Manure in combination with optimal topdressing with nitrogen fertiliser improved growth, grain yields and the efficiencies of water and nitrogen use in winter wheat in the Xinjiang Oasis drylands

Grain yield and yield components

At the winter wheat mature stage (stage 94 of Zadoks’ scale), three representative quadrats covering 1 m² (1 m × 1 m) were randomly harvested from each plot and air-dried for grain separation. Grain from the three sample squares in the same plot were mixed and weighed to determine grain yield. At physiological maturity (stage 92 of Zadoks’ scale), spike number was counted from the two one m double rows of each plot. Grain number per spike and 1,000-grain weight were determined by counting the grains of 50 randomly selected spikes from the two, one m double rows.

Soil water content and storage

Soil cores four cm in diameter were randomly collected from each soil layer (each layer was 20 cm thick, for a total of 100 cm) at three locations in each plot at the pre-sowing, anthesis (stage 64 of Zadoks’ scale), and maturity (stage 94 of Zadoks’ scale) stages in each growing season using a hand-held soil auger, and the values were averaged. The soil samples were packed in aluminum boxes, brought to the laboratory immediately, and dried in an oven at 105 °C for 24 h to a constant weight, to determine the soil moisture content (soil water content (SWC), %). Undisturbed soil cores (100 cubic centimeters) were collected from soil profile pits at depths of 0–100 cm to determine the soil bulk density following Finn et al. (2015). Soil water storage (SWS) was calculated as follows (Mo et al., 2016): (1)${\displaystyle SWS\left(mm\right)=SWC\left(\text{\%}\right)\times \rho b\left(\mathrm{g}{\mathrm{cm}}^{-3}\right)\times SD\left(\mathrm{mm}\right)}$ where ρb (g cm⁻³) is the dry soil bulk density of a given soil layer, and SD (mm) is the given soil depth.

Evapotranspiration and water use efficiency

The actual evapotranspiration (ET_c, mm) of winter wheat was estimated using the soil water balance equation. ET_c was calculated using the following formula (Li et al., 2023): (2)${\displaystyle {ET}_c=P+I+U+\left({SWS}_2-{SWS}_1\right)-R-D}$ where SWS₁ (mm) is the soil water storage at the time of sowing, SWS₂ (mm) is the soil water storage at the end of the growing season, and P is the total rainfall during the wheat growing season. I, R, U, and D are the amounts of irrigation (mm), surface runoff (mm), upward capillary flow into the root zone (mm), and drainage (mm), respectively. As the experimental site is rainfed farmland with a deep groundwater table and no irrigation was provided over the growing season, I, R, U, and D are negligible.

Finally, water use efficiency (WUE) was calculated using the following formula (Zhao et al., 2023): (3)${\displaystyle WUE\left(\mathrm{kg}{\mathrm{ha}}^{-1}{\mathrm{mm}}^{-1}\right)={Y/ET}_c}$ where Y (kg ha⁻¹) is the grain yield and ET_c (mm) is the actual evapotranspiration during the crop growing season.

Leaf area index and photosynthetic parameters

At the jointing (stage 31 of Zadoks’ scale), booting (stage 43 of Zadoks’ scale), anthesis (stage 64 of Zadoks’ scale), early milk (stage 73 of Zadoks’ scale) and early dough (stage 83 of Zadoks’ scale) stages, the leaf area index (LAI) was measured from 10 randomly selected destructive plant samples. Leaf length (L) and the widest width (W) of each leaf of all 10 sampled plants were measured with a ruler, and the LAI was calculated as follows (Gao et al., 2010): (4)${\displaystyle LAI=0.8\times \frac{\sum _{i=1}^m\sum _{j=1}^n\left(Lij\times Wij\right)}{m}\times N/S}$ where, Lij is the leaf length (cm) of the jth leaf on ith plant, Wij is the largest width (cm) of the jth leaf on ith plant, m (m =10) is the measured number of plants, n is the number of leaves per plant, N is the plant numbers of a plots, S is land area of a plot (cm²).

