wishingrad et al. 2015 behaviour.pdf

Behaviour 152 (2015) 635–652 brill.com/beh

Risk in a changing world: environmental cues driveanti-predator behaviour in lake sturgeon

(Acipenser fulvescens) in the absence of predators

Van Wishingrad a,∗, Annessa B. Musgrove a,

Douglas P. Chivers a and Maud C.O. Ferrari b

a Department of Biology, University of Saskatchewan, 112 Science Place,Saskatoon, SK, Canada S7N 5E2

b Department of Biomedical Sciences, WCVM, University of Saskatchewan,52 Campus Drive, Saskatoon, SK, Canada S7N 5B4

*Corresponding author’s e-mail address: [email protected]

Accepted 4 October 2014; published online 6 November 2014

AbstractRapidly changing climates and habitats represent challenges faced by the majority of animalspecies on our planet, and are leading to rapid declines in global biodiversity. However, the degreeto which behaviour is influenced by changing environmental cues is not well understood. Specifi-cally, environmental cues that have been correlated with predator abundance or performance overevolutionary history may have significant effects on prey behaviour. In the present study, we inves-tigated the role of water clarity on foraging activity in lake sturgeon (Acipenser fulvescens) in theabsence of predators. Foraging activity was significantly higher during the night than the day andwas higher in turbid environments versus clear environments, indicating that decreased turbidityalone, may in part drive anti-predator behaviour and constrain foraging activity. Future work ex-ploring the interconnectedness of environmental cues and behavioural changes will help us betterunderstand the many ways rapidly changing environments can influence behavioural ecologicalprocesses.

Keywordsenvironmental change, behavioural plasticity, anti-predator behaviour, foraging, evolutionarytrap.

1. Introduction

Human-induced rapid environmental change is putting animals in evolution-arily novel situations to which they may or may not be adapted. A recent

© Koninklijke Brill NV, Leiden, 2015 DOI 10.1163/1568539X-00003246

http://www.brill.com/behaviour

mailto:[email protected]

http://dx.doi.org/10.1163/1568539X-00003246

636 Risk in a changing world

meta-analysis of more than 3000 rates of phenotypic change suggests thatmost of the responses of animals to rapid environmental change involve phe-notypic plasticity rather than immediate genetic evolution (Hendry et al.,2008). In face of such challenges, animals often display alterations in be-haviour, as it is one of the fastest phenotypic changes, and because behaviourcan usually be much more plastic than other phenotypic traits.

Behavioural plasticity refers to an animal’s ability to vary its behaviouraccording to different conditions (West-Eberhard, 1989) and may be thequickest way animals adjust to a changing environment. For example, lizardssubjected to high temperatures will alter their basking behaviour to regulatetheir physiological performance (Huey et al., 2003). Similarly, changes topredator abundance or performance can have a tremendous impact on be-havioural decisions regarding foraging intensity or reproductive investment(Sih, 1987; Lima, 1998; Stankowich & Blumstein, 2005). Behavioural plas-ticity can therefore be invaluable for populations subjected to rapidly chang-ing climates and habitats. Alternatively, in a rapidly altered environment,a formerly adaptive behaviour may become non-adaptive and have fitnessconsequences. Environmental cues that have been correlated with predatorabundance or performance over evolutionary history may have significanteffects on present-day prey behaviour even in the absence of predators. How-ever, how changing environmental conditions influence behaviourally plastictraits is not well understood (Ghalambor et al., 2010).

Organisms inhabiting aquatic ecosystems are ideal for testing the fitnessconsequences of behaviour in response to environmental changes. Aquaticecosystems have been subjected to a wide range of environmental changesincluding changes to temperature, decreased pH, contamination and habitatfragmentation (Schindler, 1988; Winder & Schindler, 2004; Orr et al., 2005).Moreover, there is increasing concern that anthropogenic activities are lead-ing to changes in turbidity (Chivers et al., 2013). Turbidity (water hazinessor cloudiness), results in benthic smothering, changes rates of photosyn-thesis, and can often result in substantial changes to community structure(Liljendahl-Nurminen et al., 2008). River turbidity levels are rapidly chang-ing worldwide (Kundzewicz et al., 2009) and in some areas river turbidityfluctuates with water-flow regimes, where turbidity levels are linked to flowrates (Dewson et al., 2007). Decreases in turbidity can be attributed to bothbiotic agents, such as changes in faunal community structure (Skubinna etal., 1995), or anthropogenic factors, such as dam construction along large

V. Wishingrad et al. / Behaviour 152 (2015) 635–652 637

rivers (Johnson et al., 1995). Predator–prey interactions, in particular, areinfluenced by turbidity levels (Bonner & Wilde, 2002; Chivers et al., 2013).

