CONCEPTS & SYNTHESIS EMPHASIZING NEW IDEAS TO STIMULATE RESEARCH IN ECOLOGY

Ecology, 00(0), 0000, pp. 000–000 0000 by the Ecological Society of America

SOCIAL INFORMATION USE IS A PROCESS ACROSS TIME, SPACE, AND ECOLOGY, REACHING HETEROSPECIFICS

1,4 2,5 1 3 JANNE-TUOMAS SEPPA¨NEN, JUKKA T. FORSMAN, MIKKO MO¨NKKO¨NEN, AND ROBERT L. THOMSON 1Department of Biological and Environmental Science, POB 35, FIN-40014, University of Jyva¨skyla¨, Jyva¨skyla¨, Finland 2Department of Ecology, EBC, Uppsala University, Norbyva¨gen 18D, SE-752 36 Uppsala, Sweden 3Section of Ecology, Department of Biology, University of Turku, 20014 Turku, Finland

Abstract. Decision making can be facilitated by observing other individuals faced with the same or similar problem, and recent research suggests that this social information use is a widespread phenomenon. Implications of this are diverse and profound: for example, social information use may trigger cultural evolution, affect distribution and dispersal of populations, and can involve intriguing cognitive traits. We emphasize here that social information use is a process consisting of the scenes of (1) event, (2) observation, (3) decision, and (4) consequence, where the initial event is a scene in such a process of another individual. This helps to construct a sound conceptual framework for measuring and studying social information use. Importantly, the potential value of social information is affected by the distance in time, space, and ecology between the initial observation and eventual consequence of a decision. Because negative interactions between individuals (such as direct and apparent competition) also depend on the distance between individuals along these dimensions, the potential value of information and the negative interactions may form a trade-off situation. Optimal solutions to this trade-off can result in adaptively extended social information use, where using information gathered some time ago, some distance away, and from ecologically different individuals is preferred. Conceivably, using information gathered from a heterospecific individual might often be optimal. Many recent studies demonstrate that social information use does occur between , and the first review of published cases is provided here. Such interaction between species, especially in selection, has important consequences for community ecology and conservation. Adaptively extended social information use may also be an important evolutionary force in guild formation. Complex coevolutionary patterns may result depending on the effect of information use on the provider of information. Key words: autocorrelation; habitat selection; heterospecific attraction; public information; social information; social learning.

QUESTIONS ON SOCIAL INFORMATION USE This phenomenon, generally termed social informa- Organisms continuously face situations where a tion use or social learning, has attracted a lot of research interest in recent years and appears to be widespread decision has to be made with imperfect information across the animal kingdom (Danchin et al. 2004). Social and thus with uncertainty about the consequences of information use is an exciting topic for several reasons. that decision. Fortunately, fellow beings are usually It is a mechanism allowing cultural evolution—a faced with at least a similar conundrum, and the possible, potentially important, but often hotly debated decisions made by—and the repercussions for—these pathway of inheritance. Important aspects of a species’ individuals can be used as information to aid in one’s ecology, such as foraging, spatial distribution, and own decision-making process. dispersal, may be profoundly affected by social infor- mation use, leading to population-level consequences. Manuscript received 18 October 2006; revised 3 January Questions on the acquisition, processing, and usage of 2007; accepted 12 December 2006. Corresponding Editor: B. P. social information bring in intriguing cognitive issues in Kotler. 4 E-mail: janseppa@jyu.fi communication, social learning, and imitation. Studying 5 Present address: Department of Biology, University of a phenomenon linking such diverse aspects of evolu- Oulu, POB 3000, FI-90014 Oulu, Finland. tionary, behavioral, and cognitive ecology as well as JANNE-TUOMAS SEPPA¨NEN ET AL. Ecology, Vol. 00, No. 0 YNTHESIS &S

FIG. 1. Schematic illustration of three linked sequences of information use. Alice observes (personal sample) the environment directly. Bob observes the publicly observable manifestation of Alice’s decision (social cue). Carol observes the consequence of Bob’s information use (vicarious sample). Not illustrated is the case of observing another’s broadcasted observation (evolved signal

ONCEPTS or semantic communication). The hypothetical example underlines that even if the source accurately reflects resources at the time of

C observation and the decision is correct in the light of that information, information value is not necessarily positive if the consequence is realized temporally, spatially, or ecologically far from the observation.

evolutionary psychology and anthropology (Boyd and linked to an information use process of another Richerson 1985, Whiten and Ham 1992) is likely to yield individual (Fig. 1). This definition firmly connects the important, integrating insights. concept of social information to events and entities Recent reviews of social information use cover amenable to empirical research, and it invokes a natural foraging (Galef and Giraldeau 2001), performance of measure of the value of information: fitness. Also, a others and cultural evolution (Danchin et al. 2004), and categorization of the types of social information, if such implementing statistical decision theory (Dall et al. is sought, logically follows from this framework. 2005). The categorization (Zentall 2006) and strategies Recognizing the entire sequence instead of just the (Laland 2004) of acquisition of social information are sources or mechanisms in the process of social informa- better understood, but there is little agreement on central tion use provides novel insights. Importantly, informa- definitions and concepts of social-information use in tion use is necessarily an extended process in temporal, ecology. What is social information use exactly? Where, spatial, and ecological dimensions, and this is likely to how, and between whom is it expected to occur? Can and have far-reaching consequences. The second part of this should specific types of social information be distin- paper elaborates our view that the effect of autocorre- guished, and if so, what are they (Bednekoff 2005, Dall lation on the trade-offs between competition and 2005, Danchin et al. 2005, Lotem and Winkler 2005)? information use along temporal, spatial, and ecological The first part of this paper outlines the argument that dimensions can affect how social information is used. information use in general is a sequential process The extended nature of social information use can also (McNamara and Houston 1980, Dall et al. 2005), and determine what, and perhaps more importantly whom, social information use is formed when such a process is organisms should use as a source of social information. Month 2007 SOCIAL INFORMATION FROM HETEROSPECIFICS

Plausibly, individuals of most animal species are uncertainty about the environment is reduced. In other constantly surrounded by diverse sights, sounds, and words, for information to have value, it must change the scents that mirror decisions of someone else, while some functioning of an organism in an evolutionarily relevant fewer scenes reveal the success of the decisions made. manner (Dall et al. 2005, Danchin et al. 2005). And, Colonial species aside, most of the observable events from a practical point of view, we can rarely perceive shall involve individuals of other species. We argue that information possessed by an animal until it alters its social information use is not necessarily restricted to a actions. It seems then that the most salient quantifica- conspecific setting and that heterospecifics are possible, tion of information for purposes of evolutionary ecology and sometimes the preferred, or the only, sources of is the difference in resulting fitness between the chosen information. In the third part of this paper we provide decision and the decision that would have been made the first brief review of the published cases of without the observation. This does not require the interspecific social information use. chosen decision to be an ‘‘adaptive response’’ (Jablonka Finally, we discuss the implications of our arguments. 2002) or lead to ‘‘a superior outcome’’ (Weinberger Ecological consequences of interspecific social informa- 2002). tion use have impact on community ecology and We suggest that in addition to ‘‘adaptive response’’ conservation. Adaptively extended social information decisions yielding increased fitness, information use can use gives rise to interesting evolutionary scenarios in also involve a decision that is ‘‘misinformed,’’ a phenology, social learning, and guild formation. Co- ‘‘gamble,’’ or ‘‘assured.’’ As long as consequences of a evolution between providers and users of information is decision have not been realized and there is uncertainty also conceivable, leading to arms races, or perhaps to a regarding the true state of the world, there also remains mutual ‘‘Information Market.’’ uncertainty (a probability distribution) of eventual

