The Dynamics of Carbon–Nutrient Balance: Effects of Cottonwood Acclimation to Short- and Long-Term Shade on Beetle Feeding Preferences

Abstract

The carbon–nutrient balance hypothesis (CNBH) predicts that shading should increase leaf palatability to herbivores by decreasing concentrations of carbon (C) -based chemical defenses and increasing nitrogen (N). We measured cottonwood (Populus deltoids) growth, leaf chemistry, and beetle (Plagiodera versicolora) feeding preferences on saplings grown in either continuous high (HH) or low (LL) light, and saplings switched from high to low (HL) or low to high (LH) light for nine days. As expected, based on the CNBH, shading increased total N and decreased total phenol glycoside (C-based secondary metabolites) concentrations in plants from all shade treatments (LL, HL, and LH), relative to HH plants, with plant growth and gross leaf chemistry being affected by initial and final light regime. In contrast, while specific phenol glycoside concentrations were affected by the initial and final light regime, they also showed an initial × final light interaction. Beetles tended to prefer LL to HH plants. Beetles unexpectedly preferred HH to either HL or LH switched plants, most likely because high concentrations of a specific phenol glycoside – salicin – occurred in both switched treatments and inhibited beetle feeding. Plant chemical allocation during light acclimation led to unpredictable changes in specific C-based compounds, even though plant growth and gross chemistry conformed to expectations for shading effects and the CNBH. The response of this herbivore to altered concentrations of a specific compound confounded predictions based on average dynamics of suites of chemicals. Our findings may help explain why relationships between light availability and herbivory in field studies, where light varies on many time scales, can differ from those predicted by the CNBH. Understanding both dynamic plant chemical responses to altered resource availability and controls over allocation to specific compounds would likely enhance future predictability of specific environment-plant-herbivore interactions.