at SciVerse ScienceDirect

Animal Behaviour 83 (2012) 1505e1510

Contents lists available

Animal Behaviour

journal homepage: www.elsevier .com/locate/anbehav

Habitat change influences mate search behaviour in three-spined sticklebacks

Jan Heuschele a,b,c,*, Tiina Salminen a, Ulrika Candolin a

aDepartment of Biosciences, University of Helsinki, Helsinki, FinlandbNational Institute of Aquatic Resources, Technical University of Denmark, Charlottenlund, DenmarkcDepartment of Biology, University of Bergen, Bergen, Norway

a r t i c l e i n f o

Article history:Received 8 November 2011Initial acceptance 7 December 2011Final acceptance 13 March 2012Available online 27 April 2012MS. number: 11-00899R

Keywords:eutrophicationGasterosteus aculeatushabitat complexitymate choicemultiple cuessexual selectionthree-spined stickleback

* Correspondence: J. Heuschele, National Institute oUniversity of Denmark, Kavalergaarden 6, 2920 Charl

E-mail address: janheuschele@gmail.com (J. Heusc

0003-3472/$38.00 � 2012 The Association for the Studoi:10.1016/j.anbehav.2012.03.027

Mate choice is one of the main mechanisms of sexual selection, with profound implications for individualfitness. Changes in environmental conditions can cause individuals to alter their mate search behaviour,with consequences for mate choice. Human-induced eutrophication of water bodies is a global problemthat alters habitat structure and visibility in aquatic ecosystems. We investigated whether changes inhabitat complexity and male cue modality, visual or olfactory, influence mate search behaviour of femalethree-spined sticklebacks, Gasterosteus aculeatus. We allowed gravid females to search for mates inexperimental pools that contained two nesting males and one social female, under low and highstructural complexity (created from green Plexiglas sheets), with access to either visual or olfactory cuesof the individuals. We found increased habitat complexity reduced the number of visits to nesting males,while a switch from visual to olfactory cues reduced the time spent searching for males, the number ofvisits to nesting males, the time spent evaluating males, and the relative time spent associating withmales rather than females. Thus, females decreased mate searching and mate evaluation in the absenceof visual stimulation. This reduced the rate of mate encounters and probably also the opportunity forchoice. Our results show that changes in habitat structure and visibility can alter female mate searching,with potential consequences for the opportunity for sexual selection.� 2012 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

Studies on the impact of anthropogenic activities on ecosystemsfocus mainly on ecological processes and how changes in theseinfluence the abundance and distribution of species. Less attentionhas been given to human-induced changes in evolutionaryprocesses and how alterations in the strength and direction ofselection influence population dynamics, despite growing evidencefor contemporary evolution (e.g. Hairston et al. 2005; Candolin &Heuschele 2008; Hendry et al. 2008; Pelletier et al. 2009).

Sexual selection is one of the mechanisms of evolution, and oneof its main components is mate choice. Mate choice depends onmate encounter rate, which is influenced by environmentalconditions (Real 1990, 1991; Crowley et al. 1991). Changes in theenvironment, such as degraded visibility or reduced populationdensity, can alter the costs and benefits ofmate searching and causeindividuals to change their search behaviour (Candolin & Wong2012). Female fiddler crabs, Uca mjoebergi, for instance, reducetheir investment in mate searching when the distance to moreattractive males is large (Booksmythe et al. 2008), while treeswallows, Tachycineta bicolor, choose extrapair mates closer to theirown nest when the cost of mate searching is increased

f Aquatic Resources, Technicalottenlund, Denmark.hele).

dy of Animal Behaviour. Published

experimentally (Dunn &Whittingham 2007). In general, the effectsof altered search costs on mate choosiness are well documented.

