Optimizing Stay-Green Level Synergistically Enhances Grain Yield and Nitrogen Use Efficiency in Chinese Maize Hybrids

Abstract

Stay-green (SG) breeding has been widely adopted to improve maize grain yield, but its impact on nitrogen use efficiency (NUE) remains poorly understood. We evaluated the impacts of SG on yield and NUE in Chinese maize hybrids over six decades with multi-location field trials, combined with physiological and transcriptome analysis. Hybrids were classified by SG level into normal senescence (SE, <35%), optimal SG (opt-SG, 35%–55%), and over SG (ov-SG, >55%). Across 138 hybrids, grain yield increased linearly with SG up to ~45% and then plateaued, while N remobilization efficiency (NRE) declined linearly. Ov-SG impaired NRE and grain N concentration without further yield gain, whereas opt-SG achieved both high yield and efficient N remobilization. The opt-SG hybrid XY335 had higher leaf NRE and greater post-silking N uptake than ov-SG hybrid ZD958. The superior NRE in XY335 involved early N remobilization from middle and upper leaves during silking to 30 days after silking (DAS), preferential degradation of soluble proteins and cell wall-associated N while maintaining photosynthesis. Transcriptome analysis and WGCNA (Weighted Gene Co-expression Network Analysis) identified a module highly expressed in XY335 at 30 DAS, which is enriched in many N-remobilization genes, including cysteine/serine proteases and amino acid transporters. These findings define opt-SG as an ideotype that synergistically enhances yield and N remobilization, which offers targets for breeding N-efficient maize for sustainable agriculture.

Keywords

maize / nitrogen remobilization efficiency (NRE) / nitrogen use efficiency (NUE) / stay-green (SG) / transcriptome

1 INTRODUCTION

Maize is the most productive crop globally, yet nitrogen (N) remains a key limiting nutrient for its production (Marschner, 2011; Han et al., 2020). China is now the largest producer and consumer of N fertilizer, consuming about one-third of the global supply, with its share projected to reach half by 2050 (Good and Beatty, 2011; Cui et al., 2018). However, N use efficiency (NUE) in Chinese maize is only 30%, resulting in a serious environmental footprint (Fan et al., 2012; Cui et al., 2018). Therefore, improving NUE while maintaining yield is essential. Both genetic improvement and precise N management are critical for enhancing yield and NUE. In particular, breeding high-yielding and N-efficient hybrids offers a key strategy to increase productivity with reduced N input (Duvick, 2005; Chen et al., 2013a).

Two key processes determine NUE in maize: N uptake efficiency (NUpE) and N utilization efficiency (NUtE) (Good et al., 2004; Mueller and Vyn, 2016). Studies comparing Old Era (1940–1990) and New Era (1991–2010) hybrids have demonstrated increases in yield, and total N accumulation, particularly post-silking N accumulation (Ciampitti and Vyn, 2012; Mueller and Vyn, 2016). However, grain N concentration (GNC) has decreased with the year of release (Duvick, 2005; Chen et al., 2013b; Haegele et al., 2013; DeBruin et al., 2017). Old cultivars depend more on remobilized pre-silking N (Ciampitti and Vyn, 2013; DeBruin et al., 2017). The SG trait, which is characterized by increased leaf longevity, is associated with maize yield improvement (Duvick and Cassman, 1999; Pommel et al., 2006; Mueller and Vyn, 2016). Modern hybrids exhibit a greater level of SG, linked to delayed leaf senescence and sustained post-silking photosynthesis (Chen et al., 2014). Early-senescing hybrids show lower NUpE and higher N remobilization efficiency (NRE), while late-senescing hybrids exhibit higher NUpE and lower NRE (Ciampitti and Vyn, 2012). The trade-off associated with delayed senescence is decreased GNC rather than increased plant N uptake (Chen et al., 2013b). SG has been considered a superior characteristic in commercial maize since the 1970s–1980s, with its prevalence continuously increasing in both the USA and China (Duvick et al., 2003; Fan et al., 2012; Thomas and Ougham, 2014). However, it remains unclear whether breeding for SG affects NUE in Chinese hybrids.

SG maize genotypes delay leaf senescence, which extends photosynthetic activity during grain filling, thereby enhancing carbon assimilation and yield. However, this prolonged greenness creates a physiological trade-off: while maintaining photosynthesis, SG plants often retain more N in vegetative tissues at maturity, which potentially reduces N remobilization to grains (Du et al., 2024; DeBruin et al., 2017). Studies comparing SG and non-SG hybrids reveal that excessive leaf N retention does not always translate into superior photosynthesis. This is because photosynthetic N use efficiency (PNUE) may decline due to reduced enzyme activity (Yang et al., 2023). Under stress conditions, SG hybrids can exhibit a lower N harvest index, with less remobilized N from leaves (Liu et al., 2022). Efficient N remobilization is critical, as 60%–70% of grain N derives from vegetative tissues accumulated before silking (Masclaux-Daubresse et al., 2010). Improving the SG trait could balance prolonged photosynthesis in leaves with efficient N remobilization to grain, thereby maintaining both yield and GNC.

N remobilization and delivery to grain in maize are subject to complex genetic regulation. Chloroplastic GS2 assimilates NH₄⁺ derived from nitrate reduction in young leaves, supporting continued N metabolism during the functional SG phase (Masclaux-Daubresse et al., 2008). During senescence, cytosolic GS1 converts ammonium released from protein degradation into glutamine for transport to grain (Martin et al., 2006; Thomas and Ougham, 2014). For example, ZmGS1.4 expression is lower in late-senescing hybrids during grain filling, which correlates with reduced remobilization efficiency (Gallais and Hirel, 2004; Martin et al., 2005). Recent studies have reported that ZmTHP9 (asparagine synthetase) enhances seed protein content and NUE (Huang et al., 2022). Beyond these enzymatic activities, several N transporters contribute to N remobilization in maize, such as amino acid transporters ZmAAP4, ZmVAAT3, and ZmAAP6 which facilitate amino-N mobilization within the ear (Pan et al., 2015; Wang et al., 2022), and nitrate transporter ZmNRT1.5/ZmNPF7.9 which delivers nitrate to the developing endosperm (Wei et al., 2021). However, the spatiotemporal expression patterns of senescence-associated N transport and metabolism genes across different leaf positions during the grain-filling period, and how SG status modulates those patterns, remain poorly understood.

