Two on-farm field experiments, one with 12 short maturity maize hybrids (from 100 to 105-day) and one with 11 long maturity maize hybrids (from 110 to 115-day), were conducted in central Iowa, USA, in 2021 and 2022, We evaluated in total 23 Bayer Crop Science maize hybrids commercialized from 1980 to 2020 ( Table 1 ). Hybrids were selected based on their commercial relevance and were all conventional with no GMO traits.

The experimental design was a randomized complete block with two replications. In the short maturity hybrids experiments, the plot size was 15 m² (4-row plot), while in the long maturity hybrid experiments, 60 m² (8-row plot). Maize hybrids were planted at near optimum planting dates for this region (Apr 30 to May 13; Baum et al., 2020), at current plant densities of 7.9 to 8.5 pl m^-2, and received sufficient N fertilizer application ( Table 2 ). Maize hybrids were sown with precision planters used in breeding programs. Experiments were all rainfed, with a row spacing of 76 cm, and the previous crop was soybean.

The soil at all experimental fields was clay loam, rich in soil organic matter (3.4% in 0-30cm) and with a shallow water table (mean 1.3 m, range: 0.3 to 3 m; Archontoulis et al., 2020; Elli and Archontoulis, 2023). Weather conditions for the experiments were provided by Bayer Crop Science (gridded weather dataset ERA5-Land, Muñoz-Sabater et al., 2021). The cumulative growing season (April through September) rain was 591 mm in 2021 and 537 mm in 2022, with the 30-yr average being 677 mm ( Table 2 ). Crops experienced 16% below-normal rainfall ( Supplementary Figure S1 ), and slightly above normal radiation (<5%) in both years. The mean growing season temperature was 1.3°C above normal in 2021 and around normal in 2022 (normal = 18.6°C). The growing season average vapor pressure deficit (VPD) averaged 0.82 kPa, which was 25% above normal across locations and years.

Plant counts were determined early in the season, showing an average plant density of 8.3 pl m^-2 with very small variability among plots within an experiment (average standard deviation of 0.39 pl m^-2 and coefficient of variation of 4.8%). At the start of grain-fill (R2 stage) and at physiological maturity (R6 stage, Abendroth et al., 2011), one plant was harvested from the center rows of each plot at the ground level. Following plant harvest, the brace roots, leaf blades, ears, and tassel were removed from the plant, and the stem was cut into four equal-length sections ( Figure 1 ). Samples were dried in a forced-air oven at 60°C until constant weight and weighed separately. After grounding the samples, stem N and C concentrations were determined using a CHNS Elemental Analyzer (Elementar Americas). Plant and ear heights were measured during plant harvest as the distance from the soil surface to the insertion of the flag leaf blade and from the soil surface to the base of the main ear, respectively. Prior to mechanical harvest, five consecutive ears from the center row were hand harvested for grain N content (CHNS Elemental Analyzer). Plot-level grain yield estimates were derived from the middle rows of each plant using combine harvesters. Grain yield data are reported here at 150 g kg^-1 moisture, while dry matter and N content data are reported on dry basis.

Stem dry matter (DM) and N remobilization were determined as:

Plant trait genetic gain was calculated using a linear fixed-effect model using the lm function in R (R Core Team, 2022). We analyzed separately the genetic gain for each trait, stem section, and relative maturity combination. The statistical model included the hybrid year of release, season, replicates within a season, and the interaction hybrid year of release with season as fixed effects. Genetic gain significance (hybrid year of release effect) was tested using a type III F-test (emmeans package; Lenth et al., 2023). We found weak evidence of year of release and season interaction across most traits. Thus, a single genetic gain was reported across seasons. Mean comparisons between R2 and R6 stem dry matter, N concentration, and N content were made by paired t-test ( Figure 1 ). Regression analyses in R (R Core Team, 2022) were used to explore stem sections dry matter and N remobilization relationships.