Surprisingly little attention has, however, been paid to the effectof anthropogenic activities on mate search behaviour and itsconsequences for adaptive mate choice. Considering the profoundimpact that humans are having on the environment, and the crucialimportance of mate encounter rate in mate choice (Kokko &Mappes 2005), more research is clearly needed. Mate searchingprecedes mate choice and determines the number and quality ofmates encountered. Changes in mate searching because of envi-ronmental change could thus have profound implications for matechoice, irrespective of the direct effect of the environment on matechoice. Moreover, human-induced environmental changes areusually more rapid than natural changes and can preclude geneticadaptation, which forces individuals to rely on earlier evolvedbehavioural reaction norms to survive the changes (Kinnison &Hairston 2007; Tuomainen & Candolin 2010). These behaviouralresponses are not necessarily adaptive and can drive populationsinto decline (Ghalambor et al. 2007; Candolin & Wong 2012).Thus, changes in mate search behaviour can influence both theprobability of mating and the quality of themates acquired, and canimpinge on population persistence.

A growing human-induced problem in aquatic ecosystems iseutrophication. Increased input of nutrients into the water triggers

by Elsevier Ltd. All rights reserved.

J. Heuschele et al. / Animal Behaviour 83 (2012) 1505e15101506

primary production, which alters habitat structure and reducesvisibility (Cloern 2001). This has been found to influence matechoice in a range of aquatic species (Heuschele et al. 2009; Sundinet al. 2010; Maan et al. 2010; van der Sluijs et al. 2010), but effectson mate search behaviour have received little attention. In theBaltic Sea, the three-spined stickleback, Gasterosteus aculeatus,breeds in habitats that currently are changing because of increasedgrowth of micro- and macroalgae. Shallow bays that previouslywere sparsely vegetated have become densely vegetated withfilamentous algae, particularly with Cladophora glomerata, whilethe water has become more turbid because of the growth ofphytoplankton. These changes have been found to reduce mateencounter rate of males in the field (Candolin & Voigt 2001),impede the ability of females to discriminate among males andprolong mate evaluation time (Engström-Öst & Candolin 2007;Wong et al. 2007; Candolin et al. 2007). In addition, female stick-lebacks switch from a predominant use of visual cues to anincreased use of olfactory cues when visibility is poor (Heuscheleet al. 2009). The switch could influence mate choice, as olfactorysignals provide information mainly on genetic compatibility(e.g. Aeschlimann et al. 2003), while visual signals convey infor-mation about both direct and indirect benefits (McLennan 2006).

The use of multiple cues in different sensory modalities canfacilitate the detection and recognition of potential mates, provideadditional information on mate quality and improve the evaluationof cues (Møller & Pomiankowski 1993; Johnstone 1996; Candolin2003; Hebets & Papaj 2004; Bro-Jørgensen 2010). Changes in envi-ronmental conditions that reduce the number of cues that can beused, or the efficiency and information content of these cues, couldrequire an increased mate search effort to maintain a high mateencounter rate. We investigated whether female sticklebacks adjusttheir mate search behaviour to changes in habitat complexity andavailability of male cues, visual or olfactory, to maintain a high mateencounter rate. Dense vegetation has been found to reduce mateencounter rate in the field (Candolin & Voigt 2001), but the mech-anism underpinning the reduction is unknown. A reduction in mateencounter rate could result in reduced mate choice, which in turncould influence offspring production and the genetic constitution ofthe population (Neff & Pitcher 2004; Candolin & Wong 2012).

METHODS

Fish Collection and Maintenance

Adult three-spined sticklebacks were caught withminnow trapsfrom a shallow bay on Långskär Island in the Baltic Sea, close to

Opaque separator

PlexiglastubeNestingdish

Low structural complexity

Release ar

m1 m2

f2f1

Figure 1. Experimental pools with low

Tvärminne Zoological Station in southern Finland (60� N, 23� E).The fish were caught in May and June 2008. They were transportedto the station within 30 min, in buckets (ca. 1 fish/litre) filled withwater from the site. All fish were healthy and active on arrival.The fish were housed in an outdoor facility under natural lightconditions. Females were housed in large flow-through holdingtanks (150 litres) at a density of 0.25 fish/litre. Males were placed inindividual flow-through aquaria (10 litres) that contained a nestingdish (11 cm in diameter), which was filled with sand and filamen-tous algae, C. glomerata, as nesting material (Candolin 1997).A gravid female, enclosed in a transparent perforated Plexiglascylinder, was presented to each male for 30 min daily to stimulatenest building. All fish were fed frozen bloodworms (chironomidlarvae) ad libitum. After the experiment, the fish were released atthe site of origin. The experiment was conducted in June and earlyJuly and all trials were performed between 0900 and 1700 hours.The experimental procedures were approved by the Animal CareCommittee of the University of Helsinki (86-06 and STH421A).

