Regulation of maize growth, nutrient accumulation and remobilization in relation to yield formation under strip-till system

ABSTRACT A two-year field experiment was conducted to elucidate the adaptive growth mechanism of maize under strip-till (ST) compared with conventional-till (CT). The biomass accumulation of ST plants was significantly lower than CT until V14 (14th leaf), but restored thereafter with one below-ear-node leaf reduced. At silking, the accumulation of nitrogen (N) was reduced by 8.3–10.7% compared to CT. During post-silking, vegetative-N remobilization was reduced by 20.4%, post-silking N uptake increased by 33.9% in ST compared to CT. Leaf senescence was delayed and more green leaf area at physiological maturity in ST. It is concluded that ST plants have the mechanism of ‘Recovery Growth Adaptation’ to get the similar yield as in CT plants: (1) facilitating growth rate at around V14 when the soil temperature was greatly improved to stabilize ear growth and grain number; (2) getting to silking the same time as in CT plants, so as to ensure the duration of grain filling; (3) increasing post-silking N uptake to fulfill the demand of grain development and reduce leaf N remobilization, so as to maintain leaf function and increase thousand-grain weight, which compensate for the loss of grain number per ear.


Introduction
China is the world's second largest maize producer after the United States. The total production of maize in Northeast China Plain accounts for 32% of China (National Bureau Statistics 2021). For the past decades, farmers in this area have been adopting conventional tillage (CT) in which maize residues are removed for various purpose and the barren soils are cultivated before maize planting. As a result, the black soil in Northeast China Plain is facing serious degradation, and is becoming 'thining, barren and hardening' (Li et al. 2021). To alleviate soil degradation, conservation tillage such as no-till (NT) with straw mulching has been suggested as alternatives to CT (He et al. 2021). However, as found in the literature, NT normally have a negative effect on maize yield in the black soil in Northeast China, depending on years and soil conditions, due to poor sowing quality (Liu et al. 2017;Chen et al. 2021), low soil temperature (Ma and Rivero 2010;Page et al. 2019) and compacted soil (Afzalinia and Zabihi 2014;Peigne et al. 2018), etc.

Experimental design and field management
The experiment was a single-factor that completely random design with three replications. Two tillage treatments were set up in the spring of 2018: strip tillage (ST) and conventional tillage (CT). The size of each plot was 76 m in length and 7.8 m in width, with 12 rows spaced 0.6 m apart. According to the nutrient requirements of the local spring maize (Wang et al. 2018) and the optimal fertilizer recommendation (Feng et al. 2017), the following fertilizers were applied before planting: 180 kg N ha −1 (as urea), 75 kg P 2 O 5 ha −1 (as superphosphate), 90 kg K 2 O ha −1 (as potassium chloride).
The field was covered with maize residues that left from the previous maize crop. In 27 April 2019 and 2 May 2020, strip tillage was conducted using a strip-tiller (Yetter-2984, Illinois, USA). A clean soil strip about 12-15 cm in depth and 22-25 cm in width was created for planting and the residues were moved to the inter-row area (between the rows) and were not disturbed through maize growth period ( Figure 1S-A). For conventional-till, the residues were removed from the field in 26 April 2019 and 1 May 2020, next day the soil was tilled using a rotary tiller to a depth around 15 cm ( Figure 1S-B). Maize was sown on May 9 and May 7 in 2019 and 2020, respectively, using a no-till planter (KangDa, Jilin, China). The planting density was 70,000 plants per ha with the distance between plant of 0.24 m, and the inter-row distance of 0.60 m. Maize hybrid De-Mei-Ya-3 (KWS, Heilongjiang, China) was used, which GDUs (Growing Degree Units) is about 2200°C from emergence to maturity of 113 days. Herbicide and pesticide were used to control the weeds and pests. No irrigation was applied. Maize was harvested on September 30 and September 27 in 2019 and 2020, respectively.

Seedling emergence
Emergence rate (ER) was determined when 60% of seeds had emerged (VE) in a treatment. An emergence rate index (ERI) was determined using a method outlined by Erbach (1982) in which two rows, 10 m in length, were marked before maize emergence and monitored the emergence rate (ER) each day for 10 consecutive days after the first emergence. ERI was calculated using the following equation (Erbach 1982): where n is the number of days after planting, first is the number of days after planting when the first plant emerged, last is the number of days after planting when emergence was completed, % n is the percentage of plants emerged on day n, and % (n − 1) is percentage of plants emerged on day n − 1.

