Boldness is sometimes measured by how well an animal can tolerate human proximity (Atwell et al. 2012). In this study, Boldness was measured following a similar procedure as the one used by Atwell et al. (2012), with the flying initiation distance (FID). The FID is defined as the distance between the place where a person that was approaching an animal was, when the animal escaped and the exact point where the it fled (Ydenberg & Dill, 1986).

Most bird’s activity peaks are at the sunrise or at the sunset (Ralph, 1993). Therefore, the data was obtained between 16:30 and 19:00, hours inside of the activity peak (Ralph, 1993). All data was obtained on sunny days, as most of the studies with FID have used this condition to avoid the effect of rain in the results (e.g. Fernández-Juricic & Schroeder, 2003, Rodríguez-Prieto et al. 2008; Carrete & Tella, 2011). It has been demonstrated that the responses of birds to predators vary in the reproductive season (Colombelli-Négrel et al. 2010). Therefore, the data was collected before the reproductive season of the Chiguanco Trush, between July and November 2016 (Fjeldså & Krabbe, 1990; Baptista et al. 2013). Only adult individuals were approached as some behavioural traits might be affected by age (Greenberg and Mettke-hofmann, 2001; Griffin and Guez, 2014). Meanwhile adults and juveniles are easy to differentiate due to the characteristic juvenile individuals’ mottled plumage, males and females are not. Sexual dimorphism is not consistent due to a strong interindividual variation that happens both in females and males. Therefore, no discrimination was made between sexes. The data was obtained performing walks along the selected study places and approaching the birds of interest when spotted. To avoid measuring two times the same individual three different walks were defined for each study place and never repeated. In total, twelve walks were performed in the four different study places, six inside the urban area and six inside the rural area.

The FID was only measured on birds that were on the ground (resting, foraging or grooming) as the FID may vary if the bird is on a tree or a bush, allowing smaller FID (Krams, 1996). Only solitary birds were approached as it has been demonstrated that the size of the group may also affect the FID, bigger groups normally have bigger FID (Fernández-Juridic et al. 2002).

The bird was always approached by the same person, in straight line and avoiding the interception of vegetation on the way so the bird could detect the person. The approach was performed at a constant speed of ca 1.0 m/s and maintaining the visual contact with the bird (Blumstein et al. 2003). When the bird escaped, the person measured the distance between itself and the place where the bird took flight with a measuring tape. This was considered as the boldness measurement (centimeters). In total, the FID of 46 birds, 25 in the urban area and 21 in the rural area was measured.

Neophobia was measured following a similar procedure as the one described by Mettke-Hofmann et al. (2002). It was tested on the morning of day three at 09:00 after the food deprivation. A novel object with food inside (“drawer test”) was placed at the same spot where the normal food dish used to be placed with fresh food in the mornings of days one and two. This was done so that the willingness to feed competed with the phobia to the new object (Greenberg and Mettke-hofmann, 2001). The latency of the bird to approach the object from the moment it was introduced in the enclosure was considered as the neophobia measurement (seconds) (Greenberg and Mettke-hofmann, 2001). An observer was located behind the door of the enclosure, looking through a hole. The observer measured the latency to approach the novel object with a chronometer. In addition, all the tests were video recorded with a GoPro HERO and afterwards reanalysed to ensure that the data was correctly collected.

Innovation and problem solving abilities can be studied using problem solving boxes (Griffin and Guez, 2014). In this study the problem solving tests were based on the problem boxes used by Audet et al. (2015). Two different problem boxes were used. On the third day, the first problem test (“drawer test”) was presented. The birds were presented with a drawer (7 cm height × 7 cm width × 7cm length) made from plastic (1 transparent side and 4 matt sides). In both tests there was food placed, the food used to motivate the birds was fresh bread (considered to be their favourite food, according to pilot study observations). The birds had the opportunity to gain access to food by pulling the drawer from a hook or from the upper part of it. On fourth day, the second problem test (“tube test”) was presented. The birds were presented with a tube made from transparent plastic with an opening on one of the extremes (5 cm height × 5 cm width × 20 cm length), inside the tube there was a wooden stick (25cm) with a hook attached to the extreme of the stick that was outside the tube. Food was placed on the extreme of the stick which was inside the tube. The birds needed to realize that the only way to solve the problem and reach the food was by pulling the stick with their beak using the hook. The birds were given a maximum of 20 minutes to solve the tests, after that, the test was retired and the bird was considered to be unable to solve the test. The time was decided according to pilot study observations in which, on average, the birds lost interest in the tests (stop trying to solve it and started to ignore it) after 13 minutes. The time to solve the test only started counting from the moment the bird pecked the problem test for the first time, in order to avoid neophobia effects (in the case of the drawer test this neophobia effect was considered as the neophobia measurement, see 3.5). An observer measured the ability to solve or not the test and the latency to solve it (if success) looking through a hole behind the enclosure door. In addition, the tests were video recorded with a GoPro HERO and afterwards reanalysed to ensure that the data was correctly collected. Both the ability to solve the test or not, plus the latency (in case of success) to solve it were considered to be the problem solving abilities.