In this study, we first conducted field trials to investigate the relationships between SG and NUE in Chinese maize hybrids. Field Experiment 1 (26 hybrids, 2010–2011) classified SG types and assessed yield and N-related traits. Experiments 2 and 3 (138 hybrids, across multiple locations) validated the correlation between SG level and NRE. Furthermore, using the selected hybrids XY335 and ZD958, we quantified N remobilization dynamics through physiological and transcriptomic analyses. We hypothesize that advanced hybrids can achieve optimal levels of SG to balance grain yield and NUE by spatiotemporally regulating N remobilization from vegetative tissues.

2 RESULTS

2.1 Changes in SG level of Chinese maize hybrids over six decades

To investigate the changes in SG in maize, we selected 26 representative hybrids that were widely adopted in China from the 1950s to the 2000s (Table S1). Based on hierarchical cluster analysis using SG level at maturity and the onset of senescence (Ts), these 26 maize hybrids were classified into three types: normally senescent (SE, SG level < 35%, average 24%), optimal SG (opt-SG, 35%–55%, average 48%), and over SG (ov-SG, > 55%, average 65.8%) (Fig. 1–A). Among the different SG types, Ts did not differ significantly between the ov-SG and the opt-SG types, but both were significantly later than the SE type (P < 0.05) (Table 1). While the pre-silking days were similar across the three types, the maturity and post-silking days were significantly longer in the opt-SG type and the ov-SG type than in the SE type (P < 0.05). The senescence rate parameters Vm (the decreasing rate of senescence) and Vmax (the maximum reduction rate of senescence) decreased significantly in the opt-SG and the ov-SG type compared with those in the SE type (Table 1). Collectively, SG hybrids delayed the initiation of senescence and prolonged the post-silking period. Furthermore, the SG level increased over time, from approximately 20% in hybrids promoted in the 1950s to over 65% in those promoted in the 2000s, and was significantly and positively correlated with the number of years since the 1950s as a continuous variable (R² = 0.49, P < 0.001). Most of the modern maize hybrids were the opt-SG and the ov-SG types, especially the opt-SG type (Fig. 1–B). The SE type showed the earliest and fastest leaf senescence, while the SG types had delayed senescence onset and slower rates, with the ov-SG type remaining largely green at maturity (Fig. 1–C).

2.2 Biomass accumulation and NUE among different SG types

To investigate how modern maize hybrids have increased biomass and NUE, we measured aboveground dry mass and N concentration at silking and maturity in 2010 and 2011 for the 26 hybrids. Biomass, NUE and related traits differed significantly among the three SG types (Fig. 2). At silking, the SG types had significantly higher stem and leaf dry matter than the SE type, increasing by 29%–33% and 44%–57%, respectively (Fig. 2–A and –B). Similarly, at maturity, the SG types had significantly higher stem dry matter, leaf dry matter and grain yield (GY) than the SE type, increasing by 43%–57%, 59%–91%, 55%–68%, respectively (P < 0.05) (Fig. 2–C to –E). Within the SG types, the opt-SG and the ov-SG types showed no difference in biomass and GY at both silking and maturity (Fig. 2–A to –E). Therefore, SG breeding significantly increased the biomass and GY of modern maize hybrids.

For N-related traits, stem and leaf N concentrations were similar between the SE and the SG types at silking (Fig. 2–F and –G). At maturity, leaf N concentration was significantly higher in the ov-SG type, while grain N concentration (GNC) was higher in the SE type (P < 0.05) (Fig. 2–H to –J). For N uptake, opt-SG and ov-SG types showed no significant differences during the pre-silking, post-silking, and whole periods, but both types had higher N uptake than SE type across all periods (Fig. 2–K to –M). For N utilization, although there was no significant difference in NUtE and N harvest index (NHI) among them, NRE of the ov-SG type was significantly lower than that in the SE type (P < 0.05) (Fig. 2–N to –P). These results indicated that increasing SG level in modern hybrids significantly increases maize yield and N accumulation without affecting NUtE. In contrast to the ov-SG hybrids, which significantly increased N residues in leaves and reduced NRE and GNC, the opt-SG hybrids achieved high yield without negative effects on NRE and GNC.

2.3 Correlation of SG level with yield and NUE

To understand how SG breeding affected yield and NUE in modern maize hybrids, 138 hybrids at three experimental locations across two years (24 hybrids at Shunyi in 2010 and 2011, 57 hybrids at Shangzhuang in 2011, and 57 hybrids at Quzhou in 2011) (Table S1 and S2) were used to investigate the correlations between SG, yield and NUE. Total N accumulation (R² = 0.16, P < 0.001), pre-silking N accumulation (R² = 0.14, P < 0.01), and post-silking N accumulation (R² = 0.11, P < 0.01) showed significant positive correlations with SG level (Fig. 3–A). As the SG level increased, the rate of increase in post-silking N accumulation (0.89) was much higher than that in pre-silking N accumulation (0.42). In addition, we observed a significant positive correlation between leaf N concentration (LNC) and SG level (R² = 0.18, P < 0.01), and a significant negative correlation between GNC and SG level (R² = 0.13, P < 0.01) (Fig. 3–B). Furthermore, the NRE decreased linearly with increasing SG level (R² = 0.26, P < 0.001) (Fig. 3–C). There was a linear-plateau correlation between the relative grain yield and SG level. Specifically, grain yield increased linearly with increasing SG level from 0 to 45% (R² = 0.30, P < 0.01) (Fig. 3–C). Beyond this point, the grain yield reached a plateau when stay-green level exceeded 45%. These results suggested that SG breeding not only significantly increased GY and N accumulation, but also increased residual N in leaves at maturity, resulting in decreased GNC and NRE. Thus, maintaining the SG level at about 45% achieved high yield and high NUE simultaneously.

Data were collected from 24, 57, and 57 hybrids grown at Shunyi, Shangzhuang, and Quzhou, respectively, except that NRE was not evaluated at Quzhou. Hybrids marked with dotted circles were considered outliers and excluded from correlation analysis.

2.4 ¹⁵N labeling reveals superior N remobilization in XY335

We then selected two representative hybrids, XY335 (opt-SG, 45% SG level) and ZD958 (ov-SG, 64% SG level), to investigate mechanisms underlying SG-dependent N remobilization. A greenhouse pot experiment was conducted using ¹⁵N double-labeling (labeled at the V6 and silking stages, respectively) to quantify post-silking N uptake and remobilization (Yang et al., 2016). Between the two genotypes, no significant difference in biomass, total N accumulation and GY was observed (Table 2). Although total N uptake did not differ between the two hybrids, XY335 had lower pre-silking N uptake and higher post-silking N uptake. As a result, XY335 had lower stem/leaf N concentrations, but higher grain N accumulation than ZD958.

