Decision making process is an important component of information use by animals and has already been studied in natural situations. Decision making takes time, which is expressed as a cost in evolutionary explanations of decision making abilities of animals. However, the duration of information assessment and decision making process has not been measured in a natural situation. Here, we use responses of wild magpies (Pica pica) to predictably approaching humans to demonstrate that, regardless of whether the bird perceived high (decided to fly away) or low (resumed foraging) threat level, the bird assessed the situation faster when approaching humans looked directly at it than when the humans were not directly looking at it. This indicates that prey is able to extract more information about the predator’s intentions and to respond sooner when the predator is continuously (“intently”) looking at the prey. The results generally illustrate how an increase of information available to an individual leads to a shorter assessment and decision making process, confirming one of central tenets of psychology of information use in a wild bird species in its natural habitat.
Citation: Lee S-i, Hwang S, Joe Y-e, Cha H-k, Joo G-h, Lee H-j, et al. (2013) Direct Look from a Predator Shortens the Risk-Assessment Time by Prey. PLoS ONE 8(6): e64977. https://doi.org/10.1371/journal.pone.0064977
Editor: Daniel Osorio, University of Sussex, United Kingdom
Received: January 6, 2013; Accepted: April 19, 2013; Published: June 5, 2013
Copyright: © 2013 Lee et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was funded by the following grants of the National Research Foundation of Korea (www.nrf.re.kr); 313-2007-2-C00747, 2010-0025546, 2010-0009006 and 2011-0030744. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Processes of assessment and decision making – have been studied in natural situations in various contexts including mate choice –, habitat choice – and foraging –. Prey escape responses to predators also provide opportunities to study the assessment (predation risk assessment) and the consequent decision-making processes. The studies of prey escape reactions are dominated by cost-benefit analyses in the tradition of behavioral ecology (e.g. –), or by proximate analyses of simple and fast escape responses (e.g. –) motivated by neurophysiological and neuroethological research. Some studies attempted to combine these two approaches trying to determine how simple responses (unlike more complex cognitive processes of assessment) of prey can lead to optimization of prey survival (e.g. –).
The studies of prey decision making based on more complex assessment processes (cognitive processes rather than simple mechanisms such like prey escape responses to looming stimuli –) have mostly focused on determining the assessment rules (e.g. –), perhaps because the assessment rules are at the center of attention in neurobiology , evolutionary psychology  or cognitive ecology . They are also important for adaptive interpretations of prey flight initiation distances (e.g. , –). The costs of decision making process, such as the time required for risk assessment, are important for the theory of information use in ecology . However, the time required for assessment and its influence on decision making process have not been determined in natural situations and have not been explicitly incorporated in the classical theoretical models of escape behavior (e.g. , , ).
The process of assessment and decision making requires time , and it may be shorter or longer depending on the factors that affect the amount of information available to the prey , , . Some of these factors may be the very same factors that normally indicate higher predation risk or stress in general. Direct gaze, looking at the prey, is one of them. For example, prey seem to treat the direct gaze or direct “looking” of the predators as an indicator of higher predation risk –. However, it has been documented that gaze, face and head, as well as their movements, may also contain important clues that may allow humans or computer algorithms to predict the actions of the subject – regardless of how threatening the action may be. These findings suggest that gaze and face may provide crucial information to the prey about predator’s future actions. If animals use this extra information associated with the predator looking at them in their risk assessment, this may lead to the shorter assessment duration when a predator looks directly at the prey (regardless of the outcome of the assessment). This idea has never been tested because previous experiments were based on the notion that direct look at the prey, as well as the face/head orientation, solely indicate higher predation risk to the prey and that prey adjusts its behavior accordingly. Furthermore, the previous experiments usually resulted in only one behavioral outcome: fleeing from the predator (e.g. ). In this situation it is impossible to study the duration of assessment that is independent from the outcome of assessment.
