Acceleration or deceleration of litter decomposition by herbivory depends on nutrient availability through intraspecific differences in induced plant resistance traits

• Published in: The Journal of ecology 2018-11, Vol.106 (6), p.2380-2394
• Main Authors: Burghardt, Karin T., Bradford, Mark A., Schmitz, Oswald J.
• Format: Article
• Language: English
• Subjects: Acceleration; Availability; community genetics; Deceleration; Decomposition; Ecosystems; Environmental gradient; Fertility; Genotypes; goldenrod; Growth rate; Herbivores; Herbivory; Heterogeneity; induced defenses; Inoculum; intraspecific trait variation; Leaf litter; Leaves; microbes; Microcosms; Microorganisms; Mineral nutrients; Nutrient availability; Nutrient content; Nutrient cycles; Nutrient loss; Organic chemistry; Palatability; Patchiness; Plant resistance; Plant-herbivore interactions; Soil; Soil conditions; Soil fertility; Soil nutrients; Soils; Spatial heterogeneity; Spatial variations; Species; Substrates; Tissue
• ISSN: 0022-0477; 1365-2745
• DOI: 10.1111/1365-2745.13002

1. Herbivores often induce changes in plant defensive chemistry or nutrient content that may respectively inhibit or promote microbial decomposition of senesced litter. Often the directional impact of herbivores on decomposition is considered to be a property of a species or ecosystem. While rarely explored, plasticity in the induction of defensive strategies across environmental gradients may also result in divergent impacts of herbivores on decomposition (deceleration vs. acceleration). 2. Here, we examined how soil nutrient conditions determine legacy effects of herbivory, using nine goldenrod genotypes grown across four levels of nutrient supply and with or without grasshopper herbivory. In this species, herbivory induces defensive traits in genotypes grown in high nutrient conditions but induces tolerance (compensatory growth) in low nutrient conditions. We combined senesced litter from each treatment with a common soil inoculum in microcosms and measured soil respiration and litter mass loss over 100 days as estimates of decomposition. 3. Plant genotype, nutrient environment, and herbivory each altered decomposition. The legacy effect of herbivory overwhelmed the positive effect of high soil nutrient supply on decomposition. This significant herbivory × nutrient environment interaction meant that herbivore-induced plants grown in high nutrient environments produced litter that was more resistant to microbial breakdown than litter from the same genotype not exposed to herbivores. But the opposite occurred at low nutrient levels where litter from herbivore-induced plants was most readily decomposed. Furthermore, we mechanistically tie treatment-induced changes in leaf traits to decomposition rates. Lastly, we demonstrate a significant correlation between herbivore growth rates on living tissue and decomposition efficiency by the microbial community of senesced tissue, suggesting that herbivores and microbes perceive the "quality" of the induced substrate similarly. 4. Synthesis. Herbivore-induced changes in leaf palatability and trait expression due to defense induction or compensatory growth can cascade through to either inhibit or promote the decomposability of leaf litter within a single species. Here, the key determinant of which occurs is the soil fertility environment of the goldenrod clone. These findings offer mechanistic understanding of how spatial heterogeneity in ecosystem process rates may be generated by spatial variation in herbivory and