Turbidity has been shown to decrease foraging success for predatorybanded kokopu (Galaxias fasciatus), common galaxias (G. maculatus) andsablefish (Anoplopoma fimbria) (Rowe & Dean, 1998; De Robertis etal., 2003), and impede learned predator recognition in fathead minnows(Pimephales promelas) (Ferrari et al., 2010; Chivers et al., 2013). How-ever, turbidity has also been found to reduce predation risk for northern pike(Esox lucius) (Lehtiniemi et al., 2005), white sturgeon (Acipenser fulvescens)(Gadomski & Parsley, 2005) and damselflies (Ischnura elegans) (Van deMeutter et al., 2005), allow fathead minnows to use feeding grounds occu-pied by a predator (Abrahams & Kattenfeld, 1997; Chiu & Abrahams, 2010),and allow lake sturgeon (Acipenser fulvescens) to curtail costly escape be-haviours (Wishingrad et al., 2014a). Furthermore, Ferrari et al. (2014) foundthat delta smelt (Hypomesus transpacificus) that evolved in turbid-water en-vironments are decimated by predators in clear-water environments as theyrely exclusively on turbidity to avoid predators. Turbidity can therefore eitherhave a positive or negative impact on a species, depending on their particu-lar life-history, their level of exposure to visually-guided predators, and theenvironmental conditions in which they evolved.

In the present study, we examined the effect of turbidity on foraging ac-tivity in lake sturgeon in the absence of predators to investigate the degreeto which environmental cues alone can affect foraging activity. Nocturnalforaging may reflect an adaptive balance between foraging activity and anti-predator behaviours (Helfman, 1993). However, due to the severe cost offailing to avoid a predator, this balance may be shifted away from activeforaging activity and toward anti-predator inactivity. Thus, given the oppor-tunity, nocturnal animals may forage actively during the day under the coverof turbidity. We predict that foraging activity will be lowest during the day,and highest during the night, owing to the increased risk of visual predationduring the day. Second, we predict that due to the anti-predator benefit of tur-bidity, foraging activity levels will be higher in turbid environments than inclear-water environments. Third, we predict that as a result of growth and de-velopment, the difference between clear- and turbid-water foraging activitywill be higher in young fish than in older fish.

Sturgeon are particularly good candidates to investigate the role of turbid-ity on anti-predator behaviour because sturgeon are a non-visual feeding fish


(Loew & Sillman, 1993; Boglione et al., 1997; Rodríguez & Gisbert, 2002).Sturgeon forage primarily by using four sensory barbels on the ventral sideof their body immediately anterior to the mouth (Peterson et al., 2007). Thesebarbels allow sturgeon to decode the chemical composition of potential preyitems, allowing them to forage without the use of sight. Furthermore, stur-geon electrosense has recently been described, whereby sturgeon can detectthe electrical signals of potential prey items, allowing sturgeon to recoverprey hidden beneath sandy substrates and forage effectively in turbid riversand lakes (Zhang et al., 2012). Thus, while turbidity may impair the ability ofvisual predators to detect them, it does not detrimentally influence sturgeonforaging ability. In the wild, most fish spend part of their day in an activeforaging state, and the other half in an inactive state that is intimately linkedwith predator avoidance (Helfman, 1993). This is especially true of youngsturgeon, that forage actively during the night and reduce activity during theday to avoid visual predators (Richmond & Kynard, 1995). Furthermore,Lake Sturgeon metabolic rate has been linked to decreases in light levels(Svendsen et al., 2014). Turbidity may therefore provide cover from visually-guided predators, allowing sturgeon to increase foraging activity in darkerturbid environments relative to brighter clear-water environments. Therefore,even in the absence of predators, water clarity levels may exert a strong effecton the trade-off between foraging and anti-predator behaviour.

2. Materials and methods

2.1. Study species

We obtained lake sturgeon from Sustainable Sturgeon Culture, Ontario, inMay 2012 where they were spawned from wild broodstock endemic to theRainy River, ON, Canada. They were housed at the RJF Smith Centre forAquatic Ecology, held in a 640-l flow-through tank, and fed every 6 h withlive brine shrimp (Artemia spp.) and crushed frozen bloodworm (Chirono-midae spp.). They were maintained in dechlorinated tap water between 16–18°C under a 14:10 h artificial light/dark cycle. However, the RJF SmithCentre for Aquatic Ecology is designed in such a way that sunlight illu-minates the facility. Thus, a 17:7 natural light/dark cycle (resembling locallight-patterns) predominated during the course of the study.


