**2. Materials and Methods**

A micro-plot field experiment was done in Changsha (28◦11 N, 113◦04 E), Hunan Province, China in 2017. The soil in the experimental field was a Fluvisol (FAO taxonomy) with the following chemical properties at the upper 20 cm layer: pH 5.75, 34.2 g organic matter kg−1, 81.6 mg available N kg<sup>−</sup>1, 34.4 mg available P kg−1, and 56.7 mg available K kg−1.

Thirty-two bottomless plastic boxes (length × width × height = 40 cm × 40 cm × 30 cm) were pushed into the soil of the experimental field to a depth of 20 cm to establish micro-plots. Four rice cultivars (Huanghuazhan, Liangyoupeijiu, Longliangyou 97, and Xiangliangyou 396) were grown and each micro-plot was treated with either earthworm castings (17 kg m−2, EC17) or no castings (0 kg m−2, EC0). The level of earthworm castings for EC17 was chosen according to the product of duration of oilseed rape-growing season (218 days) and daily production rate of earthworm castings (78 g m−<sup>2</sup> day<sup>−</sup>1) in a no-tillage rice-oilseed rape rotation field in Nanxian (29◦21 N, 112◦25 E), Hunan Province in 2015. The daily production rate of earthworm castings was obtained from an investigation conducted on 10 randomly selected 1-m<sup>2</sup> areas in the field on the first day after oilseed rape harvest. The earthworm castings (pH 7.89, 61.4 g organic matter kg−1, 128 mg available N kg−1, 44.2 mg available P kg−1, and 254 mg available K kg−1) used in the experiment were collected from several no-tillage rice-oilseed rape rotation fields in Nanxian after oilseed rape harvest in 2016. The micro-plots were arranged in a split-plot design with four replicates, using the level of earthworm castings as the main plots and cultivars as the subplots.

Rice seeds were pre-germinated and then sown in a seedbed on 10 May. Rice seedlings were transplanted on 5 June with one seedling per hill and four hills per micro-plot. Earthworm castings were applied and incorporated in the upper 20 cm of the soil on the day before transplanting. Urea was split in three applications: 6.0 g N m−<sup>2</sup> at 1 day before transplanting (basal), 3.6gNm−<sup>2</sup> at 7 days after transplanting, and 2.4 g N m−<sup>2</sup> at panicle initiation. Superphosphate was applied as basal at a rate of 4.8 g P2O5 m<sup>−</sup>2. Potassium chloride (8.4 g K2O m−2) was split equally in two applications: one at basal and the other at panicle initiation. The micro-plots were flooded with a water depth of about 5 cm until 7 days before maturity. Agrochemicals were used to control diseases, insects, and weeds.

At maturity, rice grains were sampled from each micro-plot. The sampled grains were oven-dried at 70 ◦C to a constant weight and then hulled, ground, and sieved (100 mesh). About 1.0 g of the sieved sample was weighed, hydrolyzed, and derivatized to determine total amino acid content with high-performance liquid chromatography method [15]. Approximately 0.5 g of the sieved sample was digested by H2SO4-H2O2 method to determine the total N content. The N analysis was carried out using a segmented flow analyzer (SAN Plus, Skalar Inc., Breda, The Netherlands). The ratio of amino acid to N (total amino acid content/total N content) was calculated.

Data analysis including analysis of variance and linear regression analysis was performed using Statistix 8.0 (Analytical Software, Tallahassee, FL, USA).