XY335 and ZD958 differed in their capacity for pre-silking N remobilization from vegetative tissues to grains and post-silking N uptake into grains (Fig. 4). At silking, 54.3% (0.71 g) and 33.5% (0.44 g) of pre-silking N were distributed in leaves and stems of XY335, respectively, compared with 50.4% (0.77 g) and 39.3% (0.60 g) in ZD958. At maturity, vegetative N remobilized from leaves was higher in XY335 (44.1%, 0.58 g) than that in ZD958 (35.3%, 0.54 g). Residual N in leaves of XY335 (10.2%, 0.13 g) was lower than that of ZD958 (15.1%, 0.23 g) (Fig. 4). By contrast, vegetative N remobilized from stems in ZD958 (27.0%, 0.41 g) was slightly higher than that in XY335 (23.2%, 0.31 g). The proportion of N distributed to grain from vegetative N was higher in XY335 (69.5%) than in ZD958 (63.8%). At maturity, 74.5% (0.75 g) post-silking N was distributed to grains of XY335, which was higher than that of ZD958 (71%, 0.44 g). Therefore, XY335 exhibited higher grain N accumulation due to higher pre-silking N remobilization capacity and post-silking N uptake and remobilization.

2.5 Spatial and temporal dynamics of leaf N remobilization

To investigate spatial variation in leaf N remobilization, each leaf was sampled at silking and maturity to determine LNC and ¹⁵N changes. At silking, LNC did not differ significantly between XY335 and ZD958, with values of 24.1 g kg⁻¹ and 24.2 g kg⁻¹, respectively. (Fig. 5–A). At maturity, in the middle and upper leaves but not in the lower leaves, XY335 had significantly lower N concentration than ZD958 (Fig. 5–B). As a consequence, XY335 remobilized more ¹⁵N from middle and upper leaves than ZD958. These results indicate that the middle and upper leaves are crucial for the leaf NRE difference between these hybrids.

To examine the temporal dynamics of pre-silking leaf NRE, three representative leaves were analyzed: the third leaf above the ear (upper), the ear leaf (middle), and the third leaf below the ear (lower). LNC followed a linear-plateau trend, across all three leaves, remaining stable until 20 days after silking (DAS) and then decreasing significantly thereafter for both hybrids (Fig. 5–C to –E). In the ear leaf, XY335 exhibited lower N concentration at 40 and 50 DAS (Fig. 5–D). In the upper leaf, XY335 consistently displayed lower N concentration than ZD958 throughout the sampling period. In the lower leaf, no marked difference in N concentration was observed between the two hybrids (Fig. 5–E). NRE dynamics revealed no difference in the lower leaf either (Fig. 5–H). However, XY335 exhibited higher N remobilization in the ear leaf after 30 DAS. In the upper leaf, XY335 initiated remobilization immediately after silking, whereas ZD958 only started after 20 DAS (Fig. 5–F and –G).

Photosynthetic efficiency in the ear leaf did not differ between XY335 and ZD958, both declined after 30 DAS (Fig. 5–J). XY335 showed higher photosynthetic efficiency in the upper three leaves (especially at silking and 20 DAS), whereas ZD958 was superior in the lower three leaves (Fig. 5–I and –K). Analysis of N components in the ear leaf from silking to 30 DAS revealed that XY335 exhibited greater decreases in soluble protein, free amino acids, Rubisco, PEPC, and cell wall-associated N than ZD958, while PPDK and thylakoid-associated N declined less (Fig. S1). Therefore, XY335 mobilized redundant N fractions more efficiently to enhance N remobilization without compromising photosynthesis in the upper and ear leaves.

2.6 Transcriptome and weighted gene co-expression network analysis (WGCNA) identify a key module for N remobilization

To investigate the transcriptional dynamics during the onset and progression of senescence and N remobilization between XY335 and ZD958, we performed RNA-seq using the ear leaf at the non-N-remobilization phase (0 DAS), the N remobilization onset phase (15 DAS), and the phenotypically visible N remobilization phase (30 DAS). By comparing the gene expression at 15 DAS and 30 DAS to 0 DAS, we identified 3971 and 4264 DEGs in ZD958 and XY335, respectively (Fig. 6–A). Among them, 1556 were unique to ZD958, 1849 were unique to XY335, and 2415 were common (Fig. 6–B). The GO enrichment analysis of the unique DEGs in ZD958 showed that these DEGs were mainly related to protein phosphorylation, oxidation-reduction processes, and purine ribonucleoside biosynthetic process, and unique DEGs in XY335 were enriched in plastid organization, translation, and photosynthesis (Fig. 6–C). A total of 4414 DEGs were identified by comparing the transcriptome between ZD958 and XY335 at three time points, of which 2232 were up-regulated and 2182 were down-regulated in XY335 (Fig. 6–D). Differential gene expression peaked at 30 DAS, coinciding with higher NRE in XY335 than in ZD958. These results suggest that ZD958 and XY335 employ distinct regulatory programs for post-silking senescence and N remobilization.

To further examine the gene expression modes among different genotypes, we performed WGCNA of genotypic-DEGs. A total of 4333 DEGs were grouped into five modules (Fig. 7–A). Pearson correlation analysis between these modules and N remobilization stages revealed genotype- and stage-specific correlations (Fig. 7–B). The “brown”, “green” and “blue” modules were positively correlated with 0 DAS (r = 0.9, P = 3 × 10⁻⁷), 15 DAS (r = 0.9; P = 4 × 10⁻⁷) and 30 DAS (r = 0.86, P = 5 × 10⁻⁶) in ZD958, respectively. In XY335, “yellow”, and “turquoise” modules were positively correlated with 0 DAS (r = 0.91; P = 2 × 10⁻⁷) and 30 DAS (r = 0.77, P = 2 × 10⁻⁴). Most genes (1982, 45.7%) in the “turquoise” module were highly expressed in the 30 DAS of XY335, suggesting a potential link to its high NRE. The enriched GO categories in this module included N metabolism and amino acid metabolism, such as response to histidine/leucine/phenylalanine, extracellular-glutamate-gated ion channel activity and cysteine-type endopeptidase activity (Fig. 7–C). In addition, the enriched categories in this module included iron ion binding, electron carrier activity, monooxygenase/oxidoreductase activity, chloroplast thylakoid membrane, and responses to salicylic acid.