Mate Search Experiment

To investigate whether habitat structural complexity and malecue modality influence female mate search behaviour, we allowedfemales to search for males under two artificial habitat complex-ities, low and high, created from Plexiglas sheets, andwith access toeither visual or olfactory cues of males. We transferred tworandomly selected males, which had completed nest building andhad entered the courtship stage, together with their nesting dishes,to a pool with a diameter of 1.5 m and a water depth of 20 cm. Themales were enclosed in separate Plexiglas cylinders (diameter30 cm; Fig. 1). A nongravid female was added, enclosed in her owncylinder, to control for the interest of the female in shoaling withconspecifics rather than searching for nesting males. The threecylinders were evenly distributed within the pool (Fig. 1). Tomanipulate cue availability, the cylinders were either opaque withsmall holes, allowing the transmission of olfactory cues but not ofvisual cues, or transparent and sealed, allowing the transmission ofvisual cues but not of olfactory cues. All individuals were trans-ferred to the pools in the evening before the experimental trial tookplace to give them time to acclimatize.

During the day of the experimental trial, we presented a gravidfemale in a water-filled glass jar to each male for 5 min to stimulatecourtship (Rowland 2004). After the stimulus females had beenremoved, we placed a clear Plexiglas cylinder with a gravid female(diameter 12 cm, open at both ends) next to the social female. After5 min of acclimatization, we gently removed the cylinder and

High structural complexity

ea

Malem1 m2

f1f2

‘Social’ femaleFocal female

and high structural complexity.

J. Heuschele et al. / Animal Behaviour 83 (2012) 1505e1510 1507

released the female. The behaviour of the female was recorded for20 min. The female experienced two treatments, low and highhabitat complexity, with a 2 h break in between when she wasgently removed from the pool. We manipulated habitat complexityby adding or removing green Plexiglas sheets (Fig. 1). The sheetshampered visibility and increased the need of the female to searchfor males, which resembles the effect of the dense curtains offilamentous algae found in the shallow breeding bays, while stan-dardizing the treatments. The order of the two habitat treatmentswas randomized among replicates. The female was exposed to thesame individuals (two males and one female) in the two habitatcomplexity treatments. The location of the two males was alter-nated between treatments to eliminate an effect of position on theresults. The water in the pool was mixed between treatments todiffuse any olfactory cues remaining in the water and to allowa new gradient to build up. There was no effect of the order of thetreatments on the results, which suggests that any olfactory cuesremaining in the water did not influence female behaviour. Eachfemale had access to only olfactory or only visual cues of the threeindividuals. One female spawned before the end of the trial andwasexcluded from the analysis, leaving 25 visual and 25 olfactoryreplicates. No males or females were reused among replicates.

Female search behaviour was recorded using a top-mountedcamcorder. From the videos, we recorded the total time spentswimming, the number of visits to each enclosed individual withina 5 cm zone of its cylinder and the duration of the visits, using thesoftware JWatcher 1.0 (http://www.jwatcher.ucla.edu/). Total timespent swimming reflects search activity, encounter rate (number ofvisits/swimming time) reflects search efficiency, and number andduration of visits reflect female preferences.

The wet mass and standard length of the individuals weremeasured after the experimental trial. Body size, swimming time,and number and duration of visits were log transformed prior toanalyses.

