Results
Our food-choice experiment detected consistent variation among individuals in food preferences over a short period (Fig 1 ). The niche centre of each individual in the resource space—defined using the two first axes of a principal coordinate analysis (Fig. 1D ; Table S2 )—was highly consistent over the trials (Table 1; Fig. 2 ) regardless of age, sex, trial order or aviary (Table S3 ). Despite striking differences in nutritional contents and size of the seeds used for this experiment (Fig 1B,C ), individuals only coincided in avoiding one food type (i.e. peas).