Free-living urban pigeons at 12 sites on the island of Montreal
served as subjects for the field experiments in the summer
of 2001. Distance between the sites ranged from 0.4 to
6.4 km, with an average of 2.2 km. Flock size varied from
seven to 175 pigeons depending on the site, but day-to-day
fluctuations in numbers for a given site were minimal. For 2
days before the tests began, we visited each site at a specific
time and offered 50 g of mixed seeds at the feeding area to
habituate the pigeons to the time and place of feeding. On
subsequent days, we presented the Plexiglas box used in the
captive innovation test at each site for a maximum of two
20 min trials over 2 consecutive days. Latencies to first contact
and to successful opening of one of the lids or drawers
were taken from the video records of each trial, as well as
flock size. When no birds opened the box or when the opening
could have been accidental, we assigned the maximum
latency of 2402 s to this flock. To analyse the results from
the field tests, we used non-parametric statistics, as no transformation
could correct for the non-normal distributions of
the data.