Introduction

Wheat (Triticum aestivum L.), is one of the world’s most widely cultivated crops and plays an important role in ensuring national food security [1]. Global wheat productivity has been increasing gradually due to the continuous development of technology. However, wheat cultivation still faces a myriad of challenges such as lodging risks that threaten crop yield [2]. Wheat lodging is mainly caused by extreme and destructive weather events, which have been recently occurring more frequently. Stem lodging in wheat is a major agronomic problem that has far-reaching economic consequences by limiting the yields of cereals in both developed and developing countries [3, 4]. It is estimated that lodging leads to a loss of up to 60%–80% in the yield of wheat crops due to a reduced photosynthetic capacity resulting due to the destruction of the canopy structures [5].

Stem lodging usually occurs at the basal second internode, and it is correlated with plant height and stem strength [6]. For overcoming this problem, the wheat breeders have improved the lodging resistance (LR) through the technique of stem dwarfing. The “Green Revolution,” which saw an enhanced application of dwarf-plant-based breeding methods, has contributed to a sharp reduction in the lodging risk and an increase in the average yield of wheat [7]. Recent progress in Green Revolution has been made through the identification of the Reduced height (Rht) genes, including both the gibberellic acid (GA)-sensitive and GA-insensitive loci [8, 9]. In addition to reducing the plant height (PH), certain Rht loci also affect the architecture of the inflorescence and thus may influence grain yield (GY) [10, 11]. Quantitative trait loci (QTL) have also been reported for stem strength, culm wall thickness, pith diameter and stem diameter, which are associated with wheat lodging resistance [4, 12]. A single solid stem QTL was identified on chromosome 3BL, which also contributes to lodging resistance [13]. A common genomic region affecting overall stem strength, including internode material strength, internode diameter, and internode wall width, is located in the interval 278–287 cM of chromosome 3B [14]. Many QTL for plant height have also been identified with most of them located close to those reported Rht genes [15, 16]. The results of a comparative test between the winter wheat cultivars developed in the UK since the twentieth century revealed that, the PH was reduced and the GY was enhanced through genetic improvement [17]. Similarly, the PH was observed to gradually decrease with the application of breeding through the years since the 1960s in China [18].

The enhancement in the GY is hampered by the commonly occurring problem of crop lodging, which is perhaps underestimated and is constantly aggravating as the yield increases [19]. Improving the LR in crops by reducing the PH is limited only up to a certain extent in the currently used methods of crop production [20]. Several studies have suggested that the yield is reduced markedly when the PH is restricted to below a certain level using dwarfing genes [21, 22]. An unscientific approach to dwarfing can limit the GY, hence other methods of improving the LR need to be considered. One such method is the modification of the structure of the stem tissues [23]. Previous studies on the microstructure of stems using the cross-sections of the basal regions of the stems revealed that the tissues were highly lignified and the stems organized with a significant number of large- and small-sized vascular bundles of increased area thus improving the mechanical strength in the lodging resistant wheat cultivars [13]. Fiber which mainly consists of cellulose, hemicelluloses, and lignin plays an important role in the mechanical support of the cell wall [24]. Among these components, cellulose is the main one and has contributed to maintaining culm mechanical strength. Several qualitative characteristics such as crystallinity contribute to the strength of cellulose fibers [25]. Furthermore, the levels of certain mineral nutrients such as K and Si were closely correlated to LR [26, 27]. The anatomical structure and the mineral nutrient levels in the stem have significant impacts on the LR; however, the causes behind these variations which occur during the breeding process are currently ambiguous.