DEFINING VALUE AND THE PROCESS OF INFORMATION USE fitness values. Any change in the probability distribution of fitness values that results from the decision can be C The information concept of communication theory seen to constitute potential information value, even if ONCEPTS (Shannon 1948) measuring uncertainty has been widely the mean expected fitness is lowered (‘‘misinformation’’) applied across disciplines, but it was originally devel- or if just the variance of fitness increases (‘‘gamble’’) or oped to understand propagation of information in decreases (‘‘assurance’’). cables and gadgets. Being about communication chan-

Potential information value is thus a change in &S nels, it explicitly excludes the value of the observation to distribution, not necessarily definable as a single the receiver and information content or semantic quantity. For realized information value to be deter- meaning of an observation. For a behavioral or YNTHESIS mined, the eventual consequence needs to be realized. evolutionary ecologist seeking to understand how Realized information value equals the difference be- biological organisms use information to guide their tween the average fitness that would have been without behavior or how this affects evolution, a narrow, information use and fitness that resulted after informa- technical definition of information is rarely useful alone (Dall 2005). tion use. As such, it can be quantified by a single Even if ‘‘information’’ in isolation is difficult to define measure of fitness. The ethereal concept of information adequately, we do have concepts to specify the use of then becomes more tangible: we would say that the value information. Information concepts from economics aim of information in a particular case was 0.75 more to understand the use of information in a biological offspring or a 105 g loss of mass or a 17% increase in organism, and thus are perhaps more useful to survival probability. evolutionary and behavioral ecology than definitions Linked information: use sequences of two individuals from telecom technology. The value of information define social information use (Gould 1974) is an important concept obviously relevant to evolutionary ecology (Stephens 1989). And, the We wish to highlight the importance and usefulness of integrated picture of information use in statistical recognizing information use as a process. Individual’s decision theory (for an introduction see Pratt et al. information use consists of four scenes in sequence: (1) 1995), involving prior knowledge, observation, decision an observable event or state (internal or external), (2) the among alternative actions available, and the utility of observation of that event or state, (3) a decision consequences as well as the costs involved, is a useful manifested as altered action, and (4) the consequence conceptual framework for behavioral ecology (McNa- of that action. mara and Houston 1980, Dall et al. 2005). In social information use, separate sequences of different individuals become linked. A scene in the The value of information: positive, negative, gambled, information use sequence of individual A (Alice) is or assured observed by individual B (Bob), who in turn is observed Stephens (1989) highlighted that the value of infor- by individual C (Carol), and so on (Fig. 1). A natural mation is defined by the fitness gain from outcomes, not classification of three sources of social information is the content of observation per se or how much defined by the origin of this link. JANNE-TUOMAS SEPPA¨NEN ET AL. Ecology, Vol. 00, No. 0

First, the link between two individuals may originate environment. Finally, it might be difficult for researchers from an observation, but this requires that Alice to discriminate between success and decisions on broadcasts the observation. Regardless of whether Bob investment. is the intended recipient or inadvertent eavesdropper of All of these sources can provide discrete (e.g., the broadcast, this differs fundamentally from the other survival, breeding failure, prey unit capture) as well as types of social information use described later because a graded (e.g., duration, intensity, number of actions) ‘‘meaning’’ shared by Alice and Bob is involved, and information. Also, the information produced by the communication, with all its intricate co-evolutionary ‘‘informant’’ can be inadvertent just as well as inten- processes, may occur. A link originating from observa- tional. Thus the categories are delineated here based tion can perhaps be further divided (Marler et al. 1992) solely on the origin of the link, not the scale of the into a motivational ‘‘evolved signal’’ if Alice signals information or its supposed cognitive complexity or something internal (such as fear, dominance status, or intentionality (cf. Danchin et al. 2004, Dall et al. 2005). toxicity) and referential ‘‘semantic communication’’ The important difference between decisions and conse- (e.g., Zuberbuhler 2000) if Alice’s observation is of quences as a source of social information is that something in the external environment (such as food decisions of others may be misinformed, and thus can resource or threat or type of predator). be intrinsically inaccurate or unreliable sources (Tem- Second, and possibly the most common link in social pleton and Franklin 1992, Koops 2004, Dugatkin 2005), information use of non-human , is the kind potentially resulting in maladaptive informational cas- where the decision is observed (the link from Alice to cades (Giraldeau et al. 2002). However, the actions Bob in Fig. 1; partially corresponding to ‘‘social cue’’ in revealing decisions of others are likely to occur Dall et al. 2005). Most of the publicly observable events spatiotemporally closer to the cause (resource) of the by other individuals reflect their decisions. For example, success, are likely to be commonly publicly observable, a decision to stay vs. leave, a decision to call vs. be silent, and thus may sometimes be preferable, or the only a decision to breed vs. wait, a decision to mate vs. not: available, source of information. all are overtly manifested as presence, vocalization, nest or clutch or litter, and copulation, respectively. It should VALUE OF INFORMATION IS A FUNCTION OF DISTANCE be noted that generally the decisions and actions of Alice Because of the inherently sequential nature of YNTHESIS and Bob need not be the same, and thus animal information use, there is a time lag between the event traditions and imitation are special cases where the observed by Bob and the consequence of any decision, &S actions themselves are transferred (e.g., Whiten et al. regardless of the type of event observed. Furthermore, 1999, Krutzen et al. 2005). However, semantically blind, the scenes in Bob’s information use sequence are very nonimitative social information use may well be a likelytobesomedistanceawayfromeachother precursor to these more intricate forms of social spatially and ecologically as well. This is because Alice information (Lotem et al. 1999). is likely to occupy the spatial location of the event Bob ONCEPTS

