Crop and nutrient harvest indices for spring wheat genotypes grown with different fertiliser and carbon dioxide levels, under field and controlled environments : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Abstract

Historically, genetic gains in wheat (Triticum aestivum L.) yield have been attributed to breeding advances which resulted in increased crop harvest index (CHI: the ratio between harvested grain and total above-ground biomass), and improved agronomic practices, such as better nitrogen (N) fertiliser management. In this study, two experiments were carried out in the field in 2017—2018 (Experiment 1), and in a glasshouse in 2018—2019 (Experiment 2). These were used to quantify the CHI, nutrient harvest indices (nitrogen, NHI and other nutrients, NuHIs) and N use efficiency (NUE: ratio between grain yield and amount of N supplied)) for six spring wheat genotypes (‘Discovery’, ‘Duchess’ ‘Reliance’, PFR-2021, PFR-3019 & PFR-3026) grown at low and optimum N fertiliser supply. Then two controlled environment experiments (3 and 4) were used to quantify the CHI, NHI and NuHIs for ‘Discovery’ under ambient and elevated carbon dioxide (CO₂) (aC0₂ and eCO₂; respectively), at low and optimum phosphorus (P) and potassium (K) fertiliser supply. The aims of this study was to investigate the influence of fertiliser supply and growth environment on crop growth, nutrient accumulation, partitioning and harvest indices of spring wheat genotypes. Overall, the CHI values depended on carbon remobilisation to the grain, but there was no relationship between CHI and grain yield. ‘Duchess’ had a lower CHI (0.33±0.04) than the average of the other genotypes (0.56±0.01; 0.44±0.04) in Experiments 1 and 2, respectively. This was explained by its lower translocation of straw biomass to the grain component, as reflected in its low thousand-grain weight (TGW, g) and high screenings. The lower overall CHI in Experiment 2 (∼0.42±0.03), compared with 0.50—0.55 in Experiments 1, 3 and 4, could have been caused by high temperature episodes near anthesis, which resulted in reduced number of grains per unit area (grain density). The CHI was also reduced by P deficiency, which reduced the area of individual leaves (carbon source) and number of fertile tillers (carbon sink), hence the low above-ground biomass (AGB) and grain yield. In all four experiments, the relationship between TGW and grain density, showed that the highest yielding genotype ‘Discovery’ was positive and above the regression line. This higher TGW for ‘Discovery’ was due to greater total AGB, because it remobilised the same proportion of total carbon to the grains as the other genotypes. These results confirm that the relationship between TGW and grain density can be used to explain yield differences. The greater AGB and grain yield in Experiments 1 and 2, for ‘Discovery’ resulted from the fastest pre-anthesis leaf area expansion rate, higher maximum green leaf area index (GLAI) and longer GLA duration above the critical GLAI during the grain-filling period, which resulted in greater intercepted radiation. In contrast, all genotypes had a conservative specific leaf N (SLN) content above the critical threshold of 1.1 g N/m², which meant their photosynthetic capacity and therefore the radiation use efficiency were not different. Furthermore, in Experiments 3 and 4, the optimally fertilised crops had larger flag leaf area and more fertile tillers per tube, thus a higher leaf area that lead to higher intercepted radiation. The AGB for these crops increased by 11—23% with eCO₂ and grain yield increased by 6—14%. The AGB decreased by ∼90.0%, from 57.4±1.28 g/tube for the optimally fertilised crops to 5.29 g/tube for the P deficient crops. The NHI (ratio between N accumulated in the grain and N accumulated in the AGB at harvest maturity) differences among the genotypes were small, ≤ 6.40% in Experiments 1 and 2, and lower for ‘Discovery’ and ‘Duchess’, compared with the other genotypes. However, NHI was severely reduced (37%) by P deficiency compared with the optimally fertilised crops. The NuHIs (ratio between nutrients accumulated in the grain and nutrients accumulated in the AGB at harvest maturity) differed among the genotype across the experiments, but were inconsistent among genotypes in the different environments. There was no relationship between NuHIs and the proportion of nutrients at anthesis. Therefore, individual NuHIs were a function of remobilisation efficiency, rather than timing of nutrient uptake. High NHI and PHI across the environments, were due to their efficient translocation from the vegetative to the grain component, while the low CaHI and KHI showed these nutrients were not readily translocated to the grain, with 60—100% of Ca, K, N and P having been accumulated by anthesis. There was a strong, positive relationship between grain N concentration and grain sulphur (S) and zinc (Zn) concentration, and a negative relationship to grain K concentration in Experiments 1 and 2. This has implications for human health, with increased concentration of S and Zn, being a positive result. In Experiments 1 and 2, NUE differed among the genotypes at optimum N fertiliser supply, higher for ‘Discovery’ compared with ‘Duchess’ and ‘Reliance’. Similar NUE at low N fertilier supply shows that the selected genotypes had no differentiating traits that could be used for breeding, specific to low N fertility conditions. This was despite fact that these genotypes were recommended by wheat breeders on the basis that they had a range of attributes that could enhance NUE. In Experiment 1, NUE was explained by the N uptake efficiency (NupE) at both N fertiliser rates, while in Experiment 2, NUE was explained by the N utilisation efficiency (NutE). These inconsistent results meant that NUE was not a stable trait, and therefore is limited as a criterion for future breeding selection. The contribution of this research to future breeding was premised on the confirmation that CHI has plateaued at ∼0.50. Results suggest that further grain yield increases must come from genetic enhancements that either increase total AGB at the current CHI levels, or with increased CHI, as some genotypes had higher CHI values of up to 0.59. The effects of N and P fertiliser supply on CHI, NHI and NuHI highlights the importance of fertiliser management on wheat production, and could be used alongside breeding to increase grain yield. At the agronomic level, the insights from this research can improve our understanding of the importance of P fertiliser on yield. At the research level, such understanding shows the physiological mechanisms to inform breeding and improve predictive models of wheat production. In this study, CHI did not respond to increasing CO₂ level. However, response of CHI to increasing CO₂ level has been reported, under water or N stress. Therefore, the interaction of P fertiliser supply, water and/ or N stress needs to be investigated in future.