Effects of Selenium-Enriched Rape Returning Amount on Available Selenium Content in Paddy Soil and Selenium Accumulation in Rice

Selenium-rich rape “Selenium Ziyuan No.1” was used as green manure to study the effects of different amounts of green manure returned to the field on the release characteristics of available selenium in acidic paddy soil in southern China, and to analyze the absorption and transformation of selenium in rice, so as to provide a theoretical basis for planting natural selenium-rich rice in acidic areas of southern China. Six treatments with different amounts of selenium-enriched rapeseed returning (0, 5, 10, 15, 20, and 25 t/hm2) were set up. Two rice varieties (selenium-rich rice variety Meixiangzhan 2 and common rice variety Zhongguangxiang 1) were selected. The results showed that (1) with the increase of selenium-rich rapeseed returning amount, the available selenium in soil showed an increasing trend. Over time, soil available selenium showed a significant increasing trend, and the content of soil available selenium reached the maximum at tillering stage, and then decreased. (2) For selenium-rich varieties, when the amount of selenium-rich rapeseed returned to the field was less than 15 t/hm2, the selenium content in rice grains increased significantly with the increase of the amount of selenium-rich rapeseed returned to the field, then remained basically stable. For conventional varieties, with the increase of the amount of selenium-rich rapeseed returned to the field, the selenium content of rice grains showed an increasing trend, but the overall selenium content was much lower than that of selenium-rich variety. (3) With the increase of the amount of rapeseed returned to the field, the rice yield had an increasing trend, but the maximum rice yields appeared when the amount of selenium-rich rapeseed returned to the field was 15 t/hm2. Therefore, Se-enriched rape returning could promote the release of available selenium in soil and the enrichment of selenium in rice plants, and significantly increase the selenium content in rice. According to the selenium content and yield of rice, it is suggested that the selenium-rich rice variety Meixiangzhan 2 was chosen and the amount of Se-rich rape returning is 15 t/hm2.