Male Coloration

To determine whether female behaviour correlated with malered nuptial coloration and whether this depended on the treat-ment, the coloration of the two enclosed males was assessed in theevening before the experimental trial and immediately after thetrial (Candolin 1999). The male was placed into a small glass boxcontaining a black sponge that fixed him in position, after whichthe glass box was placed in a nonreflective dark box containinga camera, Canon EOS D350. The lateral side of the male was pho-tographed under standardized settings (aperture f10.0, shutterspeed 1/200, ISO 200) using the inbuilt flash. Constant colourrepresentation was confirmed by measuring RGB values of white,red, green and blue colour markings attached to the surface of theglass box. The similarity between human and stickleback percep-tion allowed us to use the camera’s trichromatic colour system toanalyse the red coloration (Rowe et al. 2006). The area of red wasmeasured using the imaging software ImageJ version 1.38w (http://rsb.info.nih.gov/ij/) with the plug-in ‘Threshold Color’ (http://www.dentistry.bham.ac.uk/landinig/software/software.html). We chosethe HSB colour model and selected red areas using the followingvalues: hue 0e32 and 218e255, saturation 75e255 and brightness0e255. The images were converted to 8-bit grey scale, and theselected areas were measured using the ‘Image/Adjust/Threshold’and the ‘Analyze/Measure’ commands. We measured the total areaof the male to calculate the percentage red area. The quality of thered colour of the selected area was estimated using the red‘intensity’, calculated as R/(RþGþB) (Barber et al.1999) using valuesfrom the ‘Measure RGB’ Plugin of ImageJ. We omitted two replicatesfrom the analyses because of missing images.

Statistical Analysis

We determined the effect of habitat complexity and cuemodality on female search behaviour using linear mixed-effectmodels with habitat complexity and cue modality as fixed factorsand focal female as a random factor. We used the statistical soft-ware R version 2.9.1 (R Development Core Team 2010) and the lmefunction for normally distributed data (nlme package; Pinheiroet al. 2010) and the lmer function (lme4 package; Bates et al.2011) for binomially distributed data.

The initial models included all interactions between the fixedeffects, which were reduced to the minimum adequate modelsusing stepwise backward deletion. We verified the reliability of theresults of the final model by inspecting the distribution of theresiduals visually (qeq plot, histogram of residuals) and statistically(ShapiroeWilk). The order of the two habitat treatments had noeffect on search behaviour and is not considered further.

To examine whether habitat complexity and cue modalityinfluenced the proportion of visits to males compared to the female(number and duration of visits) and, thus, mate search behaviour,we included the sex of the visited individual as a fixed factor in themodels. We first determined the probability of an individual beingvisited using a mixed model with a binominal link function. Wethen investigated the influence of the independent factors (habitatcomplexity, cue modality and sex of the individual) on the numberand duration of the visits for the individuals that were visited.

To investigate whether habitat complexity and cue modalityinfluenced female mate preferences, we randomly designated onemale as Male 1 and calculated for this male the proportion of totalfemale visits (visitsMale1/(visitsMale1 þ visitsMale2)) and the propor-tion of total visit duration (timeMale1/(timeMale1 þ timeMale2)).Proportions were arcsine square-root transformed. We used theproportion in the first habitat treatment as the dependent factorand the proportion in the second treatment as a covariate, keepingcue modality as a fixed factor. To determine whether habitatcomplexity and cuemodality influenced female preference formalecharacteristics (or correlates thereof), we correlated for the twomales the ratio of body size (lengthMale1/lengthMale2), area of redcoloration (Rmale1/Rmale2) and red colour quality (RIMale1/RIMale2)with the proportion of both female visits and duration of visits,keeping both structure and cue modality as fixed factors. Theseproportions were log transformed.

RESULTS

Search Behaviour

Females spent less time swimming when they had access toolfactory cues instead of visual cues of the three enclosed individ-uals (F1,48 ¼ 9.882, P ¼ 0.003; Fig. 2), and less time swimming in themore complex habitat (F1,49 ¼ 6.250, P ¼ 0.016). Females visited theenclosed individuals less often with olfactory cues (z ¼ 2.574,P ¼ 0.010; total number of visits: 300 in 50 pools) and in the morecomplex habitat (z ¼ 2.319, P ¼ 0.020). Calculated per minute spentswimming, females paid fewer visits to the enclosed individualswith olfactory cues (F1,48 ¼ 21.204, P < 0.001) and in the morecomplex habitat (F1,47 ¼ 7.405, P ¼ 0.009; Fig. 3).

