Investigating the context-dependent effects of plant mutualists on plant and herbivore performance in an agricultural legume.

摘要

Nutritional symbioses between plants and plant-root mutualists are not only important players in nutrient acquisition by plants but contribute to nutrient cycling, biological community composition and even plant defense elicitation. Furthermore, several studies have purported that anthropogenic nutrient loading may disrupt these symbioses by effectively removing plant dependence on symbionts for nutrient acquisition, while some have evidenced that the presence of another symbiont may ameliorate the negative effects of nutrient loading on said first symbiont. In addition, less is known on how nutrient loading effects may influence higher trophic levels, such as insect herbivores. This study investigated how arbuscular mycorrhizal fungi (AMF) and rhizobacteria interact under nitrogen or phosphorus amendment to influence each other's abundance, plant growth and defense traits, and herbivore performance using a factorial greenhouse experiment and herbivory assay. Results illustrated no support for the negative influence of nutrient loading on symbiont abundance and fitness, nor, as a corollary of that, evidence of ameliorating effects of one symbiont on the other in dual inoculation treatments. In addition, there was evidence of nodulation in the absence of rhizobial inoculation, which calls into question the use of nodule counting as a metric of rhizobial abundance and fitness. In addition, the insignificant response of alfalfa across nutrient and symbiont treatments for several response variables may illustrate that the domestication of alfalfa has reduced its dependence on belowground mutualists for nutrient acquisition, such as the fact that plant phosphorus and protein content were found not to be dependent upon symbiont treatment. There was also no influence of nutrient and symbiont treatments on herbivore performance, despite significant differences in foliar defensive chemistry across symbiont and/or nutrient treatments. These results do not support nutrient loading hypotheses, and thus highlight the need to consider species identities as an influence in interaction outcomes.