Undoubtedly, the genetic improvement of crops and the addition of chemical-based fertilizers especially nitrogen (N) fertilizer are the two most important factors affecting agricultural output [28]. Previous studies have revealed the mechanisms behind the influence of N in improving the yield and quality of crops from the perspective of agronomic traits, photosynthesis and physiology, ecological effects, and nutrient absorption and transport [29, 30]. Intensive methods of crop management, such as excessive application of N fertilizers and a high planting density enhanced the susceptibility of plants to lodging, which was mainly due to an increased stem length, decreased stem diameter, and reduced cell-wall thickness (WT) that may diminish the parameters of flexural rigidity and breaking resistance which indicate physical strength [31]. Additionally, N luxury significantly reduced the generation of H, G, and S monomers of the lignin (related to stem hardness) as well as its total content in crop stem [32]. Reduced breaking strength and higher lodging index in wheat under higher N were associated with reduced cellulose contents in the culm [33]. Some microelements (Si, Ca, Mg, Fe, Zn, Cu, and Mn) in the culm can also be affected by N application, resulting in lower breaking strength and higher lodging index in rice plants, in particular breaking strength of culm which plays a key role in lodging index [34].

The production of biomass and GY were concomitantly reduced with the decreasing application of N [35]. Achieving a simultaneous increase in both the GY and LR seems to be difficult in wheat. However, studies have been conducted to understand the trade-off between the two and assess the possibility of managing the lodging risk without an associated reduction in yield [13]. In this study, the differences in the GY and LR in selected wheat varieties obtained through genetic improvement under varying regimens of N fertilizer were undertaken. The main objective was to explore the effect of agronomic characters and physiological mechanisms (in terms of stem anatomical structure, mineral element and cellulose) on the trait of LR caused due to genetic improvement and N regimes. The findings can help in understanding the adaptability of genetically improved crops to N fertilizer and provide a scientific reference that can be accessed before initiating crop breeding for the efficient utilization of resources in the future.

Results

Improvement of yield and yield-related parameters of the varieties developed over the decades

Two-way ANOVAs revealed that the grain yield was markedly affected by the variety (V), N rate (N) and their interaction effect (V × N) during 2019–2020 and 2020–2021 (Fig. 1). Compared with the cultivar developed in the 1950s, the yield of the cultivars developed in the 1960s, 1970s, 1980s, 1990s, 2000s, 2010s, and 2020s demonstrated an increase of 9.7%, 35.5%, 68.5%, 60.9%, 70%, 71.3%, and 75.6% during the experimental period of 2019–2020 respectively; and of 3.9%, 33%, 55.4%, 54.9%, 47.1%, 55.4%, and 61.6% during 2020–2021, respectively. The mean decadal rate of increase in the yield of the cultivars from the 1950s–1980s was 1.81 kg ha−1 decade−1 during the experimental period of 2019–2020, and 1.56 kg ha−1 decade−1 during 2020–2021. The mean decadal rate of increase in the yield of the cultivars from the 1990s–2020s was 3.32 kg ha−1 decade−1 during 2019–2020, and 2.78 kg ha−1 decade−1 during 2020–2021. The application of N fertilizer significantly increased the GY in all eight cultivars, but excessive input of N did not further improve the GY but inhibited it. In comparison with the N0 treatment, the GY at N180 increased by 65.1% (the 1950s), 54.7% (1960s), 40% (1970s), 33.8% (1980s), 26.8% (1990s), 46.9% (2000s), 54.3% (2010s), and 38% (2020s) during the experimental season of 2019–2020; and by 65% (1950s), 81.6% (1960s), 24.7% (1970s), 20.6% (1980s), 40.8% (1990s), 64% (2000s), 74.5% (2010s), and 45.5% (2020s) during the experimental season of 2020–2021. At N360, the GY of all the cultivars improved less rapidly or even declined when compared to that at N180.

Fig. 1
figure 1

Yield of the different wheat varieties during 2019–2020 (A) and 2020–2021 (B). ** indicate that the yield is significantly influced by the variety, N rate and their interactions at 0.01 levels

Full size image

Two-way ANOVAs revealed that the number of panicles per m2 (NPM), the number of grains per panicle (NGP) and 1000-grain weight (TGW) were markedly affected by variety (V) and N rate (N). Also, the effects of V and N interactions (V × N) on NPP and NGP were highly significant (Table 1). NPM, NGP and TGW are crucial indicative parameters indicating the yield of a crop. The NGP and TGW increased with era processing, whereas the NPM decreased obviously. The NPM in all the cultivars enhanced significantly with the increasing input of N (Table 2). Except in the case of the cultivars from the 1950s and 1970s, the NGP was markedly elevated at N180 over the N0 treatment but then reduced at the N360 treatment. The TGW was marginally affected by the addition of N fertilizer, the order being N180 > N360 > N0.

