et al. 2003 Hatchery selection promotes boldness in newly hatched brown trout (Salmo trutta): implications for dominance Behavioural Ecology 15 (2): 192-198

Wilson, A. & Stevens, E. D. 2005 Consistency in Context-specific Measures of Shyness and Boldness in Rainbow Trout, Oncorhynchus mykiss Ethology 111: 849-862

Yamamoto, T., Ueda, H & Higashi, S. 1998 Correlation among dominance status, metabolic rate and otolith size in masu salmon Journal of fish biology 52: 281-290

Överli, Öivind, Korzan, Wayne, J. et al. 2004 Stress coping style predicts aggression and social dominance in rainbow trout Hormones and behaviour 45: 235-241

References

Related documents

Here, we investigated whether over-winter loss of biomass and lipid content of juvenile roach (Rutilus rutilus) was affected by the physiologically relatively small (2-5 uC) changes

The POL syndrome is a framework connecting life-history traits (traits describing the characteristics of the life cycle of an organism) with behavioural and physiological traits,

observations of breeding, to see if the breeding among the brown trouts in the sample match any distribution. Thus the censored observations are excluded. Figure 3: The

Farbridge KJ, Leatherland JF (1992) Temporal changes in plasma thyroid hormone, growth hormone and free fatty acid concentrations, and hepatic 5'-monodeiodinase activity, lipid

Together the results presented in this thesis shows that elevated, suboptimal temperature and nutritional may have propound effects on important processes as growth, food intake

My results also show there was no difference in the spatial distribution between night and day in the presence of ice cover, but in the absence of ice cover trout were more

The studies included in this thesis provide, for the first time to my knowledge, empirical data on performance of hatchery produced fish and their progeny under

Here, we investigated if reef fish assemblages in the tropical south-western Atlantic demonstrate ontogenetic habitat connectivity and identify possible nurseries on