Here, we follow Blumenstein and Bouskila ’s suggestion that “observing the resultant behavior may permit inferences about assessment only when there are different behavioral responses to the stimuli”. Although originally this referred only to the inferences about assessment rules rather than costs or durations of this process, we apply their idea to test the prediction that direct look by a predator at a prey shortens the risk assessment and decision-making processes in the prey regardless of the level of predation risk estimated by the prey. We used a situation where both types of behaviors were observed in the urban population of the Eurasian Magpie, Pica pica: the behaviors that indicate high perceived predation risk (flee from an approaching predator), and those that indicate low perceived predation risk (ignore predator’s approach). Birds, like magpies, show behaviors that indicate when they became alert and therefore it is possible to measure the assessment time between the moment of becoming alert and the initiation of behavioral response. If the direct look of the predator provides the prey with some additional information useful in the risk assessment, then the assessment process should be shorter ,  regardless of whether the outcome of the assessment indicates high or low risk, that is regardless of whether a bird flies away or resumes foraging. On the other hand, if the effect of approaching humans on the bird’s timing of behavioral response is mostly due to variation in the perceived level of threat  or stress , rather than due to the amount of information available to them, then a bird whose behavioral response indicates higher perceived threat/stress level should take the decision sooner than a bird whose behavioral response indicates lower level of perceived threat/stress. Hence, the bird who flew away should take the decision sooner than the bird who decided to resume foraging.
In this study, we compare the effect of gaze on the duration of the assessment, between situations when a bird decided to fly away (outcome of an apparent assessment of high threat level) and the situations when a bird decided to cease alertness and to ignore the walking human. Additionally, we examine the effect of direct gaze on distance variables, including the classical flight initiation distance (FID).
The research has been conducted according to relevant national and international guidelines.
Experiments were conducted at 13 territories (8 in 2007 and 5 in 2008) of the Eurasian Magpies on the campus of Seoul National University, South Korea. The campus of Seoul National University represents a semi-urbanized environment where magpies are exposed to human pedestrians and are accustomed to human presence. The experiment was conducted on sunny days in mid-May 2007 and 2008, which corresponds to early nestling stage in the breeding of magpies. We conducted the experiments on the magpies that were not individually marked. However, magpies are highly territorial and immediately respond to any intruders by chasing them away. Thus, it is highly unlikely that any other bird than the breeder takes time to forage on the ground in the territory. We were aware of the boundaries of breeding territories, and we made sure that the testing sites were located well within the territory boundaries. Hence, we could be fairly sure that the foraging individuals were the territory owners. In early feeding period, the breeding female stays most of the time in the nest brooding the nestlings. Thus, we could easily observe one single magpie nearby the nest and we assumed that it was a male. We conducted experiments on these single magpies foraging near known active nests. Since the magpies were very accustomed to human presence, they did not show any vigilance behavior towards passing humans. The experimental procedure began after checking that the focal bird was continuously foraging for more than five minutes. We aimed at conducting at least a one gazing (up to four) and one non-gazing (up to five) experiment at each site in a randomized order. The order of experiments among the sites (gaze followed by non-gaze or non-gaze followed by gaze condition), and the dates of experiments for each site, were randomized within each year. For a given territory, the trials were done with at least two-day interval. Considering that the magpies see hundreds of people passing through their territories every day, and it is not uncommon to find people looking at magpies in the campus while walking, we do not think that habituation to the repeated testing might have occurred. In addition, the experimenters changed their clothes every day, and we believe that the experimental procedures were done in a manner that fully imitates the normal behavior of passers-by in the campus.
We use the term “gaze” for convenience to describe the situation where both the gaze per se and the facial direction are strictly correlated during the natural behavior of predators. Thus, in gaze condition, both eye gaze and facial direction were towards the bird; in non-gaze condition, both were away from the bird. In 2008, we carefully chose five experimental sites that were not examined in the previous year to avoid pseudo-replication (magpies are long-lived, stay in the same territories for many years and defend the territories year-round in our study area ).
We avoided situations with more than one magpie around because any social interaction between the magpies were likely to affect their responses to the experimental condition. We used the same walking speed of the experimenters across all the trials (approximately 1 step per second, which is similar to the pace of two people walking slowly together and talking to each other on the campus) and we conducted the experiments when there were no people around who can potentially affect the response of magpies. However, it is important to keep in my mind that numerous people frequently walked along those paths throughout the day and our experimental approaches were designed to imitate any such group of two young people walking on the path. In both years, the two walking experimenters were female students who performed the same role across the trials. The third experimenter, who recorded the trials in non-gaze condition (see below), were a female student in 2007 and a male student in 2008. None of the persons conducting the experiments took part in other research activities at the nests of magpies to assure that the birds do not individually recognize the experimenters .