2.2. Experimental set-up

Trials took place in 100-l polyurethane D-ended simulated river mesocosmsmeasuring 1 m long × 50 cm wide × 30 cm high (Figure 1). Each mesocosmwas divided in half longitudinally with an opaque acrylic panel measuring60 cm long to allow for the circular flow of water with a submersible 15 l/minpowerhead which provided a flow of 2–3 cm/s. A rectangular section of thelong-edge of the mesocosm measuring 55 cm long × 20 cm wide was seg-regated using vinyl-coated fiberglass mesh mounted on PVC frames. Insidethe mesocosm four boxes were placed, each measuring 20 cm long × 13.5 ×cm wide × 1.5 cm deep which held substrates used to simulate river micro-habitats. A rocky microhabitat was created by layering cobble (grain size:10–20 mm) under rocks (grain size 30–70 mm) stacked to a peak 5 cm high.A sandy microhabitat was created by filling a substrate box with sand (grainsize 0.25–1.2 mm) 1.5 cm deep. Sturgeon were able to move freely betweenthe rocky and sandy microhabitats. Two boxes of each microhabitat typewere randomly arranged in the mesocosm before fish were introduced. Therectangular section of the mesocosm containing four microhabitat boxes, twoof each type, between the mesh barriers, comprised the experimental area.Each mesocosm was filled with 25 l of water to a height of 5.5 cm over thesubstrate boxes to control for use of the water column by the fish. Water waseither left clear or turbidity was adjusted by adding 2.25 g of bentonite (aninert colloidal clay, Sigma-Aldrich; CAS number 1302-78-9) immediately

Figure 1. Overhead view of mesocosms used to simulate river microhabitats during experi-ments, showing the size and position of the grid used to score activity levels (quantified basedon number of grid lines crossed).


before the fish were added producing a concentration of 0.09 g/l, approx.22.5 NTU (Nephelometric Turbidity Units), equivalent to approx. 17 cmSecchi depth (Shoup & Wahl, 2009). Turbidity measurements were taken us-ing a LaMotte 2020e portable turbidity meter (LaMotte, Chestertown, MD,USA); Secchi depth measurements were taken using a 120 cm turbidity tubefor intermediate turbidity levels (Wildlife Supply, Yulee, FL, USA). Eachmesocosm was reconditioned with 1.12 g of bentonite every 6 h to accountfor loss of suspended clay with time, yielding an NTU range of 18.2–27.8and a Secchi depth range of 20–15 cm, respectively, during the experiment.In clear-water mesocosms, NTU was equal to 0.1. Thus, the clear- and turbid-water mesocosms provided very different physical environments for the fishduring the course of the study.

2.3. Experimental protocol

We investigated how turbidity (clear vs. turbid environments), time-of-day(day vs. night) and age-class (young vs. old) influenced diurnal foraging ac-tivity patterns. Young fish were between 42 and 54 days post-hatch (dph)(between 28 and 40 days after the onset of exogenous feeding); old fishwere between 57 and 69 dph (between 43 and 55 days after the onset ofexogenous feeding). These two age-classes were selected in order to eval-uate if any effects due to turbidity or time-of-day were weaker in olderfish, since previous studies found that young Lake Sturgeon were morenocturnally active (Peake, 1999), and much more reactive to alarm cues(Wishingrad et al., 2014b) relative to older conspecifics. Four combina-tions resulted (young/clear-water, young/turbid-water, old/clear-water, andold/turbid-water), and 20 replicates were run per treatment combination.Prior to the onset of trials three fish were arbitrarily selected from the holdingtank placed in the experimental area of each mesocosm and left to accli-mate for 24 h. During the night, incident light at an approximate intensityof 0.25 lux (equivalent to the light of a full moon on a clear night) providedsome illumination. Trials were conducted during June and July 2012 by us-ing overhead cameras mounted with infrared lights to record both diurnaland nocturnal sturgeon behaviour over a 24 h period. All fish were fed every6 h during the acclimation and observation periods to control for food-drivenchanges in activity. At the completion of trials, fish were placed in a sep-arate holding tank and not used for subsequent trials to maintain statisticalindependence among trials.


2.4. Video analysis

Foraging activity was scored as the number of times a randomly-selectedfocal fish passed over a superimposed 2 × 4 grid (grid size: 10 × 13.5 cm)during a 3-min interval each hour for 24 h. The side of each square wasequivalent to approx. 3–5 body lengths of the fish. Since sturgeon forage byactively cruising along the substrate surface, activity level is a good proxy offoraging activity, which includes feeding, moving between foraging patches,and exploring their environment to seek out resources.

2.5. Statistical analysis

We analysed our data using the R environment for statistical program-ming (version 2.15.2; R Development Core Team, 2012). Due to a loss ofvideo files from two old age-class/clear-water mesocosms (48 3-min in-tervals) we based our analysis on the remaining 480 young/clear-water,480 young/turbid-water, 432 old/clear-water, and 480 old/turbid-water be-havioural observations, comprising a total of 1872 3-min observation in-tervals. Each mesocosm was considered the sampling unit. All statisticalanalyses were run using raw values, however, plots and reported results ofactivity level are presented in units per hour. Values are means ± standarderror.

To quantitatively describe foraging activity through time within each en-vironment, and to compare foraging activity through time between environ-ments, polynomial regression models were fit to mean hourly activity valuesfor each treatment (N = 40 replicates/hour in turbid environments; N = 38replicates/hour in clear environments). This was accomplished by (1) iden-tifying an initial model that described the data well (i.e., contained onlysignificant terms), (2) deriving best-fit parameter estimates for two separatemodels (one for each treatment) based on the number of parameters in theinitial model, and (3) systematically combining coefficients (i.e., refitting themodels) to identify a more parsimonious set of model parameters. This sim-plified parameter set would provide a better balance between model accuracyand complexity (determined using Akaike’s Information Criterion with a cor-rection for small sample sizes (AICc), where models with lower AIC valueswere considered more parsimonious; Burnham & Anderson, 2002).