The distribution of visits among the two males and the socialfemale depended on cue modality but not on habitat complexity:females visited the individual males more often than the socialfemale when they had access to visual cues, while there was nodifference in the distribution of visits with olfactory cues (interac-tion cue * sex: F1,171 ¼ 5.998, P ¼ 0.015; Figs 4, 5). The visits to themales were longer than visits to the female with visual cues, butshorter than visits to the female with olfactory cues (interaction

Cue typeM

ean

nu

mbe

r of

vis

its

0

1

2

3

4

5

6

Visual Olfactory

SexFemaleMale

Figure 4. Mean � SE number of visits to each enclosed individual with access to visualor olfactory cues. The two habitat treatments are combined.

200

Sex

Female

Cue type

Mea

n t

ime

spen

t sw

imm

ing

(min

)

0

5

10

15

Visual Olfactory

Complexity

Low

High

Figure 2. Mean � SE time spent swimming with access to visual or olfactory cues ofthe enclosed individuals under the two habitat complexities.

J. Heuschele et al. / Animal Behaviour 83 (2012) 1505e15101508

cue * sex: F1,171 ¼7.797, P ¼ 0.006; Fig. 5). The proportion of visits toeach male correlated with the proportion of time spent with themale (F1,42 ¼ 792.565, P < 0.0001), independent of habitatcomplexity and cue modality.

Female Preference for Male Characteristics

The ratio of body size, area of red and red colour quality (bothbeforeandafter the trial) betweenthe twomalesdidnot correlatewiththe proportion of visits each male received or the proportion of timethe females spent with eachmale, independent of habitat complexityand cue modality (all P> 0.05). Changes in male red coloration overtheexperimentdidnot correlatewith femalepreferences independentof habitat complexity and cue modality (all P> 0.05).

Cue type

Mea

n n

um

ber

of v

isit

s p

er s

wim

min

g m

inu

te

0

0.5

1

Visual Olfactory

ComplexityLowHigh

Figure 3. Mean � SE number of visits per minute spent swimming to all enclosedindividuals, with access to visual or olfactory cues of the enclosed individuals underthe two habitat complexities.

DISCUSSION

We found female mate search behaviour depended on themodality of the cues of the enclosed individuals and, to a lesserdegree, on the structural complexity of the habitat. Females spentless time swimming and visited the enclosed individuals less oftenwith access to olfactory rather than visual cues, and visited eachindividual less often in themore complex habitat. Thus, females did

Cue type

Mea

n d

ura

tion

of

visi

ts (

s)

50

100

150

Visual Olfactory

Male

Figure 5. Mean � SE duration of visits to each enclosed individual with access tovisual or olfactory cues. The two habitat treatments are combined.

J. Heuschele et al. / Animal Behaviour 83 (2012) 1505e1510 1509

not compensate for reduced visibility and increased structuralcomplexity by increasing search activity. Instead, their interest insearching for males declined, probably because of the lack of visualstimulation. This resulted in a lower mate encounter rate, both inthe more complex habitat and with the access to olfactory ratherthan visual cues. This suggests that the documented lower mateencounter rate in dense vegetation in the field (Candolin & Voigt2001) could arise, at least partly, from reduced female searchactivity when visibility is obstructed by dense curtains of fila-mentous algae.

The modality of the cues of the nesting males and the socialfemale influenced the distribution of the visits. The female visitedand spent more time with the males than with the social femalewhen she had access to visual cues, but showed no preference foreither sex with olfactory cues, although olfactory cues are known toinfluence mate choice (Aeschlimann et al. 2003). Thus, the lack ofvisual cues of the nesting males and their replacement witholfactory cues decreased the interest of the female in searching formales. It is possible that the confinement of the males in theperforated Plexiglas cylinders hampered the spread of olfactorycues and resulted in an unnaturally low gradient of olfactory cues,which could partly explain the low search activity with these cues.Yet, low activity with olfactory cues has also been found in otherfishes that reproduce in shallow water: guppy, Poecilia reticulata,females show no mate preference when only olfactory cues arepresent (Shohet & Watt 2004; Archard et al. 2008), and broad-nosed pipefish, Syngnathus typhle, males do not distinguishbetween males and females if only olfactory cues are accessible(Lindqvist et al. 2011). Thus, visual cues appear essential for matesearching to occur in several fishes that spawn in shallowwater andexpress mate choice. Their loss cannot be fully compensated byolfactory cues.

