Experimental design and procedure
To examine whether great tits were able to learn the specific volatiles of the plant species, and associate it with the food, we performed a two-choice experiment in the same aviaries where the training was conducted. We offered all possible pairwise combinations of plant treatments, except both Control plants, to birds with all types of training. This comprised five plant combinations and three bird training treatments (Table 1). All birds participated in all five experiments, i.e. we conducted 145 trials in total.
All experimental trials were performed during 5 days (20 - 28 Aug 2018), between 09:00 and 17:00, under sunny and warm conditions to avoid variation in the emission of volatiles due to differences in ambient conditions such as temperature (Vallatet al. 2005). On each experimental day, a new pair of saplings was placed in the aviary prior to the first trial, and then replaced by a new pair of saplings after every four hours. In contrast to training trials, there were no mealworms pinned to any of the saplings during the experiment. Apart from that, the saplings were prepared the same way as for the training. We conducted two trials in two aviaries simultaneously. Prior to each experimental trial, the birds were starved in their housing cages for 90 minutes. This ensured that the birds were motivated to search for larvae on the saplings in the experimental aviary. After each training (or experimental trial), the bird was captured with a net and returned to its cage. We recorded the behaviour of birds in the aviary during the 20 min long trials using a video camera (Panasonic Full HD V180EP-K). One observer, unaware of the treatment, analysed the video tapes, recorded the time spent on each of the saplings (i.e., on the trunk inspecting the foliage form there, or in the foliage) and the time spent in the proximity of the sapling (i.e., within 50 cm from the sapling, on the netting of the cage, or on the pot, but only when the bird paid visible attention to the sapling). A new experimental trial started every 30 minutes, while 5 minutes were allowed for the catching and replacement of birds, and the first 3-5 minutes of the video recording were discarded to account for the acclimatization of the bird to the cage.