We tested the effects of time-of-day, age, and treatment on foraging ac-tivity using a 3-way repeated-measures ANOVA to control for spatial andtemporal autocorrelation within each mesocosm (i.e., activity levels were


measured in each mesocosm-replicate at two times: during the day and dur-ing the night). Due to a high proportion of zeros in the dataset (approx.15% as a result of the absence of foraging activity at certain times) we ranthe analysis on values of average activity level during either day or nightin each mesocosm to simplify the analysis. Factors used in the repeatedmeasures ANOVA were: water clarity (clear/turbid), time-of-day (day/night)and age (young/old). Model residuals satisfied assumptions of normality andhomoscedasticity, no data points had undue leverage, and no outliers wereidentified.

3. Results

A single quadratic model was initially fit to combined clear- and turbid-watermean activity data since all terms in this model were significant, whereaslower- and higher-order models both contained insignificant terms (Table 1).Two quadratic models were subsequently fit, one for each treatment, all termsof which were significant (Table 2). Combining coefficients b, vertex posi-tion (time of minimum activity level), and c, rate of change (speed of increaseand decrease in activity level) yielded the most parsimonious set of models(Table 3). However, coefficient a, y-intercept (maximum activity level), wasnot combined. The top model had a model weight of 0.52, which can be in-terpreted as meaning there is a 52% chance that it is the best model in the

Table 1.Model parameter estimates comparing the relative fit of linear, quadratic and cubic models tothe full data set (i.e., both clear and turbid treatments).

Estimate SE t p

Linear model a 579.138 64.35 9 <0.001b −5.528 4.794 −1.153 0.255

Quadratic model a 930.038 56.881 16.35 <0.001b −101.228 11.456 −8.836 <0.001c 4.161 0.481 8.65 <0.001

Cubic model a 826.11026 66.73124 12.38 <0.001b −40.45162 25.67195 −1.576 0.1223c −2.58767 2.62635 −0.985 0.3299d 0.19561 0.07499 2.609 0.0124

The quadratic regression model was selected as a starting point because all parameterestimates were significant.


Table 2.Quadratic model parameter estimates for separate clear and turbid treatments.

Estimate SE t p

Clear water a1 904.645 37.007 24.45 <0.001b1 −100.924 7.453 −13.54 <0.001c1 3.837 0.313 12.26 <0.001

Turbid water a2 1072.1317 44.7664 23.949 <0.001b2 −95.2706 9.0158 −10.567 <0.001c2 3.4373 0.3786 9.079 <0.001

These separate models were subsequently refit to identify a more parsimonious set of modelparameters (i.e., identify which parameter estimates clear and turbid treatments share). Termdefinitions: a = y-intercept; b = vertex position; c = rate of change.

set, given the data (Symonds & Moussalli, 2011). Furthermore, all three topmodels had a combined model weight of 0.92, indicating that combining ver-tex position and rate of change, but not maximum activity level (y-intercept),yields a strong model fit.

These results indicate that in both clear- and turbid-water treatments for-aging activity decreased to a minimum at the same time-of-day (aroundmid-day), and rate of increase and decrease in foraging activity was simi-lar between the two treatments. Therefore, fish in turbid environments were

Table 3.AICc values for models sharing common terms, indicating that combining coefficients b

(vertex position) and c (rate of change), but not a (y-intercept) yields the most parsimoniousmodel set (i.e., fish from turbid treatments were consistently more active during both the dayand at night).

Model adjustments k AICc �AICc wi

Combined b and c terms 5 600.79 0 0.52Combined b term 6 602.68 1.89 0.20Combined c term 6 602.75 1.96 0.20None (no combined terms) 7 605.42 4.64 0.05Combined a 6 606.89 6.10 0.02Combined a and c terms 5 609.79 9.00 0.01Combined a and b terms 5 613.89 13.10 0.00All combined terms (single model) 4 618.12 17.33 0.00

Term definitions: a = y-intercept; b = vertex position; c = rate of change; k = number ofparameters in the model; wi = model weight.


Figure 2. Mean (±SE) foraging activity for each hour in either clear N = 38 (white points);or turbid N = 40 (grey points) environments. Polynomial models are fit to either clear- (blackline) or turbid-water (grey line) mean activity values (see text for details).

consistently more active during both the day and at night (Figure 2). For-aging activity was significantly influenced by the main effects of turbidity(F1,74 = 5.849, p = 0.018) and time of day (F1,74 = 148.173, p < 0.001)(Figure 3; Table 4). No significant interactions were identified, but the inter-action between age and time of day approached significance (F1,74 = 3.572,p = 0.063).