C The third type of social information use occurs when observes (remember, Alice’s observation, decision, or the consequence is observed (link from Bob to Carol in consequence is the event observed by Bob), and Alice is Fig. 1; partially corresponding to ‘‘public information’’ likely to be different from Bob. In short, the conse- or ‘‘vicarious sampling’’ in Templeton and Giraldeau quence of social information use will often be tempo- 1996). While a decision by someone else might well be rally, spatially, and ecologically some distance away misinformed or reflect resources inaccurately, a conse- from the source of information. quence (success) is reliable evidence about resources. Although many theoretical (e.g., Doligez et al. 2003) as Distance erodes the value of information well as empirical (Doligez et al. 2002, Danchin et al. Distance, in temporal, spatial, and ecological dimen- 2004) studies have concentrated on conspecific public sions, between the observed event and the eventual information, this type of social information use may consequence of a decision, affects the value of informa- nevertheless be relatively infrequent outside the con- tion. To illustrate, an alarm call about a predator cept’s original setting in social foraging (Templeton and happens at a certain time, at a certain location, and is Giraldeau 1996). This is because success of others (e.g., about only certain, specific things in the world. It has survival or growth) will be difficult to use as information little value for decision making a few years later, a if it becomes publicly observable some time after or kilometer away, or for an individual with a completely some distance away from its cause, especially in more different set of predators. spatially extended associations or if individuals associate With long enough distances, the relationship between temporarily. In situations involving synchronous phe- the state of the world at the point of the observed event nology (e.g., seasonal breeding), success of others can be and at the point of the realized consequence is simply used to gain the same kind of success (e.g., offspring) in random, and therefore the average value of information the next opportunity at the earliest, not the current. is zero. Fortunately, this erosion usually happens Both difficulties are further exacerbated if the species in gradually. At smaller distances the real world tends to question is time limited or if it faces an unpredictable be positively autocorrelated (e.g., Nekola and White Month 2007 SOCIAL INFORMATION FROM HETEROSPECIFICS

1999, Roy et al. 2005). Hence, the value of information is a decreasing function of the distance between the event observed and eventual consequence of the decision. Information can have a nonzero average value if the consequences occur sufficiently close in the temporal, spatial, and ecological dimensions to the original observed event, i.e., if the world is still positively autocorrelated at that degree of separation (Boulinier and Danchin 1997, Doligez et al. 2003). The erosion in time has been shown to be taken into account by some animals: nine-spined sticklebacks progressively devalue personally gathered information with the time lag since the observation and eventually switch to using current social information (van Bergen et al. 2004).

Trade-off between the value of information and cost of competition So, information is more valuable the shorter the distance between the source and the consequence in information use sequence. But social information use may incur costs at short distances. Alice is unlikely to be

a mere obliging informant; she may exploit the same C resources, attract shared predators, interfere with, or ONCEPTS even kill Bob. This reduction in fitness tends to be stronger the closer Bob gets to Alice in time, space, and ecology. Conversely, increasing distances reduce the costs of competition, as depleted food will eventually &S grow back, an aggressive competitor will not attack beyond some range, and predators of a small species will not harm a large species. YNTHESIS The magnitude of the two opposing effects and shapes of their decrease along the respective dimensions then define the net value of information. It is a decreasing or unimodal function of temporal, spatial, and ecological distance between the first and last scene of social information use. The process of information use may have fixed costs also (e.g., lost opportunity, locomotion), and the window within which using social information is adaptive is the range of distances where the net information value exceeds those baseline costs (Fig. 2A). An interesting situation arises if the autocorrelation of the cost of competition decreases sufficiently more FIG. 2. Visualizing the value of social information as a rapidly than the autocorrelation of the resource (Fig. function of distances. (A) Potential information value (thin 2A). Net value of information may then be maximized at solid line) decreases with distance (here Gaussian autocorrela- some optimal distance between the original event tion). Cost of competition (dotted line) also decreases (here observed and the eventual consequence of the decision. exponential autocorrelation), and eventually both have zero effect on expected fitness. Subtracting the cost of competition Using social information at shortest distances may even from potential information value yields net information value become unfavorable. The value of information can be (bold line). Information use is adaptive within the window (shaded interval) where information value exceeds the baseline ‹ cost (dashed line) of information use (e.g., neural processing the window of profitable information use is the region where and locomotion). Here, net information value is a unimodal the surface exceeds the upper plane. (C) Including all three function of distance, and peaks at some distance away from the dimensions along which competition and information value observed event. (B) With two dimensions (here time and space), may vary, the window of profitable information use is the the height of the surface represents the net information value. volume enclosed by the outermost isosurface. Outside of this The lower plane is the expected fitness without information use, volume the baseline costs exceed the net information value. and the upper plane is the baseline cost of using information. Here, information value peaks at some distance along each of Information value peaks here at some distance along both axes; the three dimensions (core of the volume). JANNE-TUOMAS SEPPA¨NEN ET AL. Ecology, Vol. 00, No. 0

such a unimodal function of distance simultaneously in to demonstrate the phenomenon when in fact it was in two (Fig. 2B) or all three (Fig. 2C) of the distance progress. dimensions. In that case the range of, e.g., spatial Interestingly, the most similar individuals, those of the distances over which information use is beneficial, is same size, age, sex, or species, might often be suboptimal different depending on the given temporal distance. sources of information. Furthermore, ecological simi- Thus, for example, if Bob’s decision is limited to some larity might preclude information use at the optimal minimum distance from Alice (e.g., due to Alice’s spatial and temporal distance. Spatially, territoriality is territoriality) and to some minimum temporal difference likely to be stronger with similarity: individuals less alike from Alice’s decisions (e.g., due to Bob’s timing of are often tolerated at closer proximity than equal rivals. breeding) and Bob is ecologically rather different from In the temporal domain, close ecological similarity can Alice, Bob might be just outside the window in which impose a long time lag between the observation of using Alice as information would be beneficial (i.e., success and consequence of a decision based on such behind the outermost isosurface in Fig. 2C). observation. This happens if everyone’s success is determined within a short, periodic frame of time, while Adaptively extended social information use decisions have to be made before this time frame, as in Evolutionary adaptation or developmental or behav- seasonal breeding (Boulinier et al. 1996, Nocera et al. ioral plasticity shifting our Bob-organism’s position on 2006). A decision based on observing the success of any of the dimensions could bring it within the window others in the current time frame can then affect the of profitable information use. For example, increased observer’s success in the next time frame at the earliest. fighting ability (allowing shorter spatial distance), earlier This reasoning suggests that using individuals of onset of breeding (shortening temporal distance), or a other, slightly differing species as sources of information change in diet (character convergence, more ecological should be common and sometimes preferred over overlap) could allow social information use. Further conspecific sources. In addition, if some other resource changes bringing Bob closer to Alice should be selected is strongly limiting and heterospecifics do not compete as severely for it, then using heterospecifics instead of for, until the more familiar forces of competition, conspecifics might pay, even if the potential information resulting in character displacement, halt the change. value from conspecifics would be higher. An obvious YNTHESIS Naturally, the same optimum could also be reached case is seeking information about a food resource while from other starting points. In any case, the result is avoiding competition for mates: heterospecific individ-