Introduction
Selenium is a trace element that the human body needs but cannot synthesize, and must be supplemented from the outside, and all organs of the human body need a certain amount of selenium to maintain its function [1]. Selenium defciency to a certain extent will lead to hypertension, diabetes, coronary heart disease, asthma, and more than 40 kinds of acute and chronic diseases [2]. Selenium defciency is an important reason for the high incidence of various cancers and chronic diseases [3]. Selenium in rice, vegetables, and fruits is the main source of selenium absorbed by the human body [4]. Selenium in rice is relatively stable and easy to preserve, which is regarded as an important source of selenium-rich food [5]. However, rice likes to absorb and accumulate heavy metal cadmium, which is not an essential element for the human body, and is harmful to human health [6]. Previous studies have shown that there is a signifcant negative correlation between selenium and cadmium in rice [7]. Increase in the selenium content in rice means reducing the content of cadmium, a heavy metal element. Because the selenium content of rice, vegetables, and fruits is relatively low under natural growth conditions, the traditional method is crop-added exogenous selenium by increasing selenium fertilizer or spraying selenium on the leaves [8][9][10]. Tis practice not only increases the economic cost but also has the risk of environmental pollution, which is not conducive to the safe production of agriculture [11].
Green manure is usually used to improve soil organic matter and selenium content, so as to increase the selenium content in rice. It can also be used as a high efcient material to active selenium in the soil [12]. In recent years, in order to improve the system of farmland rotation and fallow, the Ministry of Agriculture and Rural Afairs of China has jointly promoted this work with the Ministry of Finance every year. Te implementation scale will be expanded to 240 million acres in 2021 [13]. In the south, the planting modes of Rice-Rape or Rice-Rice-Rape are mainly promoted [13]. As the selenium-rich rape is returned to the feld, the soil organic matter is improved, and the activation of the soil selenium element is promoted. In addition, the selenium absorbed by the selenium-rich rape could be directly used as a source of activated organic selenium for rice to absorb and utilize. However, little was known about the efects of selenium-enriched rape returning on the release dynamics of soil available selenium, and the characteristics of absorption and accumulation of selenium in rice [14]. Terefore, it is of great signifcance to study the efects of the amount of selenium-rich rape returned to the feld on the available selenium content in soil and the accumulation of selenium in rice.
At present, research on selenium-enriched rice is mainly refected in the impact of selenium addition on rice yield and selenium content, while the research on activating selenium in soil to achieve natural selenium-enriched rice is relatively rare [12,15]. Rape, as a kind of green manure, is one of the important traditional methods to improve the content of soil organic matter. It is a key measure to improve the soil and increase grain production. Turning over rape can increase the content of humus and large aggregates in soil, increase the ratio of loose and stable soil, increase the enrichment coefcient of >5 mm particle size carbon, and increase the contribution rate of aggregates to soil organic matter [16]. Rape rotation can increase the content of organic matter, total nitrogen, and alkali-hydrolyzable nitrogen in the soil surface, and intercropping population structure also plays a certain role in improving soil fertility [17]. According to Zhang's research [18], in terms of improving soil nutrient content, the efect of ploughing green manure rape is more signifcant, which not only reduces the amount of chemical fertilizer but also improves soil nutrient content and increase production.
Tere was a signifcant positive correlation between the amount of rapeseed returned to the feld and the yield of rice and the number of soil bacteria, and a signifcant positive correlation between the amount of rapeseed returned to the feld and the soil alkali-hydrolyzable N, available P, organic matter, pH, total number of culturable microorganisms, and B/T value [19]. Te comprehensive metabolic activity of soil microbial carbon source was the strongest, the biodiversity was the highest, and Shannon index and evenness index are the highest in the treatment of the amount of rape returning is 15.0 t/hm 2 , which is more conducive to the stability of soil microecosystem [19]. According to Fan et al. [20], the return of rape to the feld can increase the organic matter content and rice yield, mainly by increasing the soil dissolved organic carbon (DOC) and enhancing the activities of soil catalase and cellulase.
Wang et al. [21] investigated the correlation between soil and selenium-rich rice grains in the main rice-producing areas in eastern and southern Guangxi, and found that there was a signifcant positive correlation between soil organic matter content and selenium in paddy felds. Soil organic matter content plays a special role in the transportation of selenium between soil and plants, afecting the absorption and accumulation of selenium by plants, especially the selenium content in rice grains. Te research of Han et al. [22] shows that such a rule also exists in dry land, and the change of soil organic matter can explain the variation of total selenium in surface soil >60%. Tis shows that under the same climatic conditions, the organic matter content of soil developed from similar parent material is the main factor afecting the total selenium content of soil [23][24][25]. However, there are also contrary views. Huang et al. [26] showed that soil available selenium was negatively correlated with soil organic matter and soil cation exchange capacity, and positively correlated with soil acidity and alkalinity. Selenium content in rice was positively correlated with available selenium and total selenium in soil. Diferent rice genotypes have signifcant diferences in the enrichment of selenium in soil: Zhao et al. used the method of applying selenium fertilizer to soil in a pot experiment, and the selenium-rich genotype Wuyou 308 was 4.69 times the selenium content of the lowest variety [27]; Zhang et al. [28] collected 80 rice varieties and planted them in natural selenium-rich soil in Fengcheng City, Jiangxi Province, and evaluated the selenium content in brown rice. A series of selenium-rich rice varieties were screened out and used as high selenium-rich rice germplasm resources for the breeding of new seleniumrich rice varieties. According to Zhang et al.'s research [29], rice genotype and the proportion of available selenium in soil are the main factors for the accumulation of organic selenium in rice grains. However, the distribution of selenium in diferent parts of diferent rice genotypes was completely consistent, which was root > stem and leaf-> grain. Selenium content in diferent parts of rice and organic selenium content in brown rice increased with the increase of the soil selenium application level. Te selenium content of high accumulation rice varieties was signifcantly higher than that of low accumulation rice varieties.
So far, there are many studies on using green manure to return to the feld to improve soil organic matter, thereby increasing the absorption of selenium by rice. However, there is no relevant paper on the use of selenium-rich rapeseed returned to the feld to improve soil organic matter content, and also to directly improve soil available selenium content without artifcial application of selenium fertilizer. Te release process of available selenium in soil, the distribution law of selenium in rice plants, and its efect on yield need to be further studied after the selenium-rich rape was returned to the feld.
In the blooming stage of selenium-rich rape, after picking the fowering Chinese cabbage, the aboveground part of rape was cut and directly returned to the feld after cutting. Te release dynamics of available selenium in soil and the enrichment characteristics of selenium in roots, stems and leaves, and grains of diferent rice varieties under diferent amounts of selenium-rich rape returning were analyzed, and the appropriate amount of rape returned to the feld was determined according to the selenium content of rice in each treatment. At the same time, the corresponding feld demonstration and technology promotion work were carried out to provide theoretical basis and technical support for the large-scale production of seleniumrich rice.