Table 1 Yield parameters in the wheat varieties after growth under different N regimens during the experimental seasons of 2019–2020 and 2020–2021. The least significant difference (LSD) test was performed to compare the mean values of the various treatments. The lowercase letters after the numbers in the same column indicate that the differences between the three N levels for the same variety were significant at the level of 0.05, and the capital letters after the numbers in the same column indicate that the differences between the eight varieties were significant at the level of 0.05. ** indicate that the yield components were significantly influenced by variety, N rate, and their interactions at 0.01 levels, and ns indicates ‘not significant’
Full size table
Table 2 Mechanical tissue thickness, wall thickness, cellulose content, and anatomical structure of the stems of the wheat varieties after growth under different N regimens. The number of small vascular bundles, NSVB; the number of large vascular bundles, NLVB; the area of small vascular bundles (μm2), ASVB; the area of large vascular bundles (μm2), ALVB; cellulose content, CC; mechanical layer thickness (μm), MLT; wall thickness (μm), WT. The least significant difference (LSD) test was performed to compare the mean values of the various treatments. The lowercase letters after the numbers in the same column indicate that the differences between the three N levels for the same variety were significant at the level of 0.05, and the capital letters after the numbers in the same column indicate that the differences between the eight varieties were significant at the level of 0.05. ** indicate that the anatomic parameters were significantly influenced by variety, N rate, and their interactions at 0.01 levels, and ns indicates ‘not significant’
Full size table

Lodging rate and lodging resistance index (LRI)

During both the experimental seasons, natural lodging was observed in the three cultivars- ND2419, BJ8, and XY4 but not in others (Fig. 2). The plants of these three cultivars did not demonstrate lodging at N0 but at the N360. This suggested that the LRI increased through genetic improvement; the LRI of the cultivars from the 2000s, 2010s, and 2020s was 138.6% (2019–2020) and 27.4% (2020–2021) higher than those from the 1950s–1990s (Fig. 3). Furthermore, the LRI was observed to be negatively correlated with the amount of N fertilizer.

Fig. 2
figure 2

Pictures of the wheat field taken at the grain-filling stage during 2020–2021. Yellow values represent the two-year average wheat lodging incidence

Full size image
Fig. 3
figure 3

Lodging resistance index (LRI) of the different wheat cultivars at the grain filling stage during the experimental seasons of 2019–2020 (A) and 2020–2021 (B). ** indicate that the LRI is significantly influced by the variety, N rate and their interactions at 0.01 levels

Full size image

PH, PCG, and the CC of the stems

The PH and PCG gradually decreased with the ear processing (Fig. 4A–F). The application of N fertilizer increased the PH and elevated the PCG in all the cultivars. Compared to the control, the increase in the height of the plants of the cultivars from the1950s, 1960s, 1970s, 1980s, 1990s, 2000s, 2010s, and 2020s at N180 was 8.5%, 11.2%, 11.4%, 9.1%, 2.2%, 9.8%, 7.2%, and 2.3% during 2019–2020, respectively; and was 4.2%, 10.9%, 13.3%, 6.5%, 3.3%, 3.2%, 5.7%, and 5.0% during 2020–2021, respectively (Fig. 4B, E). Similarly, in comparison to N0, the elevation in the PCG at N180 was enhanced by 8%, 15.5%, 21.1%, 7.9%, 11.6%, 8.2%, 8.9%, and 11.4% during 2019–2020, respectively; and by 7.4%, 7.3%, 14.2%, 20.3%, 4%, 2.3%, 4.5%, and 15.7% during 2020–2021. It is worth noting that the cultivar from the 1970s had the highest increase in the PH and elevation of the PCG with the addition of N fertilizer. This effect was lower in the N360 vs. N180 group than in the N180 vs. N0 group (Fig. 4C, F). At N180, the maximum increase in the PH and elevation in the PCG were 6.6% and 11.5% during 2019–2020; and 9.8% and 12.3% during 2020–2021, respectively.

