Birds and their training
We used 30 naïve adult great tits as our model insectivorous bird species. All birds were collected from nest boxes when 10-13 days old; between 22nd and 24th of May 2018. All chicks from a given nest (4-6 chicks per nest), were kept together in one cage until fledging. Cages had dimensions 0.7 × 0.4 × 0.5 m with three perches. The birds were kept under natural winter daylight (8:15 to 17:00) augmented with fluorescent light tubes and automatically opened/closed window shades. After fledging, birds were moved to cages in groups of two to three birds, irrespective of their sex. Up to that period, chicks were individually hand fed every half‐hour with hand rearing food for all bird chicks (Nutribird a21), a bread-like diet (made from Nutribird a21, commercial chicken food (Country’s Best Show 1 crumble), eggs, wheat flower, sugar and sunflower margarine, mixed and baked for ~40 minutes at ~180°C temperature) supplemented with mealworms. This was repeated for a period of 13.5 h (6:45 am–8:15 pm) every day. Hand feeding was continued after fledging, but the intervals were prolonged to 1h and later 4h, to stimulate birds to feed by themselves. At independence, about 35 days after hatching, birds were relocated to individual cages. From then on, the baked diet, mixed seeds, commercial dried insects (Nobby Orlux Insect Patee Premium) and water were provided ad libitum. Mealworms were offered only as supplementary food. During the whole period of captivity, chicks from different nests could hear and see each other, since they were raised in the same room.
None of the birds used in the experiment had previous direct experience with foliage or plant volatiles except that they could have been indirectly exposed to some plant cues associated with caterpillars and other herbivorous insects fed to them by their parents in the first 10-13 days after hatching. As Cattley guava naturally occurs in South America, its volatiles are evolutionary unfamiliar to the birds used in our experiment. Moreover, the bird chicks came from nest boxes placed in mixed oak-spruce forest, with no Scotch elm trees present in the surroundings. As HIPVs are plant-specific, we can consider that the birds used in our experiment are naïve to HIPVs emitted by both sapling species used in the experiment. While Cattley guava represented a completely novel signal to the birds, Scotch elm represented a signal to which the birds could be evolutionarily adapted. Birds were randomly assigned into three training groups: (1) to associate food with the HIPVs of the Cattley guava, (2) to associate food with the HIPVs of the Scotch elm, and (3) not trained to associate food with any green foliage.
We performed all training, and experiments, in two Y-shaped aviaries built with mesh screens (Figure S1). Two 2m tall dead Pedunculate oaks in a large pot were placed in the corners of the aviary and served as a perch. The birds whom we conditioned to find food in induced Cattley guava were offered one induced Cattley guava with 5 mealworms pinned to the leaves and one control cattle guava without any mealworms. The position of the induced and control cattle guava (left or right corner) in the aviary was selected randomly. Birds were allowed to habituate to the aviary and search for food, thus associate the volatile compounds of guava with the reward, for 1.5 hour every day. The birds were released to the aviary in pairs to make the learning (from each other) faster in first 3 trials. We conditioned birds to the volatile compounds of elm the same way. To condition the control birds to the cage, and initiate their interest in searching for food, we pinned 5 mealworms directly on one of the dead Pedunculate oaks. We conducted one training trial per bird daily. After each training trial, we checked how many meal worms were eaten. We considered the birds to be conditioned to the volatile compounds of the specific plant after they repeatedly ate at least 4 out of the 5 mealworms from the induced sapling.
It took 15 trials to condition all the birds to the induced saplings successfully. Some (40%) birds did not pay too much attention to the plants during the first five trials but investigated mostly the aviary netting and surroundings. After first five training trials, we started to record some missing mealworms in most of the bird individuals, while some (20%) bird individuals took at least 10 trials until they started eating mealworms. After 15th trial, all birds successfully searched for meal worms on the induced saplings. This contrasted with previously described similar training, during which birds were able to find the larvae on sapling after fifth acclimatization (Amo et al. 2013a).