Experimental Site and Soil
Characterization. Te experiment was carried out in the rice planting base of Zhangmu Town, Fumian District, Yulin City, Guangxi, from December 2020 to July 2021. Before the experiment, the tested paddy feld was a typical selenium-rich soil, with available selenium content of 83.89 μg/kg and total selenium content of 1.07 mg/kg. Other basic physical and chemical properties were as follows: pH is 5.18, total nitrogen content is 1.98 g/kg, total phosphorus content is 1.15 g/kg, total potassium content is 1.23 g/kg, alkali-hydrolyzable nitrogen content is 89.27 mg/kg, available phosphorus content is 44.06 mg/kg, and the content of organic matter is 31.29 mg/ kg. Figure 1 shows the layout plan of the rice planting community.

Test Materials.
Selenium-rich rapeseed variety Selenium Ziyuan No.1 was selected as green manure in winter fallow feld. Selenium Ziyuan No. 1 was cultivated by Wang Hanzhong, a member of the Chinese Academy of Engineering, and was the frst new hybrid rapeseed variety in the world. Te seeds were sown on December 10, 2020, and the seeding rate was 30 kg/hm 2 . After that rape was harvest on March 10, 2021, the overground part of the rape was cut, the green manure stems and leaves are cut into the length of 10∼20 cm, and then, the green manure stems and leaves are scattered on the ground and ploughed into surface soil about 20 cm deep. Irrigation and retting was done frst, and then transplanting of rice seedlings on April 10.

Test Design.
Six amounts of rapeseed returned to the feld, and two rice varieties were set up. Six amounts of selenium-rich rape returned to the feld were F1 (0 t/hm 2 ), F2 (5 t/hm 2 ), F3 (10 t/hm 2 ), F4 (15 t/hm 2 ), F5 (20 t/hm 2 ), and F6 (25 t/hm 2 ). Two rice varieties were selenium-rich rice variety V1 (Meixiangzhan 2) and common rice variety V2 (Zhongguangxiang 1), where F1 is set to CK. Meixiangzhan 2 was bred by the Rice Research Institute of Guangdong Academy of Agricultural Sciences. Zhongguangxiang 1 was bred by the Institute of Crop Science, Chinese Academy of Agricultural Sciences. Tere were 12 treatments, and each treatment was repeated 3 times. Randomized block arrangement was used in the experiment, and each plot was 12 m long, 5 m wide, and 60 m 2 in area. Te experimental plots were separated by ridges and ditches with a width of 50 cm left to avoid cross infuence. Te water and fertilizer management, pest control, and other feld management work after transplanting are the same as the traditional methods.
After being cut, the selenium-rich rape was weighed according to the design requirements and the area of the test area, and then moved to each test area to be turned over and returned to the feld. Soil samples were collected from each test area by the fve-point sampling method on March 10 (D1: the day rape returned to the feld), April 10 (D2: the day rice was transplanted), May 8 (D3: tillering stage), May 26 (D4: jointing stage), June 12 (D5: poplar fower heading stage), and July 28 (D6: maturity stage) in 2021. Ten, they measure the available selenium content of the soil. On July 28, rice yield was measured, and rice plants were collected. Ten, the content of selenium in rice roots, stems and leaves, and grains was measured.

Determination Items and Methods.
Te available selenium content of the tested soil was measured by the NaH 2 PO 4 extraction method [30]. After rice roots, stems and leaves, and grains are dried and crushed, the total selenium content is determined according to the National Food Safety Standard Determination of selenium in Food (GB 5009.93-2017) [31]. Te method of rice yield measurement is based on the National Measures for Acceptance of Grain Yield Measurement for High Yield Establishment (Trial) issued by the General Ofce of the Ministry of Agriculture of the People's Republic of China in June, 2008 [32].

Statistical Analysis.
Te experimental data were analyzed by two factor analysis of variance and was used in IBM SPSS Statistics 22 software, and the model was designated as the interaction between the variety and the amount of rape returned to the feld [33]. Te main efects were compared, and the confdence interval was adjusted by LSD method. Te signifcance level was 0.05, and the confdence interval was 95%. Duncan's method was used to compare the diferences in one-way analysis of variance. WPS Ofce 2021 was used to sort out the test data and make charts.