Across experiments and hybrids, the grain yield significantly increased with the year of hybrid release from 8.3 to 16.9 Mg ha^-1 ( Figure 2a ). Long maturity hybrids had consistently greater yields than the short maturity hybrids. However, the genetic gain was similar across relative maturities (134 kg ha^-1 year^-1; Figure 2a ). The grain N amount significantly increased with the year of hybrid release ( Figure 2b ) but at different rates depending on relative maturity. The grain N content genetic gain in long maturity hybrids was 2-fold greater than that of short maturity hybrids (1.02 vs. 0.53 kg N ha^-1 year^-1; Figure 2b ).

Across all hybrids, the plant height ranged from 227 to 256 cm ( Figure 2c ), the ear height from 92 to 136 cm ( Figure 2d ), and the ear-to-plant height ratio from 0.40 to 0.53 ( Supplementary Figure S2 ). In long maturity hybrids, plant and ear heights were significantly (p<0.05) reduced with the year of release by 0.26 and 0.22 cm year^-1, respectively ( Figures 2c, d ). New long maturity maize hybrids had 9 cm shorter plant height and 8 cm lower ear height compared to old hybrids. The ear-to-plant height ratio did not change with the hybrid year of release in long maturity products. In short maturity hybrids, the plant height, ear height, and ear-to-plant height ratio did not change with the year of release ( Figures 2c, d ; Supplementary Figure S2 ).

Across hybrids, the total stem dry matter ranged from 75 to 104 g plant^-1 at R2 stage and from 55 to 83 g plant^-1 at R6 ( Figure 3 ). On average, the whole stem dry matter decreased by 18% from R2 to R6 stage ( Table 3 ; Figure 1 ). The average vertical distribution of the stem dry matter along the plant height was: 8% (top), 18% (middle-top), 30% (middle-bottom) and 44% (bottom, Figure 1 ). The relative distribution of dry matter changed little from R2 to R6 stage ( Figure 1 ).

In both relative maturities, newer hybrids had less total stem dry matter at R2 stage than older hybrids (-0.28 g plant^-1 year^-1, p>0.05). The overall reduction from 1980 to 2020 hybrids was 10% and occurred in the top three-quarters, with no significant change in the bottom stem portion ( Figure 3 ). Because of this, new hybrids had 7% (long maturity) and 13% (short maturity) increased relative allocation of dry matter at the bottom stem portions at R2 stage ( Table 3 ). At R6 stage, the patterns were similar to those in R2 stage, with no change in stem dry matter at the bottom with the year of release and a reduction in top-plant stem dry matter, which led to an overall insignificant change in total stem dry weight in both relative maturities ( Figure 3 ; Table 3 ).

Across hybrids, whole stem average N concentration ranged from 0.67 to 0.88% at R2 stage and from 0.45 to 0.68% at R6 stage ( Figure 3 ). On average, the whole stem N concentration decreased by 24% from R2 to R6 stage ( Figure 1 ; Table 3 ). The average vertical distribution of the stem N concentration was relatively uniform along the plant height at both growth stages, with a slightly higher N concentration at the bottom of the stem at R2 ( Figure 1 ; Table 3 ). In long maturity hybrids, whole stem average N concentration was similar between old and new hybrids at R2 stage, but there was a significant decrease of -0.0023% year^-1 (p<0.05) at R6 stage. Most of this reduction was driven by the bottom stem portion ( Figure 3 ). In short maturity hybrids, whole stem average N concentration slightly decreased (p>0.05) with the year of hybrid release for both R2 and R6 stages ( Figure 3 ; Supplementary Table S1 ).

Across hybrids, the total stem N content ranged from 0.51 to 0.82 g N plant^-1 at R2 stage and from 0.32 to 0.55 g N plant^-1 at R6 ( Figure 3 ). On average, the whole stem N content decreased by 37% from R2 to R6 stage ( Table 3 ; Figure 1 ). The average vertical distribution of the stem N content along the plant height was 8% (top), 17% (middle-top), 28% (middle-bottom), and 47% (bottom, Figure 1 ), which is nearly identical to the stem dry matter distribution. The relative distribution of the stem N content along the plant remained unchanged from R2 to R6 stage ( Table 3 ; Figure 1 ).