Type of Response and Timing
The main goal in our study was to measure the duration of the assessment time, which is the time between the moment of becoming alert and the moment of taking a behavioral action of a focal bird. The basic procedure was as follows. We selected the experimental sites based on the location of the pedestrian path and the lawn where the focal magpie was usually found foraging. When the magpie was approximately 1.5 m away from the pedestrian path (tangential distance) and about 15 m away from the experimenters, and there was no noticeable place for the magpies to hide, two experimenters started walking towards the foraging magpie while facing it. In ‘Gazing’ condition, both experimenters constantly looked at the foraging magpie. In ‘Non-gazing’ condition, both experimenters did not look at the magpie, and their behavior imitated typical human passers-by in the campus. One of the two experimenters who walked towards the magpie carried a stopwatch and measured the response time defined as the interval between the moment when the magpie showed the first response (i.e. recognition of our presence indicated by disruption of foraging and alert posture with the head up) and the moment when the magpie showed the second response, which was categorized as one of the following: ignore (resume foraging), walk away (and eventually resume foraging at a distance to the path), hop away (and eventually resume foraging) or fly away (escape from human predator). The duration of hopping or walking away was not recorded, but it seemed that during that time the magpies did not stop watching humans. The other walking experimenter marked her own position on the path with the chalk bound to the tip of the umbrella at the moment when the magpie showed the second response, which marked the end of the experiment (for the estimation of distance variables, see below). In ‘Non-gazing’ condition when the walking experimenters could not directly and constantly observe the magpie, the response time and second response distance were measured from the video recorded by the third experimenter who filmed the experimental sessions from a 15–20 m distance (the walking experimenters and the focal magpie were visible in each video). From a sample of video recordings of gazing treatment, we confirmed that both methods of estimating the time and distance gave similar results.
Additionally we measured several distance variables, which are similar to variables already measured in other studies. We defined the distance between the experimenters and the location where the magpies showed the second response (second response distance). In 2008, we additionally measured the distance between the magpie and the two experimenters at the moment of the first response (first response distance). The distance variables are indices rather than absolute measures of distance between the human and the bird, because they were measured along the path of approach (Fig. 1), between the observer and the point of the shortest distance from the path to the bird’s initial location. Additionally, for those birds that hopped or walked away (Fig. 1), these distances are less accurate because unavoidably the bird moved away during the observers’ approach (but rarely more than by 1–2 m).
In order to examine the effect of gaze on the type of response of foraging magpies, we conducted multinomial regression on the four types of responses (ignore, walk away, hop away and fly away) with generalized linear mixed model. Identity of the foraging magpie (breeding territory identity, assumed from the proximity to bird’s nest) was treated as a random factor.
Because birds have a wide field of view – and because they might have only briefly looked strictly away from the moving experimenters, we believe that while walking or hopping away (Fig. 1; walk away → forage and hop away→ forage), the birds could still monitor the information from the approaching humans until the point where they resumed foraging. Therefore the time to the second response for these birds represented the duration of the risk assessment processes less precisely. But, the two extreme responses, fly away and ignore, more clearly showed that the bird abandoned careful (or any–in case of flying away) monitoring of the approaching human from the moment of decision to either forage again or to fly away. Therefore, for the main analysis, we only used the fly away and ignore responses to clearly determine whether gaze decreases the assessment time in prey (cognitive processing time) regardless of whether the approaching predator is assessed as dangerous (fly away response) or not (ignore response).
We used general linear mixed models to analyze the response time and second response distance (data collected in 2007 and 2008) as well as the first response distance (2008 data only) between gazing and non-gazing conditions, where the identity of the focal magpie was treated as a random factor. Year was included as an additional explanatory variable in all analyses. Throughout the text, averages values were given with standard errors. Our full dataset was unbalanced, and among our 13 magpies, 3 magpies were tested in only one treatment (gaze or non-gaze). In order to examine the robustness of our results, we repeated the same statistical procedures on 10 magpies that were tested in both treatments. Raw data can be delivered upon request. The analyses were conducted in SAS ver 9.3 (SAS Institute,Cary, USA).
