Consequences of plant density and natural versus simulated herbivory for common milkweed

Background/Question/Methods

Common milkweed, Asclepias syriaca, is a relevant study organism for two reasons: (1) it is the preferred host plant for the monarch butterfly, Danaus plexippus and (2) it exhibits inducible defensive responses to herbivory. These defenses include the production of trichomes, toxic cardiac glycosides (cardenolides), and latex. As the host plant for monarchs, a better understanding of the ecology of milkweed is increasingly significant given recent declines in monarch populations. Previous studies have demonstrated that an escalation in defensive compounds in milkweed has deleterious effects on herbivore (monarch) fitness. To gain a better understanding of the drivers of escalation in defenses we tested two hypotheses: (1) increased plant density results in competition which lowers plant defenses to herbivory and (2) Natural and simulated herbivory should induce the same defensive responses in milkweed as long as the simulated herbivory is done in the same relative amount and to the same plant tissue as natural herbivory. Our study consisted of common milkweed grown at low and high densities, which were then subjected to natural herbivory by monarch caterpillars or simulated herbivory done by ripping of foliar tissue. Plant size and defenses were sampled three times within a 72 hour timeframe.

Results/Conclusions

Our density treatments showed that competition has significant effects on all measures of size. As expected, plants in the low-density treatments were consistently larger than those in the high-density treatment. For example, leaf area in the low-density was 44.47± 2.84 cm² versus the high-density which was 27.81± 2.57 cm² (P = 0.003). Height and basal stem diameter showed similar results. Our preliminary results also indicate that there is an interaction between plant density and sampling time on total cardenolide concentration in foliar tissue. Plants in the high-density treatment had higher cardenolide concentrations than those in the low-density treatment after 24 and 48 hours, but this pattern reversed at 72 hours (P = 0.04). We also found that simulated herbivory induced significantly higher cardenolide concentrations than natural herbivory across both density treatments after 72 hours: Simulated = 7.20 ± 0.31, Caterpillar = 6.44 ± 0.31 μg mg^-1 (P = 0.04). This differs from what other studies have found likely because our methods for simulating herbivory mimicked both the amount of tissue damaged as well as the type of tissue damaged.