Nitrogen's Role in Changing Kernel Weight in Maize: Relevant Physiological Mechanisms During Reproductive Stages

Although grain yield (GY) in maize (Zea mays L.) is the product of both kernel number (KN) per unit area and kernel weight (KW), the latter has historically been considered the least variable component. However, as a result of sink strength enhancement by genetic improvement, KW in modern genotypes has become more responsive to changes in environmental and management conditions. The prospect of KW becoming a more important driver behind GY variability warrants embarking on more intensive research into the physiological mechanisms underlying when post-flowering stress conditions can limit KW. In pursuit of that goal, this dissertation evaluated the effects of N availability on: 1) the sources of dry matter (DM) and N assimilates for the growing kernels during the reproductive period; 2) the determination of potential KW (i.e., potential kernel sink capacity) during the lag phase of grain filling; and 3) the realization of final KW (i.e., actual sink capacity) during the linear phase of grain filling.
To investigate how N availability affected source capacity during reproductive growth, a 2-year field study that combined N timing and N rate treatments (under a common density of 8.3 plants m^-2) was conducted at the Purdue Rice Farm (LaCrosse, IN) in 2016 and 2017. The timing applications included: all N applied at planting, split application between planting and early sidedress, and split application between planting and V12 (for the last 56 kg N ha^-1). The five N rates tested were 0, 112, 168, 224, and 280 kg N ha^-1. Biomass samples of plant components at R1, R3, and R6 enabled DM and N sources (i.e., post-silking DM production, post-silking N uptake, DM and N remobilization) to be calculated separately for the two main grain-filling phases: from R1 to R3 (i.e., lag phase) and from R3 to R6 (i.e., linear phase). In both seasons, lag-phase DM production (PostDM_R1.R3) was much less responsive to N rates than that during the linear-phase (PostDM_R3.R6). In the lag phase, substantial DM gains in leaves and stems occurred in both years, while N content gains were mostly detected in reproductive tissues. Differential seasonal patterns in post-silking N uptake were observed, with plants either achieving net above-ground N content gains only during the lag phase (PostN_R1.R3) in 2016, or during the linear phase (PostN_R3.R6) in 2017). Both GY and KW were gradually increased by N supply, with reproductive tissues proving to be relatively stronger sinks for N than for DM during the lag phase.
To understand how N availability affected the determination of potential KW, three field experiments testing N rates, plant densities and N timing applications on a single commercial hybrid (DKC63-60) were conducted over a 3-year period. The second season (2017) of the previously described study was used as Experiment 1, considering the original three N timings and a sub-set of N rates (0, 112, and 224 kg N ha^-1). Experiments 2 (2018) and 3 (2019) were conducted at the Purdue Agronomy Center of Research and Education (West Lafayette, IN) and each one involved four N rates (0, 84, 168, and 224 kg N ha^-1, all applied at planting) and two plant densities (7.9 and 10.4 plants m^-2). Endosperm cell number (ECN), an indicator of potential KW, was determined at different timings (from 9 to 17 days after silking -DAS-). Biomass samples taken at V12, R1 and R3 enabled calculations of plant and ear growth and N accumulation rates were calculated. High GY (15.7-16.6 Mg ha^-1) were achieved at maximum N in the first two years (2017-2018), and average GY increased 8.1 Mg ha^-1 in response to N (i.e., from 0N to 224N) over the 3-year period. In addition, GY variability was largely explained by KW in all experiments. Low N treatments consistently reduced ECN at 9, 10, 13, and 17 DAS. Final KW responses to N rates were always explained by ECN, though the strength of the relation changed with the experiment and the relative DAS sampling time. For each sampling date, ECN was highly correlated with ear N allocation rate during the lag phase. The N rate effects on potential KW were not associated with plant growth rate per kernel during the critical period bracketing silking, and a positive relationship (rather than a trade-off) was found between KW and KN. We concluded that N played a direct role in potential KW determination as this process seemed to be highly dependent on N assimilates. Under higher plant N availability, individual kernel sink capacity was found to increase via gains in ECN (independently of N timing and plant density).
To investigate how N availability affected the realization of final KW at maturity, intensive data were collected during the linear grain-filling period from same three field experiments. Since kernels accumulate both DM and N assimilates during the linear phase, kernels were removed from ears of all treatments on a weekly basis over the entire 9-10-week interval from the early R3 stage. Linear plateau models were then fitted to the resulting kernel DM and N values on a thermal time basis to obtain characterizing parameters. Increases in N supply, regardless of N application timing or plant density, improved final DM accumulation in kernels by either increasing both the effective grain-filling rate (EGFR) and grain filling duration (GFD), or by increasing GFD alone. Kernel N content (KNC) increased consistently under higher N availability because of gains in both kernel N accumulation rate (KNAR) and duration (KNAD). Kernels actively accumulated N until late in the season, as shown by the similar durations to peak DM and peak KNC (averaging ~1140ºCd^-1 for GFD and ~1120ºCd^-1 for KNAD). While EGFR was less impacted by N rate differences, KNAR was much more responsive, showing a strong correlation with final KW (r=0.96). Furthermore, KNAR was highly correlated with whole-plant N uptake by R3 (PNU_R3) (r=0.80). We concluded that kernel DM accumulation during the linear phase, and therefore final KW, was limited by N assimilate remobilization to kernels from N reserves accumulated prior to R3.
Overall, this dissertation provides new evidences for the distinct indirect and direct roles that N plays in the physiological mechanisms that determine final KW in maize at high potential GY. Indirectly, leaf N is responsible for post-silking photosynthesis, the major source of carbohydrates for the kernels, while stem N works as a reserve buffer to delay leaf senescence resulting from an early N remobilization to developing kernels. Directly, N assimilates are highly demanded by reproductive tissues during the lag phase to fulfill the endosperm cell division requirements that establish potential KW. Similarly, during the linear phase, N assimilate availability may limit DM deposition in kernels as differences in final KW were strongly associated with the kernel N accumulation rate.