Available Selenium in Soil.
It can be seen from Table 1 that the efect of each treatment on the soil available selenium content on the day of rape returning to the feld was not signifcant (0.05 level). Tere was no signifcant difference in available selenium content of soil sampled at International Journal of Analytical Chemistry diferent time in F1 treatment. Te return amount of selenium-enriched rape (F2-F6) had a signifcant efect on the soil available selenium content at other sampling times (0.01 level).
It can be seen from Figure 2 that with the increase of the amount of selenium-rich rape returned to the feld, the available selenium in the soil showed an increasing trend. On the day of rice transplanting, the available selenium It can also be seen from Figure 2 that when the amount of selenium-rich rape returned to the feld is 0, there is no signifcant diference in the soil available selenium content in diferent periods. For other treatments, the soil available selenium showed a signifcant increasing trend with the passage of time, and the soil available selenium content reached the maximum at the tillering stage, and then showed a decreasing trend [13]. When the amount of selenium-rich rape returned to the feld was 5 t/hm 2 , the contents of available Se in soil increased signifcantly by 14.80%, 55.87%, 54.20%, 47.05%, and 41.67% on D2, D3, D4, D5, and D6, respectively, compared with D1. Tere was no signifcant diference in soil available selenium content between D3 and D4, D5 and D6, and there was signifcant diference in the soil available selenium content in other periods [34]. When the amount of selenium-enriched rape returned to the feld was 10 t/hm 2 , compared with D1, the available selenium content of D2, D3, D4, D5, and D6 increased by 23.57%, 82.23%, 80.09%, 73.45%, and 71.84%, signifcantly. Tere was no signifcant diference in soil available selenium content between D3 and D4, D4 and D5, D5 and D6, but there was signifcant diference in the soil available selenium content in other treatments [35]. When the amount of selenium-enriched rape returned to the feld was 15 t/hm 2 , the available selenium content in soil increased by 30.12%, 107.24%, 96.09%, 89.54%, and 85.52% in D2, D3, D4, D5, and D6, respectively, compared with D1, and the diferences were signifcant. Tere was no signifcant diference in the soil available selenium content between D3 and D4. Tere was also no signifcant diference in the soil available selenium content between D4 and D5, D6, but there was signifcant diference in soil available selenium content in other periods [20]. When the amount of selenium-enriched rape returned to the feld was 20 t/hm 2 , compared with D1, the available selenium content in soil increased signifcantly by 35.03%, 124.79%, 115.53%, 105.35%, and 98.79% in D2, D3, D4, D5, and D6, respectively [6]. When the amount of selenium-rich rape returned to the feld was 25 t/hm 2 , the available Se content in soil increased by 36  International Journal of Analytical Chemistry diference in soil available selenium content between D5 and D6, but there was signifcant diference in soil available selenium content in other periods.

Selenium Distribution in Rice.
It can be seen from Table 1 that the return amount of selenium-rich rape and rice varieties had signifcant efects on the selenium content in the rice roots, stems and leaves, and grains, but the interaction between the rice varieties and the return amount of selenium-rich rape only had a very signifcant efect on the selenium content in grains (0.01 level), a certain efect on the content of selenium in stems and leaves (Sig. 0.053).

Root.
It can be found out from Figure 3 that compared with the selenium-rich variety V1, the root selenium content of V2 has an increasing trend. Compared with V1, the root selenium content of V2 signifcant increased by 9.92% and 6.32%, respectively, when the amount of selenium-enriched rape returned to the feld was F3 and F6. It can also be seen from Figure 3 that with the increase of the amount of rape returned to the feld, the selenium content in rice roots has an increasing trend. For seleniumrich variety V1, the root selenium content of F2, F3, F4, F5, and F6 increased signifcantly by 70.35%, 81.73%, 93.52%, 97.05%, and 101.67%, respectively, compared with F1. Tere was no signifcant diference between F2 and F3, and among F4, F5, and F6, while there was signifcant diference among other treatments. For V2, compared with F1, the Se contents in roots of F2, F3, F4, F5, and F6 were increased by 80.61%, 99.39%, 105.55%, 106.96%, and 114.03%, respectively. Te diferences among F3, F4, and F5 were not signifcant, and the diferences among F4, F5, and F6 were also not signifcant, while the diferences among other treatments were signifcant. Figure 4 that compared with selenium-rich variety V1, the selenium content in stems and leaves of general variety V2 has an increasing trend. Compared with V1, the selenium content in the stems and leaves of V2 increased by 21.26%, 22.64%, 33.20%, 17.32%, 14.66%, and 15.69%, respectively, when the selenium-enriched rape returned to the feld for F1 to F6, and the diferences were signifcant.

