Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Jun 8.
Published in final edited form as: J Comp Physiol B. 2009 Aug 25;180(2):211–220. doi: 10.1007/s00360-009-0395-8

Behavioral and respiratory responses to stressors in multiple populations of three-spined sticklebacks that differ in predation pressure

Alison M Bell 1,✉, Lindsay Henderson 2, Felicity A Huntingford 3
PMCID: PMC4899084  NIHMSID: NIHMS789479  PMID: 19705129

Abstract

Individual animals of the same species inhabiting environments which differ in the frequency and magnitude of stressors often exhibit different physiological and behavioral responses to stressors. Here, we compare the respiratory response to confinement stress, and behavioral responses to ecologically relevant challenges among sticklebacks from 11 different populations varying in predation pressure. We found that sticklebacks from high predation populations breathed faster in response to confinement stress and were, on an average, more behaviorally responsive to a pike behind glass compared with sticklebacks from low predation populations. These patterns differ from the results of studies on other species, highlighting the need for a conceptual framework to understand the proximate and ultimate factors shaping variable responses to stressors over developmental and evolutionary time. Moreover, physiological and behavioral responses were integrated with each other, both at the individual and population levels. In general, fish that were more aggressive and bold in the presence of a predator breathed faster, independent of body size. These results are consistent with the growing body of evidence that individuals differ in a suite of physiological and behavioral mechanisms for coping with challenges in the environment.

Keywords: Aggression, Antipredator behavior, Behavioral syndrome, Geographic variation, Personality, Ventilation rate

Introduction

Not every individual within a species responds to challenges in the same way. When confronted by a predator, for example, some individuals rapidly mobilize a physiological stress response—levels of catecholamines increase in the brain, the animal breathes faster and glucocorticoids are released—while other individuals show a weaker physiological response, or recover faster once the stressor is removed (Cockrem and Silverin 2002b; Ebner et al. 2005; Koolhaas et al. 1999; Koolhaas et al. 1997; Wada et al. 2008). In some cases, intraspecific variation in stress responses can be attributed to differences in body size (e.g., Bell et al. 2007), sex (e.g., Pottinger et al. 1995), age (Heidinger et al. 2006; Otte et al. 2005) and experience (e.g., Bartolomucci et al. 2005; Sloman et al. 2001), or genetic background (e.g., Evans et al. 2006; Pottinger and Carrick 1999). There is also growing evidence that geographic variation in the magnitude and predictability of stressors in the environment is associated with population differences in stress responses. For example, chickadees from disturbed versus undisturbed areas (Lucas et al. 2006), or lizards at the periphery versus the center of the range (Dunlap and Wingfield 1995) differ in physiological responses to stressors (see also Muller et al. 2007).

Three-spined stickleback fish (Gasterosteus aculeatus) are especially well-suited for testing the hypothesis that animals inhabiting environments that vary in risk show contrasting physiological responses to stressors. Marine three-spined sticklebacks colonized freshwater environments in the northern hemisphere at the retreat of the last glaciers ~ 12,000 years before present. The radiation created a naturally replicated experiment of independently derived freshwater populations that evolved similar phenotypes in parallel in response to similar selective pressures (Bell and Foster 1994). The replication, therefore, provides an opportunity to compare not just two phenotypically divergent populations, but multiple, independent populations.

Predators are one of the most important selective pressures for sticklebacks, especially when they are small or young. Evolutionary responses to selection by predators have shaped morphological (Reimchen 1994, 2000) and behavioral (Huntingford et al. 1994) traits. Sticklebacks inhabiting water bodies where there are abundant predators (‘high predation populations’) have more pronounced defensive morphology such as lateral plate armoring and long, serrated spines (reviewed in Reimchen 1994) as well as heightened antipredator behavioral responses compared with sticklebacks from populations with fewer predators in the environment (‘low predation populations’) (reviewed in Huntingford et al. 1994). Given the important selective pressure exerted by predators, and that predators are stressors (Bell et al. 2007; Blanchard et al. 1998; Carere et al. 1999; Cockrem and Silverin 2002a; Eilam et al. 1999; Scheuerlein et al. 2001), we hypothesize that sticklebacks from high predation populations show contrasting physiological and behavioral responses to stress compared with sticklebacks from low predation populations.

But how should we expect physiological indicators of stress to differ between high and low predation populations? On the one hand, we might hypothesize that sticklebacks from high predation populations will be chronically stressed, and therefore, show continuously increased indicators of stress compared with sticklebacks from low predation populations, even in a ‘nonstressed’, baseline condition (as in Bonier et al. 2007; Boonstra et al. 1998; Hik et al. 2001). However, chronic exposure to elevated stress hormones such as cortisol has deleterious consequences for health (Sapolsky et al. 2000), so chronic activation of the hypothalamic–pituitary–interrenal (HPI) axis would be selected against in high predation populations (McEwen and Wingfield 2003).

It is more likely that sticklebacks from high predation populations experience a higher frequency of acutely stressful encounters than fish from low predation populations. That is, considering that predators come and go (Lima and Bednekoff 1999), the frequency of encounters with predators will be greater in high predation environments and thus the frequency with which the HPI axis (and sympathetic system) is activated will also be greater than in a low predation environment. Therefore, natural selection by predators might have favored the ability to mount an appropriate acute response to risk. Indeed, several studies have shown that animals from more dangerous environments show higher acute stress-induced levels of glucocorticoids (Dunlap and Wingfield 1995; Lucas et al. 2006).

Finally, it is possible that animals that have repeatedly been exposed to stressors in the environment (such as in high predation populations) show an attenuated physiological response to stress due to habituation (Brown et al. 2005a; Caldji et al. 2000; Espmark and Langvatn 1985; Romero 2004). Indeed, we know that in trout, repeat exposure to mild stressors can result in acclimation, whereby the animal no longer responds to the stressor because it no longer perceives the stressor as threatening (Pickering and Pottinger 1985). However, if the stressor remains a perceived threat, repeated exposure may result in facilitation, where the response to the original and novel stressors is enhanced (Romero 2004). Our aim in this paper is to generate information that bears on these alternatives by comparing multiple high and low predation populations of sticklebacks. We wish to know whether sticklebacks from high predation populations, which presumably have been repeatedly exposed to predator-induced stress, show a greater or lesser response to challenges compared with sticklebacks from low predation populations.

