The study aimed to explore effects of different fertilization on paddy soil aggregate characteristics and potassium distribution under rapeseed-rice rotation and winter fallow by the located experiment test. It was hoped that the study would provide a basis for the sustainable utilization of potassium resources in paddy-upland rotation areas of Southern China.

Based on the location experiment of potassium fertilizer in different rotation patterns started in 2016, five treatments were selected, including CK(F) (no fertilization and winter fallow), NPK(F) (NPK fertilizer and winter fallow), NPK(R) (NPK fertilizer and winter rapeseed), NPK (SR+R) (NPK fertilizer with rice straw returning and winter rapeseed), and NP50%K (SR+R) (50% reduction of K fertilizer with rice straw returning and winter rapeseed). The soil samples were collected from the 0-20 cm soil layer of each treatment after rape harvest of the third year for analysis of physical and chemical properties, such as soil aggregates distribution, and exchangeable and non-exchangeable potassium content variation in soil aggregates. The stability of soil aggregates and the distribution of potassium in aggregates were further investigated.

The soil aggregate fractions of all treatments was the highest in <0.053 mm. Compared with NPK(F), NPK(R) increased the proportion of aggregates of 1-2 mm, 0.5-1 mm and 0.25-0.5 mm by 26.2%-82.6% under the same fertilization treatment. Also the stability of soil aggregates was increased, the proportion of aggregates of >0.25 mm, mean weight diameter (MWD), and geometric mean diameter (GMD) were significantly increased by 30.6%, 31.2% and 82.0%, respectively. Under rapeseed-rice rotation, the proportion of aggregates of >2 mm was increased under NPK(SR+R), which was 69.7% higher than that under NPK(R). The exchangeable potassium content in soil aggregates decreased with decreasing particle size under all treatments. NPK(SR+R) significantly increased exchangeable potassium content in all aggregate fractions by 22.2%-46.0% compared with NPK(R) under rapeseed-rice rotation. NP50%K(SR+R) significantly reduced the exchangeable potassium content in aggregates of >0.5 mm by 19.4%–20.6% than NPK(SR+R). Compared with NPK(F), three fertilization treatments under rapeseed-rice rotation all reduced the non-exchangeable potassium content in aggregate fractions. Under all treatments, the contribution rate of different fractions of soil aggregates to the bulk soil in potassium was the highest in <0.053 mm. Compared with NPK(F), NPK(R) significantly increased the contribution rates of 1–2 mm and 0.5–1 mm aggregates to bulk soil in potassium by 82.6%, 52.1% (exchangeable potassium) and 105.5%, 36.9% (non-exchangeable potassium), respectively.

The rapeseed-rice rotation could increase MWD, GMD, macroaggregate proportion and the contribution rate of macroaggregate to the bulk soil in potassium and improve soil structure. Under this rotation pattern, the chemical fertilizer combined with rice straw returning could increase the exchangeable potassium content in all aggregate fractions, which improved the potassium supply in paddy soil. However, the rapeseed-rice rotation reduced the content of non-exchangeable potassium in soil aggregates due to high potassium demand, and the input of potassium fertilizer should be appropriately increased.

China has rich wheat and maize stalk resources. It is necessary to clarify the spatial-temporal characteristics of the wheat and maize stalk resources, the corresponding nutrient resources and chemical fertilizer reduction potential through stalk incorporation, which can provide decision basis for promoting the utilization of stalk resources and reducing chemical fertilizer application.

Based on the soil long-term monitoring data of the Ministry of Agriculture and Rural Affairs in wheat and maize planting areas of China from 1988 to 2019, this study analyzed the amount of wheat and maize stalk and its nutrient resources and fertilizer reduction potential through stalk incorporation in different years in each region of China.

The annual average amount of wheat and maize stalk resources reached 1.62×10⁸ t and 4.23×10⁸ t, respectively in 2010s, which were increased by 0.16×10⁸ t and 2.04×10⁸ t compared with 1990s, respectively. The annual average amount of wheat and maize stalk NPK resources reached 278.19×10⁴ t and 901.08×10⁴ t, respectively, which were increased by 27.97×10⁴ t and 434.82×10⁴ t compared with 1990s, respectively. Both of them increased most in North China (NC). The annual average amount of wheat stalk resources and NPK resources decreased first and then increased with planting years, while maize increased. In the first stage (1990s to 2000s) and the second stage (2000s to 2010s), the annual variation rate (AVR) of wheat stalk resources were -42.47×10⁴ t·a^-1 and 205.10×10⁴ t·a^-1, and the AVR of nutrient resources were -0.26×10⁴ t·a^-1 of N, -0.03×10⁴ t·a^-1 of P, -0.44×10⁴ t·a^-1 of K and 1.27×10⁴ t·a^-1 of N, 0.14×10⁴ t·a^-1 of P, and 2.11×10⁴ t·a^-1 of K, respectively. In the first stage and the second stage, the AVR of maize stalk resources were 397.82×10⁴ t·a^-1 and 1 643.60×10⁴ t·a^-1, and the AVR of nutrient resources were 3.46×10⁴ t·a^-1 of N, 0.56×10⁴ t·a^-1 of P, 4.46×10⁴ t·a^-1 of K and 14.30×10⁴ t·a^-1 of N, 2.30×10⁴ t·a^-1 of P, and 18.41×10⁴ t·a^-1 of K, respectively. There were more than 80% of wheat stalk and its nutrient resources distributed in NC and Middle and lower reaches of Yangtze River (MLRYR), with the highest in NC (0.93×10⁸ t, 160.31×10⁴ t of NPK), and the lowest in Southwest China (SW) (0.09×10⁸ t, 16.05×10⁴ t of NPK). About 70% of maize stalk and its nutrient resources were distributed in Northeast China (NE) and NC, with the highest in NE (1.39×10⁸ t, 296.96×10⁴ t of NPK), and the lowest in MLRYR (0.21×10⁸ t, 44.40×10⁴ t of NPK). The annual average nutrient-release amount of wheat stalk incorporation per unit area were 21.1 kg·hm^-2 of N, 3.0 kg·hm^-2 of P and 62.3 kg·hm^-2 of K in China, with the highest in NC, and the lowest in SW. The annual average nutrient-release amount of maize stalk incorporation per unit area were 48.6 kg·hm^-2 of N, 10.6 kg·hm^-2 of P and 97.7 kg·hm^-2 of K in China, with the highest in Northwest China (NW), and the lowest in SW. The annual average nutrient-release amount of wheat and maize stalk incorporation per unit area increased continually during 1988-2019. The proportion of wheat stalk returning nutrients to the annual chemical fertilizer application were 9.13%-10.82%, 4.26%-6.43% and 88.02%-111.86%, respectively, and that of maize stalk were 22.99%-24.37%, 16.04%-28.67% and 150.29%-171.95%, respectively.

In general, using wheat and maize stalk properly could satisfy the basic potassium requirement for crop production and reduce the application of about 10%-20% nitrogen fertilizer and 5%-20% phosphorus fertilizer. Making full use of stalk resources was an important guarantee for reducing fertilizer application and increasing efficiency.