N remobilization in leaves is mainly mediated by autophagy and vacuolar proteases. In the “turquoise” module, a large number of proteases and amino acid transport-related genes were significantly upregulated. For example, the expression of several cysteine protease genes (Zm00001D000327, Zm00001D005828, and Zm00001D049109), serine protease genes (Zm00001D003678, Zm00001D007073, Zm00001D009178, Zm00001d013865, Zm00001D025891, Zm00001D027355, and Zm00001D049055), and ZmATG12 (Zm00001d018259) in XY335 were significantly higher than ZD958 (Fig. S2). The amino acid and peptide transporters play an important role in N remobilization to the seeds. For instance, two amino acid transporters (Zm00001d003422 and Zm00001d044387) were significantly up-regulated from silking to 30 DAS in XY335 in the “turquoise” module. Several amino acid permeases (Zm00001d012231, Zm00001d023592, Zm00001d025437, and Zm00001d053143) were also strongly induced in XY335 at 30 DAS (Fig. S3). Transcriptome analysis further confirmed that the upregulation of proteases and amino acid transporters in XY335 mechanistically underpins its early and efficient N remobilization from middle and upper leaves without compromising photosynthesis.

3 DISCUSSION

3.1 Excessive SG fails to increase yield and impairs NRE in Chinese maize hybrids

SG is recognized as a valuable trait that can modify N uptake and remobilization in crops (Gregersen et al., 2013; Thomas and Ougham, 2014). However, the relationship between SG and NUE remains poorly understood (Rajcan and Tollenaar, 1999; Ciampitti and Vyn, 2013), particularly with respect to how SG traits affect NRE in maize breeding programs. Here we show that SG breeding has significantly increased grain yield and N accumulation in modern Chinese hybrids. Nevertheless, the ov-SG compromises NRE and GNC with no further yield improvement. In contrast, the opt-SG achieves high yield while maintaining efficient N remobilization, as demonstrated by the opt-SG hybrid XY335 compared to the ov-SG hybrid ZD958 (Fig. 2 to 4). The superiority of XY335 in NRE is attributed to the early initiation of N remobilization from middle and upper leaves during the critical period from silking to 30 DAS, the efficient degradation of soluble proteins and cell wall-associated N while maintaining photosynthetic capacity, and the coordinated upregulation of genes involved in N remobilization, including glutamine synthetases, proteases, and amino acid transporters (Fig. 5 to 7). These findings establish opt-SG as an ideotype trait for breeding hybrids with high yield and efficient NUE.

Modern hybrids have been selected for delayed senescence, increased specific leaf N, and higher kernel number under high planting density (Mueller and Vyn, 2016; DeBruin et al., 2017). Newer hybrids with functional SG phenotypes maintain photosynthetic capacity and retain more leaf N at maturity. This results in high NUpE (Ciampitti and Vyn, 2012; Mueller and Vyn, 2016; DeBruin et al., 2017; Yang and Udvardi, 2018). However, the relationship between SG and productivity is not always positive (Borrell et al., 2001). Studies on drought tolerance found no correlation between green leaf longevity and grain yield (Bolanos and Edmeades, 1996). At high planting densities, delayed senescence of middle and upper leaves in SG hybrids did not increase yield (Antonietta et al., 2014). Our results demonstrate that yield failed to increase when the SG level exceeded a threshold of approximately 45% (Fig. 3–C). The Ov-SG hybrids did not exhibit higher photosynthetic rates compared to opt-SG hybrids (Chen et al., 2014), possibly because of altered N allocation within leaves that compromises photosynthetic efficiency (Mu et al., 2016). Consistently, a recent study in tropical maize germplasm found no significant correlation between SG and grain yield under low N conditions. This suggests that chlorophyll retention does not guarantee N remobilization for grain filling under stress (Abu et al., 2025). Thus, SG benefits are conditional, and exceeding an optimal level may impose metabolic costs with no yield improvement.

The negative impact of SG on GNC has been previously documented (Thomas and Ougham, 2014). We also observed that GNC decreased significantly from 1.6% to 1.4% with increasing SG level (Fig. 2 and 3). Late-senescing hybrids exhibited lower NRE but higher post-silking N uptake compared to early-senescing hybrids (Martin et al., 2005; Rajcan and Tollenaar, 1999; Pommel et al., 2006; Ciampitti and Vyn, 2012). Grain N is derived from both pre-silking N remobilization and post-silking N uptake, with newer hybrids deriving more grain N from post-silking uptake than older hybrids (Ciampitti and Vyn, 2013; Mueller and Vyn, 2016). In this work, however, excessive SG reduced both GNC and NRE. The ov-SG hybrids showed a 2.26% reduction in GNC compared to the opt-SG hybrids (Fig. 2–J), with more N retained in leaves at maturity. Therefore, the opt-SG cultivars can maintain grain yield while reducing straw N residue and selecting high-yielding hybrids with high NRE can increase GNC and grain yield simultaneously (Chen et al., 2015; Mi et al., 2016).

3.2 Optimal SG achieves high yield and high grain N via spatiotemporal N remobilization

XY335 (opt-SG) and ZD958 (ov-SG) are two widely planted hybrids in China. They represent high-yielding hybrids with high and low grain N concentration (GNC), respectively. Using ¹⁵N labeling, we found that leaves were the primary organ for N remobilization. The opt-SG exhibited approximately 10% higher leaf N remobilization NRE than the ov-SG (Fig. 4). The post-silking period is critical for dry matter accumulation and N transfer in maize (Lee et al., 2007; Hirel et al., 2007). The opt-SG not only achieved higher NRE but also had 5% greater post-silking N uptake compared to the ov-SG, explaining its higher GNC (Fig. 5). Importantly, the higher leaf NRE in opt-SG did not compromise photosynthetic efficiency despite lower leaf N concentration, maintaining photosynthetic NUE, as reported by Chen et al. (2014). This maintenance of photosynthetic capacity during active N remobilization may reflect superior coordination of carbon (C) and N metabolism. In C4 plants such as maize, the C concentrating mechanism enables more efficient N use, and key transcription factors like ZmDof1 and ZmM28 have been shown to orchestrate the balance between carbon assimilation and N utilization (Kurai et al., 2011; Wu et al., 2019).