A sensitive measure of the physiological response to stress is ventilation rate. Ventilation rate is a particularly useful index of stress because it can be measured noninvasively, and, therefore, repeatedly on the same individuals, although its sensitivity as a measure of disturbance means that caution must be exercised in its use (Barreto and Volpato 2004). Ventilation rate is a sympathetic response, which quickly increases in response to stressors (Priede 1985), including predators (Barreto et al. 2003; Cooke et al. 2003; Hawkins et al. 2004; Hojesjo et al. 1999; Johnsson et al. 2001; Metcalfe et al. 1995; Sundstrom et al. 2005), possibly in preparation for a flight response. Baseline ventilation rate, stress-induced ventilation rate and metabolic rate have been linked to different behavioral tendencies (Johnsson et al. 2001; Sloman et al. 2000; Verbeek et al. 2008). For example, individual Atlantic salmon with higher standard metabolic rates tend to be more dominant (Cutts et al. 1998), while ‘reactive’ personality types have higher parasympathetic responses in great tits (Carere et al. 2003; Carere and Van Oers 2004) and mice (Koolhaas et al. 1999; Veenema et al. 2003). Indeed, individual differences in metabolic rate have recently been proposed as an anchor underlying personality variation (Careau et al. 2008).

Here, we assess whether sticklebacks from a series of populations exposed to high and low predation pressure by fishes exhibit systematic differences in key behavioral and respiratory responses to stressful stimuli. At the same time, we ask whether individual differences in all of our measures are correlated with each other, i.e., when individuals that are especially behaviorally responsive to the threat of predation or aggression are also particularly physiologically responsive to a different stressor at a different point in time. Finally, we assess whether the observed behavioral reactions (and correlated physiological response) are likely to represent a stable characteristic of each individual.

Methods

Overview

We measured the stress-induced ventilation rate on individual subadult three-spined sticklebacks from 11 different populations varying in predation pressure by fishes. Individual behavioral responses to an unfamiliar conspecific and to a predator were measured 1 day later.

Sticklebacks were collected in Scotland between 11 August and 25 September 2004 (see Malhi et al. 2006 for geographic information about the sites). The intensity of predation by piscivorous fish (pike, perch, brown trout, rainbow trout, eels and arctic charr) was assessed using historical records and whole-loch seine surveys. We expect that all of the freshwater populations have some shared evolutionary history with pike because pike are widely distributed throughout the region. Eleven different freshwater populations were classified as ‘low predation’ if they did not contain any piscivorous fishes, and were classified as ‘high predation’ if piscivorous fishes were detected (Table 1). Although sticklebacks are also subject to predation by birds and invertebrate larvae, those predators can move between water bodies, making it difficult to determine predation pressure by those predators. Previous work has shown that the populations are genetically differentiated from each other and are approximately 15,000 years (Malhi et al. 2006).

Table 1.

Description of the populations

Population Location Confirmed fish predators
Carbeth Near Glasgow Brown trout, rainbow trout
Balmaha Near Glasgow
Balvormie Near Edinburgh
Blackwaterfoot Isle of Arran Brown trout
Eliburn Near Edinburgh Pike, perch
Greenan Isle of Bute
Machrie Isle of Arran
Mealt Isle of Skye Arctic charr
Mugdock Near Glasgow
Quien Isle of Bute Brown trout
Storr Isle of Skye Brown trout, eel
Open in a new tab

The ‘location’ refers to the general geographic location of the population in Scotland

Experiments were carried out at the Glasgow University Scottish Centre for Ecology and the Natural Environment (SCENE), Rowardennan. Groups of fish (n = 10–40) were maintained in separate flow-through stock tanks (210 L, 90-cm diameter × 33-cm-deep). All of the fish were allowed to adjust to the laboratory for at least 1 month before their behavior was observed. The water temperature in the stock tanks was 9 ± 2°C. The photoperiod was 14L:10D. Fish were fed with frozen bloodworms ad lib daily except on the day of observation, when they were unfed.

Behavioral observations took place between 2 October and 15 December, 2004, when the fish were approximately 3–5 months of age. None of the fish were sexually mature. Glass aquaria were located inside in a U-shaped flume containing water from Loch Lomond at 10 ± 1°C.

Four aquaria in the flume were used for behavioral observations (‘observation tanks’, 40 L, 48 × 32 × 26 cm). Exterior lines on the tanks were divided into 16 equally sized areas. All the observation tanks were surrounded by opaque screens on three sides and were situated next to a window in the flume so that the behavior of the fish could be observed. The windows were covered by a blind with a small opening which allowed the observer to see through the window with minimal disturbance to the fish. Each observation tank contained a plastic plant and a length of pipe (12 cm diameter, 36-cm-tall) that stood vertically on one side of the tank and allowed fish to be introduced into the tank with a minimum of disturbance.

The observation tanks were situated in different compartments (402 L, 220 × 63 × 29 cm) on either arm of the flume. Each of the two compartments contained one of two live pike (Esox lucius, 46, 41 cm standard length) and cloth plants that served as hiding places for the pike. The compartments were fitted with a removable opaque cover that created a dark, shaded area for the pike. The pike rarely left the sheltered area unless the cover and hiding places were removed. The pike were caught by hook and line on 5 February 2004 in the Dubh Lochen near the SCENE Field Station. The two pike were fed dead minnows and dead sticklebacks ad libitum.

Measuring stress-induced ventilation rate

A stickleback was removed from the stock tank and placed in a 1-L beaker filled with water. Opercular beats were immediately counted from an overhead position every 15 s for 1 min producing four repeated measures of stress-induced ventilation. For comparison with baseline ventilation rate, we also report overall stress-induced ventilation per minute. Then, the fish was placed into an observation tank for one night to acclimate to the flume and behavioral observations commenced the following morning.

Measuring baseline ventilation rate

To verify that the observed differences in ventilation rate reflected a response to stress, we measured baseline ventilation rate on undisturbed sticklebacks from a subset of the populations. During normal holding conditions (10°C), five individuals from each of six of the populations (Balmaha, Balvormie, Blackwaterfoot, Carbeth, Eliburn and Machrie) were measured for ventilation rate (beats/min) by eye.

