Cycads: their evolution, toxins, herbivores and insect pollinators.

Palaeobiological evidence indicates that gymnosperms were wind-pollinated and that insect pollination began in angiosperms in the Lower Cretaceous (ca. 135 mya) leading to close associations between higher plants and their pollinators. Cycads, which were widespread and pervasive throughout the Mesozoic (250-65 mya) are among the most primitive living seed-plants found today. Because pollination by beetles and by thrips has now been detected in several modern cycads, it is attractive to speculate that some insects and cycads had already developed similar mutualistic interactions in the Triassic (250-205 mya), long before the advent of angiosperms. We also draw attention to another key factor in this insect-plant relationship, namely secondary, defensive plant substances which must always have controlled interspecific interactions. Cycads mainly produce toxic azoglucosides and neurotoxic non-protein amino acids (e.g. BMAA), which apparently are crucial elements in the development and maintenance of mutualism (pollination) and parasitism (herbivory) by cycad-linked herbivores. We now add new results on the uptake and storage of the main toxin, cycasin, of the Mexican cycad Zamia furfuracea by its pollinator, the weevil Rhopalotria mollis, and by a specialist herbivore of Zamia integrifolia, the aposematic Atala butterfly Eumaeus atala.