Type of Response and Timing
The second response type differed significantly between gazing and non-gazing conditions (multinomial analysis, χ2 = 5.34, P = 0.021; Fig. 2). While magpies more often reacted by flying away in response to the gaze of experimenters, they more often resumed foraging or hopped/walked away from the path when the experimenters did not look directly at them. This indicates that humans, who walked along the standard campus paths and gazed directly at the birds foraging near the paths, were perceived by the birds as threatening more often than humans who did not gaze directly at the birds. However, regardless of the type of the second response, the response time was shorter when the experimenters looked directly at the birds: there was no significant interaction between the binary type of response (fly away or ignore) and the gaze treatment (gazing versus non-gazing), and there was no effect of the type of response (Fig. 3 (A), Table 1). Similar results were obtained when hopping away and walking away were pooled with ignore into one category “remain” and compared with the category “flee” (see Methods, Table S1 and Figure S1). This indicates that, regardless of the action taken by the bird in response to approaching humans, it took shorter time for the birds to initiate a behavioral action when humans directly looked at them, as if it was easier for the birds to estimate the degree of threat from an approaching human who directly gazed at them.
28 and 27 tests were conducted for gaze and non-gaze conditions respectively.
(A) the effect of gaze on the response time. (B) the effect of gaze on the second response distance. Among the four types of second responses, fly away and ignore responses were compared. Grey bars represent data from gaze condition and white bars are for non-gaze condition.
All the distance-based variables showed trends consistent with previous studies of FID or with the trends observed in response time. The first response distance did not differ between gazing and non-gazing conditions (9.08±0.69 m in gazing condition and 9.53±0.81 m in non-gazing condition; F1,18 = 0.18, P = 0.68). This indicates that the distance at which the bird stopped foraging and became alert did not depend on the presence of gaze.
During the response time, the experimenter covered the first-to-second-response distance (estimated in 2008 only as first response distance minus second response distance). Because the response time was shorter in gaze condition, this distance was also shorter in gaze (3.84±0.47 m) than in non-gaze (5.18±0.55 m) conditions, albeit only marginally non-significantly so (F1,18 = 3.44, P = 0.08).
Consequently, the second response distance was also affected. Although the second response distance did not differ significantly between experimental conditions in a simple analysis (Fig. 3 (B), Table 1), after including the first response distance as a covariate (for 2008 data only), the regression coefficient between first response distance (X) and the second response distance (Y) tended to be steeper in gazing (β = 0.85±0.18) than in non-gazing (β = 0.46±0.16) condition (interaction between the gazing type and first response distance is marginally non-significant; F1,16 = 2.67, P = 0.12; Fig. 4). All these trends in the distance variables appeared to be simple consequences of the difference in the response time between gazing and non-gazing treatments.
The second response distance for those birds that flew away (rather than resumed foraging with or without walking/hopping away to a new foraging spot) corresponds to the classical Flight Initiation Distance (FID; distance between predator and prey at the moment of prey escape initiation). In accordance with the classical literature, the second response distance tended to be larger in gazing (6.27±0.43 m) than in non-gazing (5.04±0.58 m) treatment but the effect was not significant (F1,12 = 1.34, P = 0.27).
We have shown here, for the first time, that the time to reach a decision and perform an action by a prey was shortened significantly when the predator directly looked at the prey, regardless of the type of prey’s reaction (to fly away or to ignore the predator approach), i.e. regardless of the perceived level of risk. These results suggest that, when an approaching predator directly looks at the prey, it reveals more information about itself and this extra information appears to be used by the prey to speed up the risk assessment process that leads to the choice of one of the several behavioral actions. The effect of direct look at the prey on the timing of prey behavioral response was independent from the effect of direct look on the level of perceived risk, because, similar to previous studies , , the birds flew away more frequently (a behavior indicating higher perceived threat) in response to humans looking directly at them.