The visits to the social female were longer than visits to themales with olfactory cues, but not with visual cues. Thus, femaleswere more interested in associating with females than searchingfor males when visual cues of nesting males were absent. Femalesticklebacks form shoals during the breeding season (Mori 1995),probably as protection against predators and to facilitate foragingand nest raiding (Whoriskey & Fitzgerald 1985). This could explainthe preference of the focal female for the social female when nonesting males, or nest activities, were visible. This furtherstrengthens the conclusion that females are less interested insearching for males when visual cues are lacking.

No consistency in mate preferences for certain males or theirtraits was detected between the two habitat complexity treat-ments, either with visual or with olfactory cues.Whether this is dueto a lack of preferences or to preferences changing depending onhabitat complexity cannot be determined. Female preferences didnot correlate with differences between the males in body size orred nuptial coloration, in agreement with earlier studies on thesame population (e.g. Heuschele et al. 2009).

Our results show that female sticklebacks decrease their matesearch activity and mate evaluation time when visual cues of malesare lacking, probably because of the absence of visual stimulation.Olfactory cues did not make up for the loss of stimulation, and,hence, mate encounter rate decreased. This suggests thatenvironmental changes that increase habitat complexity andreduce visibility have the potential to reduce mate encounter rate.This could, in turn, hamper female mate choice and, ultimately,influence population viability. Earlier research on the investigatedpopulation shows that the variation among males in the field innumber of eggs in their nests is lower in denser vegetation, whichindicates that the variation in mating success is also lower(Candolin 2004). The present results suggest that the reductioncould be caused by reduced mate searching and mate evaluation in

dense vegetation. The consequences that this could have for pop-ulation dynamics deserve closer attention. Relaxed mate choice canhave both positive and negative effects on population viability(Candolin & Heuschele 2008), as more random mate choice canreduce population viability, while reduced search activity canfavour population growth by reducing energy expenditure andpredation risk (Møller & Alatalo 1999; Whitlock 2000; Lorch et al.2003, Candolin & Wong 2012). Whether the positive or negativeeffects predominate in the investigated population is unknown.

Multimodal signals are assumed to facilitate female choice,particularly by improving the detection and recognition of poten-tial mates and providing more information on mate quality(Candolin 2003; Hebets & Papaj 2004; Bro-Jørgensen 2010). Thisstudy shows that a communication channel lost because of changesin the environment is not necessarily replaced by another. Instead,the loss can result in reduced mate searching and mate evaluation.More studies are now needed on the ability of animals to modifytheir use of mate choice cues when adapting to changing condi-tions. We also need more information on the consequences thataltered mate search behaviour has for individual fitness and pop-ulation dynamics.

Acknowledgments

We thank L. Vlieger for assistance, C.W. Berg and P. Gienapp forstatistical advice, B. Diaz Pauli for discussions and two anonymousreferees for insightful comments. Tvärminne Zoological Stationprovided working facilities. This work was supported by theAcademy of Finland to U.C. J.H. is currently financed by a DFGResearch Fellowship (grant number HE 6050/1).

References

Archard, G. A., Cuthill, I. C., Partridge, J. C. & van Oosterhout, C. 2008. Femaleguppies (Poecilia reticulata) show no preference for conspecific chemosensorycues in the field or an artificial flow chamber. Behaviour, 145, 1329e1346.

Aeschlimann, P., Haberli, M., Reusch, T., Boehm, T. & Milinski, M. 2003. Femalesticklebacks Gasterosteus aculeatus use self-reference to optimize MHC allelenumber during mate selection. Behavioral Ecology and Sociobiology, 54,119e126.

Barber, I., Arnott, S. A., Braithwaite, V. A., Andrew, J., Mullen, W. &Huntingford, F. A. 1999. Carotenoid-based sexual coloration and body condi-tion in nesting male sticklebacks. Journal of Fish Biology, 57, 777e790.

Bates, D., Maechler, M. & Bolker, B. 2011. lme4: Linear Mixed-Effects Models Using S4Classes. R package version 0.999375-42. http://CRAN.R-project.org/package¼lme4.