Grain yield and its components
At physiological maturity, two center rows (10 m length and 1.2 m width) of each plot were harvested by hand. The number of ears obtained from yield measured area was converted into the ear number per ha (EN). From among the harvested ears, 20 were chosen to measure the grain number per ear (GN). Grains were threshed by a grain thresher, and three 100-grain samples were weighted to determine the 1000-grain weight (TGW). Harvest index (HI) was calculated according Wnuk et al. (2013). HI (%) = (grain yield/ biological yield) × 100 (2)

Plant sampling, dry matter measurement and growth analysis
Plants were sampled at V6, V9, V14, R1, R2, R3 and R6 growth stages based on CT treatment (D'Andrea et al. 2008). For each sampling, three successive plants from each plot were cut at the soil surface (not the yield rows). At R6 stage, the plants were separated into straw (including leaves, stalk, and sheaths) and grain. All plant samples were dried in an oven (DHG-9420A; Bilon Instruments Co. Ltd., Shanghai, China) at 85 ± 5°C (three or four days for V6, four or five days for V9, six or seven days for V14 and R1, more than seven days for R2-R6) after heating at 105°C for 30 min, to a constant weight and the dry weight was then measured. The length and the maximum width of each leaf were measured using a ruler. Green leaf area and the related parameters were calculated as follows (Gallais et al. 2006): Leaf area index (LAI) was the total green leaf area per unit land area.

Soil temperature and soil water content
A portable thermometer (TP -101, XinTai wei, China) was used to measure soil temperature at different soil depths. Soil temperature at both intra-row and inter-row was recorded at 9 am and 2 pm from 13 April to 21 July in 2019 and 23 April to 19 July in 2020. Also in the intra-row and the interrow, a manual soil drill was used to sample soil cores at 0-5 cm, 5-10 cm, 10-15 cm, 15-20 cm, 20-30 cm, 30-40 cm depths. Soil water content was measured as Li et al. 2020.

Statistics
All data were analyzed with analysis of variance using SPSS 19.0 (SPSS Inc., Chicago, IL, USA) to examine the effects of tillage systems on maize growth, nutrients, soil physical properties, grain yield and its components. Comparisons among different treatments were performed with the Least Significant Difference (LSD). A P-value ≤ 0.05 was considered significant. Correlations between staygreen degree at R6 and nutrient remobilization efficiency, nutrient accumulation after R1 were tested with a Pearson rank correlation. All the graphics were made using SigmaPlot 14 (Systat Software Inc., CA, USA).

Soil temperature and soil water content
As shown in Figure 2, before tilling (April 27) in 2019, soil temperature under ST was 9.8°C and 6.7°C lower than that of CT at 5 cm and 10 cm soil depths, respectively. After strip tillage, the soil temperature in the planting strip of ST was increased rapidly, because the straw residue was removed into the inter-row. During the vegetative growth stage (VE -R1), the soil temperature at 5 cm and 10 cm soil depths in the intra-row of ST was only 0.8 and 0.9°C lower than that of CT, respectively. In the inter-row soil, however, the temperature of ST was 4.5 and 2.4°C lower on average compared to CT at 5 cm and 10 cm soil depths, respectively. The spring in 2020 was colder than in 2019 ( Figure 1), which led to lower soil temperature in both tillage systems. But the trend in the difference of soil temperature between the two tillage treatments was similar as in 2019. The accumulated soil temperature (>10°C) during VE -R1 stage was calculated as the growing degreedays (GDD soil ). It was found that, at 5 cm soil depths, the GDD soil in the intra-row of ST decreased by 51-55°C compared to that of CT, and the GDD soil in the inter-row of ST decreased by 144-275°C compared to that of CT.
Soil water content under ST was higher than that of CT, especially in the topsoil ( Figure 2S). In 0-5 cm soil layer, during the observation date, the average soil water content in intra-row and inter-row of ST increased by 24.1% and 17.4% compared to CT, respectively. In 5-10 cm soil layer, the average soil water content in the intra-row and the inter-row of ST increased by 9.4% and 7.8% compared to CT, respectively.

Grain yield and its components
The interaction effect of tillage × year was not significant for grain yield and its components (Table 1). Across 2019 and 2020, there was not significant difference in grain yield between ST and CT. Grain number (GN) per year of ST was 4.3% less compared to that of CT. In contrast, the 1000-grain weight (TGW) of ST was 3.1% higher than that of CT. Ear number (EN) per hectare and harvest index (HI) were not affected by tillage systems.

