Disadvantages of sowing methods on soil water content root distribution and yield of wheat (Triticum aestivum L.) in the Loess Plateau of South Shanxi, China Open Access

With the current growth rate, the global population will reach about 10 billion in 2050 (Alexandratos & Zhu 2012). Wheat (Triticum aestivum L.) is the third most grown crop globally and feeds about 30% of the world's population (FAO 2016). Increasing crop yield to maintain food security while reducing environmental impacts of agriculture is the dual challenge for humans in the future (Xue et al. 2019). Precipitation in the arid area of the Loess Plateau is the sole water source for wheat production. The average annual rainfall is 501.6 mm over the last 30 years, and approximately 55% (275 mm) occurs in summer fallow from July to September, and evaporation in summer fallow is large. Approximately 45% (225 mm) of annual precipitation occurs during the winter wheat growth period. Inadequate and variable rainfall often results in inadequate soil water at sowing, leading to drought stress in the growing period. This is the main factor that results in low and unstable winter wheat yield (Du et al. 2015). China is the largest producer of wheat. The North China Plain (NCP) is one of the most vital cereal production regions in China, with 25% of the national food production (Shao et al. 2011).

Several sowing techniques have been used in summer fallow for rain harvesting in rain-fed regions of China (Wang et al. 2016). Drill sowing, and wide-space sowing could break the compacted layers underneath the plough layer, decrease soil bulk density (0–40 cm soil layer), effectively retain rainwater in soil and increase soil water content in the wheat field compared with no or reduced sowing, leading to the increase in grain yield of wheat by 31% in the Loess Plateau (Jin et al. 2007; Temesgen et al. 2009; Wang et al. 2012). Furthermore, excessive N input also leads to N fertilizer residue, which causes many environmental problems, such as soil acidification, N₂O emissions, and decreased soil microbial activity (Wang et al. 2014; Sang et al. 2016). Our long term study in the Loess Plateau showed that drill sowing and wide-space sowing in summer fallow mitigates soil compaction by loosening shallow soil and breaking plough pan, and increased the soil water storage at 0–300 cm layers by 8% (Sun et al. 2018; Xue et al. 2019).

In the dryland of wheat production, wheat yield has been found to be linearly correlated to the soil water content at planting (Musick et al. 1994). Water storage during summer fallow also plays a major role in yield production in the dryland of the Loess Plateau (Zhang et al. 2008; Wang & Shangguan 2015). Furthermore, rainfall during the critical growth stages of wheat plays an important role in yield production as well. For instance, rainfall in the wintering and jointing stages has been shown to be essential for strong growth of winter wheat (Hatfield & Christian 2018). Rainfall before anthesis is critical for high yield while precipitation after anthesis shows a low impact on yield.

The growth and yield of wheat depend on the absorption of water and nutrients through the root system, while the limiting water condition restricts root growth and nutrients uptake (Spedding et al. 2004; Zhang et al. 2014). Root growth could be improved by modifying physical conditions and moisture content of soil by sowing practices (Costa et al. 2010; Dusserre et al. 2012). Therefore, a proper sowing practice can increase soil water retention and water storage capacity and thus improve root growth (Lampurlanés et al. 2001). Sowing can increase the length of finer roots leading to the larger absorption area of roots (Karunatilake & van Es 2000). Thus, we conducted a field experiment to examine the effect of sowing method on winter wheat fallow on soil water storage and growth and distribution of winter wheat roots in the soil profile.

Sowing methods for root study are destructive due to the need to excavate roots from the field, extract roots from soil, and quantify roots by scanning and image analysis (Gregory et al. 2009). It is time-consuming and tedious. Now, non-destructive techniques such as minirhizotron are available. Minirhizotron allows rapid dynamic measurements of root growth in the field (Chen et al. 2020). Our objectives were: (1) to examine effects of drill sowing on soil water content at sowing and drill sowing growth in sowing seasons; (2) to assess effects of drill sowing on growth and distribution of roots in 0–100 cm depth during plant growth in sowing seasons; (3) to evaluate in-season rainfall on growth and distribution of roots in 0–100 cm depth in the dry seasons; and (4) to evaluate effects of drill sowing on yield and water use efficiency in sowing seasons, and the contributions of soil water content at sowing to yield.

