Changes in host plant quality, including foliar amino acid concentrations, resulting from global climate change and attack from multiple herbivores, have the potential to modify the pest status of insect herbivores. This study investigated how mechanically simulated root herbivory of lucerne (Medicago sativa) before and after aphid infestation affected the pea aphid (Acyrthosiphon pisum) under elevated temperature (eT) and carbon dioxide concentrations (eCO2). eT increased plant height and biomass, and eCO2 decreased root C:N. Foliar amino acid concentrations and aphid numbers increased in response to eCO2, but only at ambient temperatures, demonstrating the ability of eT to negate the effects of eCO2. Root damage reduced aboveground biomass, height, and root %N, and increased root %C and C:N, most probably via decreased biological nitrogen fixation. Total foliar amino acid concentrations and aphid colonization success were higher in plants with roots cut early (before aphid arrival) than those with roots cut late (after aphid arrival); however, this effect was counteracted by eT. These results demonstrate the importance of amino acid concentrations for aphids and identify individual amino acids as being potential factors underpinning aphid responses to eT, eCO2, and root damage in lucerne. Incorporating trophic complexity and multiple climatic factors into plant–herbivore studies enables greater insight into how plants and insects will interact in the future, with implications for sustainable pest control and future crop security.
The importance of phosphorus (P) in regulating ecosystem responses to climate change has fostered P-cycle implementation in land surface models, but their CO 2 effects predictions have not been evaluated against measurements. Here, we perform a data-driven model evaluation where simulations of eight widely used P-enabled models were confronted with observations from a long-term free-air CO 2 enrichment experiment in a mature, P-limited Eucalyptus forest. We show that most models predicted the correct sign and magnitude of the CO 2 effect on ecosystem carbon (C) sequestration, but they generally overestimated the effects on plant C uptake and growth. We identify leaf-to-canopy scaling of photosynthesis, plant tissue stoichiometry, plant belowground C allocation, and the subsequent consequences for plant-microbial interaction as key areas in which models of ecosystem C-P interaction can be improved. Together, this data-model intercomparison reveals data-driven insights into the performance and functionality of P-enabled models and adds to the existing evidence that the global CO 2 -driven carbon sink is overestimated by models.