The alkaloidal responses of wild tobacco to real and simulated herbivory.

I compared the induced alkaloidal response in undamaged leaves of plants subjected to herbivory by the larvae of Manduca sexta and to different simulations of this herbivory; all herbivory treatments removed similar amounts of leaf mass. Although larval feeding induced a significant increase (2.2x) in alkaloid concentrations compared to undamaged plants, the alkaloid responses to larval feeding were significantly lower than the responses to an herbivory simulation (4x controls) which involved removing the same amount of leaf area from the same positions on the leaf, over a similar time period. Moreover, another herbivory simulation, identical in amount of leaf mass removed and duration of damage to the larval feeding, but without regard to spatial array of leaf damage, resulted in an alkaloidal response (5.5x controls) higher still than the previous herbivory simulation. In a second experiment the importance of leaf vein damage on the induced alkaloidal response was examined. Here, leaf removal that involved cutting leaf tissues from between secondary veins before removing the midrib, resulted in alkaloidal responses that were significantly lower (1.7x controls) than responses from leaf removal that involved cutting both veins and midribs along with the intervein tissues (2.6x controls). Vein damage alone did not produce a significant response. These results indicate that herbivory is difficult to simulate: that how a leaf is damaged can be as important as the magnitude of leaf damage in determining a plant's response to damage.