In maize, middle and upper leaves exhibit higher photosynthetic rates, specific leaf N, and PNUE than lower leaves (Chen et al., 2016). NRE is genotype-specific (Gregersen, 2011), and senescence of upper leaves is primarily controlled by genotype (Pommel et al., 2006). Our results revealed spatial differences in leaf N remobilization: XY335 showed significantly higher N remobilization capacity in upper leaves compared to ZD958 (Fig. 5–C), while no significant difference was found in lower leaf NRE during reproductive growth (Fig. 5–E). Notably, N remobilization from upper and middle leaves of XY335 was rapid yet proceeded without compromising photosynthesis (Fig. 5–I to –J). This accounts for the higher PNUE in opt-SG hybrids.

The C-N transition point, at which leaves shift from C capture to N export during senescence, is delayed in SG genotypes (Thomas and Ougham, 2014). Delayed senescence can enable high yield but may result in inefficient N remobilization, whereas accelerated senescence leads to efficient remobilization but lower yield (Gregersen, 2011). Opt-SG hybrids achieved both high yield and high grain N content by optimizing the carbon–nitrogen transition point. In upper leaves, opt-SG initiated N remobilization immediately after silking, whereas ov-SG delayed initiation until 20 DAS. This difference arises because SG traits delay the onset of senescence, and ov-SG additionally exhibits a slower senescence rate than opt-SG (Thomas and Howarth, 2000). Compared to ov-SG, opt-SG performs rapid N remobilization from middle and upper leaves without compromising photosynthesis, resulting in higher GNC.

Our WGCNA further identified a "turquoise" module specifically upregulated in XY335 at 30 DAS. This module was enriched in genes encoding cysteine proteases, serine proteases, the autophagy-related gene ZmATG12, and amino acid transporters. This coordinated transcriptional program, absent in ZD958, explains the efficient N remobilization in XY335. The enrichment of senescence signaling pathways in this module links protease activation to developmental cues, which is consistent with salicylic acid-mediated senescence promotion (Abdelrahman et al., 2017). Notably, the cytosolic glutamine synthetase gene ZmGS1.4 (Martin et al., 2006) is also co-expressed within this module, suggesting a coordinated regulatory network from proteolysis to N export. Together with known targets such as ZmTHP9 (Huang et al., 2022) and ZmAAP6 (Wang et al., 2022), these components represent promising candidates for breeding high-NUE maize hybrids.

In conclusion, excessive SG fails to increase grain yield and compromises NRE, thereby disrupting the coordination between grain yield and grain protein content. In contrast, opt-SG achieves high yield while maintaining efficient N remobilization through enhanced PNUE in middle and upper leaves, together with the timely initiation of post-silking N remobilization. Importantly, identifying an optimal SG threshold and elucidating its spatiotemporal coordination with post-silking N dynamics provides a physiological and molecular framework for breeding high-yielding, N-efficient maize hybrids. As genomic and phenomic resources continue to expand, this framework is well positioned to support the development of sustainable agricultural practices that simultaneously deliver high yield, high grain N accumulation, and low straw N residue.

4 METHODS

4.1 Plant materials and experiment locations

A total of 26 maize hybrids (Table S1) released from the 1950s to 2000s in China were used in field experiment 1. This experiment was conducted at the Shunyi Experimental Station of the Chinese Academy of Agricultural Sciences in Beijing (40°13′ N, 116°65′ E) in 2010 and 2011. Fifty-seven maize hybrids (Table S2) released since the 1990s were used in field experiment 2 at Shangzhuang Experimental Station of China Agricultural University in Beijing (39°57′ N, 116°17′ E) and in field experiment 3 at Quzhou Experimental Station of China Agricultural University in Hebei (36°78′ N, 114°92′ E) in 2011. Two typical hybrids, XY335 and ZD958, were selected for a greenhouse pot experiment at China Agricultural University in 2011 and 2012. They were also used for field experiment 4 at Shangzhuang in 2019. The soil (0–30 cm) chemical characteristics at each location are shown in Table S3.

4.2 Field experimental design

A randomized complete block design was used in this study. Experiment 1 at Shunyi had three and four replicates in 2010 and 2011, respectively. Each plot was 4 m in length, with 0.6 m between rows and 0.28 m between plants (plot area 12 m²). A total of 240 kg N ha⁻¹ was applied, half at sowing and the other half at the V8 stage. Before sowing, 85 kg P₂O₅ ha⁻¹ and 90 kg K₂O ha⁻¹ were applied. Experiment 2 at Shangzhuang and Experiment 3 at Quzhou were designed with three replicates in 2011. The plot size, plant spacing, and N application rate were the same as in Experiment 1, with base fertilizers of 120 kg P₂O₅ ha⁻¹ and 150 kg K₂O ha⁻¹. Experiment 4 at Shangzhuang had four replicates. Each plot was 4 m long with ten rows per hybrid, at a spacing of 50 cm between rows and 20 cm between plants (plot area 20 m²). N was applied at 180 kg N ha⁻¹, with 30% applied before sowing and the remainder at the V6 stage. Basal fertilizers were applied at 120 kg P₂O₅ ha⁻¹ and 72 kg K₂O ha⁻¹.

For the greenhouse pot experiment, we used pots that were 30 cm wide and 60 cm high. Each pot contained 15 kg of soil and was planted with one maize plant. Before planting, urea, calcium superphosphate, and potassium sulfate were added at rates equivalent to 180 kg N ha⁻¹, 85 kg P₂O₅ ha⁻¹, and 90 kg K₂O ha⁻¹ based on the pot surface area. Plants were transferred on 13 April 2011 and 30 March 2012, and harvested on 10 August 2011and 27 July 2012, respectively. At silking, each pot was watered with 2 L of Hoagland's solution (Yang et al., 2016). All pots were arranged in a randomized design and re-randomized every two weeks.

Detailed information on plot size, plant density, N rate, and replicates for each experiment is provided in Table S4.

4.3 ¹⁵N labeling

In the greenhouse pot experiment, 2.5 mg of ¹⁵N (as K¹⁵NO₃, 10.2% ¹⁵N atom excess) in 1 L of water was applied to the pot surface at the stem elongation (V6) stage and at the silking stage, respectively. Control pots received 2.5 mg of N as KNO₃ at the same time. The ¹⁵N labeled at the V6 stage was assumed to be completely absorbed and assimilated before silking (Gallais et al., 2007; Yang et al., 2017).