&S social information use that is adaptively extended in one uals with some ecological overlap provide some infor- or more dimensions to solve the trade-off between the mation but do not compete for mates. value of information and cost of competition. A number of correlative and experimental studies We know just one experiment demonstrating this have recently demonstrated that other species are indeed ‘‘adaptively extended social information use.’’ White and used as sources of social information. We provide the ONCEPTS Galef (2000) demonstrated that male Japanese Quail

C first review of published cases here. Coturnix japonica show mate-choice copying, but not before 48 hours after the observed mating, when the HETEROSPECIFICS AS A SOURCE OF INFORMATION demonstrator male’s advantage in sperm competition Individuals of other species may themselves generate a has eroded. Similarly, Witte and Massman (2003) valuable resource, as when a more aggressive, more suggested that the ability of female sailfin mollies vigilant, or more vulnerable species provides protection Poecilia latipinna to retain their socially learned male for others (e.g., Slagsvold 1980, FitzGibbon 1990, preferences for at least 24 hours allows them to use that Bshary and Noe¨1997, Bogliani et al. 1999). But information later, thus avoiding sperm depletion. obtaining information about external resources may We would expect to see adaptively extended social also be an important factor in the formation of mixed- information use in other contexts as well. For example, a species associations. In most natural communities, predator might be inclined to hunt at a location a day heterospecifics comprise the majority of observable after observing a successful kill, instead of immediately individuals, and together they are likely to provide a (prey is alert or hidden) or a month later (reliability of richer source of potential information than conspecifics finding prey is low). Or, a migrant arriving to alone. We review here an illustrative set of studies breeding grounds would not try to build its nest adjacent describing interspecific social information use with to a previously established bird (exploitative and different types of behaviors used as a source of interference competition), nor in another forest patch information. (chance level of resources), but instead at a short distance away within the same patch. Temporally and Location decisions as a source spatially extended social information use can conceiv- Informants do not need to come from a closely related ably be difficult to recognize observationally. Further- species or even from the same class, as long as they are more, experiments investigating social information use similar enough in the relevant ecological parameter. Flat without sufficient temporal and spatial scale might fail Platysaurus broadleyi and several bird species in Month 2007 SOCIAL INFORMATION FROM HETEROSPECIFICS

South Africa feed on ripe cordata fruits. The Buchanan-Smith 2000) are among the most stable lizards use the presence of as information about associations in animals, and information use within the fruits and were experimentally attracted to birds these groups probably differs little from intraspecific even when cages containing the birds were placed away situations. from the trees (Whiting and Greeff 1999). While using Anemonefishes (Amphiprion spp.) depend on a patchy each other’s signals as information is not rare between resource, the anemones, for reproduction. Hattori birds and mammals (see ‘‘Heterospecifics as a source of (2000) observed that the juveniles of a smaller, less information: Vocalization decisions as a source’’), most mobile A. perideraion always co-occurred with larger, published cases of interspecific social information use more mobile A. clarkii, although the latter suppresses involve less distant species pairs. reproduction of A. perideraion (Hattori 2002). This Many studies on interspecific social information use suggests a potential case of interspecific information use. are done with birds. Clergeau (1990) conducted exper- Possibly, the better ability of adult A. clarkii to migrate iments with stuffed European Starling Sturnus vulgaris and decide between anemones makes their presence a models. He suggested that the attraction of other, valuable source of information for A. perideraion larvae. smaller species to the models can be partly explained Eventually A. clarkii migrate to a vacant (never by use of information about resources. Information occupied) anemone whereupon resident A. perideraion about resource locations is particularly crucial in mature and defend their anemone against subsequent environments where an individual has to cover large immigrations (Hattori 2002). areas to find a resource patch. For example, it has been Social information use occurs in invertebrates, too, suggested that other pelagic seabirds have used Black- but interspecific cases have been suggested rarely since browed Albatrosses (Thalassarche melanophrys)as Darwin (Chittka and Leadbeater 2005). Two examples indicators of food location (Silverman et al. 2004). come from foraging stingless bee (Apidae) species: three

When choosing a breeding location, many bird species species were more attracted to certain heterospecific C are in quite a hurry. Individuals of earlier established species than to their own nestmates (Slaa et al. 2004) and ONCEPTS species can be used as informants by later establishing Trigona snipipens was more attracted to odor marks of species to save time. Elmberg et al. (1997) showed that Melipona rufiventris than to those of nestmates when Common European Teals (Anas crecca) tended to be foraging in new locations (Nieh et al. 2004). Spiders attracted to lakes occupied by earlier arriving Mallards Hypochilus thorelli are attracted to webs of a hetero- (Anas platyrhynchos). Our studies have demonstrated specific spider Achaearanea tepidariorum (Hodge and &S that presence and density of resident tits (Parus spp.) are Storfer-Isser 1997), although other explanations than used by migrant birds as information about breeding information about resources are possible: A. tepid- YNTHESIS patch and nest-site quality (Mo¨nkko¨nen et al. 1990, ariorum might be important prey for H. thorelli (Hodge 1997, Timonen et al. 1994, Forsman et al. 1998, and Uetz 1996), or webs might provide structural Thomson et al. 2003). At least the Pied Flycatcher support (but see Schuck-Paim and Alonso 2001). (Ficedula hypoleuca) derives direct fitness benefits from using this information (Forsman et al. 2002), and Vocalization decisions as a source possibly can also get information benefiting foraging Responding to predator-elicited signals of other (Seppa¨nen et al. 2005). Further, Pied Flycatchers seem species is quite common, at least in vertebrates (e.g., to heed not only the risk, but also the Zuberbu¨hler 2000, Johnson et al. 2003, Pollock at al. protection gained from Sparrowhawks (Accipter nisus) 2003, Sullivan et al. 2003), and even between birds and in nest-site choice and breeding investment decisions mammals (Rainey et al. 2004, Randler 2006) or plant (Thomson et al. 2006). species (Kessler et al. 2006). As these signals are often Among mammals, mixed-species associations are similar between species and can elicit communal defense, common in New World primates, and Norconk (1990) utilizing them can be seen as communication. This is a suggested that the more detailed information possessed special case of social information use, where the link is by a species with a smaller home range could be formed directly from observation to observation, as parasitized by a species having less accurate informa- demonstrated in the mammals (e.g., Zuberbuhler 2000) tion, essentially the same situation thought to occur and birds (e.g., Templeton et al. 2005) that convey to between resident and migrant birds (Mo¨nkko¨nen et al. others the type or threat of the predator they observe. 1999). Intriguingly, the idea of a window of ecological Intriguingly, a recent paper by Parejo et al. (2007) difference for mixed-species groups was suggested by suggests that avian brood parasites may eavesdrop on Heymann (1997): tamarins (Sanguinus spp.) readily form the signals of other species revealing parenting ability. mixed-species groups, but not with species with a small Our interest here is whether other species’ decision to size difference. He suggested that there should be an vocalize can be used as information without a shared upper limit to size difference as well, beyond which the ‘‘meaning’’ of the vocalization. species are too different ecologically to benefit from Predation risk may be revealed by signals other than association (Heymann 1997). The permanent mixed- alarms as well. Møller (1992) demonstrated that hearing species troops of callitrichine monkeys (Heymann and the song of Black Wheatear (Oenanthe leucura) increases JANNE-TUOMAS SEPPA¨NEN ET AL. Ecology, Vol. 00, No. 0

