Breeding in winter wheat (Triticum aestivum L.) can be further progressed by targeting previously neglected competitive traits

Abstract

Breeders work to adapt winter wheat genotypes for high planting densities to pursue sustainable intensification and maximize canopy productivity. Although the effects of plant-plant competition at high planting density have been extensively reported, the quantitative relationship between competitiveness and plant performance remains unclear. In this study, we introduced a shoot competitiveness index (SCI) to quantify the competitiveness of genotypes and examined the dynamics of nine competitiveness-related traits in 200 winter wheat genotypes grown in heterogeneous canopies at two planting densities. Higher planting densities increased shoot length but reduced biomass, tiller numbers, and leaf mass per area (LMA), with trait plasticity showing at least 41% variation between genotypes. Surprisingly, genotypes with higher LMA at low density exhibited greater decreases under high density, challenging expectations from game theory. Regression analysis identified tiller number, LMA, and shoot length as key traits influencing performance under high density. Contrary to our hypothesis, early competitiveness did not guarantee sustained performance, revealing the dynamic nature of plant-plant competition. Our evaluation of breeding progress across the panel revealed a declining trend in SCI (R² = 0.61), aligning with the breeding objective of reducing plant height to reduce individual competitiveness and increase the plant-plant cooperation. The absence of historical trends in functional traits and their plasticities, such as tiller number and LMA, suggests their potential for designing ideal trait-plasticity for plant-plant cooperation and further crop improvement.

Keywords: breeding progress; canopy productivity; intergenotypic competition in plant; phenotypic plasticity; plant-plant competition; plant-plant interaction; planting density.

Figure 1

Effects of planting density on functional traits in winter wheat. Each data point represents the estimated marginal means of a genotype that were grown in low (T1) and high (T2) planting density at the booting stage. Plasticity was calculated as the percentage deviation of T2 trait values from T1. Traits shown include shoot biomass (A–C), shoot length (D–F), total tiller number (G–I), total green leaves on the main stem (J–L), and leaf mass per area [LMA, (M–O)]. Boxplots of trait distributions in T1 and T2, with solid lines showing genotype responses (A, D, G, J, M) and scatter plots of trait values in T1 vs. T2, with fitted linear regression lines, formulas, adjusted R², and p-values, with dashed lines indicating 1:1 relationship (B, E, H, K, N). Comparison between plasticity and T1 trait values, with fitted linear regression lines, formulas, adjusted R², and p-values, with dashed lines marking zero plasticity (C, F, I, L, O).

Figure 2

Pearson correlation coefficients between functional traits and shoot competitiveness index (SCI). (A) Trait values at low density at booting stage; (B) trait values at high density at booting stage; (C) plasticity of traits at booting stage; and (D) trait values at high density at anthesis. The correlation coefficient was calculated between pairs of traits in the 200 studied genotypes. The absolute growth was defined as the difference in shoot dry mass between booting stage and anthesis. X indicates insignificant correlation.

Figure 3

Relationship between shoot biomass at anthesis (H2) and shoot competitiveness index (SCI) at booting. Each point represents marginal mean of a genotype. The regression is shown as solid blue line. A SCI greater than 0 indicates higher competitiveness compared to neighboring plants.

Figure 4

Breeding progress of shoot biomass (B), shoot length (D), total tiller (F), leaf mass per area (LMA, H) and their plasticity (A, C, E, G) between low (T1) and high (T2) planting density. Using a sliding-window approach, each data point represents the mean values of a subset group of 10 genotypes, with the shaded area indicating the standard deviation. The black line represents the linear regression with the formula and the adjusted R² reflecting absolute breeding progress.

Figure 5

Breeding progress of shoot competitiveness index (SCI). Using a sliding-window approach, each data point represents the mean SCI of a subset group of 10 genotypes, with the shaded area indicating the standard deviation. The black line represents the linear regression with the formula and the adjusted R² reflecting absolute breeding progress.