4.4 Green leaf area measurement

In field experiments 1, 2, and 3, five uniform plants per hybrid were randomly selected for green leaf area measurement. Green leaf area was recorded every 7 days from silking to physiological maturity in Shunyi in 2010, and at silking and maturity in Shangzhuang and Quzhou. Green leaf area was calculated using the equation:

(1)\mathrm{G}\mathrm{L}\mathrm{A}=\mathrm{L}(\mathrm{l}\mathrm{e}\mathrm{a}\mathrm{f}\mathrm{l}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h})\times \mathrm{W}(\mathrm{m}\mathrm{a}\mathrm{x}\mathrm{i}\mathrm{m}\mathrm{u}\mathrm{m}\mathrm{l}\mathrm{e}\mathrm{a}\mathrm{f}\mathrm{w}\mathrm{i}\mathrm{d}\mathrm{t}\mathrm{h})\times 0.75

The senescence process was studied using the curvilinear equation described by Chen et al. (2013b), calculated as follows:

(2)\mathit{y}={\mathit{a}\mathit{e}}^{\mathit{b}\mathit{-}\mathit{c}\mathit{x}}/(1+{\mathit{e}}^{\mathit{b}\mathit{-}\mathit{c}\mathit{x}})

where y is the relative green leaf area (RGLA) at a given growth stage relative to silking; x is days after silking; a is the initial RGLA value, fixed at 1; b relates to the onset of senescence; and c relates to the senescence rate. The onset of senescence (Ts) is defined as the time when relative leaf area declines to 95%, and the SG level is defined as the proportion of green leaf area at maturity relative to that at silking.

4.5 Photosynthetic rate

In the greenhouse pot experiment, six uniform plants per hybrid were selected. Photosynthetic rate was measured on the ear leaf, the third leaf below, and the third leaf above the ear leaf every 10 days from silking to maturity, between 9:00 a.m. and 11:00 a.m. on sunny days using a LI-6400XT portable photosynthesis system (LI-COR Inc., Lincoln, NE, USA).

4.6 Plant sampling

In field experiments 1 and 2, three plants per block were harvested at silking and maturity and separated into leaves, stems (including sheaths and tassels), and grains. Grain yield was determined from two rows per plot and reported at 14% moisture. Nitrogen concentration was determined by the Kjeldahl method.

In the greenhouse pot experiment, six uniform plants per hybrid were harvested at 0, 10, 20, 30, and 40 days after silking (DAS) and at maturity, and divided into leaves (each leaf separated at silking and maturity), stems, roots, cobs, husks, and grains. Samples were oven-dried at 70 °C and ground, and a 3.5 mg aliquot was used for ¹⁵N analysis by isotope ratio mass spectrometry (DELTAplus XP, Thermo-Finnigan). For leaf physiological analysis, the ear leaf and the third leaf above and below the ear were collected between 9:00 a.m. and 12:00 p.m. at 0, 10, 20, 30, and 40 DAS, immediately frozen in liquid N₂ after midrib removal, ground, and stored at −80 °C.

In field experiment 4, three ear leaves per plot were sampled between 9:00 a.m. and 10:00 a.m. at 0, 15, and 30 DAS, and mixed as one biological replicate, with three biological replicates per treatment. After removing the midribs, leaves were immediately frozen in liquid nitrogen; half of each sample was used for biochemical measurements and the other half for RNA extraction.

4.7 Biochemical Measurements

Soluble protein, free amino acids, Rubisco, PEPc, PPDK, thylakoid N, and cell wall N were measured in ear leaf samples collected at 0 and 30 DAS in field experiment 3, following the methods of Mu et al. (2016). Briefly, soluble protein was extracted using a phosphate buffer and quantified by the Bradford method; Rubisco, PEPc, and PPDK contents were determined by ELISA; thylakoid membranes were isolated by differential centrifugation; and cell wall fractions were obtained after sequential extraction with 80% ethanol and 1 M NaCl.

4.8 RNA extraction, library construction, and sequencing

Total RNA was extracted from ear leaf samples (0, 15, 30 DAS) from field experiment 4 using TRIzol reagent. cDNA libraries were constructed and sequenced on an Illumina HiSeq 4000 platform.

4.9 Differential expression analysis and WGCNA

Clean reads were mapped to the B73 reference genome (V4). Gene expression was normalized to FPKM. Differentially expressed genes (DEGs) were defined as those with |log₂FC| ≥ 1 and FDR ≤ 0.001. WGCNA was performed using the WGCNA package in R (Langfelder and Horvath, 2008). The networks were visualized using Cytoscape v3.7.1.

4.10 Calculations

(3)\begin{array}{rl}& \mathrm{N}\text{utilization efficiency (NUtE) = grain yield /}\\ & \mathrm{N}\text{accumulation at maturity}\end{array}

(4)\begin{array}{rl}& \mathrm{N}\text{harvest index}(\mathrm{N}\mathrm{H}\mathrm{I})=\text{grain}\mathrm{N}\text{content}/\\ & \mathrm{N}\text{accumulation at maturity}\times 100\text{\%}\end{array}

(5)\begin{array}{rl}& \text{NRE}=\text{(whole-plant}\mathrm{N}\text{accumulation at silking}-\\ & \text{whole-plant}\mathrm{N}\text{accumulation in the stover at}\\ & \text{maturity) / whole-plant}\mathrm{N}\\ & \text{accumulation at silking}\times 100\text{\%}\end{array}

(6)\begin{array}{rl}& {}^{15}\mathrm{N}\text{content}=\mathrm{N}\text{content}\times \left(\mathrm{A}\mathrm{T}{\text{\%}}_{\text{sample}}-\mathrm{A}\mathrm{T}{\text{\%}}_{\text{control}}\right)\times \\ & 15\mathrm{g}{\mathrm{m}\mathrm{o}\mathrm{l}}^{-1}/14\mathrm{g}{\mathrm{m}\mathrm{o}\mathrm{l}}^{-1}\end{array}

where AT% is the ¹⁵N abundance, control is the natural ¹⁵N abundance.

The PreN (¹⁴N + ¹⁵N) is stored in vegetative organs and remobilized to grain (RemN) after silking. The RemN was calculated by the change in ¹⁵N abundance (V6 labeling) and was not influenced by PostN.