The fag leaf from the main tillers of five random plants was tagged at anthesis. The net photosynthetic rate (Pn) and transpiration rate (Tr) of flag leaves of winter wheat at the anthesis stage was measured by the TARGAS-1 Portable Photosynthesis System (PP Systems) between 9:00 and 11:00 on sunny days. The CO₂ concentration of the leaf chamber was set as 400 µmol mol⁻¹ and the other gas exchange parameters (1,000 µmol⋅ m⁻²⋅ s⁻¹, 25 °C). The average value was computed from three flag leaves for each replicate.

Aboveground biomass, crop N uptake and nitrogen use efficiency

Fifteen plants were sampled from each plot at the maturity (stage 94 of Zadoks’ scale) stage. The root system was cut off and the plants were processed further. Plants at the maturity stage were separated into stem + sheath, leaf, glumes, and grains. The plant samples were dried at 105 °C for 30 min, and then at 80 °C until they reached a constant weight. Then, the samples were ground so that they could pass through an 80-mesh sieve and digested using H₂SO₄⋅ H₂O₂. The nitrogen content was determined using Nessler’s colorimetric method (Bao, 2000). The N uptake (kg ha⁻¹) by wheat was quantified by multiplying the biomass of each organ by its respective N concentration, calculated as follows (Li et al., 2021): (5)${\displaystyle CropNuptake=\sum N_c\times O_d\times 10^{-3}}$ where N_c and O_d are the N concentration (g kg⁻¹) and dry weight (kg ha⁻¹) in different organs of the wheat plant, and 10⁻³ is the conversion coefficient. Nitrogen use efficiency (NUE, %) was determined using the recovery efficiency of nitrogen, calculated as follows Liu et al. (2023): (6)${\displaystyle NUE\left(\text{\%}\right)=\left({TA}_N-{TA}_0\right)/N}$ where TA_N is the total aboveground nitrogen accumulation by wheat with nitrogen fertiliser applied; TA₀ is the total nitrogen accumulation by wheat without nitrogen fertiliser applied, and N is the amount of nitrogen fertiliser applied.

Leaf area index (LAI) and aboveground biomass

Manure, topdressing N fertiliser, and year had significant effects on the wheat LAI and aboveground biomass (Table S1). Adding manure fertiliser increased LAI values by 33.5% and aboveground biomass by 25.9% (2-year averages) compared with no manure fertiliser (Fig. 2). With an increased topdressing N rate but without addition of manure, LAI values increased by 35.9% in 2021–22 and aboveground biomass increased by 29.6% (Fig. 2). In 2022–23, the LAI and aboveground biomass values were highest with the M0N150 treatment, significantly increasing by 66.90% and 27.8%, respectively (P < 0.05), compared to the M0N0 treatment and by 7.00% and 5.8%, respectively, compared to the M0N300 treatment (P > 0.05, Fig. 2). Under manure addition, the highest LAI and aboveground biomass values were with M1N150 in both years (Fig. 2). Compared with the M1N150 treatment, the LAI values of the M1N0 and M1N300 treatments decreased by 35.5% and 9.2% (2-year average, Fig. 2A), respectively, and the aboveground biomass values decreased by 33.3% and 9.9%, (2-year average, Fig. 2B), respectively.

Photosynthetic parameters

Nitrogen application had a significant effect on the net photosynthetic rate (Pn); manure and N had significant effects on the transpiration rate (Tr); and year had a significant effect on both Pn and Tr. However, there was no significant interaction among the three factors (Table S1). Addition of manure reduced Pn values by 6.8% and Tr values by 3.3% in 2021–22 compared to without manure (Fig. 3). Notably, in 2022–23, addition of manure increased Pn values by 6.74% and Tr values by 17.13% compared to without manure (Fig. 3). Regardless of whether manure was added, the N150 treatment resulted in higher Pn and Tr values compared to the N0 and N300 treatments in both years (Fig. 3). Comparing the Pn and Tr values of each treatment, those of the M1N150 treatment had the highest values. The M1N150 treatment increased the respective Pn values in the M0N0, M0N150, M0N300, M1N0, and M1N300 treatments by 7.5%, 0.1%, 4.4%, 14.3%, 22.6% (2021–22), and by 28.8%, 11.8%, 12.5%, 20.9%, and 10.5% (2022–23); and increased the respective Tr values by 6.9%, 3.3%, 5.7%, 9.3%, 18.8% (2021–22), and by 35.1%, 16.7%, 22.1%, 1.8%, and 18.0% (2022–23) (Fig. 3).