Fig. 4
figure 4

Plant height and plant center of gravity of the wheat varieties at different ages under the three N regimens during 2019–2020 (A–C) and 2020–2021 (D–F). The different lowercase letters indicate significant differences (p < 0.05) among different varieties at the same N condition. The different uppercase letters indicate significant differences (p < 0.05) among different N rates for the same variety

Full size image

In the cultivars from the 1950s–1990s, the CC did not change significantly during 2019–2020, but it began to decline in those from the 2010s (Table 2). The average CC of the cultivars from the 1950s–2000s was 7.0% (2019–2020) and 8.0% (2020–2021) higher than those from the 2010s–2020s. The CC demonstrated a reciprocal relation with N. Compared to N0 treatments, the cellulose content in the cultivars from the 1950s, 1960s, 1970s, 1980s, 1990s, 2000s, 2010s, and 2020s at N360 was reduced by 4.8%, 7.3%, 6.5%, 7.9%, 7.0%, 13.1%, 8.8%, 10.2% during 2019–2020, respectively; and enhanced by 6.1%, 10.3%, 11.3%, 13.6%, 12.1%, 11.1%, 6.6%, and 12.8% during 2020–2021, respectively.

SBS and microstructure of the second internode of the stems

Two-way ANOVAs revealed that the SBS and microstructure were markedly affected by the variety (V) and N rate (N), and there was interaction effect on the SBS, NSVB, MLT and WT. No significant changes in the SBS among the wheat cultivars from the 1950s–1980s were observed in both the experimental seasons (Fig. 5). However, in comparison to these, the SBS of the cultivars from the 1990s–2020s substantially increased by 45.7% (2019–2020) and 40.1% (2020–2021) on an average, respectively. Application of N fertilizer resulted in a reduction of the SBS, the mean SBS at N180 reduced by 13.3% and 16.5% in comparison to that at N0; and by 13.7% and 14.6% at N360 in comparison with that at N180.

Fig. 5
figure 5

Stem breaking strength (SBS) of the second internode of the different wheat varieties during 2019–2020 (A) and 2020–2021 (B). ** indicate that the SBS is significantly influced by the variety, N rate and their interactions at 0.01 levels

Full size image

The mean WT of the second internode of the cultivars from the 2000s, 2010s, and 2020s was 42.7% (2019–2020) and 52.4% (2020–2021) higher than that of the cultivars from the 1950s–1990s (Table 2). Furthermore, the WT demonstrated an inverse correlation with N fertilizer. Compared with that at N0, the mean WT at N180 decreased by 15.8% (2019–2020) and 16.5% (2020–2021); while that at N360 decreased by 25.4% (2019–2020) and 23.4% (2020–2021). Similarly, the mechanical layer thickness (MLT) of the cultivars from the 2000s, 2010s, and 2020s was on average 22.7% (2019–2020) and 23.0% (2020–2021) higher than that of the others. However, the response of MLT to N fertilizer was opposite to that of the WT. Compared with that at N0, the mean MLT at N180 increased by 14.1% (2019–2020) and 11.9% (2020–2021); and that at N360 enhanced by 28.7% (2019–2020) and 27.8% (2020–2021).

There appeared to be no clear trend in the number of small vascular bundles (NSVB) and large vascular bundles (NLVB) among the different wheat cultivars (Table 2). However, a remarkable increase in both was observed after the addition of N. Compare to N0, the mean NSVB at N180 increased by 23.5% (2019–2020) and 20.9% (2020–2021); at N360 enhanced by 35.2% (2019–2020) and 31.8% (2020–2021). The mean NLVB at N180 increased by 7.9% (2019–2020) and 11.8% (2020–2021); at N360 enhanced by 10.4% (2019–2020) and 9.6% (2020–2021).

