We recorded significant differences in whether the focal squirrel would move away or not between our four experimental treatments (Chi-test: χ23 = 18.49, P < 0.001) (Fig. 1). When the observer remained see more on the footpath, only 22% of individuals (9

of n = 41 approached in this manner) moved away from the observer, contrasting with 63% (26 of n = 41) of squirrels approached when the observer left the footpath (χ21 = 8.20, P = 0.004). Squirrels were also more reactive when he was looking at them (χ21 = 10.24, P = 0.001). When the observer was looking at them, 40 and 90% (on footpath and off footpath, respectively) of squirrels moved away compared with only 5 and 35% when the observer was not looking at them (the remainder of individuals remained where they were). There was a significant treatment effect upon alert distance (Median test: χ23 = 30.66, P < 0.001), FID (χ23 = 32.49, P < 0.001) and distance fled (χ23 = 33.32, P < 0.001). For alert distance (Fig. 2), squirrels showed little change in behaviour when the pedestrian remained on the footpath and did not look at them (median alert distance 0 m, 0–4 m); there was no significant difference in alert distance for the other three treatments (5 m; 0–8 m). For FID and distance fled (Fig. 2), squirrels were most reactive (longer

Small molecule library distances) when approached by a pedestrian that moved off the footpaths and looked at the squirrel as he approached (FID: 4 m; 0–6.5 m; distance fled: 6 m; 0–13 m); the other three treatments were not significantly different from each other (FID: 0 m;, 0–5.5 m; distance fled: 0 m; 0–10 m). In the urban environment, the mark of a successful animal species is likely to be its ability to distinguish between innocuous stimuli and genuine risk. Eastern grey squirrels have established populations in the many cities across the globe. In this study, we show that selleck products eastern grey squirrels living in a Manhattan park show behavioural flexibility towards pedestrian activities. Squirrels allow pedestrians to approach closely compared with conspecifics in rural areas (e.g. FID 10.4 ± 6.65 m; Cooper et al., 2008). PCVST squirrels are mostly exposed to pedestrians

acting in a benign, predictable manner, but show different responses towards pedestrians showing more irregular behaviour, presumably because they perceive it as more risky. First, PCVST squirrels were more reactive when the pedestrian was looking directly at them as he approached, as has been observed in urban birds (increasing FID or other escape reactions, e.g. ceasing foraging or alarm calling, Bateman & Fleming, 2011, Clucas et al., 2013, Lee et al., 2013). Second, PCVST squirrels were more reactive towards a person that diverged from ‘normal’ behaviour by not keeping to the footpaths. Similar responses have been recorded in Alpine marmots Marmota marmota (Mainini, Neuhaus & Ingold, 1993) and for American robins Turdus migratorius (Eason et al.