Measuring behavioral responses to an unfamiliar conspecific intruder

We employed a procedure that was designed to simulate a challenge to the resident fish by an intruding conspecific (Bell et al. 2007). A live conspecific (approximately 10% smaller than the resident fish) was removed from a holding tank (49 L, 61 × 31 × 26 cm) containing the 6–7 different fish that were used as intruders. Each intruder was allowed at least 30 min to recover between tests. Different intruders reacted similarly to this situation, and we observed no effect of repeated testing on the behavior of intruders and no trends in behavioral outcomes were evident in successive trials with the same intruder (Bell, personal observation). The intruder was placed into the observation tank with the resident fish. The number of times the resident nipped at the intruder (‘nips’), the total time orienting to the intruder (‘time orienting’) and the number of times the resident raised its dorsal spines (‘spines’) was recorded for 5 min after the resident first oriented to the intruder. After the behavioral observation, the intruder was removed from the tank and the resident fish was presented with the pike 4 h later.

Measuring behavioral responses to a predator

This procedure was designed to simulate a potential predatory threat by a live pike. A chamber to contain the pike (61 × 32 × 22 cm) was added to the flume and positioned lengthwise next to the observation aquarium. The chamber had an opaque top and bottom and clear sides. One of the sides had a removable door. A removable opaque divider was situated between the observation aquarium and the predator chamber. Prior to observing an individual stickleback’s behavior, the pike was transferred into the chamber by removing the cover over the pike compartment of the flume and opening the door to the chamber. In general, the pike willingly swam into the chamber, seeking cover. After entering the chamber, the door was closed and 500 mL of water from the pike’s holding area was added to the aquarium, providing olfactory cues of predation risk.

The divider separating the observation aquarium from the chamber was gently removed allowing the stickleback to have a clear view of the pike. The behavior of the individual stickleback was observed for 5 min after the divider was removed and the following behaviors were recorded: the number of times the focal fish moved into a different area of the tank (moves), the number of times the focal fish raised its dorsal spines (spines), the number of times the focal fish swam up to the mouth of the pike (inspection) and total time orienting toward the pike (time orienting). After the behavioral observation, the opaque divider separating the chamber from the observation aquarium was replaced. To eliminate any olfactory cues that might affect subsequent behavioral observations, the water in each of the observation tanks was replaced after each behavioral observation.

After behavioral testing, the fish were measured for weight and a small amount of tissue from the tail fin was placed in 80% ethanol for DNA extraction for sex determination. DNA was extracted from each fin clip and genetic sex was determined by genotyping each individual for a male-specific genetic marker validated for sticklebacks (Peichel et al. 2004).

Estimating repeatability

To test whether individual differences in behavioral reactions to the pike were stable across time, a small subset of individuals (n = 4–7) from one of the populations (Balmaha) was assayed for their reaction to the pike on several occasions (3–7) over 4 months with at least 7 days between observations.

All of the procedures were carried out according to the institutional guidelines and in accordance with the UK Animals (Scientific Procedures) Act of 1986.

Data analysis

To compare the time course of stress-induced ventilation rate, a repeated measures model was used with the following factors: weight, predation pressure (high or low predation) and population (nested within predation pressure). General linear models were used to compare behavior. Many of the behavioral variables were not normally distributed or were heteroscedastic, but were resistant to transformation. Although GLM is relatively robust to violations of normality and homoscedasticity (Zar 1999), we also report the results of the nonparametric Mann–Whitney U test comparing high and low predation populations. Because we did not have a priori predictions about causal relationships between stress-induced ventilation rate and behavior, we used partial correlations (controlling for population and length) to look for relationships between the different responses across all the individuals, and Spearman correlations to detect relationships among population means. The sequential Bonferroni procedure was applied to correct for multiple comparisons (Rice 1989). Repeatability of each of the behaviors was estimated as in Lessells and Boag (1987). All analyses were carried out in SPSSv16, and all tests are two-tailed.

Results

On an average, 16 individuals (range = 3–20) per population were measured for stress-induced ventilation frequency, and 19 individuals (range = 18–20) per population were measured for behavior. Fish from low predation populations were larger than fish from high predation populations (weight: 0.50 ± 0.19 g, n = 98 vs. 0.38 ± 0.15 g, n = 120; length: 33.76 ± 5.76 mm vs. 30.44 ± 3.83 mm). There were no detectable sex differences in stress-induced ventilation rate, behavior or body size, therefore sex was not considered in subsequent analyses. Stress-induced ventilation rate was not related to body size (Table 2).

Table 2.

Between-subjects effects on repeated measures of stress-induced ventilation rate

Source df F Sig.
Intercept 1 476.01 0.000
Weight 1 0.59 0.442
Population (predation pressure) 9 7.21 0.000
Predation pressure 1 6.84 0.010
Error 159
Open in a new tab

The effect of population was tested by nesting it within high or low predation pressure

Do high and low predation populations differ in baseline ventilation rate?

Baseline ventilation rate was 72 beats/min (standard deviation = 6.21, n = 30). We did not detect population variation in baseline ventilation rate (population nested within predation pressure: F1,24 = 2.055, P > 0.05), or a significant difference between high and low predation populations (F1,4 = 0.265, P > 0.05).

Does stress-induced ventilation rate differ between high and low predation populations?

Stress-induced ventilation rate was highest during the first 15 s following transfer to the beaker, and then decreased over the next three 15-s intervals (Fig. 1). The average number of opercular beats during the 1-min observation period was 108 beats, with a maximum of 174 beats/min, which is comparable to stress-induced ventilation rates in other fishes (see Brown et al. 2005a).

Fig. 1.

Fig. 1

Open in a new tab

The number of opercular beats in the four different 15-s intervals (op1-4) as a function of population type. Low predation populations are in gray, high predation populations are in black. Values show mean ± SE

There was substantial among-population variation in stress-induced ventilation rate, some of which could be attributed to predation pressure: fish from populations with fish predators breathed faster than fish from low predation populations (Fig. 1; Table 2). There was not a significant between-population difference in the time course of the response (sphericity assumed, time × predation pressure F3,477 = 0.114, P = 0.952). Therefore, for simplicity, we focus on the number of opercular beats during the first 15-s interval (peak rate).

