Plant defenses can affect herbivory directly, by impacting herbivore physiology or behavior, and indirectly, by modulating predator-prey interactions.  Plant defenses generally influence predation by affecting the quality of herbivores as prey or by altering their vulnerability to enemies. With regards to herbivore suppression, plant defenses work synergistically with predators in some situations and antagonistically in others. In order to investigate interactive effects of plant defenses, predation, and herbivore density on plant productivity and herbivore success, we conducted a factorial experiment in which the presence of predators (Podisus maculiventris) and levels of chemical defense in tomato plants (Lycopersicon esculentum) were manipulated in mesocosms across a range of naturally observed densities of beet armyworm caterpillars (Spodoptera exigua).  To manipulate defense levels, we used JAI-1 mutant plants that are unable to produce inducible defenses (low level of defense) and wild type plants (var. Castlemart) sprayed with jasmonic acid (high level of defense).  We stocked cages with 4 plants of one defense type, either 8, 16, 32, or 48 caterpillars, and either 0 or 1 adult predator.  After 4 days, we searched cages and recorded the number and condition of surviving herbivores as well as the damage that the plants had suffered. 

Results/Conclusions

Plant defenses increased predation rates, which in turn decreased the amount of damage that plants suffered, according to significant paths in a structural equation model (SEM).  A non-significant SEM path between predation and plant damage indicated that predators affected plant damage through consumptive but not non-consumptive effects.  Per-capita predation (kills/predator/herbivore) decreased with increasing herbivore density when plants were highly defended, but it was not density-dependent when defenses were low. As a result, higher levels of plant defense increased predation much more at the lowest herbivore density (71% increase) compared to the highest herbivore density (18% increase).  We hypothesize that increased rates of contact between predators and prey and compensatory feeding by predators were responsible for strengthened predator-prey interactions on highly defended plants.  We base this conclusion on the finding that leaf damage was more dispersed on highly defended plants, which suggests increased herbivore movement, and the fact that plant defenses caused herbivores to develop more slowly and to feed less, which lowered their quality as prey.  Our results indicate that induced plant defenses can affect the strength and density dependence of predation, and that this interaction has implications not only for herbivore population dynamics, but for plant productivity as well.