(7)\begin{array}{rl}& {\text{RemN}}_{\text{organ}}=\mathrm{N}\text{content of organ}{}_{\text{silking}}\times \\ & ({}^{15}\mathrm{N}\text{content of organ}{}_{\text{silking}}-\\ & {}^{15}\mathrm{N}\text{content of organ}{}_{\text{maturity}})/\\ & {}^{15}\mathrm{N}\text{content of organ}{}_{\text{silking}}\text{or}=\\ & ({}^{15}\mathrm{N}\text{content of organ}{}_{\text{silking}}/\\ & {}^{15}\mathrm{N}\text{content of plant}{}_{\text{silking}}-\\ & {}^{15}\mathrm{N}\text{content of}{\text{organ}}_{\text{maturity}}/\\ & {}^{15}\mathrm{N}\text{content of plant}{}_{\text{maturity}})\times \\ & {}^{15}\mathrm{N}\text{content of plant}{}_{\text{silking}}\end{array}

(8)\begin{array}{rl}& {\mathrm{N}\mathrm{R}\mathrm{E}}_{\text{organ}}=({}^{15}\mathrm{N}\text{content of organ}{}_{\text{silking}}-\\ & {}^{15}\mathrm{N}\text{content of organ}{}_{\text{maturity}})/\\ & {}^{15}\mathrm{N}\text{content of organ}{}_{\text{silking}}\times 10\text{\%}\end{array}

where ¹⁵N content of plant_silking = ¹⁵N content of plant_maturity, because¹⁵N labeling at V6 was no longer absorbed by plants after silking. Organ is leaf, stem or root.

(9){\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{N}=\mathrm{R}\mathrm{e}\mathrm{m}\mathrm{N}}_{\mathrm{l}\mathrm{e}\mathrm{a}\mathrm{f}}{+\mathrm{R}\mathrm{e}\mathrm{m}\mathrm{N}}_{\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}}{+\mathrm{R}\mathrm{e}\mathrm{m}\mathrm{N}}_{\mathrm{r}\mathrm{o}\mathrm{o}\mathrm{t}}+\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{N}

4.11 Statistical analysis

All data were analyzed using SPSS 22.0 (IBM Corp., Armonk, NY, USA). Two-way ANOVA (genotype × stage or genotype × year) was performed followed by Tukey’s HSD test for multiple comparisons (P < 0.05). Correlation analysis was performed using Pearson’s correlation coefficient. The linear-plateau model for grain yield versus SG level was fitted using SigmaPlot 14.0. For WGCNA, Pearson correlations and module-trait relationships were assessed with P-values adjusted by false discovery rate (FDR).

References

[1] Abdelrahman M , Papadopoulos A , Kunert K . et al. The role of salicylic acid in plant senescence and stress responses. Plant Physiol. Biochem., 2017, 119: 269–277

[2] Abu P , Badu-Apraku B , Ifie B E . et al. SeqSNP-based genic markers reveal genetic architecture and candidate genes for low nitrogen tolerance in tropical maize inbred lines. Front. Plant Sci., 2025, 16: 1558741

[3] Antonietta M , Fanello D D , Acciaresi H A . et al. Senescence and yield responses to plant density in stay green and earlier-senescing maize hybrids from Argentina. Field Crops Res., 2014, 155: 111–119

[4] Borrell A , Hammer G , Van Oosterom E . Stay-green: A consequence of the balance between supply and demand for nitrogen during grain filling. Ann. Appl. Biol., 2001, 138: 91–95

[5] Bolanos J , Edmeades G O . The importance of the anthesis-silking interval in breeding for drought tolerance in tropical maize. Field Crops Res., 1996, 48: 65–80

[6] Chen F , Fang Z , Gao Q . et al. Evaluation of the yield and nitrogen use efficiency of the dominant maize hybrids grown in North and Northeast China. Sci. China Life Sci., 2013a, 56: 552–560

[7] Chen X C , Chen F J , Chen Y L . et al. Modern maize hybrids in Northeast China exhibit increased yield potential and resource use efficiency despite adverse climate change. Glob. Change Biol., 2013b, 19: 923–936

[8] Chen Y , Xiao C , Chen X . et al. Characterization of the plant traits contributed to high grain yield and high grain nitrogen concentration in maize. Field Crops Res., 2014, 159: 1–9

[9] Chen Y L , Mu X H , Chen F J . et al. Dry matter partitioning and nitrogen accumulation in modern maize hybrids. Field Crops Res., 2015, 172: 1–10

[10] Chen Y , Wu D , Mu X . et al. Vertical distribution of photosynthetic nitrogen use efficiency and its response to nitrogen in field-grown maize. Crop Sci., 2016, 56: 397–407

[11] Ciampitti I A , Vyn T J . Physiological perspectives of changes over time in maize yield dependency on nitrogen uptake and associated nitrogen efficiencies: A review. Field Crops Res., 2012, 133: 48–67

[12] Ciampitti I A , Vyn T J . Grain nitrogen source changes over time in maize: A review. Crop Sci., 2013, 53: 366–380

[13] Cui Z , Zhang H , Chen X . et al. Pursuing sustainable productivity with millions of smallholder farmers. Nature, 2018, 555: 363–366

[14] DeBruin J L , Schussler J R. , Mo H . et al. Grain yield and nitrogen accumulation in maize hybrids released during 1934 to 2013 in the US Midwest. Crop Sci., 2017, 57: 1431–1446

[15] Du K , Zhao W , Lv Z . et al. Optimal rate of nitrogen fertilizer improves maize grain yield by delaying the senescence of ear leaves and thereby altering their nitrogen remobilization. Field Crops Res., 2024, 310: 109359

[16] Duvick D N , Cassman K G . Post-green revolution trends in yield potential of temperate maize in the North-Central United States. Crop Sci., 1999, 39: 1622–1630

[17] Duvick D N, Smith J S, Cooper M. 2003. Long-term selection in a commercial hybrid maize breeding program. In: Janick J, ed. Plant Breeding Reviews. New Jersey: John Wiley & Sons, 24.