Does behavior vary between high and low predation populations?

When presented with an unfamiliar conspecific, some individuals nipped at the intruder as many as 39 times in 5 min, while other individuals did not nip at all. Similarly, individuals differed in the frequency with which they raised their spines and oriented toward the intruder. Overall, there were significant between-population differences in the aggressive responses toward the intruder, independent of body size (Table 3). However, there was not a detectable systematic difference in the one aggressive behavior of fish from populations with or without piscivorous fishes according to either the GLM or Mann–Whitney U test (Tables 3, 5).

Table 3.

Effects of population (predation pressure), predation pressure and body size on aggressive behaviors

Source Dependent variable df F Sig.
Intercept Spines 1 8.32 0.004
Nips 1 2.31 0.131
Time orienting 1 60.44 0.000
Weight Spines 1 0.57 0.453
Nips 1 0.20 0.659
Time orienting 1 5.24 0.023
Population (predation pressure) Spines 9 2.95 0.003
Nips 9 4.91 0.000
Time orienting 9 4.82 0.000
Predation pressure Spines 1 1.57 0.212
Nips 1 0.16 0.686
Time orienting 1 0.06 0.812
Error Spines 204
Nips 204
Time orienting 204
Open in a new tab

The effect of population was tested by nesting it within high or low predation pressure

Table 5.

Nonparametric tests of behavioral differences between fish from high and low predation populations

Aggression: spines Aggression: nips Aggression: orienting Antipredator: moves Antipredator: spines Antipredator: inspection Antipredator: orienting
Mann–Whitney U 5,194.5 5,713.5 5,618.5 4,559.5 4,779.5 4,966.5 5,584.0
Z −1.53 −0.46 −0.57 −2.87 −2.40 −2.44 −0.62
P 0.127 0.648 0.571 0.004 0.016 0.015 0.538
Open in a new tab

sn = 97 individuals from low predation populations, n = 121 individuals from high predation populations

Individuals also showed variable responses to the pike predator. Although some (n = 63) individuals went up to the mouth of the pike and inspected it, other individuals hardly moved at all in the presence of the predator behind glass. Some of the variation among individuals could be attributed to body size (among all populations, bigger fish moved less: r = −0.27, P < 0.0001) and some to the population of origin of the fish, independent of body size (Table 4). Unlike aggressive behavior, there was a statistically significant difference in the antipredator behavior of fish from high versus low predation populations: overall, fish from high predation populations moved more, raised their spines more, and may have inspected more than fish from low predation populations (Fig. 2; Tables 4, 5).

Table 4.

Effects of population (predation pressure), predation pressure and body size on antipredator behaviors

Source Dependent Variable df F Sig.
Intercept Moves 1 42.56 0.000
Spines 1 20.29 0.000
Predator inspection 1 14.14 0.000
Time orienting 1 54.20 0.000
Weight Moves 1 4.75 0.030
Spines 1 1.46 0.228
Predator inspection 1 2.35 0.126
Time orienting 1 0.22 0.643
Population (predation pressure) Moves 9 3.82 0.000
Spines 9 5.27 0.000
Predator inspection 9 1.88 0.056
Time orienting 9 1.90 0.053
Predation pressure Moves 1 6.83 0.010
Spines 1 7.92 0.005
Predator inspection 1 3.22 0.074
Time orienting 1 0.56 0.457
Error Moves 204
Spines 204
Predator inspection 204
Time orienting 204
Open in a new tab

The effect of population was tested by nesting it within high or low predation pressure

Fig. 2.

Fig. 2

Open in a new tab

Variation among types of populations in behavioral responses to a live predator. Bars show means ± SE. Sample sizes are low n = 97, high n = 118

Are differences in behavior related to stress-induced ventilation rate?

Individual differences in behavior were related to differences in stress-induced ventilation rate measured in the same fish on the previous day. In general, fish that breathed faster in response to confinement stress were also more aggressive toward the conspecific and bold toward the pike 1 day later, an effect that persisted when differences in body size and between populations were accounted for (Table 6).

Table 6.

Correlations between stress-induced ventilation rate and behavior

Aggressive behaviors
Antipredator behaviors
Peak rate Spines Nips Time orienting Moves Spines Inspections Time orienting
Aggressive behaviors
 Peak rate – 0.175* 0.195* 0.213** 0.101 0.143 0.186* 0.155*
 Spines 0.483 – 0.205** 0.226** 0.300*** 0.228** 0.230** −0.019
 Nips 0.297 0.444 – 0.555*** 0.222** 0.067 0.192* 0.074
 Time orienting 0.551 0.614* 0.843** – 0.286*** 0.051 0.191* 0.150*
Antipredator behaviors
 Moves 0.161 −0.143 −0.003 0.284 – 0.719*** 0.383*** 0.232**
 Spines 0.173 −0.188 −0.147 0.177 0.950** – 0.107 0.088
 Inspections 0.786** 0.257 0.331 0.505 0.101 −0.011 – 0.253***
 Time orienting 0.276 0.529 −0.378 −0.019 −0.029 0.021 0.059 –
Open in a new tab

Partial correlations across individuals controlling for weight and population are on the top diagonal, n = 164 individuals

Correlations between mean values for the different populations are on the bottom, n = 11 populations

Correlations that are statistically significant at the P < 0.01 level are still significant after the sequential Bonferroni correction

*

P < 0.05

**

P < 0.01

***

P < 0.001

A similar pattern was detected at the between-population level: fish from populations with higher average stress-induced ventilation rates were also more aggressive and bold in the presence of a predator (Fig. 3; Table 6).

Fig. 3.

Fig. 3

Open in a new tab

Populations that were bold toward a predator had higher stress-induced ventilation rate. Each data point represents the mean of a population. Low predation populations are in open circles, high predation in filled circles. R2 is 0.618

Are individual differences in behavior stable over time?

Individual behavioral reactions to the pike were repeatable over a period of several months. For example, the repeatability estimate of moves in the presence of the pike was 0.29 (F6,36 = 3.26, P = 0.01) and the repeatability of the number of predator inspections was 0.23 (F6,36 = 2.69, P = 0.03). These estimates are comparable to other estimates of the repeatability of behavior (Bell et al. 2009).