Seedling emergence and shoot growth
Plant growth and nutrient uptake were analyzed to explain the yield formation of ST and CT plants. ST plants emerged 2-3 days later than CT plants (Figure 3), which was equal to 14.5-32.1°C GDDs of air temperature. The emergence rate index (ERI) of ST was 47.0% (P ≤ 0.01) lower than that of CT (Table 2S). However, there was no difference in final emergence rate (ER) between the two treatments. Due to the delayed emergence, the shoot dry weight of ST was significantly lower than that of CT on the same day until V6 (in 2019) or V9 stage (in 2020) of the CT treatment ( Figure 4). The average shoot dry weight accumulation of ST at V6 and V9 were 26.2% (P ≤ 0.01), 19.9% (P ≤ 0.05) lower than that of CT, respectively. Thereafter, ST plants restored the growth rate, and the difference The values in the table were the average of two years because there was no interaction between year and tillage. The negative values of post-silking K uptake indicate net K loss from the plants. Values followed by a different letter within the same nutrient are significantly different at P ≤ 0.05. in shoot dry weight became not significant between the two treatments, although the shoot dry weight accumulation of ST was slightly lower than that of CT.

Leaf growth, leaf distribution, and leaf senescence
Leaf growth was closely related to shoot dry weight. The leaf area index (LAI) under ST was significantly lower than that of CT during the vegetative period, especially in cold 2020 ( Figure 5). The ST plants reached silking stage about two days later than CT plants. The LAI max around silking stage was 4.0 and 4.2 for ST and CT across the two years, respectively. The reason was that the total number of leaves was reduced by one under ST compared to CT, and the area of the upper leaves became smaller ( Figure 6). This missing leaf was from the nodes below the ear-node, and the number of the leaves above the ear-node was kept to 6-7 leaves as normally found. As a result, the position of the ear moved down by 1-2 nodes. During post-silking stage, the decrease of LAI of ST was slower than that of CT, especially during late grain fill stage from R3 (milk stage) to R6 (physiological maturity), suggesting leaf senescence was delayed by ST treatment. At R6 stage, the stay-green degree of ST, as indicated by the relative green leaf area (RGLA), was 33.0% (P ≤ 0.01) larger than that of CT (Figure 7).

NPK accumulation and remobilization
As shown in Figure 8 and Table 3S, N content of ST was significantly lower than that of CT during presilking stage. However, this difference disappeared during post-silking stage. At R6 stage, N content of the ST plants was even higher than that of the CT plants. Phosphorus (P) and potassium (K) contents of ST plants were lower than CT plants until as late as V9 stage, and then kept the same as in CT thereafter. During the grain filling stage (from R3 to R6), P and K content of ST plants tended to be higher than that of CT plants. At physiological maturity, the average N, P and K content of ST were 5.5%, 15.0% and 11.9% higher than that of CT, respectively (Table 3S).
Grain nutrients were contributed by both the remobilization of pre-silking accumulated nutrients in vegetative organs and nutrient uptake during post-silking stage. Compared to CT, vegetative-N remobilization and remobilization efficiency of ST plants were 20.4% and 7.0% (P ≤ 0.05) lower compared to that of CT, respectively ( Table 2). The contribution of vegetative N remobilization to the grains of ST was 14.0% (P ≤ 0.05) less compared to CT. On the contrary, the contribution of postsilking N uptake to the grains was 35.2% (P ≤ 0.01) higher in ST compared to CT. Nitrogen harvest index did not differ between ST and CT. Similarly, compared to CT, ST also reduced the remobilization of vegetative-P and vegetative-K and their contribution to the grains (by 19.9% and 10.0%, 27.4% and 65%, respectively), and increased the contribution of post-silking P and K uptake to the grains (by 27.2% and 65.9%, respectively).

Correlations between stay-green degree and NPK accumulation and remobilization
Across the 2 years and two tillage methods, there was a close negative correlation between NPK remobilization efficiency and stay-green degree of the leaves at R6 (Figure 9). On the contrary, there was a close positive correlation between post-silking NPK accumulation and stay-green degree of the leaves at R6.

Discussion
Strip-till (ST) may combine the benefits of conventional tillage (CT) and no-tillage (NT) by cultivating the intra-row soil and leaving the inter-row with residue cover (Vyn and Raimbault 1992). The planting strip of ST has similar soil temperature and moisture as in CT. While the straw mulching in the inter-row help to increase soil moisture, but decrease soil temperature (Figure 2 and 2S). In addition, the soil bulk density in the inter-row soil of ST was greater in comparison with CT, because the inter-row soil was undisturbed. The heterogeneous soil environment in ST had both advantageous and disadvantageous effects on maize growth and source and sink relationships. Maize plants have to adapt to this specific soil environment by balance different physiological process to ensure as high grain yield as possible (Lee and Tollenaar 2007).

