Corn Tiller Yield Contributions are Dependent on Environment: A Corn Tiller Yield Contributions are Dependent on Environment: A 17 Site-Year Kansas Study 17 Site-Year Kansas Study

Summary Historic breeding efforts in corn ( Zea mays L.) have resulted in uniform, single-stalked phenotypes with limited potential for environmental plasticity. Therefore, plant density is a critical yield component for corn, as corn is unable to successfully compensate for a deficit of plants. Other grass crop species can overcome plant density deficits via vegetative branching (tillering), but this trait is historically undesirable in corn. Improving corn flexibility across plant densities has potential benefits, particularly considering diverse yield environments and seasonal weather uncertainties due to climate change. The present study evaluated tiller presence with two hybrids in a range of plant densities across the state of Kansas to identify yield impacts and potential usefulness of this plasticity trait in corn. Tiller presence was identified as neutral or additive to final yields, but fine-tuning plant density was confirmed as key to maximizing grain yields. Tillers have potential to stabilize yields across plant densities in productive environments. This capability may offer a source of production stability for growers when deficits develop in plant density after planting.


Introduction
Plant density is a management strategy to optimize the balance between crop needs and resource availability (Laitinen and Nikoloski, 2019). Specifically in corn (Zea mays L.), optimal plant density has historically increased as a key driver of modern yield gains (Duvick et al., 2004). Crop plasticity, the ability of a genotype to express alternative phenotypes and adapt to contrasting environmental scenarios, is marginal in corn compared to other cultivated crops. Due to this comparatively lower plasticity, corn yields are notably dependent on plant density. This attribute is less desirable in challenging or otherwise unpredictable growing conditions (Mylonas et al., 2020).

Kansas Field Research 2022
Kansas State University Agricultural Experiment Station and Cooperative Extension Service Corn adjusts its final grain production via yield components (namely ears per area, kernels per ear, and weight per kernel). The corn yield component most easily altered via management practices is ears per area, which is adapted with plant density and prolific (multi-eared) hybrid selection. Other Poacea species, such as wheat (Triticum aestivum L.) and grain sorghum (Sorghum bicolor L. Moench), increase the number of inflorescences per area by producing additional vegetative shoots (tillers). Although genetically different from its more grass-like ancestor (Zea mays ssp. parviglumis), corn remains capable of producing tillers. This trait is often suppressed in modern hybrids (Moulia et al., 1999). Corn tillers that do appear may remain vegetative, produce harvestable grain, or develop abnormal inflorescences without harvestable grain ("tassel ears"). Due to these unpredictable, undesirable outcomes, corn tillers have been historically associated with yield reductions. For this reason, corn tillers are commonly referred to as "suckers" (Jenkins, 1941).
Growers commonly voice concerns about tiller presence in corn fields, and conclusive evidence on tillering impacts (particularly in Kansas) is lacking. Therefore, this multiseason study sought to understand 1) the impact of tiller expression on yield in varying environments, and 2) the potential of tillers as a plasticity trait in Kansas environments.

Procedures
Data presented in this report were collected during a multi-year statewide study (2019)(2020)(2021). Location characterizations are provided in Table 1.
Twelve site-years were established with a split-split-plot design, evaluating three factors: whole plot of planting density with three levels (10000, 17000, and 24000 plants/a), sub-plot of hybrid with two levels (P0805AM and P0657AM), and sub-sub-plot of tiller presence with two levels (removal at the V10 [tenth-leaf; Ritchie et al., 1997] development stage [TR], or intact throughout the season [TI]; Table 1). The remaining five site-years were established without the tiller presence factor (Table 1). In total, seventeen site-years were evaluated with at least three replications each.
Grain yields were harvested from the two central plot rows and adjusted to 15.5% standard grain moisture. Sites were clustered into three yield environments (low-, moderate-, and high-yielding; LYE, MYE, and HYE respectively) via a k-means algorithm. A linear mixed effects model was fit with grain yield as the response considering 1) fixed effects of treatment factors interacting with yield environment, and 2) random effects of site-year and design factors. The fitted model was subjected to a 3-way analysis of variance (ANOVA) and subsequent means comparison (Tukey method). A second linear mixed effects model was fit with grain yield as the response considering 1) fixed effects of observed plant density, observed tiller density, and interactions with yield environment; and 2) random effects of site-year and design factors. Predictions were generated with model coefficients based on the range of density observations across trials. All analyses and figures were generated with the R software (R Core Team, 2021).

Results
The ANOVA results for the treatment factor model are shown in Table 2. The interaction of yield environment with both plant density (P ≤ 0.001) and tiller presence (P ≤ 0.01) impacted final yields. Subsequent means comparisons are shown by yield environment in Figure 1. Plant density thresholds for grain yields within the evaluated ranges were 10,000 plants/a in the LYE, 17,000 plants/a in the MYE, and 24,000 plants/a in the HYE. Tiller presence did not reduce yields in any environment, instead tillers increased the overall yields in the HYE.
The ANOVA results for the observational analysis are shown in Table 2. The interaction of yield environment with both observed plant density and observed tiller density impacted yield predictions, in addition to the triple interaction of environment and observed densities (all significant at P ≤ 0.001). Plotted predictions are shown by yield environment in Figure 2. Overall yields both with and without tillers were stable across observed plant densities in the LYE. Overall yields were more stable with greater tiller densities across observed plant densities in the MYE and HYE. Regardless of yield environment, greatest yields were realized when plant density was optimized, minimizing tiller expression.
The results of this study support the hypotheses that 1) tiller presence alone does not reduce corn yields across environments; and 2) tillers are an indication of plant density deficits but can be useful in stabilizing these deficits in productive environments. Additional information on this study can be found in Veenstra et al. (2021).  Tested source of variation (Source), degrees of freedom (df), degrees of freedom of residuals (Residual df), F value, and the associated p value significance are presented. All sources with P-values ≤ 0.05 are shown in boldface font. Coefficient of determination values are provided. *** Significant at P ≤ 0.001. ** Significant at P ≤ 0.01, ns not significant.  Table 2 (a, plant density; b, tiller presence) by yield environment (LYE, low-yielding environment; MYE, moderate-yielding environment; HYE, high-yielding environment). Means within a panel not sharing a common letter are significant at the 0.05 probability level.