Discussion

This study shows that an acute stressor (capture and confinement) causes a rapid, dramatic, but short-lived increase in stress-induced ventilation rate in sticklebacks. While the typical ventilation rate of undisturbed sticklebacks was 72 beats per min [see also (Meakins 1975)], we found that stress-induced ventilation rate rapidly increased to as high as 174 beats per min in response to confinement stress before rapidly decreasing. This result corroborates other studies on fish that have found that ventilation rate is an indicator of acute stress (Barreto et al. 2003; Cooke et al. 2003; Hawkins et al. 2004; Hojesjo et al. 1999; Johnsson et al. 2001; Metcalfe et al. 1995; Sundstrom et al. 2005).

Remarkably, peak ventilation rate during the confinement procedure was correlated with both aggressive and antipredator behavior the following day (at both the individual and the population levels). In practical terms, these results suggest that for sticklebacks, ventilation rate is an explanatory, noninvasive and sensitive measure of response to disturbance with significant predictive content. The interrelatedness of this apparently diverse set of measures may not be so surprising when the neuroendocrine basis of both behavior and the stress response are considered.

These results add to the growing body of evidence suggesting that individuals differ in how they respond to mild environmental challenges, with behavioral differences associated with underlying differences in stress physiology (Koolhaas et al. 1999) and metabolism (Careau et al. 2008), an interpretation that is given further weight by the stability observed in individuals’ behavioral responses over time.

The present study also confirms the existence of differences in behavior between fish from populations differing in predation pressure by piscivorous fishes with individuals from high predation populations showing more movement, spine raising and inspection in response to a predator, although no differences in aggression. Studies on other fish species including bishops (Brachyraphis episcopi, Brown and Braithwaite 2004), minnows (Phoxinus phoxinus, Magurran 1990) and sticklebacks (Walling et al. 2004) have also found increased levels of ‘bold’ behaviors such as decreased latency to emerge from safety in high predation populations. Although risky behavior in a dangerous environment might seem counter-intuitive, this result is predicted by life history theory. If small individuals are especially vulnerable to predation, individuals that grow fast will be favored because they will grow out of the small vulnerable stage more quickly. Therefore, risk-taking behaviors such as active foraging and aggression that improves access to resources and therefore growth rate should be favored when predation pressure is high (Mangel and Stamps 2001).

This study also shows that an ecological context (level of predation pressure by fishes) has implications for individual physiological responses to challenge. This is reflected in the observed differences in stress-induced ventilation rate between fish from high and low predation populations, with those from high predation populations having faster rates. We did not have a priori predictions about the effects of coexisting with repeated exposure to predators on physiology because both facilitation and acclimation have been observed in other studies. For example, whereas a comparable study found that bishop fish from areas of high predation showed lower stress-induced ventilation rate (Brown et al. 2007, 2005b), a pattern consistent with acclimation (Romero 2004), we observed the opposite. This suggests a need for further work to develop a conceptual framework for understanding the proximate and ultimate factors affecting variation in stress responsiveness over developmental and evolutionary time.

We also found that although some of the observed variation in behavior and physiology could be attributed to differences in predation pressure by fishes between the populations, in all of our analyses, there was still significant population-level variation, even within a habitat type. For example, as shown in Fig. 3, the population means do not cluster in the top right and bottom left corners of the graph. In other words, the variation among populations is continuous. Other important factors that varied among populations but which were not controlled for in this study include food availability, density, parasites and gene flow with neighboring populations, etc. Although it is tempting to compare just two different populations varying in the magnitude and predictability of stressors in the environment, our results highlight the importance of comparing multiple, independent replicate populations to detect an overall effect of predation regime among all the variation.

We do not know to what extent the observed population differences in behavior and stress physiology reflect genetic and environmental influences. On the one hand, the animals in this experiment had been held in the laboratory under standard conditions for at least 1 month. This is a relatively long time considering their young age at capture and likely lifespan of 12 months, hence, it is unlikely that the observed differences were due to recent experience. On the other hand, early exposure to stressors can have long-term consequences for stress reactivity and behavior later in life (Auperin and Geslin 2008; Meaney 2001), so it is possible that early experience in the field (including with their father, Tulley and Huntingford 1987) might have contributed to the observed patterns. In addition, genetic differences between the populations cannot be excluded. Studies on other fish have shown that stress responsiveness can respond to selection (Pottinger and Carrick 1999), and behavioral reactions to both predators (Bell 2005) and con-specifics (Bakker 1986) have a heritable component in sticklebacks. Moreover, our study populations are genetically differentiated (Malhi et al. 2006) and other studies have shown that sticklebacks are capable of rapid (within 10 years) morphological evolution (Bell et al. 2004), so it is possible that the population differences reflect evolutionary responses to divergent selective pressures by predators.

Altogether, these results highlight the important role that local, habitat-specific environmental pressures play in setting up interdependent behavioral and physiological responses to challenges in the environment.

Acknowledgments

We thank Stuart Wilson, David Alvarez, Susie Coyle, John Laurie and Kate Arnold for help with the fish, and to Andy Young for catching the pike. Funding was provided by a US National Science Foundation International Postdoctoral Research Fellowship to AMB. We are also grateful to the many landowners who allowed us to collect fish on their property.

Footnotes

Communicated by G. Heldmaier.

Contributor Information

Alison M. Bell, Email: alisonmb@life.uiuc.edu, Integrative Biology, University of Illinois, 505 S. Goodwin Ave., Urbana, IL 61801, USA

Lindsay Henderson, Division of Environmental and Evolutionary Biology, University of Glasgow, Graham Kerr Building, Glasgow G12 8QQ, Scotland, UK.