The recovery growth adaptation mechanism of maize plants under ST condition
In the research of Licht and Al-kaisi (2005a), the soil temperature of ST was similar to that of CT, but higher than that of NT. However, previous studies ignored the distinction between the intra-row and inter-row. In this study, ST increased soil temperature in the seedbed (intra-row) compared to the inter-row, but the temperature was much lower than that of CT in the inter-row, which should have contributed to the 2-3 days' delay of emergence of the ST seedlings ( Figure 3). Accordingly, shoot biomass accumulation was also delayed until V14 stage (Figure 4). When maize is subjected to lowtemperature stress, the permeability of the cell membrane will increase, and reactive oxygen species will be excessively generated, resulting in an increase in the content of malondialdehyde (MDA), which will subsequently destroy more cellular structures, thereby affecting the germination and emergence of maize seeds (Miedema 1982;Theocharis et al. 2012). The late emergence of ST seedlings should have a negative impact on subsequent maize growth. However, previous studies showed that, the final biomass or yield of maize under ST were comparable to CT (Vyn and Janovicek 2001;Trevini et al. 2013;Zhang et al. 2015), but the reason is not explored. In our study, maize yield under ST was also comparable to that under CT. By investigating the maize growth process in detail, we found a 'Recovery Growth Adaptation' mechanism in ST by which ST plants get to the similar yield as CT.
Firstly, although seedling emergence was delayed under ST, the final emergence rate was not affected by ST (Table 2S), which resulted in the same ear number per ha in ST as in CT (Table 1). Cao et al. (2012) has shown that, in high-yield and super-high-yield maize production systems, the number of ears per ha has the greatest impact on yield formation. Therefore, in our study, the number of ears per ha was not the main factor affecting final yield.
Secondly, the later emergence of maize in ST had a negative effect on subsequent vegetative growth, resulting in lower shoot biomass accumulation compared to CT at the same date (Figure 4). The growth rate in ST became very similar to CT when the soil temperatures increased in the ST similar to CT (Figure 2). Grain number is mainly determined from V8 (when the floret initiation in the young ear starts) to R2 stage (when kernel development is finished) (Abendroth et al. 2011;Gonzalez et al. 2019;Mueller et al. 2019;Liu et al. 2021). D' Andrea et al. (2008) have shown that, during the critical period for kernel set of maize, the growth rate of plant biomass was significantly positively correlated with the growth rate of ear and the formation of grain number per ear. Therefore, the rapid growth (biomass accumulation) of ST plant from V14 later on facilitated the development of the younger ears and compensated the possible loss of grains per ear.
Thirdly, in order to keep up with the developmental process, ST plants had to reduce one or two leaves below the ear position ( Figure 6). Finally, ST plants reached silking at similar date. Early entry into the silking stage is essential to increase post-silking nutrient accumulation and promote biomass and yield formation . Especially in Northeast China, the frost-free period is short, ensuring the post-silking accumulated temperature is crucial to increase grain yield. Nevertheless, due to the loss of leaves, the LAI max of ST plants was lower in comparison to CT plants ( Figure 5). The loss of leaves in the early growth stage of ST plants should have some negative effect on biomass accumulation and the development of the young ear. This may explain that the number of grains per ear was significantly less (by 4.3%) in ST plants than that in CT plants.
Fourthly, about 90% of the dry weight of grain begins to accumulate two weeks after silking, the realization of this potential is determined by mostly current photosynthate supply, and less by dry matter remobilization (Johnson and Tanner 1972;Ouattar et al. 1987;Zhang et al. 2011). Green leaf area has great influence on current photosynthate production. In our research, ST plants had slower leaf senescence (Figure 7), which provided enough post-silking photosynthate to increase grain weight. However, the grain sink of CT plants was relatively large, the photosynthate produced by CT plants was relatively less distributed to each grain, which resulted in a lower grain weight (Table 1).
A previous comparison of maize yields under different tillage systems showed that maize yields under ST depended on local rainfall (drought level) and soil texture conditions, as well as years of cultivation, compared to CT (Vyn and Raimbault 1992;Licht and Al-Kaisi 2005a;Temesgen et al. 2012;Chen et al. 2021). In this study, compared with CT, the adaptation growth and yield performance of maize under ST were consistent in 2 years because of the same soil type and little difference in rainfall (Figure 1).