Booksmythe, I., Detto, T. & Backwell, P. R. Y. 2008. Female fiddler crabs settle forless: the travel costs of mate choice. Animal Behaviour, 76, 1775e1781.

Bro-Jørgensen, J. 2010. Dynamics of multiple signalling systems: animal commu-nication in a world in flux. Trends in Ecology and Evolution, 25, 292e300.

Candolin, U. 1997. Reproduction under predation risk and the trade-off betweencurrent and future reproduction in the threespine stickleback. Proceedings of theRoyal Society B, 265, 1171e1175.

Candolin, U. 1999. The relationship between signal quality and physical condition:is sexual signalling honest in the three-spined stickleback? Animal Behaviour,58, 1261e1267.

Candolin, U. 2003. The use of multiple cues in mate choice. Biological Reviews, 78,575e595.

Candolin, U. 2004. Effects of algae cover on egg acquisition in male three-spinedstickleback. Behaviour, 141, 1389e1399.

Candolin, U. & Heuschele, J. 2008. Is sexual selection beneficial during adaptationto environmental change? Trends in Ecology & Evolution, 23, 446e452.

Candolin, U. & Voigt, H. 2001. Correlation between male size and territory quality:consequence of male competition or predation susceptibility? Oikos, 95,225e230.

Candolin, U. & Wong, B. B. M. 2012. Sexual selection in changing environments:consequences for individuals and populations. In: Behavioural Responses toa Changing World: Mechanisms and Consequences (Ed. by U. Candolin &B. B. M. Wong), pp. 201e215. Oxford: Oxford University Press.

Candolin, U., Salesto, T. & Evers, M. 2007. Changed environmental conditionsweaken sexual selection in sticklebacks. Journal of Evolutionary Biology, 20,233e239.

Cloern, J. 2001. Our evolving conceptual model of the coastal eutrophicationproblem. Marine Ecology Progress Series, 210, 223e253.

J. Heuschele et al. / Animal Behaviour 83 (2012) 1505e15101510

Crowley, P., Travers, S., Linton, M., Cohn, S., Sih, A. & Sargent, R. 1991. Matedensity, predation risk, and the seasonal sequence of mate choices: a dynamicgame. American Naturalist, 137, 567e596.

Dunn, P. & Whittingham, L. 2007. Search costs influence the spatial distribution,but not the level, of extra-pair mating in tree swallows. Behavioral Ecology andSociobiology, 61, 449e454.

Engström-Öst, J. & Candolin, U. 2007. Human-induced water turbidity altersselection on sexual displays in sticklebacks. Behavioral Ecology, 18, 393e398.

Ghalambor, C. K., Mckay, J. K., Carroll, S. P. & Reznick, D. N. 2007. Adaptive versusnon-adaptive phenotypic plasticity and the potential for contemporary adap-tation in new environments. Functional Ecology, 21, 394e407.

Hairston, N., Jr., Ellner, S., Geber, M., Yoshida, T. & Fox, J. 2005. Rapid evolutionand the convergence of ecological and evolutionary time. Ecology Letters, 8,1114e1127.

Hebets, E. A. & Papaj, D. R. 2004. Complex signal function: developing a frameworkof testable hypotheses. Behavioral Ecology and Sociobiology, 57, 197e214.

Hendry, A. P., Farrugia, T. J. & Kinnison, M. T. 2008. Human influences on rates ofphenotypic change in wild animal populations. Molecular Ecology, 17, 20e29.

Heuschele, J., Mannerla, M., Gienapp, P. & Candolin, U. 2009. Environment-dependent use of mate choice cues in sticklebacks. Behavioral Ecology, 20,1223e1227.

Johnstone, R. A.1996. Multiple displays in animal communication:‘backup signals’ and‘multiple messages.’. Philosophical Transactions of the Royal Society B, 351, 329e338.

Kinnison, M. T. & Hairston, N. G. 2007. Eco-evolutionary conservation biology:contemporary evolution and the dynamics of persistence. Functional Ecology, 21,444e454.

Kokko, H. & Mappes, J. 2005. Sexual selection when fertilization is not guaranteed.Evolution, 59, 1876e1885.