4. Discussion

Activity was significantly higher in turbid environments versus clear envi-ronments, indicating that turbidity may drive foraging activity. Activity wasconsistently higher in turbid environments during the night and the day, and(while not captured by the model) activity level in clear-water decreasedabruptly in the presence of light and increased sharply in the absence oflight, whereas activity level in turbid water changed more gradually in bothcases. These results indicate that activity may be driven by available light,where foraging is immediately abandoned at the first sign of light in the earlymorning and immediately resumed once light levels drop in the evening. Fishbehaviour has long been known to be dramatically affected by the diel cy-


Figure 3. Mean (±SE) foraging activity during the day (unshaded region) and during thenight (shaded region) in clear environments (white bars) or turbid environments (grey bars).Clear-water/day activity N = 38; turbid-water/day activity N = 40; clear-water/night activityN = 38; turbid-water/night activity N = 40.

Table 4.ANOVA table showing the effect of time of day (day vs. night), age (young vs. old), andturbidity (clear vs. turbid) on foraging activity.

Source df F p

Between subjects Turbidity 1, 74 5.849 0.018∗Age 1, 74 0.026 0.872Turbidity × Age 1, 74 0.812 0.370

Within subjects Time of day 1, 74 148.173 <0.001∗Turbidity × Time of day 1, 74 0.298 0.587Age × Time of day 1, 74 3.572 0.063Turbidity × Age × Time of day 1, 74 0.185 0.669

∗ Significant p-values at a 0.05α level.


cle, where light levels are negatively correlated with foraging activity andpositively correlated with anti-predator behaviour (Helfman, 1993). How-ever, turbidity reduced this effect, likely due to the increased attenuation itimposes on visible light decreasing its penetration depth and intensity. Ourresults demonstrate that simple changes in environmental conditions whichare linked to predator performance can exert strong effects on prey fish be-haviour even in the absence of predators.

Fish in turbid environments prolonged foraging activity suggesting thatyoung sturgeon may forage at maximum anti-predator effort (Lima & Bed-nekoff, 1999). If this is the case, then turbidity may allow young sturgeonto forage beyond their minimal caloric requirement, accelerating somaticgrowth and the rate of development of morphological defences, further re-ducing the need for behavioural anti-predator effort. Consistent with thisline of reasoning, we found that the interaction between time of day andage trended toward significance, suggesting that as sturgeon grew older thediurnal foraging pattern observed became less pronounced. It may be that in-creased somatic growth (large body size), defensive morphological develop-ment (larger/sharper scutes), or some combination of both may compensatefor anti-predator behaviour in this species (i.e., trait compensation; Dewitt etal., 1999) leading to attenuation in diel activity patterns. Another possibilityis that larger sturgeon take more foraging risks since they are able to elicitstronger escape responses if and when they encounter a predator (Wishin-grad et al., 2014c). Over a longer period of time the differences in foragingactivity between day and night may become increasingly, and significantly,attenuated.

4.1. Evolution of behaviourally plastic foraging behaviour

Whether or not an animal can elicit a plastic response can give some indica-tion as to the ecological constraints that have shaped its behaviour. Adaptiveevolution of environmentally-mediated behaviourally plastic responses re-lies on four elements: (1) environments must exhibit variability; (2) fitnesswithin a population must be driven by the match between a behaviour andthe current environment; (3) individuals should be able to elicit an appropri-ate behavioural change based on a reliable cue; and (4) no single behaviourshould be useful in all environments (Via & Lande, 1985; Moran, 1992).Since behavioural plasticity tends to be favoured in variable environmentswhere the cues are reliable, then it follows that the environmental agent evok-ing a plastic response should reliably influence predation, reproduction or


some other component of fitness. On the other hand, where behaviourallyplastic responses are lacking, we may expect the environment is invariable,or that variation in the environment does not reliably predict a fitness-criticaloutcome. In the present study, the evidence that turbidity alone drives for-aging activity suggests that turbidity may likely have been a reliable cueof decreased predation risk during sturgeon evolutionary history. However,if predators are absent, decreased foraging in clear-environments becomesincreasingly non-adaptive (i.e., may be an evolutionary trap). As ecosys-tems continue to change, understanding how environmental cues interactwith predation risk will be important in order to help predict the effect ofenvironmental changes on behaviours shaped and driven by predation.

4.2. Foraging, predation, and environmental change

While foraging and avoiding predation are both necessary for survival, strate-gies shaped by predation are often under especially strong selection owing tothe high (asymmetrical) costs of losing one’s life. This high cost often shiftsthe balance toward conservative (safe) strategies that drive the probability offailure (death) toward zero (Bouskila & Blumstein, 1992). Failing to feedin any one particular instance may not result in death, but failing to avoida predator can be disastrous for prey. This idea is that of the ‘life-dinnerprinciple’ coined by Dawkins & Krebs (1979), and is generally believed tomanifest itself in predator–prey systems (but see Brodie & Brodie, 1999).Given the severe costs of failing to avoid a predator, it seems reasonable thatenvironmental variables that reliably influence predation are used as cuesto drive anti-predator behaviour. In the present study this seems to be thecase. Here, the risk of predation by visual predators was reduced in tur-bid environments relative to clear-water environments. Other environmentalvariables, such as temperature, salinity, or pH that are either negatively orpositively associated with predator activity or efficacy, may have similareffects in other systems. Future work exploring the interconnectedness of en-vironmental variables and behaviourally plastic anti-predator responses willhelp us better understand the suite of effects changing environments have onour study systems.