Similarly, no clear trend in the area of the SVB (ASVB) among different wheat cultivars was observed, whereas marked differences were found after the addition of N fertilizer (Table 2). Compare to that at N0, the mean ASVB at N180 increased by 21.9% (2019–2020) and 23.6% (2020–2021); that at N360 was enhanced by 40.7% (2019–2020) and 48.1% (2020–2021). In contrast, the mean area of LVB (ALVB) of the cultivars from the 2000s, 2010s, and 2020s was 32.6% (2019–2020) and 21.1% (2020–2021) higher than that of the cultivars from the 1950s–1990s. Compared to N0, the mean ALVB at N180 decreased by 5.8% (2019–2020) and 4.6% (2020–2021); and at N360 was reduced by 9.7% (2019–2020) and 10.6% (2020–2021).

Relationship between the agronomic traits, the microstructure, and mineral elements of the stem

Figure 6 illustrates the correlation between the lodging-related traits, stem microstructure, and mineral elements. LRI was significantly correlated in a positive manner with two agronomic traits (SD and SBS), four microstructure parameters (NLVB, ALVB, WT, and mechanical tissue thickness [MTT]), and three mineral elements (Mn, Cu, and Zn); but it was significantly negatively correlated with PH, PCG, and CC. Compared to the SVBs, the LVBs were significantly correlated with more agronomic traits and mineral element content, and the AVB was more significantly correlated with more parameters in comparison with the NVB. The elements- Mn, Cu, and Zn had a significant correlation with several agronomical and microstructure traits; while K, Ca, Fe, and Si had a significant correlation with only a handful of traits; and Mg had the least correlation with any of the traits.

Fig. 6
figure 6

The correlation between the three N regimens and the eight varieties of wheat on the lodging resistance index and related parameters in stems. LRI, lodging resistance index; PH, plant height; SD, Stem diameter; PGC, the plant center of gravity; SBS, stem breaking strength; WT, Wall thickness; Cellulose content, CC; NSVB, number of small vascular bundles; NLBV, number of large vascular bundles; ASVB, area of small vascular bundles; ALBV, area of large vascular bundles; MTT, Mechanical tissue thickness. The chemical symbols represent the mineral elements occurring in the second internodes of the stems. * and ** indicate significant differences at the levels of 0.05 and 0.01, respectively

Full size image

Discussion

Physiological mechanisms underlying the genetic improvement of wheat for enhancing the LR

Breeders have used dwarf and semi-dwarf phenotype-related genes to cultivate dwarf varieties, with a significantly reduced PH, improved harvest index, and enhanced LT; to date, 25 Rht genes have been identified in wheat, the use of which ultimately led to the development of varieties with higher GYs [8, 36]. The Green Revolution contributed to a sharp increase in the average yield of wheat over the past 50 years [37, 38]. The average wheat yield in China increased from 1840.5 kg ha−1 in 1978 to 5740.5 kg ha−1 in 2021, which was ~ 1.6 times the global average yield [28]. In this study, the yield increase in the cultivars from the 1990s–2020s was ~ 1.8 times higher than that of the cultivars from the 1950s–1980s. The yield increase achieved through genetic improvement was primarily attributed to the continuously improving grain production capacity (grain number and weight), which was accompanied by a decrease in the NPP (Table 1; Fig. 1). Wide adaptability and a strong capacity of resistance to abiotic and biotic stresses are the crucial reasons behind the significant increase in wheat yield in China after the 2000s; some dwarf, high-yielding, and disease-resistant varieties such as Zhou 8425B and 6VS/6AL play an important role when used as parents in a breeding program [39].

Lodging in plants is a crucial constraint for the stable yield and productivity of crops and is a problem that needs to be solved urgently for breeding superior, high-yielding varieties in wheat [40]. In this study, the PH remained stable in the varieties developed after the 2000s, however, the LRI was continuously improving (Figs. 4 and 5). The SBS of the modern varieties (1990s–2020s) was 45.7% higher than that of the older varieties (1950s–1980s) (Fig. 5) as it was the most dominant trait addressed in LRI improvement programs since the 2000s. Further, the SBS was significantly positively correlated with the SD and the anatomical characteristics. A previous study also revealed that 99% of the variation in the LR could be explained by the width of the MLT, thus confirming our results [18].