Stem and Leaf. It can be seen from
It can also be seen from Figure 4 that with the increase of the amount of rape returned to the feld, the selenium content in rice stems and leaves has an increasing trend. For selenium-rich variety V1, compared with F1, the stems and leaves selenium content of F2, F3, F4, F5, and F6 increased signifcantly by 37.88%, 58.15%, 86.75%, 95.34%, and 98.70%, respectively. Te diferences among treatments F4, F5, and F6 were not signifcant, while the diferences among other treatments were signifcant. For the common variety V2, the Se contents in stems and leaves of F2, F3, F4, F5, and F6 increased by 39.46%, 73.72%, 80.69%, 84.71%, and 89.58%, respectively, compared with F1. Te diferences among F3, F4, and F5 were not signifcant, and the diferences among F4, F5, and F6 were also not signifcant, while the diferences among other treatments were signifcant.

Grain.
It could be made out from Figure 5 that compared with the selenium-rich variety V1, the grain selenium content of V2 has an obviously decreasing trend. Te grain selenium content of V2 signifcantly decreased by 21.26%, 22.64%, 33.20%, 17.32%, 14.66%, and 15.69%, compared with V1, respectively, when the selenium-rich rapeseed returned to the feld for F1 to F6.
It can also be seen from Figure 5 that with the increase of the amount of rape returned to the feld, the selenium content of rice grains has an increasing trend. For seleniumrich variety V1, the root selenium content of F2, F3, F4, F5, and F6 signifcantly increased by 115.52%, 146.66%, 180.58%, 184.05%, and 189.18%, compared with F1, respectively. For V2, the selenium content in F2, F3, F4, F5,    Table 1, we can also see that for the grain selenium content, there is an interaction between the rice variety and the amount of selenium-rich rape returned to the feld. Figure 5 also shows that for the selenium-rich varieties, when the amount of selenium-rich rapeseed returned to the feld is less than 15 t/hm 2 (F4), the selenium content of rice grains increases signifcantly with the increase of the amount of selenium-rich rapeseed returned to the feld. Te selenium content in rice grain remained stable during the exceeds of this rape returning amount. For V2, with the increase of the amount of selenium-rich rapeseed returned to the feld, the selenium content of rice grains showed an increasing trend, but the overall selenium content was much lower than that of the selenium-rich variety V1.

Rice Yield.
It can be seen from the rice yield data in Table 1 that rice varieties and the amount of selenium-rich rapeseed returned to the feld have signifcant efects on rice yield, but the interaction between them has no signifcant efect on yield. It can be perceived from Figure 6 that the rice yield of V2 increased signifcantly by 13.63%, 12.53%, 14.98%, 16.74%, 15.79%, and 12.85% compared with V1 from F1 to F6.
It can also be seen from Figure 6 that with the increase of the amount of rape returned to the feld, the yield of rice has an increasing trend. For selenium-rich variety V1, the root selenium content of F2, F3, F4, F5, and F6 increased signifcantly by 8

Discussion
It is generally recognized that the application of green manure can improve soil organic matter and soil properties. In this study, there are three reasons for adopting "Selenium Ziyuan No.1" rape. First, rape is a kind of green manure; second, the soil phosphorus content in Fumian area of Yulin City is very high, and planting rape can absorb a lot of phosphorus; fnally, "Selenium Ziyuan No.1" rape is a selenium-rich variety, which can efectively absorb selenium in soil and achieve natural selenium enrichment without adding extra selenium. Te factors of applying green manure include the available selenium in soil, the selenium content in rice, and the resulted rice yield, which will be discussed in detail in the following section.