singing activity of other species. Similarly, Parejo et al. (2005) provided correlative evidence that Phelps et al. (2007) show that after a predator exposure, European Rollers (Coracias garrulous) and Kestrels tu´ngara frogs Physalaemus pustulosus resume calling (Falco tinnunculus) may observe the number of fledglings sooner and call more if they hear calls of conspecifics or of the other species and use it as social information sympatric white-lipped frogs Leptodactylus labialis, but about the quality of breeding habitat in future breeding not if they hear allopatric congener P. enesefae, despite habitat choice. In fishes, Coolen et al. (2003) showed acoustic similarity between the congeners. In mixed- experimentally that nine-spined sticklebacks (Pungitius species flocks of eastern North America, Downy pungitius) can use the foraging of three-spined stickle- Woodpeckers (Picoides pubscens) resume normal activ- backs (Gasterosteus aculeatus) as social information in ity after simulated predator presence more quickly if choosing a foraging location. Experimenters made the they hear contact calls of Black-capped Chickadees foraging more salient by lowering the food in a column (Poecile atricapillus) and Tufted Titmice (Baeolophus transparent to the three-spined stickleback demonstra- bicolor; Sullivan 1984). Information regarding the tors but not to the nine-spined stickleback observers, absence of predators was considered as the most likely invoking demonstrator pecking at the column until explanation in all of these cases. However, in Forsman receiving the food through a slot at the bottom. It seems and Mo¨nkko¨nen (2001), although songs of Willow Tit that the observer might have observed the repetitive (Parus montanus) and Redwing (Turdus iliacus) did attempts at foraging (i.e., foraging investment decision) attract more heterospecifics than control playbacks, as well as the success in foraging. proximity to nesting Sparrowhawk, and thus presum- IMPLICATIONS OF INTERSPECIFIC SOCIAL INFORMATION USE ably higher predation risk, did not increase heterospe- cific approaches to playback songs (see also Mo¨nkko¨nen Spatial distribution and conservation et al. 1996). The use of interspecific social information in habitat Similarly, to direct visual observation of presence, selection potentially has many important ecological and courtship calls given at the breeding site can be used as conservation implications. The effects of conspecific information on locations of suitable sites by other distribution on dispersal and immigration (Stamps 1988) species sharing some aspects of the breeding require- have been studied intensely. For example, conspecific ments. In the experiments of Diego-Rasilla and Luengo YNTHESIS attraction has recently been suggested to influence the (2004), marbled newts (Triturus marmoratus) headed effects of patch size and edge in fragmented landscapes, toward the courtship calls of sympatric natterjack toads in addition to dispersal success (Fletcher 2006). In &S (Bufo calamita) but not toward those of allopatric contrast, we know fairly little of the potential effects of European green toads (Bufo viridis). Lacking the stature interspecific social information use on the spatial or speed to assess its environment from afar, orientation distribution of animals. Conceivably, using the presence toward calls of heterospecifics with similar breeding or density of heterospecifics as information about requirements plausibly helps the newts in finding new

ONCEPTS habitat quality when making settlement decisions (e.g.,

C breeding ponds. Mo¨nkko¨nen et al. 1990, 1997, Elmberg et al. 1997, In Sri Lankan rainforests, playbacks of mainly Hodge and Storfer-Isser 1997, Forsman et al. 1998, insectivorous Orange-billed Babbler (Turdoides rufes- Whiting and Greeff 1999) should result in aggregated cens) and Greater Racket-tailed Drongo (Dicrurus distribution of the species involved across the landscape. paradiseus) calls attracted insectivorous birds more than Local species diversity at a given site would be expected calls of ‘‘non-nuclear’’ and mainly frugivorous Yellow- to be either higher or lower than predicted by fronted Barbet (Megalaima flavifrons), while omnivores independent settlement, depending on the presence of and frugivores responded equally (Goodale and Gota- the species used as information. On the other hand, such gama 2005). Saracco et al. (2004) showed that a strategy may leave part of suitable habitat patches specialist frugivore bird, the Puerto Rican unoccupied by the species utilizing the cues (Mo¨nkko¨- (Spindalis portoricensis), is often associated with a nen et al. 1999, Stamps and Krishnan 2005). Further, gregarious generalist bird, the Puerto Rican Tanager immigration to already established communities may be (Nesospingus speculiferus), and suggested that the loud facilitated or prohibited by the presence of heterospe- flock calls of Tanagers attract cifics, and an Allee-effect type of response to the density and other species to fruit resources. Interestingly, they of heterospecifics is possible (Forsman et al. 2002). also give anecdotal evidence that the specialist bird may Consequently, using interspecific social information be led to novel food sources by observing the foraging may have conservation implications thus far unconsid- behavior of the generalist. ered. Changes in population density or distribution of the informant species may affect the survival of Success or investment decisions as a source populations of species using them as sources of Using the breeding success, or possibly, breeding information. Environmental change may also disrupt investment decisions of conspecifics as information in the correlations between the observed event and breeding decisions has been documented for many bird resources, rendering information use maladaptive. In species (e.g., Doligez et al. 2002, 2004, Po¨ysa¨2006). short, colonization and extinction in fragmented land- Month 2007 SOCIAL INFORMATION FROM HETEROSPECIFICS scapes is no longer a function of landscape patterns and pay to exaggerate or advertise the event. For example, population parameters alone if interspecific social Dornhaus et al. (2006) show that cooperatively foraging information in dispersal and immigration decisions is social may return to the ‘‘information center’’ important. Conservation of a species involves consider- prematurely with partial loads to advance transmission ing other species in the community also, perhaps a of information. As Lotem et al. (1999) and Danchin et recommendable approach in any case. al. (2004) note, selection pressures to change inadver- tently produced information into a signal are possible, Evolutionary effects of information use perhaps even likely. This would perhaps give rise to a Because information value is a function of time lag special case of biological markets (Noe¨and Hammer- and spatial distance (Fig. 2), adaptations improving the stein 1995), the Information Market. timing and placement of decisions and consequences are CONCLUDING REMARKS expected. Such adaptations do not necessarily require altered cognitive functions, e.g., if seasonal or diurnal The examples of interspecific social information use phenology is adjusted. Naturally, cognitive adaptations presented here mostly involve vertebrates choosing a would allow more flexible adjustment of information location for foraging or breeding, and the source of use. If individuals of a given ecological difference are social information is most often the informant’s decision optimal informers, natural selection should favor to settle at a location (revealed as presence and density). discrimination abilities and, in the case of heterospe- This does not necessarily mirror the range or frequency cifics, generalizing existing intraspecific social learning to of different context of interspecific social information include individuals of the other species. use. Rather, it seems plausible that information use in Optimal distance of social information use (Fig. 2) other contexts is more difficult to observe, especially would imply a counterforce to character displacement without a dedicated experiment, and thus ends up in

