Guns and butter: a no cost defense against predation for Chrysomela confluens

Chrysomela confluens produces a salicylaldehyde-based defensive secretion which is very effective against generalist predators and apparently produced at no cost. If no cost defenses are common, then one of the basic assumptions in the plant-herbivore literature, i.e. tradeoffs among defense, reproduction, and growth, must be reconsidered. We examined the effectiveness of this defense by exposing defended larvae and larvae whose secretion had been removed to a generalist predator. Larvae which had their secretions intact were attacked by only 7% of the ants which encountered them, and none of these larvae suffered serious damage. In contrast, those which had been "milked" of their secretions immediately prior to exposure were attacked in 48% of such encounters, and two-thirds of the larvae were killed. Larvae which had been milked 24 or 72 h before exposure, then allowed to regenerate their defenses, were attacked at rates indistinguishable from larvae that had not been milked. Thus regenerated defenses are just as effective as original defenses. We also tested the hypothesis that the cost of defense production and maintainence would be reflected in reductions in developmental rates and final adult mass and increases in leaf consumption rate. We found that larvae which were milked daily of their secretions manifested no measurable cost of recharging reservoirs. Milked larvae grew and fed at the same rates as their control sibs, and became adults of equal or slightly larger size. The liberation of glucose from salicin, a precursor present in leaves of salicaceous hosts, during the production of salicylaldehyde apparently provides enough of an energetic benefit to offset the cost of maintaining an effective defense. Consistent with this hypothesis, we did not find that milked larvae compensated for increased nutritional or salicin demands by increasing their feeding rates. Although this patterns is familiar to chemical ecologists it is generally unappreciated in the plant-herbivore literature. It is likely that many arthropod herbivore defensive systems come at little or no cost, given the intimacy of association between herbivores and their food plants. Sequestration of host plant defensive chemicals which eliminates the cost of synthesis is common in arthropods. The de novo synthesis of chemical defenses may be less costly than expected if it is integrated into other parts of an insects metabolism. Calculations based on the bond energies or molecular constitution of the compounds will not yield a complete perception of cost. Tests over the life of the herbivore, coupled with an understanding of the herbivore's metabolism, are necessary.