The experiment was arranged in a split-plot design with sowing method as the main plots and N rate as subplots with three replications. Two sowing methods, wide space sowing (WS) (sowing and row width were 8 and 25 cm, respectively) and drill sowing (DS) (sowing and row width were 3 and 20 cm, respectively) (Figure 2).

Each block of land consisted of three replicates. Plot size was 6 × 50 m. The DS treatment was carried out with a ploughing machine, wheat stubble and a half of organic fertilizer (1500 kg ha⁻¹) were ploughed to 25–30 cm depth to increase soil organic carbon. At the end of August, rotary tillage and land leveling were conducted for DS and WS to conserve soil water and prepare the land for planting.

Winter wheat seeds (Cultivar Yunhan-20410) were sown in early October by drill sowing using a planting machine, and harvested in early June of the following year. The seeding rate was 90 kg ha⁻¹. Dates and details for winter wheat tillage, sowing, and pesticide is shown in Table 2. Herbicides and insecticides were applied once in spring. ‘One spray for controlling three problems’ was also applied at anthesis stage, which was a combination of pesticides, fungicides, plant growth regulators, and micro fertilizer, to prevent diseases, insects, and premature plant aging. No irrigation was applied during the whole experiment.

Stems number (the sum of main stems and tillers) of the wheat plant were counted in a typical and central row of 1-metre length at jointing, at which the wheat plants had the maximum stems number. Then the productive stem percentage was calculated as the ratio of ear number to maximum stems number. During each growing season, the wheat plants were sampled in a row of 0.5 m length at anthesis and maturity.

WUE (kg ha⁻¹ mm⁻¹) was calculated as follows: WUE = Y/ET, where Y is grain yield (kg ha⁻¹); and ET = P + ΔSWS; ET (mm) is total evapotranspiration calculated over the whole growing season (sowing to maturity); P (mm) is the amount of precipitation in the whole experimental year; ΔSWS (mm) is the difference in the soil water storage between the beginning and the end of the season.

Experimental data was statistically analyzed using Microsoft Excel 2016 and Statistix 8.0 (Analytical Software, Tallahassee, FL, USA), and the figure was generated using Origin Lab pro 2021b (OriginLab Corporation, Northampton, MA, USA). Comparisons among multiple groups were performed using Tukey's honestly significant difference (HSD) test. Probability values p < 0.05 were considered statistically significant. Statistix 8.0 software was used for variance analysis.

The average soil water content in both DS and WS gradually decreased with the progress of plant development stage. The average soil water content from sowing to the wintering stage was reduced by 1.2% in WS and 2.0% in DS, but it was reduced by 4.5% in WS and 3.9% in DS (Figure 4). At anthesis, the average soil water content reduced by 14.3% in WS and 14.9% in DS but it was reduced by 6.9% in WS and 8.1% in DS. The average soil water content at anthesis was at least 8.2% higher than that of DS. These results show that the difference in soil water consumption between DS and WS was small at wintering and anthesis stages for years. The low level of soil water consumption is found at the wintering stage in both sowing years, but soil water consumption in the DS at anthesis was almost doubled relative to that in the WS years.

The higher in-season rainfall of 169 mm in 2015/16 and 190 mm in 2016/17 significantly increased the RLSD relative to that in 2012/13 and 2009/10 seasons with the in-season rainfall of 67 and 112 mm under both WS and DS. At anthesis, the average RLSD in 2015/16 and 2016/17 seasons was 43.7% higher than that in 2012/13 and 2009/10 seasons under WS and 20.8% higher under DS. These results indicate that the higher amount of effective in-seasonal rainfall increases RLSD at anthesis.