At R2 stage, newer hybrids of both relative maturities had slightly less (p>0.05) total stem N content than older hybrids ( Figure 3 ). However, different stem portions exhibited different patterns with the year of release: the N content significantly (p>0.01) decreased in the top-middle portions of the stem and remained unchanged in the middle-bottom portions at R2 stage ( Figure 3 ).

At R6 stage, newer long maturity hybrids had less (p<0.05) total stem N content than older hybrids ( Figure 3 ). The reduction with the year of release was 2-fold larger for long maturity hybrids compared to short maturity hybrids (0.0027 vs. 0.0015 g N year^-1, Figure 3 ). All stem portions contributed to this reduction at the R6 stage in long maturity hybrids ( Figure 3 ). In short maturity hybrids, the reduction occurred at the top-middle portions of the stem only.

Across hybrids, stems remobilized 18% dry matter and 37% N content from R2 to R6 stages. Newer hybrids of both relative maturities remobilized less dry matter than the older hybrids (20% vs 17%) but slightly more N (39 vs. 36%; Table 3 ).

Across hybrids and relative maturities, the below the ear leaf stem portions contributed 80% of the total stem dry matter and N remobilization ( Figure 1 ; Table 3 ). The stem dry matter and N remobilization per stem portion were positively associated with the dry matter and N content at R2 stage ( Figure 4 ). Therefore, bottom stem portions with more dry matter and N contents remobilized the most ( Table 3 ). Furthermore, the dry matter and N remobilization patterns were similar across relative maturities and old and new hybrids ( Supplementary Figure S3 ). On average, for every 1 g of remobilized dry matter, 0.01g N were remobilized ( Supplementary Figure S4 ). In relative terms, we found 20% N remobilization (base) with no dry matter remobilization, and beyond that, for every 1% increase in dry matter remobilization, the N remobilization increased by 0.8% ( Supplementary Figure S4 ).

We found that most of the stem N is allocated at the bottom of the plant due to the greater dry matter content ( Figures 1 , 3 ; Table 3 ). This vertical pattern is different from that of leaf N distribution within the canopy, in which the most leaf N is at the middle canopy, around the ear leaf (Hirose and Werger, 1987; Bertheloot et al., 2008; Fan et al., 2023). Furthermore, we showed that most stem dry matter and N remobilization during grain-fill derive from the bottom stem ( Table 3 ). Again, this pattern differs from that of leaves, in which the largest amount of leaf dry matter and N remobilization derives from the middle of the canopy (Fan et al., 2023). These differences between stems and leaves vertical patterns are important for further understanding C and N physiological dynamics during grain-fill. Given that most previous studies reported dry matter and N vertical distribution in leaves or leaves and stems combined without focusing on the stem (Johnson et al., 2010; Fan et al., 2023), our study adds new knowledge. Our results that the bottom stem quarter had a slightly greater N concentration ( Figure 1 ; Table 3 ) than other stem portions agree with findings by DeBruin et al. (2013), the only study to our knowledge that reported vertical distributions in stem N concentration.

Our study is the first to report maize breeding effects on dry matter and N allocation and remobilization patterns by stem portion. Previous studies have examined maize breeding effects on the total stem dry matter and N in a small number of genotypes of long maturity (Chen et al., 2015; Fernandez et al., 2020). Our finding that the total stem dry matter has been affected little by breeding ( Figure 3 ) agrees with the literature (Chen et al., 2015; Saenz et al., 2025). However, we revealed that this small change was not proportional across stem portions and crop stages. The top-plant stem dry matter at R2 stage has been significantly decreased by 16 to 22% in both relative maturities, while the bottom-plant stem dry matter has remained unchanged ( Figure 3 ). Because the bottom part of the stem dry matter is large (on average 45% of total, Figure 1 ), no significant breeding effects are evident at the whole stem dry matter. Consequently, the distribution of dry matter within the stem has changed with breeding towards less dry matter at the top-portions of the stem ( Table 3 ).