Felicity A. Huntingford, Division of Environmental and Evolutionary Biology, University of Glasgow, Graham Kerr Building, Glasgow G12 8QQ, Scotland, UK

References

  1. Auperin B, Geslin M. Plasma cortisol response to stress in juvenile rainbow trout is influenced by their life history during early development and by egg cortisol content. Gen Comp Endocrinol. 2008;158:234–239. doi: 10.1016/j.ygcen.2008.07.002. [DOI] [PubMed] [Google Scholar]
  2. Bakker TCM. Aggressiveness in sticklebacks (Gasterosteus aculeatus L.): a behaviour-genetic study. Behaviour. 1986;98:145–167. [Google Scholar]
  3. Barreto RE, Volpato GL. Caution for using ventilatory frequency as an indicator of stress in fish. Behav Processes. 2004;66:43–51. doi: 10.1016/j.beproc.2004.01.001. [DOI] [PubMed] [Google Scholar]
  4. Barreto RE, Luchiari AC, Marcondes AL. Ventilatory frequency indicates visual recognition of an allopatric predator in naive Nile tilapia. Behav Processes. 2003;60:235–239. doi: 10.1016/s0376-6357(02)00127-4. [DOI] [PubMed] [Google Scholar]
  5. Bartolomucci A, Palanza P, Sacerdote P, Panerai AE, Sgoifo A, Dantzer R, Parmigiani S. Social factors and individual vulnerability to chronic stress exposure. Neurosci Biobehav Rev. 2005;29:67–81. doi: 10.1016/j.neubiorev.2004.06.009. [DOI] [PubMed] [Google Scholar]
  6. Bell AM. Differences between individuals and populations of three spined stickleback. J Evol Biol. 2005;18:464–473. doi: 10.1111/j.1420-9101.2004.00817.x. [DOI] [PubMed] [Google Scholar]
  7. Bell MA, Foster SA. The evolutionary biology of the threespine stickleback. Oxford University Press; Oxford: 1994. [Google Scholar]
  8. Bell MA, Aguirre WE, Buck NJ. Twelve years of contemporary armor evolution in a threespine stickleback population. Evol Int J Org Evol. 2004;58:814–824. doi: 10.1111/j.0014-3820.2004.tb00414.x. [DOI] [PubMed] [Google Scholar]
  9. Bell AM, Backstrom T, Huntingford FA, Pottinger TG, Winberg S. Variable behavioral and neuroendocrine responses to ecologically-relevant challenges in sticklebacks. Physiol Behav. 2007;91:15–25. doi: 10.1016/j.physbeh.2007.01.012. [DOI] [PubMed] [Google Scholar]
  10. Bell AM, Hankison SJ, Laskowski KL. The repeatability of behaviour: a meta-analysis. Anim Behav. 2009;77:771–783. doi: 10.1016/j.anbehav.2008.12.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Blanchard RJ, Nikulina JN, Sakai RS, McKittrick C, McEwen B, Blanchard DC. Behavioral and endocrine change following chronic predatory stress. Physiol Behav. 1998;63:561–569. doi: 10.1016/s0031-9384(97)00508-8. [DOI] [PubMed] [Google Scholar]
  12. Bonier F, Martin PR, Sheldon KS, Jensen JP, Foltz SL, Wingfield JC. Sex-specific consequences of life in the city. Behav Ecol. 2007;18:121–129. [Google Scholar]
  13. Boonstra R, Hik D, Singleton GR, Tinnikov A. The impact of predator-induced stress on the snowshoe hare cycle. Ecol Monogr. 1998;68:371–394. [Google Scholar]
  14. Brown C, Braithwaite VA. Size matters: a test of boldness in eight populations of the poeciliid Brachyraphis episcopi. Anim Behav. 2004;68:1325–1329. [Google Scholar]
  15. Brown C, Gardner C, Braithwaite VA. Differential stress responses in fish from areas of high- and low predation pressure. J Comp Physiol B Biochem Syst Environ Physiol. 2005a;175:305–312. doi: 10.1007/s00360-005-0486-0. [DOI] [PubMed] [Google Scholar]
  16. Brown C, Jones F, Braithwaite V. In situ examination of boldness-shyness traits in the tropical poeciliid, Brachyraphis episcopi. Anim Behav. 2005b;70:1003–1009. [Google Scholar]
  17. Brown C, Burgess F, Braithwaite VA. Heritable and experiential effects on boldness in a tropical poeciliid. Behav Ecol Socio-biol. 2007;62:237–243. [Google Scholar]
  18. Caldji C, Francis D, Sharma S, Plotsky PM, Meaney MJ. The effects of early rearing environment on the development of GABA(A) and central benzodiazepine receptor levels and novelty-induced fearfulness in the rat. Neuropsychopharmacology. 2000;22:219–229. doi: 10.1016/S0893-133X(99)00110-4. [DOI] [PubMed] [Google Scholar]
  19. Careau V, Thomas D, Humphries MM, Reale D. Energy metabolism and animal personality. Oikos. 2008;117:641–653. [Google Scholar]
  20. Carere C, Van Oers K. Shy and bold great tits (Parus major): body temperature and breath rate in response to handling stress. Physiol Behav. 2004;82:905–912. doi: 10.1016/j.physbeh.2004.07.009. [DOI] [PubMed] [Google Scholar]
  21. Carere C, Casettic R, de Acetisd L, Perrettac G, Cirullid F, Allevad E. Behavioural and nociceptive response in male and female spiny mice (Acomys cahirinus) upon exposure to snake odour. Behav Processes. 1999;47:1–10. doi: 10.1016/s0376-6357(99)00048-0. [DOI] [PubMed] [Google Scholar]
  22. Carere C, Groothuis TGG, Moestl E, Daan S, Koolhaas JM. Fecal corticosteroids in a territorial bird selected for different personalities: daily rhythm and the response to social stress. Horm Behav. 2003;43:540–548. doi: 10.1016/s0018-506x(03)00065-5. [DOI] [PubMed] [Google Scholar]
  23. Cockrem JF, Silverin B. Sight of a predator can stimulate a corticosterone response in the great tit. Gen Comp Endocrinol. 2002a;125:248–255. doi: 10.1006/gcen.2001.7749. [DOI] [PubMed] [Google Scholar]
  24. Cockrem JF, Silverin B. Variation within and between birds in corticosterone responses of great tits. Gen Comp Endocrinol. 2002b;125:197–206. doi: 10.1006/gcen.2001.7750. [DOI] [PubMed] [Google Scholar]