The accumulation of lignin and cellulose contributed to the improvement of LR [41]. In this study, the CC of the stems was also positively correlated with the SBS. It is worth mentioning that the modern varieties had a lower CC, but a higher SBS (Table 2; Fig. 5) because the increase in stem WT compensated for the enhanced risk of lodging caused by a decrease in the CC. Certain mineral nutrients mainly K, Ca, and Si were significantly associated with stem anatomical characteristics and have been widely used in improving LR (Fig. 6). A higher K content could promote the lignification of the cells of the collenchyma and sclerenchyma and increase the accumulation of structural carbohydrates to thicken the cell wall in the stem [42]. K deficiency could cause a reduction in length, diameter, and wall thickness of culms, poor root proliferation thus increasing lodging through reduced stem strength and diameter [43]. Similarly, Ca and Si deposition in epidermal cells can also increase the cellulose and lignin contents to improve the hardness, toughness and stem elasticity [42, 44]. In addition, Mn, Cu, and Zn are also associated with the SBS, which may be due to the influence of these elements on the enhancement of the MTT, WT, and ALVB (Figs. 6 and 7). However, the mechanisms underlying the effects of these three elements on the LR have not yet been elucidated.

Fig. 7
figure 7

The rationale behind the lodging of wheat in the different released years (A) under different N regimens (B). NSVB, number of small vascular bundles; NLVB, number of large vascular bundles; ASVB, area of small vascular bundles; ALVB, area of large vascular bundles; CC, cellulose content; SBS, stem breaking strength. The chemical symbols represent the mineral elements occurring in the second internodes of the stems

Full size image

Effect of N on the plasticity of LR in the different wheat varieties

The addition of N fertilizer improved the GY, but the excessive application could not further increase productivity and even negatively affected it. In this study, the application of N enhanced the yield in all the varieties and can be summarized as the 1950s–1960s > 2000s–2020s > 1970s–1990s (Fig. 1). From the perspective of yield parameters, the increased NPP was crucial in improving the yield in the older varieties, while it was both the increased NGP and NPP in the modern varieties.

Previous studies have shown that excessive application of N fertilizer increased the risk of lodging in rice [6], maize [45], and wheat [13]. The application of N led to an elevation in the levels of the endogenous GAs, which caused an elongation of the culm and hence an increase in the lodging risk [46]. In this study, it was observed that the increase in PH and the elevation in the PCG caused by N application was significantly lower in the varieties developed after the 1990s (Fig. 4). This indicates that the modern semi-dwarf varieties might be less sensitive to GA. Although the PCG increased slightly, the LRI of the modern varieties decreased significantly after N application (Fig. 3). This implies that the SBS reduced rapidly in the modern varieties after N addition. Among the various parameters that affected the SBS, the WT, the LVBA, and the CC exhibited a stronger sensitivity to N in the modern varieties. Crops with high soluble sugar content in culms are able to recover more easily from damage by strong winds and its remobilization keeps the plant alive for a longer duration during stress conditions [47, 48]. Excessive N input increased the content of soluble sugars in the leaves. However, the stems could not fully utilize these substrate to synthesize sufficient amounts of the structural carbohydrates lignin and cellulose [33]. On a molecular level, high application of N fertilizer down-regulates the expression of the genes involved in lignin and cellulose biosyntheses, causing fiber deficiency in the secondary cell walls and the weakening of mechanical tissue structure [49]. This phenomenon might be more pronounced in the modern varieties. Simultaneously, in this study, it was found that although the application of N induced the accumulation of K, Ca, and Si, the SBS significantly decreased (Fig. 5; Fig. 7). Numerous studies have confirmed that K, Ca, and Si can enhance the SBS [50, 51]. It can be concluded that the effects of mineral elements on the improvement of LR were lesser than that on the stem structure. In summary, two different types of wheat varieties are needed for the breeding-based improvement of LR in the future, the first variety with a stem anatomical structure insensitive to N fertilizer and the second variety with a higher capacity for the utilization of monosaccharides in the synthesis of cellulose and lignin polysaccharides.