Available Selenium in Soil.
Previous studies have shown that the return of rape as green manure can increase soil dissolved organic carbon (DOC), enhance soil catalase and cellulase [20], increase the content of humus and large aggregates in soil, increase the ratio of soil tightness and stability [16], increase the comprehensive metabolic activity of soil microbial carbon source, and improve biodiversity index [19] so as to increase the organic matter content of the soil surface [17,20], which could increase the yield of rice [19,20]. Organic matter is the main factor afecting the activity of soil available selenium, and there is a signifcant positive correlation between soil selenium activity and soil organic matter content [21,22]. Tis is consistent with the results of this paper. Te research in this paper shows that the amount of selenium-rich rape returned to the feld had a signifcant efect on the available selenium content of the soil. With the increase of the amount of selenium-rich rape  returned to the feld, the available selenium in the soil showed an increasing trend. At the same time, the return of selenium-rich rape to the feld not only activates soil available selenium by increasing soil organic matter but also directly provides natural selenium sources.
Tis study also showed that the amount of selenium-rich rapeseed returned to the feld afected the available selenium content in the soil. With the passage of time, the available selenium content in the soil showed a signifcant increasing trend, reaching the maximum at the tillering stage, and then decreased slowly. Te release of the selenium element of the selenium-rich rape after being pressed green and returned to the feld is afected by the decomposition degree. Until the day of rice transplanting, the increase of soil available selenium content was relatively limited in all treatments, indicating that the decomposition of rape plants was slow, which may also be related to the low temperature during spring ploughing. After rice transplanting, with the temperature rose, the decomposition rate of rape was accelerated, and a large number of organic matter and selenium elements were released into the soil, so the available selenium content of the soil increased rapidly and reached its peak at the tillering stage. After that, the remaining rape residues were relatively few. Coupled with the absorption of rice roots, the soil available selenium content began to decline. Terefore, it can be inferred that rape is easy to decay in the soil, but the release of selenium is mainly concentrated in 30∼60 days after ploughing, which is basically consistent with the previous research results on the decay law of rape after returning to the feld and the release characteristics of N, P, and K [2,36,37].
However, some people hold the opposite view. Han et al. [22] deemed that soil available selenium was negatively correlated with soil organic matter and soil cation exchange capacity. Te possible reason for the result is that the soil properties of the two places are really diferent: Shahe County is located in the central of China where the soil is mainly transformed from alluvial deposits of modern rivers and lake sediments in the southeast plain lake area, with deep soil layer, moderate texture, good ventilation, permeability, fertilizer supply, strong water and fertilizer retention, high organic matter content, and slight acidity. Located in the south of China, Fumian region is a typical acid red soil distribution area with thin soil layer and high viscosity. Meanwhile, with the aeration, water permeability, and fertilizer supply are very poor, the soil organic matter content is very low, and the soil is acidic.

Selenium Content in Rice.
Previous studies have shown that rice genotype and the proportion of available selenium in soil are the main factors for the accumulation of selenium in rice grains [25]. However, the distribution of selenium in diferent parts of diferent rice genotypes was completely consistent, which was root > stem and leaf > brown rice. Te selenium content of each part of rice and the content of organic selenium in brown rice increased with the increase of selenium application level, and the selenium content of high enrichment rice varieties was signifcantly higher than that of the low enrichment rice varieties [29,34]. Tis is consistent with the results of this paper. Te study showed that the efects of rice varieties and the amount of seleniumrich rape returned to the feld on the selenium content in roots, stems and leaves, and grains of rice were signifcant. Compared with the selenium-rich variety Meixiangzhan 2, Zhongguangxiang 1 had a tendency to increase the selenium content in roots and stems and leaves, but the selenium content in grains decreased signifcantly.
Tang used milk vetch as green manure, and the study showed [29] that when the amount of green manure returned to the feld was insufcient, the selenium supply in the soil was limited, and the selenium content in rice hovered at 0.032-0.036 mg/kg, which was difcult to achieve a breakthrough increase. With the increase of the amount of green manure returned to the feld, the grain selenium content and yield of rice showed an increasing trend. At 22.5-24 t/hm 2 , the selenium content of rice increased to 0.044-0.047 mg/kg, which meets the standard of seleniumrich rice in China (GB/T2499-2008). Tis is not consistent with the research in this paper. Te results showed that the selenium content in rice grain increased signifcantly with the increase of green manure application. When the amount of green manure application reached 15 t/hm 2 , the selenium content in rice grains remained basically stable even if it exceeded this amount. When the amount of green manure used is more than 5 t/hm 2 , the selenium content of rice grains reached the standard of selenium-rich rice. Te possible reasons for the result are as follows: frst, the differences between diferent rice varieties. Second, the difference between the efects of diferent kinds of green manures [38]. Terefore, selenium-rich rape, as a kind of green manure, plays an important role in the process of soil selenium activation.
Te study also showed that there was an interaction between the rice varieties and the amount of seleniumenriched rapeseed returned to the feld. For seleniumenriched rice variety Meixiangzhan 2, the selenium content in rice grains increased signifcantly with the increase of the amount of selenium-enriched rapeseed returned to the feld when the amount was less than 15 t/hm 2 , and remained stable when the amount was more than 15 t/hm 2 . For Zhongguangxiang 1, with the increase of the amount of selenium-rich rape returned to the feld, the selenium content in rice grains increased, but the overall selenium content was much lower than that of the selenium-rich variety Meixiangzhan 2.

