The roots of defense: plant resistance and tolerance to belowground herbivory.

Abstract

There is conclusive evidence that there are fitness costs of plant defense and that herbivores can drive selection for defense. However, most work has focused on above-ground interactions, even though belowground herbivory may have greater impacts on individual plants than above-ground herbivory. Given the role of belowground plant structures in resource acquisition and storage, research on belowground herbivores has much to contribute to theories on the evolution of plant defense. Pocket gophers (Geomyidae) provide an excellent opportunity to study root herbivory. These subterranean rodents spend their entire lives belowground and specialize on consuming belowground plant parts. We compared the root defenses of native forbs from mainland populations (with a history of gopher herbivory) to island populations (free from gophers for up to 500,000 years). Defense includes both resistance against herbivores and tolerance of herbivore damage. We used three approaches to compare these traits in island and mainland populations of two native California forbs: 1) Eschscholzia californica populations were assayed to compare alkaloid deterrents, 2) captive gophers were used to test the palatability of E. californica roots and 3) simulated root herbivory assessed tolerance to root damage in Deinandra fasciculata and E. californica. Mainland forms of E. californica contained 2.5 times greater concentration of alkaloids and were less palatable to gophers than island forms. Mainland forms of D. fasciculata and, to a lesser extent, E. californica were also more tolerant of root damage than island conspecifics. Interestingly, undamaged island individuals of D. fasciculata produced significantly more fruit than either damaged or undamaged mainland individuals. These results suggest that mainland plants are effective at deterring and tolerating pocket gopher herbivory. Results also suggest that both forms of defense are costly to fitness and thus reduced in the absence of the putative target herbivore.