We don’t know why a bird sometimes perceived the experimenters’ approach as risky and some other times not risky, but inter-individual and intra-individual variation in the risk assessment and decision making is expected –. None of the humans participating in the tests was familiar to the magpies therefore the effect of direct look (and exposing the face) cannot be explained by individual recognition, of which the magpies are capable . However, we clearly showed that some features of a human who was directly looking at the bird caused the bird to respond sooner, indicating that the bird assessed the level of threat posed by the person faster when the person was directly looking at it. These features do not seem to directly indicate higher level of threat which was presented by the “direct look” of the person, because the shortening of the assessment duration (response time) was the same for the high-threat responses (fly away) and for the low-threat responses (remain) by prey. Gaze, face and head, as well as their movements, contain important clues that may allow to predict the actions of the subject –, suggesting that they may also provide crucial information to the prey about the predator’s future actions. Hence, our results suggest that the effect of direct gaze (looking at the prey) of the predator on the behavior of the prey may extend beyond the classical issue of indicating the threat level –: the direct look from the predator may cause a general increase in information transmitted from the predator to the prey about the predator’s future actions.
Shortening of the assessment duration may be not only because of increased amount of information about the predators’ intent, but also because of increased level of stress. Stressors influencing the speed of decision may be viewed as serving an adaptive role by helping the animal to search for and scrutinize a source of danger . We cannot entirely reject the possibility that stress was higher in the gaze condition for the birds who felt more threatened and chose to flee as well as for those who felt less threatened and chose to ignore, and that this enhanced attention of birds allowing all birds to come to a quicker decision irrespective of the nature of that decision. If this were true, then the behavioral response of flying away, while indicating higher level of perceived threat, would not have been a good indicator of higher level of stress. Future experiments that manipulate level of stress and threat separately may evaluate this hypothesis. Such a total disassociation between the level of stress and the level of threat was not possible in our situation of an approach of potentially threatening humans.
We designed our study to allow for a wide range of behavioral responses. Many experiments with human predators were designed to imitate the situation where no option for ceasing the alertness was given to the prey: the experimenters often moved along trajectories that were not clearly predictable from previous experiences of the prey with humans either because human presence was not common in the area, or because the experimenter would go along a straight line towards the prey across a terrain, ignoring paths along which normally humans move (e.g. , ). In our study, we used the system that is typical for most urban situations where animals are repeatedly exposed to humans walking along designated predictable paths, and we conducted experiments imitating the normal behavior of humans in the environment. Animals tend to be less threatened by movements of humans along standard paths , , , because the prey in this situation should not expect the predator to suddenly change its path (changing path affects the escape response in prey ). Hence, a bird near the path needs only to arrive at the assessment of whether it should fear or ignore the by-passer. The sheer effect of experience with humans abundantly present on the campus also increases the likelihood that some individuals will be less threatened ,  and will perform behaviors other than escape. This situation is convenient for measuring the effect of various factors, such like predator’s gaze, on the duration of the assessment process that may result in both types of outcomes: flying away or ignoring the potential threat from the walking human.
From our results, we cannot discern whether magpies used the gaze itself or the face orientation as the cue, or whether they used other clues associated with a human looking directly at the bird. Recent studies attempted to tease apart the effect of gaze and that of the face orientation , . However, we think that discriminating the effect of gaze from that of face orientation is largely irrelevant in typical predator-prey situations, because, unlike in subtle social situations –, the gaze and the face direction of predators during their hunting behaviors are correlated such that there is no discrepancy between the gaze/look direction and the face/head direction –. The main predators of magpies in our population are the cats. Humans are the most common, although rarely lethal, threat. Avian predators are rarely found in our study area. Thus, we think that magpies in our study population developed keen alertness to the approaches and intentions of mammalian predators, and that they pay attention to the direction in which a predator looks and that direct look from the predator may, beyond indicating higher general risk, provide them with more detailed information useful in risk assessment and decision-making processes.
In summary, while our results were in agreement with the classical prediction of increased prey flight initiation distance (FID) in response to direct look from the predator, we have evidence that this effect may be present not only because the direct look indicates higher predation risk, as all previous papers assumed, but solely (or additionally) because the direct look makes the risk assessment faster for the prey. Most previous experiments always lead to the prey’s perception of high predation risk (sooner or later during the approach towards the prey the prey flew away from the predator). Therefore, they could not differentiate between the effect of direct look (or gaze) on lengthening of the FID because the direct look indicates higher perceived predation risk or because it facilitates faster risk assessment by the prey. The contribution of these two mechanisms can only be determined if experimental design promotes higher diversity in the perceived predation risk, and the behavioral response, by prey. We suggest that future studies in this area should use designs that allow distinguishing between the two mechanisms, and that theoretical models of optimal escape behaviors should incorporate the risk assessment duration, and factors affecting it, in their structure.