Stem N concentration remained unchanged at R2 stage in both relative maturities. However, at R6 stage, the average stem N concentration has significantly decreased in long maturity (111-day) hybrids in agreement with previous studies (Chen et al., 2015; Fernandez et al., 2020) and has remained unchanged in short maturity hybrids ( Figure 3 ). Interestingly, the reduction in long maturity hybrids N concentration was similar across stem sections ( Figure 3 ). Considering that stem represents up to 50% of the total maize residue (Bender et al., 2013), breeding probably has reduced maize residue quality (higher CN ratio) in long maturity hybrids, with no effect on short maturity hybrids, which could impact residue decomposition dynamics. The fact that stem N concentration has remained unchanged in short maturity hybrids is important, considering that residue decomposition constraints are more significant in the target environments of this group of hybrids (Sindelar et al., 2013; King et al., 2024).

Maize hybrids remobilized a 2-fold larger amount of stem N than dry matter ( Table 3 ), in agreement with literature (Chen et al., 2015; Archontoulis et al., 2020). Breeding slightly reduced stem dry matter remobilization but slightly increased stem N remobilization in both relative maturities, in agreement with previous findings ( Table 3 ; Chen et al., 2015; Mueller et al., 2019). The larger N requirements (grain N content, Figure 2 ) in newer hybrids, together with a lesser stem minimum N concentration boundary (Fernandez et al., 2020), might have resulted in the greater stem N remobilization. In fact, DeBruin et al. (2017) found that grain N was positively associated with hybrid leaf N remobilization at limited N availability. Furthermore, we found that the greater stem N remobilization of the new hybrids is due to greater N remobilization at the bottom of the stem, providing new evidence supporting the greater N use efficiency in newer hybrids (Mueller et al., 2019). Similar to DeBruin et al. (2013) findings for the leaves and Ciampitti et al. (2013) for the shoot, we found that stem sections with more dry matter and N remobilized the most ( Figure 4 ). However, we found that stems remobilized only 0.46g N per g N content at R2, in comparison to the 0.63g reported in leaves (DeBruin et al., 2013). These remobilization relationships were consistent across maturities and hybrids year of release ( Supplementary Figure S3 ).

This study does not account for management effects such as N fertilizer rate and plant density, which can affect N remobilization (Uhart and Andrade, 1995; Fan et al., 2023). Previous era studies found that genotypic differences in N content and remobilization between hybrids released in different years were only evidenced under sufficient/high N availabilities (Haegele et al., 2013; DeBruin et al., 2017; Fernandez et al., 2020). Besides, Chen et al. (2015) found that plant density treatments did not affect stem N remobilization. Therefore, the breeding effects on dry matter and N allocation and remobilization reported in this study might have been more moderate if hybrids had been tested under N-limiting conditions.

Newer hybrids support greater ear biomass and remobilize more N from the stem into the grain ( Figure 2 ; Mueller et al., 2019; Ruiz et al., 2023; Saenz et al., 2025), characteristics that make a crop more susceptible to lodging. However, newer hybrids are more resistant to lodging (Duvick and Cassman, 1999). Lodging resistance is typically related to plant architecture, stem node diameter and internode length, stem N remobilization, synthesis of lignin and cellulose, and disease tolerance (Rajcan and Tollenaar, 1999; Jackson-Ziems et al., 2014; Bian et al., 2016; Xue et al., 2016; Ahmad et al., 2018; Stubbs et al., 2023). However, Xue et al. (2020) showed that most genotypic differences in resistance to stalk lodging were explained by plant and ear height, stem dry weight, and average dry weight per unit length of the bottom nodes. In addition, Zhang et al. (2021, 2023) evaluated hybrids with different lodging susceptibility and found that lodging-resistant hybrids had lower plant and ear height and partitioned more dry matter to the stem base. Our study showed that new long maturity plants were shorter and had lower ear height ( Figure 2 ), even though the ear-to-plant height ratio did not change with the hybrid year ( Supplementary Figure S3 ). Besides, we found, for both maturities, a reduction in the dry matter allocation to the top stem sections with the hybrid year or release ( Figure 3 ), increasing the relative contribution of the bottom quarter to the total stem dry matter ( Table 3 ). This change in dry matter allocation reduces the height of the plants’ center of gravity, which could explain why newer hybrids are more resistant to lodging. Future research could investigate the long-term effects of breeding on other traits associated with lodging resistance, such as lignin and cellulose content in the stem.