Lindqvist, C., Sundin, J., Berglund, A. & Rosenqvist, G. 2011. Male broad-nosedpipefish Syngnathus typhle do not locate females by smell. Journal of FishBiology, 78, 1861e1867.

Lorch, P., Proulx, S., Rowe, L. & Day, T. 2003. Condition-dependent sexual selectioncan accelerate adaptation. Evolutionary Ecology Research, 5, 867e881.

McLennan, D. A. 2006. The umwelt of the threespine stickleback. In: The Biology ofthe Threespine Stickleback (Ed. by S. Östlund-Nilsson, I. Mayer & F. Huntingford),pp. 179e224. Boca Raton: CRC Press.

Maan, M. E., Seehausen, O. & van Alphen, J. J. M. 2010. Female mating preferencesand male coloration covary with water transparency in a Lake Victoria cichlidfish. Biological Journal of the Linnean Society, 99, 398e406.

Møller, A. & Alatalo, R. 1999. Good-genes effects in sexual selection. Proceedings ofthe Royal Society B, 266, 85e91.

Møller, A. & Pomiankowski, A. 1993. Why have birds got multiple sexual orna-ments. Behavioral Ecology and Sociobiology, 32, 167e176.

Mori, S. 1995. Spatial and temporal variations in nesting success and the causes ofnest losses of the freshwater three-spined stickleback, Gasterosteus aculeatus.Environmental Biology of Fishes, 43, 323e328.

Neff, B. & Pitcher, T. 2004. Genetic quality and sexual selection: an inte-grated framework for good genes and compatible genes. Molecular Ecology, 14,19e38.

Pelletier, F., Garant, D. & Hendry, A. P. 2009. Eco-evolutionary dynamics.Philosophical Transactions of the Royal Society B, 364, 1483e1489.

Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D. & R Development Core Team 2010.nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1-97. http://cran.r-project.org/web/packages/nlme/index.html.

R Development Core Team 2010. R: A Language and Environment for StatisticalComputing. Vienna: R Foundation for Statistical Computing.

Real, L. 1990. Search theory and mate choice 1. Models of single-sex discrimination.American Naturalist, 136, 376e405.

Real, L. 1991. Search theory and mate choice 2. Mutual interaction, assortativemating, and equilibrium variation in male and female fitness. AmericanNaturalist, 138, 901e917.

Rowe, M., Baube, C. & Phillips, J. 2006. Trying to see red through sticklebackphotoreceptors: functional substitution of receptor sensitivities. Ethology, 112,218e229.

Rowland, W. R. 2004. Sexual arousal increases mate selectivity in the stickleback.Behaviour, 141, 1371e1387.

Shohet, A. J. & Watt, P. J. 2004. Female association preferences based on olfactorycues in the guppy, Poecilia reticulata. Behavioral Ecology and Sociobiology, 55,363e369.

van der Sluijs, I., Gray, S. M., Amorim, M. C. P., Barber, I., Candolin, U.,Hendry, A. P., Krahe, R., Maan, M. E., Utne-Palm, A. C., Wagner, H. J., et al.2010. Communication in troubled waters: responses of fish communicationsystems to changing environments. Evolutionary Ecology, 25, 623e640.

Sundin, J., Berglund, A. & Rosenqvist, G. 2010. Turbidity hampers mate choice ina pipefish. Ethology, 116, 713e721.

Tuomainen, U. & Candolin, U. 2010. Behavioural responses to human-inducedenvironmental change. Biological Reviews of the Cambridge PhilosophicalSociety, 86, 640e657.

Whitlock, M. 2000. Fixation of new alleles and the extinction of smallpopulations: drift load, beneficial alleles, and sexual selection. Evolution, 54,1855e1861.

Whoriskey, F. & Fitzgerald, G. 1985. Sex, cannibalism and sticklebacks. BehavioralEcology and Sociobiology, 18, 15e18.

Wong, B. B. M., Candolin, U. & Lindström, K. 2007. Environmental deteriorationcompromises socially enforced signals of male quality in three-spined stickle-backs. American Naturalist, 170, 184e189.