Effect of the direct gaze and the second response on the responses of 13 foraging magpies; the effect of gaze on the response time (above) and on the second response distance (below). “Flee” includes fly away responses, and “remain” includes ignore, walk away and hop away responses. Grey bars represent data from gaze condition and white bars are for non-gaze condition. Error bars denote standard errors.
The effect of gaze on the response time and the second response distance measured in 13 foraging magpies when the type of responses were coded as “flee (fly away)” and “remain (ignore, walk away, and hop away were pooled)”.
We thank magpie team members and the members of SNULBEE for discussion. We also thank Professor Jae Chun Choe for the logistic support. This study was conducted as a part of ecology class at SNU.
Conceived and designed the experiments: SIL PGJ. Performed the experiments: HS JYE CHK GHJ HJL JWK. Analyzed the data: SIL PGJ. Wrote the paper: SIL PGJ. Prepared the drafts: HS JYE CHK GHJ HJL JWK.
- 1. Kavaliers M, Choleris E (2001) Antipredator responses and defensive behavior: ecological and ethological approaches for the neurosciences. Neurosci Biobehav Rev 25: 577–586.
- 2. Dall SRX, Giraldeau LA, Olsson O, McNamara JM, Stephens DW (2005) Information and its use by animals in evolutionary ecology. Trends Ecol Evol 20: 187–193.
- 3. Schmidt KA, Dall SRX, Gils JAV (2010) The ecology of information: an overview of the ecological significance of making informed decisions. Oikos 119: 304–316.
- 4. Trimmer PC, Houston AI, Marshall JAR, Mendl MT, Paul ES, et al. (2011) Decision-making under uncertainty: biases and Bayesians. Anim Cogn 14: 465–476.
- 5. Gibson RM, Langen TA (1996) How do animals choose their mates? Trends Ecol Evol 11: 468–470.
- 6. Luttbeg B (1996) A comparative Bayes tactic for mate assessment and choice. Behav Ecol 7: 373–377.
- 7. Castellano S, Cermelli P (2011) Sampling and assessment accuracy and mate choice: A random-walk model of information processing in mating decision. J Theor Biol 274: 161–169.
- 8. Orians GH, Wittenberger JF (1991) Spatial and temporal scale in habitat selection. Am Nat 137: S29–S49.
- 9. Railsback SF, Harvey BC (2002) Analysis of habitat selection rules using an individual-based model. Ecology 83: 1817–1830.
- 10. Yang CH, Belawat P, Hafen E, Jan LY, Jan YN (2008) Drosophila egg-laying site selection as a system to study simple decision-making processes. Science 319: 1679–1683.
- 11. Charnov E (1976) Optimal foraging: the marginal value theorem. Theor Pop Biol 9: 129–136.
- 12. Stephens DW, Krebs JR (1986) Foraging theory. Princeton: Princeton University Press. 247p.
- 13. Stephens DW, Brown JS, Ydenberg RC (2007) Foraging: Behavior and ecology. Chicago: University of Chicago Press. 576p.
- 14. Ydenberg RC, Dill LM (1986) The economics of fleeing from predators. Adv Stud Behav 16: 229–249.
- 15. Blumstein DT (2003) Flight-initiation distance in birds is dependent on intruder starting distance. J Wildl Manage 67: 852–857.
- 16. Cooper WE, López P, Martín J, Pérez-Mellado V (2012) Latency to flee from an immobile predator: effects of predation risk and cost of immobility for the prey. Behav Ecol 23: 790–797.
- 17. Santer RG, Yamawaki Y, Rind FC, Simmons PJ (2005) Motor activity and trajectory control during escape jumping in the locust Locusta migratoria. J Comp Physiol A 191: 965–975.
- 18. Burrows M (2006) Jumping performance of froghopper insects. J Exp Biol 209: 4607–4621.
- 19. Fotowat H, Gabbiani F (2007) Relationship between the phases of sensory and motor activity during a looming-evoked multistage escape behavior. J Neuroscience 27: 10047–10059.
- 20. Sutton GP, Burrows M (2008) The mechanics of elevation control in locust jumping. J Comp Physiol A 194: 557–563.
- 21. Jablonski PG, Strausfeld NJ (2000) Exploitation of an ancient escape circuit by an avian predator: prey sensitivity to a model predator display in the field. Brain Behav Evol 56: 94–106.