  25. Cooke SJ, Steinmetz J, Degner JF, Grant EC, Philipp DP. Metabolic fright responses of different-sized largemouth bass (Micropterus salmoides) to two avian predators show variations in nonlethal energetic costs. Can J Zool Revue Canadienne De Zoologie. 2003;81:699–709. [Google Scholar]
  26. Cutts CJ, Metcalfe NB, Taylor AC. Aggression and growth depression in juvenile Atlantic salmon: the consequences of individual variation in standard metabolic rate. J Fish Biol. 1998;52:1026–1037. [Google Scholar]
  27. Dunlap KD, Wingfield JC. External and internal influences on indexes of physiological stress 1. Seasonal and population variation in adrenocortical secretion of free-living lizards, Sceloporus occidentalis. J Exp Zool. 1995;271:36–46. doi: 10.1002/jez.1402710105. [DOI] [PubMed] [Google Scholar]
  28. Ebner K, Wotjak CT, Landgraf R, Engelmann M. Neuroendocrine and behavioral response to social confrontation: residents versus intruders, active versus passive coping styles. Horm Behav. 2005;47:14–21. doi: 10.1016/j.yhbeh.2004.08.002. [DOI] [PubMed] [Google Scholar]
  29. Eilam D, Dayan T, Ben-Eliyahu S, Schulman I, Shefer G, Hendries CA. Differential behavioural and hormonal responses of voles and spiny mice to owl calls. Anim Behav. 1999;58:1085–1093. doi: 10.1006/anbe.1999.1224. [DOI] [PubMed] [Google Scholar]
  30. Espmark Y, Langvatn R. Development and habituation of cardiac and behavioral-responses in young Red Deer calves (Cervus elaphus) exposed to alarm stimuli. J Mammal. 1985;66:702–711. [Google Scholar]
  31. Evans MR, Roberts ML, Buchanan KL, Goldsmith AR. Heritability of corticosterone response and changes in life history traits during selection in the zebra finch. J Evol Biol. 2006;19:343–352. doi: 10.1111/j.1420-9101.2005.01034.x. [DOI] [PubMed] [Google Scholar]
  32. Hawkins LA, Armstrong JD, Magurran AE. Predator-induced hyperventilation in wild and hatchery Atlantic salmon fry. J Fish Biol. 2004;65:88–100. [Google Scholar]
  33. Heidinger BJ, Nisbet ICT, Ketterson ED. Older parents are less responsive to a stressor in a long-lived seabird: a mechanism for increased reproductive performance with age? Proc R Soc Lond B. 2006;273:2227–2231. doi: 10.1098/rspb.2006.3557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Hik DS, McColl CJ, Boonstra R. Why are Arctic ground squirrels more stressed in the boreal forest than in alpine meadows? Ecoscience. 2001;8:275–288. [Google Scholar]
  35. Hojesjo J, Johnsson JI, Axelsson M. Behavioural and heart rate responses to food limitation and predation risk: an experimental study on rainbow trout. J Fish Biol. 1999;55:1009–1019. [Google Scholar]
  36. Huntingford FA, Wright PJ, Tierney JF. Adaptive variation and antipredator behaviour in threespine stickleback. In: Bell MA, Foster SA, editors. The evolutionary biology of the threespine stickleback. Oxford University Press; Oxford: 1994. pp. 277–295. [Google Scholar]
  37. Johnsson JI, Hojesjo J, Fleming IA. Behavioural and heart rate responses to predation risk in wild and domesticated Atlantic salmon. Can J Fish Aquat Sci. 2001;58:788–794. [Google Scholar]
  38. Koolhaas JM, Meerlo P, De Boer SF, Strubbe JH, Bohus B. The temporal dynamics of the stress response. Neurosci Biobehav Rev. 1997;21:775–782. doi: 10.1016/s0149-7634(96)00057-7. [DOI] [PubMed] [Google Scholar]
  39. Koolhaas JM, Korte SM, De Boer SF, Van Der Vegt BJ, Van Reenen CG, Hopster H, De Jong IC, Ruis MAW, Blokhuis HJ. Coping styles in animals: current status in behavior and stress-physiology. Neurosci Biobehav Rev. 1999;23:925–935. doi: 10.1016/s0149-7634(99)00026-3. [DOI] [PubMed] [Google Scholar]
  40. Lessells CM, Boag PT. Unrepeatable repeatabilities: a common mistake. Auk. 1987;104:116–121. [Google Scholar]
  41. Lima SL, Bednekoff PA. Temporal variation in danger drives antipredator behavior: the predation risk allocation hypothesis. Am Nat. 1999;153:649–659. doi: 10.1086/303202. [DOI] [PubMed] [Google Scholar]
  42. Lucas JR, Freeberg TM, Egbert J, Schwabl H. Fecal corticosterone, body mass, and caching rates of Carolina chickadees (Poecile carolinensis) from disturbed and undisturbed sites. Horm Behav. 2006;49:634–643. doi: 10.1016/j.yhbeh.2005.12.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Magurran AE. The inheritance and development of minnow anti-predator behaviour. Anim Behav. 1990;39:834–842. [Google Scholar]
  44. Malhi RS, Rhett G, Bell AM. Mitochondrial DNA evidence of an early Holocene population expansion of threespine sticklebacks from Scotland. Mol Phylogenet Evol. 2006;40:148–154. doi: 10.1016/j.ympev.2006.02.019. [DOI] [PubMed] [Google Scholar]
  45. Mangel M, Stamps J. Trade-offs between growth and mortality and the maintenance of individual variation in growth. Evol Ecol Res. 2001;3:583–593. [Google Scholar]
  46. McEwen BS, Wingfield JC. The concept of allostasis in biology and biomedicine. Horm Behav. 2003;43:2–15. doi: 10.1016/s0018-506x(02)00024-7. [DOI] [PubMed] [Google Scholar]
  47. Meakins RH. The effects of activity and season on the respiration of the three-spined stickleback. Comp Biochem Physiol. 1975;51A:155–157. doi: 10.1016/0300-9629(75)90429-6. [DOI] [PubMed] [Google Scholar]
