Experimental trees and their induction
In our experiment, we aimed to test responses of naïve chicks of great tits (Parus major ) to an evolutionarily familiar odour of herbivore-damaged Scotch elm, Ulmus glabra Huds. (Ulmaceae ), and evolutionarily novel odour of herbivore-damaged Cattley guava Psidium cattleyanum Sabine (Myrtaceae ). We used saplings (1.5-m-tall) of Cattley guava as a novel plant species, which did not have any evolutionary contact with our experimental bird species. Genus Psidium was distributed throughout tropical America and Caribbean region. However, they were adopted as a crop in subtropical and tropical Asia, tropical Africa and Oceania. SeveralPsidium cultivars are also commercially grown in southwestern Europe and Greece since the middle of the 20thcentury. Scotch elm (1.5-m-tall) was used as a sapling which naturally occurs in the great tit range of distribution, and birds thus have evolutionary contact with them. Saplings were planted in 20l pots into a standard agricultural soil two months prior to the beginning of the experiment. Control (N = 10) and experimental (induced) saplings (N = 10) were placed into two greenhouses just prior start of the experiment, so the control saplings can’t receive chemical signals from induced saplings. All saplings were watered weekly.
To prepare the induced version of a sapling (i.e. those which will produce herbivore-induced volatile compounds), we slightly scratched 10 leaves by razor and applied saliva of Locusta migratoria on them. After 30 minutes, we removed the damaged leaves. In this way, plants started to produce herbivore induced volatile compounds, but the “herbivorous” damage was not visible to the birds. For training as well as experiments, new saplings were prepared this way every 4 hours, i.e. for every 9th trial in experiment. In total, we used 30 individuals of each plant species.
We predicted that untrained, naïve birds, will not show any preferences for any saplings, with no respect whether the sapling provides olfactorial signal of insect infestation or not, and without respect whether the species is evolutionarily familiar or not. On the other hand, we predicted that birds trained to find food on saplings of experimental species will prefer that species also in pair-wise preference tests. Finally, we predicted that birds will more easily learn the volatile compounds of the tree species, which is evolutionarily familiar, and naturally occurs in their distributional range, than the volatile compounds of a completely novel, evolutionarily unfamiliar, volatile compounds.