- 22. Jablonski PG, Strausfeld NJ (2001) Exploitation of an ancient escape circuit by an avian predator: relationships between taxon-specific prey escape circuits and the sensitivity to visual cues from the predator. Brain Behav Evol 58: 218–240.
- 23. Javurkova V, Sizling AL, Kresinger J, Albrecht T (2012) An alternative theoretical approach to escape decision-making: the role of visual cues. PLOS ONE 7(3): e32522.
- 24. Bouskila A, Blumstein DT (1992) Rules of thumb for predation hazard assessment: predictions from a dynamic model. Am Nat 139: 161–176.
- 25. Blumstein DT, Bouskila A (1996) Assessment and decision making in animals: a mechanistic model underlying behavioral flexibility can prevent ambiguity. Oikos 77: 569–576.
- 26. Swaisgood RR, Rowe MP, Owings DH (1999) Assessment of rattlesnake dangerousness by California ground squirrels: exploitation of cues from rattling sounds. Anim Behav 57: 1301–1310.
- 27. Stankovich T, Blumstein DT (2005) Fear in animals: a meta-analysis and review of risk assessment. Proc Biol Sci 272: 2627–2634.
- 28. Blanchard DC, Griebel G, Pobbe R, Blanchard RJ (2011) Risk assessment as an evolved threat detection and analysis process. Neurosci Biobehav Rev 35: 991–998.
- 29. Barret HC (2005) Adaptations to predators and prey. In: Buss DM, editor. Handbook of evolutionary psychology. Hoboken: John Wiley & Sons. 210–223.
- 30. Dukas R (1998) Cognitive ecology: the evolutionary ecology of information processing and decision making. Chicago: University of Chicago Press. 430 p.
- 31. Cooper WE (1998) Direction of predator turning, a neglected cue to predation risk. Behaviour 135: 55–64.
- 32. Broom M, Ruxton GD (2005) You can run or you can hide: optimal strategies for cryptic prey. Behav Ecol 16: 534–540.
- 33. Cooper WE, Frederick WG (2007) Optimal flight initiation distance. J Theor Biol 244: 59–67.
- 34. Stankovich T (2008) Ungulate flight responses to human disturbance: a review and meta-analysis. Biol Cons 141: 2159–2173.
- 35. Chittka L, Skorupski P, Raine NE (2009) Speed-accuracy tradeoffs in animal decision making. Trends Ecol Evol 24: 400–407.
- 36. Bradbury JW, Vehrencamp SL (1998) Principles of animal communication. Sunderland: Sinauer. 882p.
- 37. Ristau CA (1991a) Cognitive ethology, the minds of other animals. Hillsdale: Lawrence Erlbaum Associates. 344p.
- 38. Ristau CA (1991b) Before mind reading: attention purposes and deception in birds? In : Whiten A, editor. Natural theories of mind: Evolution, development and simulation of everyday mind reading. Cambridge: Basil Blackwell. 209–222.
- 39. Hampton RR (1994) Sensitivity to information specifying the line of gaze of humans in sparrows (Passer domesticus). Behaviour 130: 41–51.
- 40. Watve M, Thakar J, Kale A, Puntambekar S, Shaikh I, et al. (2002) Bee-eaters (Merops orientalis) respond to what a predator can see. Anim Cogn 5: 253–259.
- 41. Eason PK, Sherman PT, Rankin O, Coleman B (2006) Factors affecting flight initiation distance in American robins. J Wildl Manage 70: 1796–1800.
- 42. Bateman PW, Fleming PA (2011) Who are you looking at? Hadeda ibises use direction of gaze, head orientation and approach speed in their risk assessment of a potential predator. J Zool 285: 316–323.
- 43. Cooper WE (2011) Influence of some potential predation risk factors and interaction between predation risk and cost of fleeing on escape by the lizard Sceloporus virgatus. Ethology 117: 620–629.
- 44. Tijerina L, Garrott WR, Stoltzfus D, Parmer E (2005) Eye glance behavior of van and passenger car drivers during lane change decision phase. Trans Res Rec 1937: 37–43.
- 45. Ji Q, Yang X (2002) Real-time eye, gaze, and face pose tracking for monitoring driver vigilance. Real-Time Imaging 8: 357–377.