  48. Meaney MJ. Maternal care, gene expression, and the transmission of individual differences in stress reactivity across generations. Annu Rev Neurosci. 2001;24:1161–1192. doi: 10.1146/annurev.neuro.24.1.1161. [DOI] [PubMed] [Google Scholar]
  49. Metcalfe NB, Taylor AC, Thorpe JE. Metabolic-rate, social-status and life-history strategies in Atlantic salmon. Anim Behav. 1995;49:431–436. [Google Scholar]
  50. Muller C, Jenni-Eiermann S, Blondel J, Perret P, Caro SP, Lambrechts MM, Jenni L. Circulating corticosterone levels in breeding blue tits Parus caeruleus differ between island and mainland populations and between habitats. Gen Comp Endocrinol. 2007;154:128–136. doi: 10.1016/j.ygcen.2007.05.031. [DOI] [PubMed] [Google Scholar]
  51. Otte C, Hart S, Neylan TC, Marmar CR, Yaffe K, Mohr DC. A meta-analysis of cortisol response to challenge in human aging: importance of gender. Psychoneuroendocrinology. 2005;30:80–91. doi: 10.1016/j.psyneuen.2004.06.002. [DOI] [PubMed] [Google Scholar]
  52. Peichel C, Ross J, Matson C, Dickson M, Grimwood J, Schmutz J, Myers R, Mori S, Schluter D, Kingsley D. The master sex-determination locus in threespine sticklebacks in on a nascent Y chromosome. Curr Biol. 2004;14:1416–1424. doi: 10.1016/j.cub.2004.08.030. [DOI] [PubMed] [Google Scholar]
  53. Pickering AD, Pottinger TG. Acclimation of the brown trout, Salmo trutta, to the stress of daily malachite green treatment. Aquaculture. 1985;44:145–152. [Google Scholar]
  54. Pottinger TG, Carrick TR. Modification of the plasma cortisol response to stress in rainbow trout by selective breeding. Gen Comp Endocrinol. 1999;116:122–132. doi: 10.1006/gcen.1999.7355. [DOI] [PubMed] [Google Scholar]
  55. Pottinger TG, Balm PHM, Pickering AD. Sexual maturity modifies the responsiveness of the pituitary–interrenal axis to stress in male rainbow trout. Gen Comp Endocrinol. 1995;98:311–320. doi: 10.1006/gcen.1995.1073. [DOI] [PubMed] [Google Scholar]
  56. Priede IG. Metabolic scope in fishes. In: Tytler P, Calow P, editors. Fish energetics: new perspectives. Johns Hopkins University Press; Baltimore: 1985. pp. 33–64. [Google Scholar]
  57. Reimchen TE. Predators and morphological evolution in three-spine stickleback. In: Bell MA, Foster SA, editors. The evolutionary biology of the threespine stickleback. Oxford University Press; Oxford: 1994. pp. 240–273. [Google Scholar]
  58. Reimchen TE. Predator handling failures of lateral plate morphs in Gasterosteus aculeatus: functional implications for the ancestral plate condition. Behaviour. 2000;137:1081–1096. [Google Scholar]
  59. Rice WR. Analyzing tables of statistical tests. Evol Int J Org Evol. 1989;43:223–225. doi: 10.1111/j.1558-5646.1989.tb04220.x. [DOI] [PubMed] [Google Scholar]
  60. Romero LM. Physiological stress in ecology: lessons from biomedical research. Trends Ecol Evol. 2004;19:249–255. doi: 10.1016/j.tree.2004.03.008. [DOI] [PubMed] [Google Scholar]
  61. Sapolsky RM, Romero LM, Munck AU. How do glucocorticoids influence stress responses? Integrating permissive, suppressive, stimulatory and preparative actions. Endocr Rev. 2000;21:55–89. doi: 10.1210/edrv.21.1.0389. [DOI] [PubMed] [Google Scholar]
  62. Scheuerlein A, Van’t Hof TJ, Gwinner E. Predators as stressors? Physiological and reproductive consequences of predation risk in tropical stonechats. Proc R Soc Biol Sci Ser B. 2001;268:1575–1582. doi: 10.1098/rspb.2001.1691. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Sloman KA, Metcalfe NB, Taylor AC, Gilmour KM. Plasma cortisol concentrations before and after social stress in rainbow trout and brown trout. Physiol Biochem Zool. 2001;74:383–389. doi: 10.1086/320426. [DOI] [PubMed] [Google Scholar]
  64. Sloman KA, Motherwell G, O’Connor KI, Taylor AC. The effect of social stress on the standard metabolic rate (smr) of brown trout, Salmo trutta. Fish Physiol Biochem. 2000;23:49–53. [Google Scholar]
  65. Sundstrom LF, Petersson E, Johnsson JI, Dannewitz J, Hojesjo J, Jarvi T. Heart rate responses to predation risk in Salmo trutta are affected by the rearing environment. J Fish Biol. 2005;67:1280–1286. [Google Scholar]
  66. Tulley JJ, Huntingford FA. Paternal care and the development of adaptive variation in anti-predator responses in sticklebacks. Anim Behav. 1987;35:1570–1572. [Google Scholar]
  67. Veenema AH, Meijer OC, de Kloet ER, Koolhaas JM, Bohus BG. Differences in basal and stress-induced HPA regulation of wild house mice selected for high and low aggression. Horm Behav. 2003;43:197–204. doi: 10.1016/s0018-506x(02)00013-2. [DOI] [PubMed] [Google Scholar]
  68. Verbeek P, Iwamoto T, Murakami N. Variable stress-responsiveness in wild type and domesticated fighting fish. Physiol Behav. 2008;93:83–88. doi: 10.1016/j.physbeh.2007.08.008. [DOI] [PubMed] [Google Scholar]
  69. Wada H, Salvante KG, Stables C, Wagner E, Williams TD, Breuner CW. Adrenocortical responses in zebra finches (Taeniopygia guttata): individual variation, repeatability, and relationship to phenotypic quality. Horm Behav. 2008;53:472–480. doi: 10.1016/j.yhbeh.2007.11.018. [DOI] [PubMed] [Google Scholar]
  70. Walling CA, Dawnay N, Kazem AJN, Wright J. Predator inspection behaviour in three-spined sticklebacks (Gasterosteus aculeatus): body size, local predation pressure and cooperation. Behav Ecol Sociobiol. 2004;56:164–170. [Google Scholar]
  71. Zar JH. Biostatistical analysis. 4. Prentice-Hall; Upper Saddle River: 1999. [Google Scholar]

RESOURCES