(Brown and Wilson 1956). Becoming more different and print only rarely. Enticing exceptions do exists in C avoiding the other individual spatially and temporally addition to the aforementioned anecdote from Saracco ONCEPTS helps to escape competition but at the same time reduces et al. (2004). Island-dwelling bird populations tend to the value of social information. Even if resource show excessive specialization at the individual level distribution and other competitors would allow further (Scott et al. 2003). Werner and Sherry (1987) suggest divergence and avoidance, natural selection may arrest that this may be achieved via interspecific copying of these at the optimal distance, if such exists. Correspond- foraging techniques. Lefebvre et al. (1997) demonstrated &S ingly, a species at the other side of the optimum would that Carib Grackles (Quiscalus lugubris) can copy converge toward its informant species. Ecologically foraging techniques of Zenaida Doves (Zenaida aurita). YNTHESIS extended social information use is predicted to generate We suspect that interspecific social information use and guilds of species, not necessarily congeners, whose its implications might be more diverse and common than ecological similarity and spatial and temporal co- previously thought or outlined here. We hope that occurrence is higher than predicted by resource avail- readers of this concept paper are encouraged to consider ability alone. the possibility of interspecific social information use in their study systems so that further discoveries and Arms races and information markets deeper understanding from greater diversity of contexts Further complexity arises if the informant suffers would surface. from the information use of another individual. The ACKNOWLEDGMENTS decisions of the user of information (e.g., to settle We wish to thank Sasha Dall for insightful comments on this nearby) may well lower the fitness of the informant paper. J. T. Seppanen was funded by Jenny and Antti Wihuri (Forsman et al. 2007). An evolutionary arms race Fund, and by Finnish Cultural Foundation. J. T. Forsman was (Dawkins and Krebs 1979) between the informant and funded by European Commission (Marie Curie Intra-European the information parasite could then result. The options Fellowship, project MEIF-CT-2003-500554). for the parasitized individual are to either cease LITERATURE CITED providing information or to attempt, e.g., by aggression, Bednekoff, P. A. 2005. Defining the concept of public closing the window of profitable information use, or to information. Science 308:354. attempt hiding the event observed by the parasite. For Bogliani, G., F. Sergio, and G. Tavecchia. 1999. Woodpigeons example, tits (Parus spp.) cover eggs with hair before, nesting in association with hobby falcons: advantages and but not during incubation, perhaps in an attempt to hide choice rules. Animal Behaviour 57:125–131. Boulinier, T., E. Danchin, J.-Y. Monnat, C. Doutrelant, and B. the presence and number of eggs from other individuals Cadiou. 1996. Timing of prospecting and the value of during the period of their habitat choice. Similarly, Piper information in a colonial breeding bird. Journal of Avian et al. (2006) suggest that Common Loon (Gavia immer) Biology 27:252–256. families ‘‘dive and scatter,’’ and chicks hide themselves Boyd,R.,andP.J.Richerson.1985.Cultureandthe evolutionary process. University of Chicago Press, Chicago, from flying adults in an attempt to hinder their Illinois, USA. prospecting. If, on the other hand, the information use Brown, W. L., Jr., and E. O. Wilson. 1956. Character of another individual benefits the informant, it might displacement. Systematic Zoology 5:49–64. JANNE-TUOMAS SEPPA¨NEN ET AL. Ecology, Vol. 00, No. 0

Bshary, R., and R. Noe¨. 1997. Red colobus and Diana monkeys use between resident and migrant birds. Behavioral Ecology, provide mutual protection against predators. Animal Behav- in press. iour 54:1461–1474. Galef, B. G., Jr., and L. A. Giraldeau. 2001. Social influences Chittka, L., and E. Leadbeater. 2005. Social learning: public on foraging in vertebrates: causal mechanisms and adaptive information in insects. Current Biology 15:R869–R871. functions. Animal Behaviour 61:3–15. Clergeau, P. 1990. Mixed flocks feeding with starlings: an Giraldeau, L.-A., T. J. Valone, and J. J. Templeton. 2002. experimental field study in Western Europe. Bird Behaviour Potential disadvantages of using socially acquired informa- 8:95–100. tion. Philosophical Transactions of the Royal Society: Coolen, I., Y. van Bergen, R. L. Day, and K. N. Laland. 2003. Biological Sciences 357:1559–1566. Species difference in adaptive use of public information in Goodale, E., and S. W. Kotagama. 2005. Testing the roles of sticklebacks. Proceedings of the Royal Society of London B species in mixed-species bird flocks of a Sri Lankan rain Sciences 270:2413–2419. forest. Journal of Tropical Ecology 21:669–676. Dall, S. R. X. 2005. Defining the concept of public information. Gould, J. P. 1974. Risk, stochastic preferences, and the value of Science 308:353. information. Journal of Economic Theory 8:64–84. Dall, S. R. X., L. A. Giraldeau, O. Olsson, J. M. McNamara, Hattori, A. 2000. Social and mating systems of the protandrous and D. W. Stephens. 2005. Information and its use by anemonefish Amphiprion perideraion under the influence of animals in evolutionary ecology. Trends in Ecology and a larger congener. Austral Ecology 25:187–192. Evolution 20:187–193. Hattori, A. 2002. Small and large anemonefishes can coexist Danchin, E., L. A. Giraldeau, T. J. Valone, and R. H. Wagner. using the same patchy resources on a coral reef, before 2004. Public information: from nosy neighbors to cultural habitat destruction. Journal of Animal Ecology 71:824–831. evolution. Science 305:487–491. Heymann, E. W. 1997. The relationship between body size and Danchin, E., L. A. Giraldeau, T. J. Valone, and R. H. Wagner. mixed-species troops of Tamarins (Saguinus spp.). Folia 2005. Response to ‘Defining the concept of public informa- Primatologica 68:287–295. tion.’ Science 308:355. Heymann, E. W., and H. M. Buchanan-Smith. 2000. The Dawkins, R., and J. R. Krebs. 1979. Arms races between and behavioural ecology of mixed-species troops of callitrichine within species. Proceedings of the Royal Society of London B primates. Biological Reviews 75:169–190. 205:489–511. Hodge, M. A., and A. Storfer-Isser. 1997. Conspecific and Diego-Rasilla, F. J., and R. M. Luengo. 2004. Heterospecific heterospecific attraction: a mechanism of web site selection call recognition and phonotaxis in the orientation behavior of leading to aggregation formation by web-building spiders. the marbled newt, Triturus marmoratus. Behavioural Ecology Ethology 103:815–826. and Sociobiology 55:556–560. Hodge, M. A., and G. W. Uetz. 1996. Foraging advantages of Doligez, B., C. Cadet, E. Danchin, and T. Boulinier. 2003. mixed-species association between solitary and colonial orb- YNTHESIS When to use public information for breeding habitat weaving spiders. Oecologia 107:578–587. selection? The role of environmental predictability and Jablonka, E. 2002. Information: its interpretation, its inheri- density dependence. Animal Behaviour 66:973–988.