Grain yield and components

Manure, N, and year significantly influenced wheat grain yields, with no observed significant interaction of these factors (Table 3). Adding manure in combination with fertiliser increased grain yields by 16.7% in 2021–22 and by 9.2% in 2022–23 compared with adding fertiliser alone (Table 3). Regardless of whether manure was added, the N150 treatment at different topdressing N rates increased grain yields over the 2 years by an average of 16.1% compared to N0, and by 5.6% compared to N300 (Table 3). Comparing the grain yields of each treatment, the highest was recorded with the M1N150 treatment, with yields of 2,516.3 kg ha⁻¹ in 2021–22 and 3,161.4 kg ha⁻¹ in 2022–23; these values were significantly higher than with the other treatments in both years, by 10.4–55.5% in 2021–22 and by 4.2–20.3% in 2022–23 (Table 3).

Manure, N, and year all had significant effects on yields (Table 3). Notably, the interaction between manure and N had very significant effects (P < 0.01) on the spike numbers (SNs) and thousand grain weights (TGWs) of wheat, and the three-way interaction had significant effects (P < 0.05) on the grain numbers per spike (GNSs) and TGWs (Table 3). Addition of manure increased SN, GNS, and TGW values by averages of 17.7% (2–year average), 4.5% (2–year average), and 0.8% (2021–22), respectively, compared to scenarios without manure, albeit with a corresponding decrease in TGW values by 8.9% in 2022–23. Under different N topdressing rates, the order of SNs (2 years) and GNSs (2022–23) of winter wheat was N0 < N300 < N150 (Table 3). The SNs (2 years) and GNSs (2022–23) of each treatment were compared; the highest values were with M1N150, with average increases of 5.9–47.5% and 2.0–17.9%, respectively, relative to the other treatments (Table 3). With topdressing N alone, there was no significant difference among the SN rates. But TGW values were significantly reduced in both years and GNS values were significantly reduced in 2021–22, with the N150 treatment resulting in the lowest values. TGWs were highest with the M0N0 treatment in both years, with significant increases of 8.7–12.2% on average over the other treatments (Table 3). In addition, the GNS values of the M1N0 treatment were the highest in 2021–22, with average increases of 2.5–11.3% compared to the other treatments (Table 3).

Crop NUA and NUE

Manure, N topdressing, and year had very significant (P < 0.01) effects on crop NUA and NUE (Table S1). The interaction between manure and topdressing N was very significant only for NUE (P < 0.01) (Table S1). Compared to N fertiliser alone, manure fertiliser significantly increased NUA by averages of 25.1% in 2021–22 and 34.4% in 2022–23 (Fig. 4A). N application increased NUA both with and without added manure (M0 and M1 conditions), with greater N application resulting in higher NUA. Both the N300 and N150 treatments differed significantly from the N0 treatment. The highest NUA value was for M1N300, which increased significantly by averages of 38.2% in 2021–22 and 31.3% in 2022–23 over the other treatments (Fig. 4A). Addition of manure promoted NUA with a corresponding increase in NUE (35.1% higher in 2021–22 and 19.8% higher in 2022–23) compared to no manure (Fig. 4). However, increased N fertiliser application led to reductions in NUE (26.3% lower in 2021–22 and 39.8% lower in 2022–23) (Fig. 4B). The highest NUE values of 31.9% and 46.5% were recorded with M1N150 in the 2021–22 and 2022–23 seasons, respectively (Fig. 4B). The lowest NUE values were with M0N300 and were 17.6% in 2021–22 and 23.9% in 2022–23 (Fig. 4B). Specifically, the M1N150 treatment resulted in significantly different values from those of the other treatments over the 2 years.