4.3. Sturgeon conservation in natural environments

It is becoming increasingly important for researchers to design studies thatnot only answer fundamental scientific questions, but also provide insights


for conservation, protection, and management of their study systems andhabitats (Caro & Sherman, 2011). All sturgeon species are now listed asendangered in part of their ranges, and most in all of their ranges (Billard& Lecointre, 2001). Behavioural plasticity may be especially important forsturgeon due to their relatively late age at maturity, long generation times,and protracted spawning periodicity (Peterson et al., 2007). Moreover, therate of molecular evolution in Acipenseriformes has been found to be nearlyhalf as fast as teleost fish (Krieger & Fuerst, 2002). In endangered sturgeonpopulations with little genetic diversity or low population growth, adaptiveevolution becomes increasingly unlikely (Carroll & Watters, 2008; Hendryet al., 2011), possibly accelerating the role that behavioural plasticity playsin mediating behavioural responses to changing environments. Changingturbidity levels in the aquatic environment are likely to influence sturgeon be-haviour and may have fitness consequences. Turbidity, therefore, should beconsidered an important element in habitat restoration programs. Increasedturbidity is likely to be highly beneficial to young sturgeon, whether preda-tors are present or not. Clear water, on the other hand, may significantlyrestrict foraging activity in lake sturgeon, even in the absence of predators.

4.4. Conclusions

Our results describe how environmental conditions alone can influence anti-predator behaviour, thereby driving foraging activity and potentially affect-ing fitness. Continuing to uncover the role of environmental variables onbehaviourally plastic traits is essential for understanding the processes thatgave rise to present-day foraging strategies. Even now the degree to whichbehaviourally plastic responses are influenced by changing environmentalcues is not well understood. Furthermore, given that rapidly changing cli-mates and habitats represent important challenges faced by the majority ofanimal species on our planet, and are leading to rapid declines in global bio-diversity, they deserve more attention.

Acknowledgements

We thank J. Hunter and Sustainable Sturgeon Culture for providing fishused in this study, and J. Sloychuk for help with fish care. Financial supportwas provided by the Natural Sciences and Engineering Research Council ofCanada to D.P.C and M.C.O.F., The Canadian Wildlife Federation to D.P.C.


and the University of Saskatchewan Department of Biology to V.W. andA.B.M. We are greatful to two anonymous reviewers for their helpful adviceon an earlier draft of this manuscript. All research herein was conducted un-der animal use protocol 20120013 issued by the University of Saskatchewanand complies with all provincial and federal legislation.

References

Abrahams, M. & Kattenfeld, M. (1997). The role of turbidity as a constraint on predator–preyinteractions in aquatic environments. — Behav. Ecol. Sociobiol. 40: 169-174.

Billard, R. & Lecointre, G. (2001). Biology and conservation of sturgeon and paddlefish. —Rev. Fish. Biol. Fisher. 10: 355-392.

Boglione, C., Bronzi, P., Cataldi, E., Serra, S., Gagliardi, F. & Cataudella, S. (1997). Aspectsof early development in the Adriatic sturgeon Acipenser naccarii. — J. Appl. Ichthyol.15: 207-2013.

Bonner, T.H. & Wilde, G.R. (2002). Effects of turbidity on prey consumption by prairiestream fishes. — Trans. Am. Fish. Soc. 131: 1203-1208.

Bouskila, A. & Blumstein, D.T. (1992). Rules of thumb in predation hazard assessment:predictions from a dynamic model. — Am. Nat. 139: 161-176.

Brodie III, E.D. & Brodie Jr., E.D. (1999). Predator–prey arms races. — BioScience 49: 557-568.

Burnham, K.P. & Anderson, D.R. (2002). Model selection and multimodal inference,2nd edn. — Springer, New York, NY.

Caro, T. & Sherman, P.W. (2011). Endangered species and a threatened discipline: be-havioural ecology. — Trends Ecol. Evol. 26: 111-118.

Carroll, S.P. & Watters, J.V. (2008). Managing phenotypic variability with genetic and en-vironmental heterogeneity: adaptation as a first principle of conservation practice. — In:Conservation biology: evolution in action (Carroll, S.P. & Fox, C.W., eds). Oxford Uni-versity Press, New York, NY, p. 181-198.

Chiu, S. & Abrahams, M.V. (2010). Effects of turbidity and risk of predation on habitatselection decisions by fathead minnow (Pimephales promelas). — Environ. Biol. Fish.87: 309-316.