&S tance, and its sharing. Philosophy of Science 69:578–605. Doligez, B., E. Danchin, and J. Clobert. 2002. Public Johnson, F. R., E. J. McNaughton, C. D. Shelley, and D. T. information and breeding habitat selection in a wild bird Blumstein. 2003. Mechanisms of heterospecific recognition in population. Science 297:1168–1170. avian mobbing calls. Australian Journal of Zoology 51:577– Doligez, B., T. Part, and E. Danchin. 2004. Prospecting in the 585. collared flycatcher: gathering public information for future breeding habitat selection? Animal Behaviour 67:457–466. Kessler, A., R. Halitschke, C. Diezel, and I. T. Baldwin. 2006. ONCEPTS Dornhaus, A., E. J. Collins, F.-X. Dechaume-Moncharmont, Priming of plant defense responses in nature by airborne C A. I. Houston, N. R. Franks, and J. M. McNamara. 2006. signaling between Artemisia tridentata and Nicotiana attenu- Paying for information: partial loads in central place ata. Oecologia 148:280–292. foragers. Behavioural Ecology and Sociobiology 61:151–161. Koops, M. A. 2004. Reliability and the value of information. Dugatkin, L. A. 2005. Mistakes and the evolution of copying. Animal Behaviour 67:103–111. Ethology, Ecology and Evolution 17:327–333. Krutzen, M., J. Mann, M. R. Heithaus, R. C. Connor, L. Elmberg, J., H. Po¨ysa¨, K. Sjo¨berg, and P. Nummi. 1997. Bejder, and W. B. Sherwin. 2005. Cultural transmission of Interspecific interactions and co-existence in dabbling ducks: tool use in bottlenose dolphins. Proceedings of the National observations and an experiment. Oecologia 111:129–136. Academy of Sciences (USA) 102:8939–8943. FitzGibbon, C. D. 1990. Mixed-species grouping in Thomson’s Laland, K. N. 2004. Social learning strategies. Learning and and Grant’s gazelles: the antipredator benefits. Animal Behavior 32:4–14. Behaviour 39:1116–1126. Lefebvre, L., J. Templeton, K. Brown, and M. Koelle. 1997. Fletcher, R. J., Jr. 2006. Emergent properties of conspecific Carib grackles imitate conspecific and Zenaida dove tutors. attraction in fragmented landscapes. American Naturalist Behaviour 134:1003–1017. 168:207–219. Lotem, A., R. H. Wagner, and S. Balshine-Earn. 1999. The Forsman, J. T., and M. Mo¨nkko¨nen. 2001. Responses by overlooked signaling component of nonsignaling behaviour. breeding birds to heterospecific song and mobbing calls Behavioural Ecology 10:209–212. under varying predation risk. Animal Behaviour 62:1067– Lotem, A., and D. W. Winkler. 2005. Defining the concept of 1073. public information. Science 308:354. Forsman, J. T., M. Mo¨nkko¨nen, P. Helle, and J. Inkero¨inen. Marler, P., C. Evans, and M. D. Hauser. 1992. Animal signals: 1998. Heterospecific attraction and food resources in reference, motivation, or both? Pages 66–86 in H. Papoucek, migrants’ breeding patch selection in northern boreal forest. U. Ju¨rgens, and M. Papoucek, editors. Nonverbal vocal Oecologia 115:278–286. communication: comparative and developmental approach- Forsman, J. T., J.-T. Seppa¨nen, and M. Mo¨nkko¨nen. 2002. es. Cambridge, University Press, Cambridge, UK. Positive fitness consequences of interspecific interaction with McNamara, J. M., and A. I. Houston. 1980. The application of a potential competitor. Proceedings of the Royal Society of statistical decision theory to animal behaviour. Journal of London B 269:1619–1623. Theoretical Biology 85:673–690. Forsman, J. T., R. L. Thomson, and J.-T. Seppa¨nen. 2007. Møller, A. P. 1992. Interspecific response to playback of bird Mechanisms and fitness effects of interspecific information song. Ethology 90:315–320. Month 2007 SOCIAL INFORMATION FROM HETEROSPECIFICS

Mo¨nkko¨nen, M., J. T. Forsman, and P. Helle. 1996. Mixed- Schuck-Paim, C., and W. J. Alonso. 2001. Deciding where to species foraging aggregations and heterospecific attraction in settle: conspecific attraction and web site selection in the orb- boreal bird communities. Oikos 77:127–136. web spider Nephilengys cruentata. Animal Behaviour 62: Mo¨nkko¨nen, M., R. Ha¨rdling, J. T. Forsman, and J. Tuomi. 1007–1012. 1999. Evolution of heterospecific attraction: using other Scott, S. N., S. M. Clegg, S. P. Blomberg, J. Kikkawa, and species as cues in habitat selection. Evolutionary Ecology 13: I. P. F. Owens. 2003. Morphological shifts in island-dwelling 91–104. birds: The roles of generalist foraging and niche expansion. Mo¨nkko¨nen, M., P. Helle, G. J. Niemi, and K. Montgomery. Evolution 57:2147–2156. 1997. Heterospecific attraction affects community structure Seppa¨nen, J.-T., M. Mo¨nkko¨nen, and J. T. Forsman. 2005. and migrant abundances in northern breeding bird commu- Presence of other species may counter seasonal decline in nities. Canadian Journal of Zoology 75:2077–2083. breeding success: a field experiment with pied flycatchers Mo¨nkko¨nen, M., P. Helle, and K. Soppela. 1990. Numerical Ficedula hypoleuca. Journal of Avian Biology 36:380–385. and behavioural responses of migrant to experi- Shannon, C. E. 1948. A mathematical theory of communica- mental manipulation of resident tits (Parus spp): hetero- tion. Bell System Technical Journal 27:379–423, 623–656. specific attraction in northern breeding bird communities? Silverman, E. D., R. R. Veit, and G. A. Nevitt. 2004. Nearest Oecologia 85:218–225. neighbors as foraging cues: information transfer in a patchy Nekola, J. C., and P. S. White. 1999. The distance decay of environment. Marine Ecology Progress Series 277:25–35. similarity in biogeography and ecology. Journal of Biogeog- Slaa, E. J., J. Wassenberg, and J. C. Biesmeijer. 2004. The use of raphy 26:867–878. field-based social information in eusocial foragers: local Nieh, J. C., L. S. Barreto, F. A. L. Contrera, and V. L. enhancement among nestmates and heterospecifics in sting- Emperatriz-Fonseca. 2004. Olfactory eavesdropping by a less bees. Ecological Entomology 28:369–379. competitively foraging stingless bee, Trigona spinipes. Pro- Slagsvold, T. 1980. Habitat selection in birds: on the presence of ceedings of the Royal Society of London B 271:1633–1640. other birds species with special regard to Turdus pilaris. Nocera, J. J., G. J. Forbes, and L. A. Giraldeau. 2006. Journal of Animal Ecology 49:523–536. Inadvertent social information in breeding site selection of Sullivan, A. M., D. M. Madison, and J. R. Rohr. 2003. natal dispersing birds. Proceedings of the Royal Society of Behavioural responses by red-backed salamanders to con- London B 273:349–355. specific and heterospecific cues. Behaviour 140:553–564. C Noe¨, R., and P. Hammerstein. 1995. Biological markets. Trends Sullivan, K. A. 1984. Information exploitation by downy ONCEPTS in Ecology and Evolution 10:336–339. woodpeckers in mixed-species flocks. Behaviour 91:294–311. Norconk, M. A. 1990. Introductory remarks: ecological and Stamps, J. A. 1988. Conspecific attraction and aggregation in behavioral correlates of polyspecific primate troops. Amer- territorial species. American Naturalist 131:329–347. ican Journal of Primatology 21:81–85. Stamps, J., and V. V. Krishnan. 2005. Nonintuitive cue use in Parejo, D., and J. Avile´s. 2007. Do avian brood parasites habitat selection. Ecology 86:2860–2867. eavesdrop on heterospecific sexual signals revealing host Stephens, D. W. 1989. Variance and the value of information. &S quality? A review of the evidence. Animal Cognition, in press. American Naturalist 134:128–140. Parejo, D., E. Danchin, and J. Avile´s. 2005. The heterospecific Templeton, C. N., E. Greene, and K. Davis. 2005. Allometry of

