Quantifying Nitrogen Use Efficiency in Wheat Using High-Precision Phenotyping

Abstract

Modern cereal cultivars rely heavily on nitrogenous inputs to reach their yield potential. However, the nitrogen use efficiency (NUE) of wheat is poor and the recovery of applied nitrogen (N) in cereal production is low, between 30-50 %. This inefficiency results in the N pollution of natural ecosystems, and an economic loss to producers. Improving NUE in wheat (Triticum aestivum L.) has so far proved difficult due to the complexity of NUE and a lack of phenotyping resolution to identify superior NUE genotypes. Three knowledge gaps are addressed in this research. Firstly, the ability of high throughput phenotyping (HTP) to help us define the NUE of Australian wheat cultivars into differences that are apparent across a range of environments and managements. Central to this is the interaction between N and water, one of the major environmental determinants of N uptake. Secondly, although shoot tissue is one of the main reservoirs of plant N, and essential to understanding N uptake and utilisation, knowledge regarding tissue N response to changing N availability is limited. These processes are known to be dynamic and must be observed temporally in order to differentiate their responses to changes in N availability. Lastly, can a combination of these two phenotyping methods help explain plant responses to variable nutrient supply? Growth responsiveness to N supply in a selection of bread wheat cultivars with varying water provision was measured using HTP. Cultivar differences were discovered in their ability to increase shoot area in response to N, in absolute growth rate response to changes in water availability as well as the ability to convert this into yield. In order to differentiate N uptake in real time, a hyperspectral reflectance method utilising a field spectrophotometer and leaf clip was adapted to Australian bread wheats using partial least squares regression. The robustness of the method was established by regressing tissue N analysis with reflectance spectra readings, giving an R2 of the predictions at 0.83. The sensitivity of the method was determined to detect changes in leaf N % in a hydroponics system with alternating high/low N availability. The cultivars responded to the change in N by readjusting their leaf-N content to an equivalent steady-state N level within two days. The final part of this project was to incorporate HTP and the hyperspectral leaf N measurements to determine how wheat growth and N uptake responded to split applications of N. When N was added at stem elongation and booting growth stages, the plants delayed their point of maximal shoot area by six days, and increases in leaf N concentration were observed the day after application. The increases in N harvest index and the grain protein content found at destructive harvest were linked to growth and leaf N concentration differences during the experiment. Overall, the research presented here has measured NUE and cultivar differences repeatedly and with high resolution. These protocols show promise for the selection of improved NUE phenotypes, which could be combined with forward genetics to differentiate NUE and its component processes and identify the underlying genetic control.