The wheat crop under wide space sowing produced higher grain yield than that under drill sowing in our present study. The yield advantage under wide space sowing and drill sowing was attributed to increased ear number. Similar effects of wide space sowing on grain yield were reported in previous studies (Liu et al. 2002), but the average annual rainfall is 278 mm (Figure 3) and insufficient for wheat growth and development. The soil water accumulation before sowing plays an important role in crop yield (Wang et al. 2018). The individual wheat plant under wide space sowing had more distance with each other than that under drill sowing (Huang et al. 2003). Additionally, it was observed that wide space sowing had a significantly positive effect on improving productive stem percentage. It could be reasonably assumed that the wheat crop under wide space sowing could uptake more N and water from soil, and then manufacture greater carbohydrates by canopy for maintaining the growth and differentiation of the huge population of stems. It was reported that wide space sowing could optimize root distribution and enhance root absorptive capacity of wheat than that under drill sowing (Sun et al. 2018; Noor et al. 2022). Roots of winter wheat sense changes in soil water content under arid environments and respond by changing metabolic activities and growth of underground and above-ground parts of the plant (Chaves et al. 2003). Root diameter and root length of winter wheat highly depend on water (Li et al. 2001). The amount of soil water absorbed by plant roots not only depends on the physical characteristics of soil but also soil water has greater influence on the growth features and distribution of roots (Yang et al. 2006). In the current study, we found that the root distribution of 0–100 cm was consistent with the water content in the soil. The low rainfall in summer fallow promotes roots growth by 20 cm deeper than that in the sowing seasons as early as the wintering stage. Similar studies have shown that most wheat's roots systems are concentrated in the upper 40 cm soil layer and play an important role in water uptake, while the depletion of water in the upper soil layer leads to the maximum growth of roots in subsoil (Xue et al. 2003; Zhang et al. 2004). Root distribution mainly occurs in the upper soil profile where the root length surface density (RLSD) was greater (Zhang et al. 2009), and about 80% of the RLSD was accumulated in the 40 cm topsoil depth under high moisture conditions (Yuhong et al. 2014). Under limiting water, plants have thicker and fewer roots in the upper layer of soil and more roots in the deeper layer of soil (Albasha et al. 2015). With the progress of plant development, soil water content reduces. The reduction of soil water in the four sowing seasons at both wintering and anthesis stages were higher than that in the four dry seasons (Figure 7). The average RLSD in the sowing seasons was 1.8–2.5-fold higher at the wintering stage and 56–71% higher at anthesis than that in the dry seasons (Figure 7(c)). In contrast, the average root diameter showed an opposite trend. Root diameter was higher in the dry seasons than in the normal/wet seasons at wintering and anthesis stages (Figure 6(e)). The lower rainfall in summer fallow reduced the soil water and RLSD but increased RD. In very dry years, root growth becomes limited due to the deprived water in the deepest horizons of water hindering the development of the root system (Verónica et al. 2010; Araki et al. 2012).

In addition, our results showed that in the dry seasons, the reduction of soil water in 80–100 cm depth was 0.8–1.2% higher from jointing to anthesis, 1.0–1.1% from anthesis to maturity than that in 0–60 cm depth (Figure 5(c)–5(e)). In the sowing seasons, soil water consumption from jointing to anthesis was similar between 0–40 and 60–100 cm depth. However, the water consumption was 4.6–4.7% more from anthesis to maturity. This shows that with the lower summer fallow rainfall, the higher soil water consumption in the deep layer of 80–100 cm proceeds from the jointing stage onwards. Araki et al. (2012) report that water scarcity affects root distribution. Our results showed that the reduction of RLSD at anthesis by drought was 29–44%, which was much lower than that in above-ground dry weight (56–67%). These results indicate that the low rainfall in summer fallow promotes root growth relative to shoot growth. It is essential to use water-saving practices to efficiently store soil water in the summer fallow period. However, the efficient utilization of rainfall during the growing period is also important to attain higher crop productivity (Shao et al. 2011). The precipitation at the key growth stages of wheat, i.e. wintering, jointing, and anthesis, is closely related to the formation of three yield components, especially in the early and mid-growth stages. Verónica et al. (2010) show that the key growth stage for developing an optimal root system is tiller, which is strongly dependent on the rainfall. Even though other later growth stages receive little rainfall, the root system can develop normally. In the current study, the lower amount of effective in-season rainfall in the two dry seasons reduced the soil water content of 0–100 cm from the wintering stages onwards, leading to a reduction of RLSD by 21–44% and RD by 0.03–0.04 mm at anthesis. As a result, the above-ground dry weight was decreased by 13–27% and grain yield by 14–24%. This shows that the less effective precipitation during the growth period impairs wheat growth, and the reduction of above-ground dry weight is higher than that in RLSD. A similar result is reported by Ali et al. (2018), who found that less rainfall (125 mm) decreases root length ratios and grain yield compared to the higher rainfall (200 and 275 mm).