Conclusion

For the improvement of the LR, the first stage of breeding mainly depended on a decrease in the PH and an increase in the contents of cellulose, Si, and Zn; the second stage mainly depended on a sharp increase in the SBS, due to enhanced SD, stem WT, K, Ca, Fe, Mn, and Cu levels (Fig. 7A). The application of N enhanced the concentrations of the three-lodging resistance-related elements K, Ca, and Si but the LR was reduced, which was mainly due to an increase in the PH and a reduction in the LVBA and CC (Fig. 7B). This study points out a direction for improving the traits associated with increased LR through breeding and the cultivation of high-yield varieties in the future.

Methods

Experimental site

Field experiments were conducted at the scientific research site of Henan Agricultural University (35°18′N, 113°95′E), Xinxiang County, Henan Province, China during the winter wheat cultivation season lasting from October 2019 to June 2021. During 2019–2020, the mean annual temperature was 12.1 °C, the mean winter temperature was 3.6 °C, and the total annual rainfall was 111.5 mm. During 2020–2021, the mean annual temperature was 11.2 °C, the mean winter temperature was 2.9 °C, and the total annual rainfall was 135.4 mm (Fig. 8). The soil texture was sandy, pH was 6.8, the organic matter content was 13.36 g kg−1 (top 0–20 cm), the total N was 1.23 g kg−1, the available K was 88.95 mg kg−1, the available P was 20.2 mg kg−1, and the bulk density was 1.34 g cm−3.

Fig. 8
figure 8

Temperature and precipitation values during the experimental seasons of 2019–2020 (A) and 2020–2021 (B)

Full size image

Experimental design

Eight popular cultivars of wheat developed from the 1950s to the 2010s in the plains of North China were selected. These included Nanda 2419 (ND2419), Beijing 8 (BJ8), Zhengyin 1 (ZY1), Xiaoyan 4 (XY4), Bainong 3217 (BN3217), Yumai 2 (YM2), Bainong 207 (BN207), and Bainong 4199 (BN4199) (Table 3). The experimental design included split plots in three replications, with the dimensions of each plot being 6.0 m × 8.0 m. A total of 187.5 kg seeds ha−1 were mechanically sown on October 15th, 2019, and 2020. The plants were grown in the main plots, and the three N regimens- 0 (no N fertilizer application), 180 (recommed rate for wheat in Henan province), and 360 kg N ha−1 (farmer practise fertilization) were applied to the subplots, with the yield performance of the different cultivars in the field determined at the grain filling stage (Fig. 9). N (as urea) was applied at two intervals, with 50% at the basal and 50% at the jointing stage (about 160 days after sowing; DAS). P (90 kg P2O5 ha−1) and K (90 kg KCl ha−1) were applied as the basal dose. The plants were harvested on June 1st, 2020, and June 3rd, 2021. Weeds, diseases, and insects were intensively controlled during the entire growing season to avoid yield loss.

Table 3 The wheat cultivars used in this study. The recommended nitrogen dosage is based on yield and nitrogen application, and fitted through a linear plus plateau equation
Full size table
Fig. 9
figure 9

Field layout of the split-plot design experiment

Full size image

Plant sampling and measurement of parameters

GY and yield parameters

At maturity, 6 m2 plants for each plot were harvested by manual, and then treshed by machine. The yield of wheat grains was adjusted to 13% moisture content. To determine the grain number per spike, twenty spikes from each plot were randomly selected. The 1000-grain weight was determined after the grains were dried to a constant weight at 80 ℃ in a forced ventilation oven.

PH and plant center of gravity (PCG)

The PH was measured from the base to the tip of the highest panicle. The center of gravity of a plant was defined as the distance from the base to the equilibrium pivot point of an entire plant with spikes, leaves, and sheaths intact [13].