Chivers, D.P., Al-Batati, F., Brown, G.E. & Ferrari, M.C.O. (2013). The effect of turbidity onrecognition and generalization of predators and non-predators in aquatic ecosystems. —Ecol. Evol. 3: 268-277.

Dawkins, R. & Krebs, J.R. (1979). Arms races between and within species. — Proc. Roy.Soc. Lond. B: Biol. Sci. 205: 489-511.

De Robertis, A., Ryer, C.H., Veloza, A. & Brodeur, R.D. (2003). Differential effects ofturbidity on prey consumption of piscivorous and planktivorous fish. — Can. J. Fish.Aquat. Sci. 60: 1517-1526.

DeWitt, T.J., Sih, A. & Hucko, J.A. (1999). Trait compensation and cospecialization in afreshwater snail: size, shape and anti-predator behaviour. — Anim. Behav. 58: 397-407.


Dewson, Z.S., James, A.B.W. & Death, R.G. (2007). A review of the consequences of de-creased flow for instream habitat and macroinvertebrates. — J. N. Am. Benthol. Soc. 26:401-415.

Ferrari, M.C.O., Lysak, K.R. & Chivers, D.P. (2010). Turbidity as an ecological constrainton learned predator recognition and generalization in prey fish. — Anim. Behav. 79: 515-519.

Ferrari, M.C.O., Ranåker, L., Weinersmith, K.L., Young, M.J., Sih, A. & Conrad, J.L. (2014).Effects of turbidity and an invasive waterweed on predation by introduced largemouthbass. — Environ. Biol. Fish. 97: 79-90.

Gadomski, D.M. & Parsley, M.J. (2005). Effects of turbidity, light level, and cover on preda-tion of white sturgeon larvae by prickly sculpins. — Trans. Am. Fish. Soc. 134: 369-374.

Ghalambor, C.K., Angeloni, L.M. & Carroll, S.P. (2010). Behavior as phenotypic plasticity.— In: Evolutionary behavioural ecology (Fox, C. & Westneat, D., eds). Oxford UniversityPress, New York, NY, p. 90-107.

Helfman, G.S. (1993). Fish behaviour by day, night, and twilight. — In: Behaviour of teleostfishes, 2nd edn. (Pitcher, T.J., ed.). Springer, London, p. 479-512.

Hendry, A.P., Farrugia, T.J. & Kinnison, M.T. (2008). Human influences on rates of pheno-typic change in wild populations. — Mol. Ecol. 17: 20-29.

Hendry, A.P., Kinnison, M.T., Heino, M., Day, T., Smith, T.B., Fitt, G., Bergstrom, C.T.,Oakeshott, J., Jørgensen, P.S., Zalucki, M.P., Gilchrist, G., Southerton, S., Sih, A., Strauss,S., Denison, R.F. & Carroll, S.P. (2011). Evolutionary principles and their practical appli-cation. — Evol. Appl. 4: 159-183.

Huey, R.B., Hertz, P.E. & Sinervo, B. (2003). Behavioral drive versus behavioural inertia inevolution: a null model approach. — Am. Nat. 161: 357-366.

Johnson, B.L., Richardson, W.B. & Naimo, T.J. (1995). Past, present, and future concepts inlarge river ecology. — BioScience 45: 134-141.

Krieger, J. & Fuerst, P.A. (2002). Evidence for a slowed rate of molecular evolution in theorder Acipenseriformes. — Mol. Biol. Evol. 19: 891-897.

Kundzewicz, Z.W., Mata, L.J., Arnell, N.W., Döll, P., Jimenez, B., Miller, K., Oki, T., Sen,Z. & Shiklomanov, I. (2008). The implications of projected climate change for freshwaterresources and their management. — Hydrol. Sci. J. 53: 3-10.

Lehtiniemi, M., Engtröm-Öst, J. & Viitasalo, M. (2005). Turbidity decreases anti-predatorbehaviour in pike larvae, Esox lucius. — Environ. Biol. Fish. 73: 1-8.

Liljendahl-Nurminen, A., Horppila, J., Uusitalo, L. & Niemistö, J. (2008). Spatial variabilityin the abundance of pelagic invertebrate predators in relation to depth and turbidity. —Aquat. Ecol. 42: 25-33.

Lima, S.L. (1998). Nonlethal effects in the ecology of predator–prey interactions. — Bio-Science 48: 25-34.

Lima, S.L. & Bednekoff, P.A. (1999). Temporal variation in danger drives anti-predatorbehavior: the predation risk allocation hypothesis. — Am. Nat. 153: 649-659.

Loew, E. & Sillman, A.J. (1993). Age-related changes in the visual pigments of the whitesturgeon (Acipenser transmontanus). — Can. J. Zool. 71: 1552-1557.


Moran, N.A. (1992). The evolutionary maintenance of alternative phenotypes. — Am. Nat.139: 971-989.