Nutrient accumulation and remobilization in relation to growth adaption of maize plants under ST condition
Leaf growth and leaf senescence are largely affected by N supply and leaf N status (Kitonyo et al. 2018;Mu et al. 2018). In this study, the low soil temperature during the early growth stage greatly inhibited N accumulation until about R2 stage in ST (Figure 8 and Table 3S). This low N accumulation should have limited leaf growth and resulted in smaller LAI (Figure 5 and 6). Low N accumulation may also directly limit ear growth and leaf initiation (Nasielski et al. 2020), and resulted in few grains number per ear (Abendroth et al. 2011;Gonzalez et al. 2019). Further study on the development of young ear and grain formation should elucidate this possibility. Interestingly, in ST plants, while one leaf was reduced from node below the ear-node, the number of the upper leaves above the ear-node was maintained ( Figure 6). The upper leaves in maize encounter high light intensity and therefore have high photosynthetic potential. Also it was reported that upper leaves in maize have high photosynthetic N use efficiency (Mu and Chen 2021). Therefore, the adaptive growth behavior of ST plants just mentioned probably meet the demand for yield formation. Stay-green is crucial for maintaining leaf photosynthesis which provides assimilates not only to the grains but also to the roots to sustain post-silking N and water uptake (Rajcan and Tollenaar 1999;Mi et al. 2003;Duvick 2005). Vice versa, more post-silking N uptake contribute to stay-green by reducing N export from the leaves to the grains (Borrell et al. 2001). In ST plants, the amount of N remobilization was reduced by 20.4% compared to CT (Table 2), which is beneficial to delay leaf senescence. Indeed, post-silking leaf senescence was significantly delayed in ST plants compared to CT plants (Figure 7). Leaf senescence was significantly and negatively correlated to N remobilization ( Figure 9). On the other hand, post-silking N uptake was increased by 33.9% in ST plants, which should compensate for the less N remobilization from leaves ( Table 2). As indicated by Licht and Al-Kaisi (2005b), there was more soil moisture conservation at late stage in ST in comparison with CT. The additional soil moisture in ST may allow for more N uptake from the soil and promote a longer grain fill period in a moisture-limiting environment. As a result, the contribution of post-silking N uptake to the grains was much higher in ST compared to CT plants. The grain weight in ST plants was higher than CT plants, which compensates for the lower grain number in ST plants, and make its grain yield similar as in CT (Table 1). To fully understand the adaptation mechanism for maize to ST system, the relationship between soil moisture dynamics, soil nutrient availability and post-silking N uptake efficiency needs further investigation.
The low soil temperature inhibited P and K accumulation at early seedling stage before V9 (Figure 8 and Table 3S). It is well established that soil P and K availability is largely affected by soil temperature (Whalen et al. 2001;Naher et al. 2019). Since V9 stage, P and K accumulation was almost the same in ST as in CT plants. Interestingly, the remobilization of vegetative-P and vegetative-K into the grains was significantly lower in ST than in CT plants (Table 2). On the contrary, the contribution of post-silking P and K uptake to the grains was increased in ST compared to CT (by 45.8 and 66.7%, respectively). These responses can be demonstrated by the increased leaf greenness in ST compared to ST at R6. As shown in Figure 9, there was a significant and negative correlation between P and K remobilization and relative green leaf area. Previous studies also found that delay leaf senescence inhibited P and K remobilization (Ning et al. 2013;Shao et al. 2021).
The current study highlights the importance of N uptake (especially post-silking N uptake) and leaf stay-green on the productivity of maize under ST system. Nitrogen plays a central role in leaf initiation, expansion and senescence (Mu and Chen 2021). Efficient pre-silking N uptake during middle growth stage should increase the recovery rate of the plants from the early inhibition, and therefore improve ear development for more grains. Efficient post-silking N uptake can sustain leaf stay-green which is essential for increasing grain fill and grain weight to compensate for the loss of grain number. For this purpose, crop and soil management in favor of root growth and N uptake, such N fertilizer sidedressing and/or deep loosening may be promising ways to improve the grain yield in ST. Stay-green cultivars may also help.

Conclusion
Seedling emergence was delayed with the same seedling emergence rate and maize growth was inhibited until around V14 stage in maize plants under ST due to the cooler soil temperature conditions. However, ST plants got the similar yield as in CT plants by taking the following 'Recovery Growth Adaptation' mechanism: (1) ST plants began to facilitate growth rate at around V14 stage when the soil temperature was greatly improved to stabilize ear growth and grain number as much as possible.
(2) ST plants reached silking stage the same time as in CT plants by reducing one below-ear node leaf, which is essential to maintain the duration of grain filling; (3) ST plants may conserve additional moisture at late stage and increase post-silking N uptake to fulfill the demand of grain development, and reduced the demand of leaf N remobilization, so as to maintain the leaves to stay green. As a result, 1000-grain weight was increased in ST which compensate for the loss of grain number per ear.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This research is supported by National Natural Science Foundation of China (No. U19A2035).