- 46. Doshi A, Trivedi MM (2009) On the roles of eye gaze and head dynamics in predicting driver’s intent to change lanes. IEEE Trans Intell Transp Syst 10: 453–462.
- 47. Heylen D (2010) Ubiquitous gaze: using gaze at the interface. In: Aghajan H, Augusto JC, Delgado RLC, editors. Human-centric interfaces for ambient intelligence. Burlington: Academic Press. 49–70.
- 48. Stankovich T, Goss RG (2005) Effects of predator behavior and proximity on risk assessment by Columbian black-tailed deer. Behav Ecol 17: 246–254.
- 49. Lee S-I, Seo K, Lee W, Kim W, Choe JC, et al. (2011) Non-parental infanticide in a dense population of the Black-billed Magpie (Pica pica). J Ethol 29: 401–407.
- 50. Lee WY, Lee S-I, Choe JC, Jablonski PG (2011) Wild birds recognize individual humans: experiments on magpies, Pica pica. Anim Cogn 14: 817–827.
- 51. Martin GR, Katzir G (1999) Visual fields, foraging and binocularity in birds. In: Adams NJ, Slotow RH, editors. Proc 22 Int Ornithol Congr, Durban. Johannesburg: BirdLife South Africa. 2711–2728.
- 52. Martin GR (2007) Visual fields and their functions in birds. J Ornithol 148: S547–562.
- 53. Fernandez-Juricic E, Gall MD, Dolan T, Tisdale V, Martin GR (2008) The visual fields of two ground-foraging birds, House Finches and House Sparrows, allow for simultaneous foraging and antipredator vigilance. Ibis 150: 779–787.
- 54. Carrete M, Tella JL (2010) Individual consistency in flight initiation distances in burrowing owls: a new hypothesis on disturbance induced habitat selection. Biol Lett 6: 167–170.
- 55. Carrete M, Tella JL (2011) Inter-individual variability in fear of humans and relative brain size of the species are related to contemporary urban invasion in birds. PLoS ONE 6(4): e18859.
- 56. Runyan AM, Blumstein DT (2004) Do individual differences influence flight initiation distance? J Wildl Manage 68: 1124–1129.
- 57. Fernandez-Juricic E, Jimenez MD, Lucas E (2002) Factors affecting intra- and inter-specific variations in the difference between alert distances and flight distances for birds in forested habitats. Can J Zool 80: 1212–1220.
- 58. Mendl M (1999) Performing under pressure: stress and cognitive function. Appl Anim Behav Sci 65: 221–244.
- 59. Rodriguez-Prieto I, Fernandez-Juricic E, Martin J (2008) To run or to fly: low cost versus low risk escape strategies in blackbirds. Behaviour 145: 1125–1138.
- 60. Mainini B, Neuhaus P, Ingold P (1993) Behavior of marmots Marmota marmota under the influence of different hiking activities. Biol Conserv 64: 161–164.
- 61. Miller SG, Knight RL, Miller CK (2001) Wildlife response to pedestrians and dogs. Wildl Soc Bull 29: 124–132.
- 62. Carter J, Lyons NJ, Cole HL, Goldsmith AR (2008) Subtle cues of predation risk: starlings respond to a predator’s direction of eye-gaze. Proc Biol Sci 275: 1709–1715.
- 63. von Bayern AMP, Emery NJ (2009) Jackdaws respond to human attentional states and communicative cues in different contexts. Curr Biol 19: 1–5.
- 64. Emery NJ (2000) The eyes have it: the neuroethology, function and evolution of social gaze. Neurosci Biobehav Rev 24: 581–604.
- 65. Langton SRH, Watt RJ, Bruce V (2000) Do the eyes have it? Cues to the direction of social attention. Trends Cogn Sci 4: 50–59.
- 66. Eaton RL (1970) The predatory sequence with emphasis on killing behavior and its ontogeny in the cheetah (Acinonyx jubatus Schreber). Z Tieprsychol 27: 492–504.
- 67. Bergstrom BJ (1985) Unusual prey-stalking behavior by a goshawk. Auk 56: 415.
- 68. Seidensticker J, McDougal C (1993) Tiger predatory behaviour, ecology and conservation. Symp Zool Soc Lond 65: 105–125.