Warming and drought differentially influence the production and resorption of elemental and metabolic nitrogen pools in Quercus rubra.

The process of nutrient retranslocation from plant leaves during senescence subsequently affects both plant growth and soil nutrient cycling; changes in either of these could potentially feed back to climate change. Although elemental nutrient resorption has been shown to respond modestly to temperature and precipitation, we know remarkably little about the influence of increasing intensities of drought and warming on the resorption of different classes of plant metabolites. We studied the effect of warming and altered precipitation on the production and resorption of metabolites in Quercus rubra. The combination of warming and drought produced a higher abundance of compounds that can help to mitigate climatic stress by functioning as osmoregulators and antioxidants, including important intermediaries of the tricarboxylic acid (TCA) cycle, amino acids including proline and citrulline, and polyamines such as putrescine. Resorption efficiencies (REs) of extractable metabolites surprisingly had opposite responses to drought and warming; drought treatments generally increased RE of metabolites compared to ambient and wet treatments, while warming decreased RE. However, RE of total N differed markedly from that of extractable metabolites such as amino acids; for instance, droughted plants resorbed a smaller fraction of elemental N from their leaves than plants exposed to the ambient control. In contrast, plants in drought treatment resorbed amino acids more efficiently (>90%) than those in ambient (65-77%) or wet (42-58%) treatments. Across the climate treatments, the RE of elemental N correlated negatively with tissue tannin concentration, indicating that polyphenols produced in leaves under climatic stress could interfere with N resorption. Thus, senesced leaves from drier conditions might have a lower nutritive value to soil heterotrophs during the initial stages of litter decomposition despite a higher elemental N content of these tissues. Our results suggest that N resorption may be controlled not only by plant demand, but also by climatic influences on the production and resorption of plant metabolites. As climate-carbon models incorporate increasingly sophisticated nutrient cycles, these results highlight the need to adequately understand plant physiological responses to climatic variables.