[18] Duvick D N . The contribution of breeding to yield advances in maize. Adv. Agron., 2005, 86: 83–145

[19] Fan M , Shen J , Yuan L . et al. Improving crop productivity and resource use efficiency to ensure food security and environmental quality in China. J. Exp. Bot., 2012, 63: 13–24

[20] Gallais A , Hirel B . An approach to the genetics of nitrogen use efficiency in maize. J. Exp. Bot., 2004, 55: 295–306

[21] Gallais A , Coque M , Le Gouis J . et al. Estimating the proportion of nitrogen remobilization and of postsilking nitrogen uptake allocated to maize kernels by nitrogen-15 labeling. Crop Sci., 2007, 47: 685–693

[22] Good A G , Shrawat A K , Muench D G . Can less yield more? Is reducing nutrient input into the environment compatible with maintaining crop production. Trends Plant Sci., 2004, 9: 597–605

[23] Good A G , Beatty P H . Fertilizing nature: a tragedy of excess in the commons. PLoS Biol., 2011, 9: e1001124

[24] Gregersen P L. 2011. Senescence and nutrient remobilization in crop plants. In: Hawkesford M J, Barraclough P, eds. The Molecular Basis of Nutrient Use Efficiency in Crops. Oxford, UK: Wiley-Blackwell: 83–102.

[25] Gregersen P L , Culetic A , Boschian L . et al. Plant senescence and crop productivity. Plant Mol. Biol., 2013, 82: 603–622

[26] Haegele J W , Cook K A , Nichols D M . et al. Changes in nitrogen use traits associated with genetic improvement for grain yield of maize hybrids released in different decades. Crop Sci., 2013, 53: 1256–1268

[27] Han Y , Li X , Yan Y . et al. Identification, characterization, and functional prediction of circular RNAs in maize. Mol. Genet. Genomics, 2020, 295: 491–503

[28] Hirel B , Le Gouis J , Ney B . et al. The challenge of improving nitrogen use efficiency in crop plants: towards a more central role for genetic variability and quantitative genetics within integrated approaches. J. Exp. Bot., 2007, 58: 2369–2387

[29] Huang Y , Wang H , Zhu Y . et al. THP9 enhances seed protein content and nitrogen-use efficiency in maize. Nature, 2022, 612: 292–300

[30] Kurai T , Wakayama M , Abiko T . et al. Introduction of the ZmDof1 gene into rice enhances carbon and nitrogen assimilation under low-nitrogen conditions. Plant Biotechnol. J., 2011, 9: 826–837

[31] Langfelder P , Horvath S . WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics, 2008, 9: 559

[32] Lee E A , Tollenaar M . Physiological basis of successful breeding strategies for maize grain yield. Crop Sci., 2007, 47: S202–S219

[33] Liu Z , Gao J , Gao F . et al. Quantifying contributions of leaf area and longevity to leaf area duration under increased planting density and nitrogen input regimens during maize yield improvement. Field Crops Res., 2022, 282: 108551

[34] Martin A , Belastegui-Macadam X , Quillere I . et al. Nitrogen management and senescence in two maize hybrids differing in the persistence of leaf greenness: agronomic, physiological and molecular aspects. New Phytol., 2005, 167: 483–492

[35] Martin A , Lee J , Kichey T . et al. Two cytosolic glutamine synthetase isoforms of maize are specifically involved in the control of grain production. Plant Cell, 2006, 18: 3252–3274

[36] Masclaux-Daubresse C , Reisdorf-Cren M , Orsel M . Leaf nitrogen remobilisation for plant development and grain filling. Plant Biol., 2008, 10: 23–36

[37] Masclaux-Daubresse C , Daniel-Vedele F , Dechorgnat J . et al. Nitrogen uptake, assimilation and remobilization in plants: challenges for sustainable and productive agriculture. Ann. Bot., 2010, 105: 1141–1157

[38] Marschner P. 2011. Marschner's mineral nutrition of higher plants. London: Academic Press.

[39] Mi G H , Chen F J , Yuan L X . et al. Nitrogen use efficiency in maize: from physiology to breeding. J. Integr. Agric., 2016, 15: 1–12

[40] Mu X H , Chen Y L , Chen J . et al. Nitrogen allocation within leaves and its effect on photosynthetic nitrogen use efficiency in maize. J. Exp. Bot., 2016, 67: 1–12

[41] Mueller M , S T J . Maize plant resilience to N stress and post-silking N capacity changes over time: a review. Front. Plant Sci., 2016, 7: 53

[42] Pan X , Hasan M M , Li Y . et al. Asymmetric transcriptomic signatures between the cob and florets in the maize ear under optimal- and low-nitrogen conditions at silking, and functional characterization of amino acid transporters ZmAAP4 and ZmVAAT3. J. Exp. Bot., 2015, 66: 6149–6166

[43] Pommel B , Gallais A . et al. Carbon and nitrogen allocation and grain filling in three maize hybrids differing in leaf senescence. Eur. J. Agron., 2006, 24: 203–211

[44] Rajcan I , Tollenaar M . Source: link ratio and leaf senescence in maize: I. Dry matter accumulation and partitioning during grain filling. Field Crops Res., 1999, 60: 245–253

[45] Thomas H , Howarth C J . Five ways to stay green. J. Exp. Bot., 2000, 51: 329–337

[46] Thomas H , Ougham H . The stay-green trait. J. Exp. Bot., 2014, 65: 3889–3900

[47] Wang T , Li Y , Huang Y . et al. Amino acid permease 6 regulates grain protein content in maize. Crop J., 2022, 10: 1536–1544

[48] Wei Y M , Ren Z J , Wang B H . et al. A nitrate transporter encoded by ZmNPF7.9 is essential for maize seed development. Plant Sci., 2021, 308: 110901

[49] Wu J , Lawit S J , Weers B . et al. Overexpression of zmm28 increases maize grain yield in the field. Proc. Natl. Acad. Sci. USA, 2019, 116: 23850–23858

[50] Yang F , Zhang Y , Zhang H . et al. Comparative physiological and transcriptome analysis of leaf nitrogen fluxes in stay-green maize during the vegetative stage. Plant Physiol. Biochem., 2023, 204: 108147

[51] Yang J , Udvardi M . Nitrogen remobilization and stay-green in crops. Curr. Opin. Plant Biol., 2018, 45: 1–8

[52] Yang L , Guo S , Chen Q . et al. Use of the stable nitrogen isotope to reveal the source-sink regulation of nitrogen uptake and remobilization during grain filling phase in maize. PLoS ONE, 2016, 11: e0162201

[53] Yang L , Guo S , Chen F . et al. Effects of pollination-prevention on leaf senescence and post-silking nitrogen accumulation and remobilization in maize hybrids released in the past four decades in China. Field Crops Res., 2017, 203: 106–113

RIGHTS & PERMISSIONS

Higher Education Press 2026