Lodging rate and lodging resistance index

The lodging period and lodging area during the growth period were observed and recorded post-lodging. The culm characteristics related to stem lodging were determined at the grain-filling stage. The second internode from the base of the freshly cut stem was collected, the leaf sheath was stripped, and the stem breaking strength (SBS) was determined using a YYD-1A collapse resistance meter (Zhejiang Top Cloud-Agri Technology Co. Ltd., Hangzhou, China) [33]. The LR and LRI were calculated as:

$$\mathrm{Lodging rate }\left(\mathrm{\%}\right)=\mathrm{actual area of lodged plants in a plot }({\mathrm{m}}^{2})/\mathrm{area of the plot }({\mathrm{m}}^{2})\times 100\mathrm{\%}$$
(1)
$$\mathrm{LRI}=\mathrm{SBS }(\mathrm{N}) /\mathrm{ PCG }(\mathrm{cm})$$
(2)

where LRI is the lodging resistance index, SBS is the stem-breaking strength, and PCG is the plant center of gravity [52].

Microstructure of the basal second internode of the stem

Three uniformly sized plants of each N-treatment group were selected at the grain-filling stage. The stem segments from the middle of the basal second internode were quickly cut with a double-sided blade and fixed in an formaldehyde-acetic acid–ethanol fixative (FAA) fixation solution (containing 70% ethanol: 5% glacial acetic acid: 3.7% formaldehyde) [53]. The images of the prepared slices were collected by the slice scanner and analyzed by using the CaseViewer slice-scanning software (3DHISTECH Kft., Budapest, Hungary). This software was used to measure the WT, tissue mechanical thickness (TMT), and the number and size of the big/small-sized vascular bundles.

Estimation of the cellulose content (CC) of the stems

The stem samples were collected at the grain-filling stage, and the total CC was determined according to an already-described method with slight modifications [54]. 1 g of dried straw from the second internode was transferred to a 50 mL test tube to which 5 mL of a solution containing acetic acid and nitric acid in a ratio of 1:1 was added. The tubes were kept in boiling water for 30 min. Subsequently, 30 mL of distilled water was added, the tubes were centrifuged at 4000 rpm for 10 min, and the supernatant was separated. This step was repeated two more times. The collected sediments were dried in an oven at 80 °C to reach a constant weight. Then 10 mL of a solution containing 10% H2SO4 and 0.1 mol L−1 of K2Cr2O7 was added, shaken adequately, and kept in boiling water for 15 min. The precipitate was then rinsed with 30 mL of distilled water, transferred to clean Erlenmeyer flasks (Shenzhen H&Q Biologics Co. Ltd., Guangzhou, China), and allowed to cool down for titration. Finally, 5 mL of 20% KI was titrated against 0.2 mol L−1 Na2S2O3 until the solution started to turn blue and retained the same color for 30 s. The control was without straws. The CC was calculated as follows:

$$\mathrm{CC}=\mathrm{Concentration of }{\mathrm{Na}}_{2}{\mathrm{S}}_{2}{\mathrm{O}}_{3} ({\mathrm{mol L}}^{-1})\cdot (\mathrm{V}-{\mathrm{V}}_{0})/({\mathrm{W}}_{\mathrm{straw}}\times 24)$$
(3)

where V is the volume of Na2S2O3 (mL) consumed by titration with the control, V0 is the volume of Na2S2O3 (mL) consumed by titration with the test sample containing straw. Wstraw is the dry weight of the straw (g).

Estimation of the levels of mineral nutrients in the stems

The plant samples were collected at the grain-filling stage, the stems were stripped of the leaf sheaths, dried at 70 °C, and then crushed to a powder. The levels of the mineral nutrients- K, Ca, Mg, Fe, Mn, Cu, Zn, and Si in the stems were determined using the MARS6 microwave digestion instrument (CEM Corporation, North Carolina, USA) and a Prodigy7 inductively coupled plasma spectrometer (Teledyne Leeman Labs, Ohio, USA).

Data analysis

Data were analyzed using the Statistical Software Package for Social Science (SPSS, ver 19.0) (IBM, New York, USA) and MS Office Excel 2010, the variance of analysis and mean value (n = 3) were compared using Fisher’s F-protected least significant difference (LSD) test at a P < 0.05. The graphs were plotted using the Origin 9.0 software (OriginLab Corporation, Massachusetts, USA).