habitat copying hypothesis: can competitors indicate habitat YNTHESIS alarm calls: black-capped chickadees encode information quality? Behavioral Ecology 16:96–105. about predator size. Science 308:1934–1937. Phelps, S. M., A. S. Rand, and M. J. Ryan. 2007. The mixed- species chorus as public information: tu´ngara frogs eaves- Templeton, J. J., and L. A. Giraldeau. 1996. Vicarious drop on a heterospecific. Behavioral Ecology 18:108–114. sampling: the use of personal and public information by Piper, W. H., C. Walcott, J. N. Mager, M. Perala, K. B. starlings foraging in a simple patchy environment. Behavioral Tischler, E. Harrington, A. J. Turcotte, M. Schwabenlander, Ecology and Sociobiology 38:105–114. and N. Banfield. 2006. Prospecting in a solitary breeder: Templeton, R. K., and J. Franklin. 1992. Adaptive information chick production elicits territorial intrusions in common and animal behaviour: why motorists stop at red traffic loons. Behavioral Ecology 17:881–888. lights. Evolutionary Theory 10:145–155. Pollock, M. S., D. P. Chivers, R. S. Mirza, and B. D. Wisenden. Thomson, R. L., J. T. Forsman, and M. Mo¨nkko¨nen. 2003. 2003. Fathead minnows, Pimephales promelas,learnto Positive interactions between migrant and resident birds: recognize chemical alarm cues of introduced brook stickle- testing the heterospecific attraction hypothesis. Oecologia back, Culaea inconstans. Environmental Biology of Fishes 66: 134:431–438. 313–319. Thomson, R. L., J. T. Forsman, F. Sarda´-Palomera, and M. Po¨ysa¨, H. 2006. Public information and conspecific nest Mo¨nkko¨nen. 2006. Fear factor: prey habitat selection and its parasitism in goldeneyes: targeting safe nests by parasites. consequences in a predation risk landscape. Ecography 29: Behavioral Ecology 17:459–465. 507–514. Pratt, I. W., H. Raiffa, and R. Schlaifer. 1995. Introduction to Timonen, S., M. Mo¨nkko¨nen, and M. Orell. 1994. Does statistical decision theory. MIT Press, Cambridge, Massa- competition with residents affect the distribution of migrant chusetts, USA. territories? Ornis Fennica 71:55–60. Rainey, H. J., K. Zuberbuhler, and P. J. B. Slater. 2004. van Bergen, Y., I. Coolen, and K. N. Laland. 2004. Nine-spined Hornbills can distinguish between primate alarm calls. sticklebacks exploit the most reliable source when public and Proceedings of the Royal Society of London B 271:755–759. private information conflict. Proceedings of the Royal Randler, C. 2006. Red squirrels (Sciurus vulgaris) respond to Society of London B 271:957–962. alarm calls of eurasian jays (Garrulus glandarius). Ethology Weinberger, E. D. 2002. A theory of pragmatic information and 112:411–416. its application to the quasi-species model of biological Roy, M., R. D. Holt, and M. Barfield. 2005. Temporal evolution. BioSystems 66:105–119. autocorrelation can enhance the persistence and abundance Werner, T., and T. W. Sherry. 1987. Behavioral feeding of metapopulations comprised of coupled sinks. American specialization in Pinaroloxias inornata, the ‘‘Darwin’s Finch’’ Naturalist 166:246–261. of Cocos Island, Costa Rica. Proceedings of the National Saracco, J. F., J. A. Collazo, and M. J. Groom. 2004. How do Academy of Sciences (USA) 85:5506–5510. frugivores track resources? Insights from spatial analyses of White, D. J., and B. G. Galef, Jr. 2000. Differences between the bird foraging in a tropical forest. Oecologia 139:235–245. sexes in direction and duration of response to seeing a JANNE-TUOMAS SEPPA¨NEN ET AL. Ecology, Vol. 00, No. 0

potential sex partner mate with another. Animal Behaviour Whiting, M. J., and J. M. Greeff. 1999. Use of heterospecific 59:1235–1240. cues by the Platysaurus broadleyi for food location. Whiten, A., J. Goodall, W. McGrew, T. Nishida, V. Reynolds, Behavioral Ecology and Sociobiology 45:420–423. Y. Yugiyama, C. Tutin, R. Wrangham, and C. Boesch. 1999. Witte, K., and R. Massman. 2003. Female sailfin mollies, Poecilia latipinna, remember males and copy the choice of Cultures in chimpanzees. Nature 399:682–685. others after 1 day. Animal Behaviour 65:1151–1159. Whiten, A., and R. Ham. 1992. On the nature and evolution of Zentall, T. R. 2006. Imitation: definitions, evidence, and imitation in the animal kingdom: reappraisal of a century of mechanisms. Animal Cognition 9:335–353. research. Pages 239–283 in P. J. Slater, J. S. Rosenblatt, C. Zuberbu¨hler, K. 2000. Interspecies semantic communication Beer, and M. Milinski, editors. Advances in the study of between two forest monkeys. Proceedings of the Royal behavior. Academic, New York, New York, USA. Society of London B 267:713–718. YNTHESIS &S ONCEPTS C