Orr, J.C., Fabry, V.J., Aumont, O., Bopp, L., Doney, S.C., Feely, R.A., Gnanadesikan, A.,Gruber, N., Ishida, A., Joos, F., Key, R.M., Lindsay, K., Maier-Reimer, E., Matear, R.,Monfray, P., Mouchet, A., Najjar, R.G., Plattner, G.-K., Rodgers, K.B., Sabine, C.L.,Sarmiento, J.L., Schlitzer, R., Slater, R.D., Totterdell, I.J., Weirig, M.-F., Yamanaka, Y.& Yool, A. (2005). Anthropogenic ocean acidification over the twenty-first century andits impact on calcifying organisms. — Nature 437: 681-686.

Peake, S. (1999). Substrate preferences of juvenile hatchery-reared lake sturgeon, Acipenserfulvescens. — Environ. Biol. Fish. 56: 367-374.

Peterson, D.L., Vecsei, P. & Jennings, C.A. (2007). Ecology and biology of the lake sturgeon:a synthesis of current knowledge of a threatened North American Acipenseridae. — Rev.Fish Biol. Fisher. 17: 59-76.

Richmond, A.M. & Kynard, B. (1995). Ontogenetic behavior of Shortnose SturgeonAcipenser brevirostrum. — Copeia: 172-182.

Rodríguez, A. & Gisbert, E. (2002). Eye development and the role of vision during Siberiansturgeon early ontogeny. — J. Appl. Ichthyol. 18: 280-285.

Rowe, D.K. & Dean, T.L. (1998). Effects of turbidity on the feeding ability of the juvenilemigrant stage of six New Zealand freshwater fish species. — New Zeal. J. Mar. Fresh. 32:21-29.

Schindler, D.W. (1988). Effects of acid rain on freshwater ecosystems. — Science 239: 149-157.

Shoup, D.E. & Wahl, D.H. (2009). The effects of turbidity on prey selection by piscivorouslargemouth bass. — Trans. Am. Fish. Soc. 138: 1018-1027.

Sih, A. (1987). Predators and prey lifestyles: an evolutionary overview. — In: Predation:direct and indirect impacts on aquatic communities (Kerfoot, W.C. & Sih, A., eds). NewEngland University Press, Hanover, NH, p. 203-224.

Skubinna, J.P., Coon, T.G. & Batterson, T.R. (1995). Increased abundance and depth ofsubmersed macrophytes in response to decreased turbidity in Saginaw Bay, Lake Huron.— J. Great Lakes Res. 21: 476-488.

Stankowich, T. & Blumstein, D.T. (2005). Fear in animals: a meta-analysis and review of riskassessment. — Proc. Roy. Soc. Lond. B: Biol. Sci. 272: 2627-2634.

Svendsen, J.C., Genz, J., Anderson, G.W., Stol, J.A., Watkinson, D.A. & Enders, E.C. (2014).Evidence of circadian rhythm, oxygen regulation capacity, metabolic repeatability andpositive correlations between forced and spontaneous maximal metabolic rates in LakeSturgeon Acipenser fulvescens. — PLoS ONE 9: e94693.

Symonds, M.R.E. & Moussalli, A. (2011). A brief guide to model selection, multimodel in-ference and model averaging in behavioural ecology using Akaike’s information criterion.— Behav. Ecol. Sociobiol. 65: 13-21.

Van de Meutter, F., De Meester, L. & Stoks, R. (2005). Water turbidity affects predator–preyinteractions in a fish-damselfly system. — Oecologia 144: 327-336.

Via, S. & Lande, R. (1985). Genotype-environment interaction and the evolution of pheno-typic plasticity. — Evolution 39: 505-522.


West-Eberhard, M.J. (1989). Phenotypic plasticity and the origins of diversity. — Annu. Rev.Ecol. Syst. 20: 249-278.

Winder, M. & Schindler, D.E. (2004). Climate change uncouples trophic interactions in anaquatic ecosystem. — Ecology 85: 2100-2106.

Wishingrad, V., Chivers, D.P. & Ferrari, M.C.O. (2014a). Relative cost/benefit trade-off be-tween cover-seeking and escape behaviour in an ancestral fish: the importance of structuralhabitat heterogeneity. — Ethology 120: 973-981.

Wishingrad, V., Sloychuk, J.R., Ferrari, M.C.O. & Chivers, D.P. (2014b). Alarm cues inLake Sturgeon Acipenser fulvescens Rafinesque, 1817: potential implications for life-skills training. — J. Appl. Ichthyol., DOI:10.1111/jai.12580.

Wishingrad, V., Ferrari, M.C.O. & Chivers, D.P. (2014c). Behavioural and morphologicaldefenses in a fish with a complex anti-predator phenotype. — Anim. Behav. 95: 137-143.

Zhang, X., Song, J., Fan, C., Guo, H., Wang, X. & Bleckmann, H. (2012). Use of electrosensein the feeding behavior of sturgeons. — Integr. Zool. 7: 74-82.

http://dx.doi.org/10.1111/jai.12580

wishingrad et al. 2015 behaviour.pdf

Documents