It is worth noting that the wheat crop under wide space sowing produced the commensurate yield compared to that under drill sowing. This result indicted that an effective sowing method could compensate the yield loss due to a reduction of 25% in N input. From this study it was found that locally recommended sowing was non-significant compared to other sowing methods where the grain yield increased due to higher drill sowing (DS) which obviously enhanced about 6% in grain nitrogen concentration. In contrast, the significant yield reduction and no significant increase in grain N concentration were observed under drill sowing due to stem lodging occurring during grain filling. It was reported in this study that wheat crops under drill sowing (DS) had high N rate and lower bending resistance of stem. A higher lodging possibility of wheat was also noticed due to the contradiction between population and individual plants. (Wang et al. 2012; Ren et al. 2019; Noor et al. 2022). The average soil water content with DS at sowing was 1.8% higher than that with WS. The soil water distribution of 0–100 cm showed that the soil layer with the highest soil water content at all growth stages was the same between DS and WS in both sowing seasons, but the reduction of the soil water content of 0–60 and 80–100 cm with DS were more than with WS in the four dry seasons. Our previous study shows that DS improves water retention by removing subsoil compaction and storing water in the subsoil layer which could be utilized during the vegetative stages of wheat plants (Sun et al. 2018).

Our results also showed that the average grain yield with DS in the four dry seasons increased by 41.4% compared with WS, which was higher than that in the sowing seasons (16.0%). A similar trend was observed for WUE. The average WUE with DS increased by 32% compared with WS in the four dry seasons, and it was higher in the four sowing seasons (15%). However, the contribution of soil water at sowing to grain yield under DS was only slightly higher than that under WS in both dry and normal/wet seasons. Furthermore, the average RLSD at all growth stages was higher with DS than with WS in both dry and sowing seasons, and the average above-ground dry weight with DS was 15–172% higher than that with WS across all growth stages. In addition, the reduction of RLSD at anthesis by drought with DS (29%) was less than that with WS (44%).

The present study compared the drill sowing and wide space sowing where DS influenced by field evapotranspiration within the same year. At a low yield level, the average field water consumption, WUE, was highest in the year with the highest yield. Wide-space sowing in the fallow period improved the precipitation, while yield components that were negatively affected by precipitation were also improved. The higher amount of effective in-seasonal rainfall in the sowing seasons increased RLSD at anthesis by 44% with WS and 21% with DS and increased grain yield by 24.1% under WS and 14.2% under DS. The average grain yield in the sowing seasons was 28–41% lower than in the normal/wet seasons. The average water use efficiency WUE with DS was 32% higher while it was 15% higher than that with WS. The average contribution rates of soil water content at sowing to grain yield under DS and WS were 23–25%. Therefore, drill sowing in summer fallow should be adopted for improving rainwater storage and yield stability in dryland areas. In high-yield years, fallow cultivation can help adjust the relationship among the components, promote a reasonable distribution, and improve yield.

This project was approved by the Shanxi Collaboration and Innovation Centre for Efficient Production of Specialty Crops in Loess Plateau. The authors are thankful to China Agriculture Research System (No. CARS-03-01-24), the Central Government guides local science and Technology Development Fund projects (No.YDZJSX2021C016), General Project of SCO Institute of Modern Agricultural Development (No.SC021B004), State Key Laboratory of Sustainable Dryland Agriculture, Shanxi Agricultural University (No.202003-2), the Basic Research Program Project of Shanxi Province (20210302123410) and the Technology innovation team of Shanxi Province (No.201605D131041). The authors are also grateful to the Engineering Key Laboratory of Shanxi Province for financial support of this study.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.