Rice Yield.
Returning green manure rape to the feld can increase soil pH value, alleviate soil acidifcation, increase total porosity of paddy soil, reduce bulk density, improve soil's physical properties [39], improve soil fertility, increase the number of grains per panicle and efective panicles, increase rice yield, improve rice appearance quality and processing quality, and improve nutritional quality and steaming quality [35,40]. At the same time, rape returning to the feld also has a certain inhibitory efect on planthoppers and sheath blight [41]. According to Zhou's et al. research [19], when the rape "Huyou 17" returning is 15-22.5 t/hm 2 , the rice yield is signifcantly increased. Especially when the amount of returning to the feld was 15 t/hm 2 , the comprehensive metabolic activity of soil microbial carbon source in the treatment of returning rape to the feld was the strongest, and the yield of rice "Qingxiang Ruanjing" reached the maximum. Tis is consistent with the results of this paper. Te results showed that the rice yield increased with the increase of the amount of rapeseed returned to the feld, but the maximum yield appeared when the amount of selenium-rich rapeseed returned to the feld was 15 t/hm 2 .
Zhu et al. [42] held that after the rape "Huyou 21" was returned to the feld as green manure, the contents of alkalihydrolyzable nitrogen, available phosphorus, and available potassium in the soil were higher than those in the feld without green manure, the plant height and chlorophyll content of rice "Qingxiangruanjing" increased, and the yield of rice increased, with the increase of rape returned to the feld. Tis is basically consistent with the results in this paper. However, he believed [42] that when the amount of rapeseed returning reached 22.5 t/hm 2 , the rice yield reached the maximum. Te main reasons for the diference were as follows: frst, the diference between diferent rape varieties; second, the diferences of soil properties in diferent regions; fnally, and perhaps most importantly, the test site is located in Guangxi. Te temperature is higher than the experimental feld of Zhu's, and the high temperature promotes the decomposition of rape returned to the feld and improves the efciency of rape returned to the feld [35,38].
Tere are diferences in the yield of diferent rice varieties, which is the driving force of human breeding eforts [43]. Te study showed that the yield of selenium-rich rice varieties was lower than that of common rice varieties. Tis indicates that the enrichment of selenium in rice gain is at the expense of yield, and it is difcult to achieve the maximum level at the same time for both yield and selenium content. Tis study also showed that although there were diferences in yield among the diferent rice varieties, the trend of yield diference was the same for the same amount of selenium-rich rapeseed returning.

Conclusion
Te results show that (1) with the increase of selenium-rich rapeseed returning amount, the available selenium in soil showed an increasing trend. Over time, soil available selenium showed a signifcant increasing trend, and the content of soil available selenium reached the maximum at the tillering stage, then decreased. (2) For selenium-rich varieties, when the amount of selenium-rich rapeseed returned to the feld was less than 15 t/hm 2 , the selenium content in rice grains increased signifcantly with the increase of the amount of selenium-rich rapeseed returned to the feld, then remained basically stable. For conventional varieties, with the increase of the amount of selenium-rich rapeseed returned to the feld, the selenium content of rice grains showed an increasing trend, but the overall selenium content was much lower than that of the selenium-rich variety. (3) With the increase of the amount of rapeseed returned to the feld, the rice yield had an increasing trend, but the maximum rice yields appeared when the amount of selenium-rich rapeseed returned to the feld was 15 t/hm 2 .
Terefore, Se-enriched rape returning could promote the release of available selenium in soil and the enrichment of selenium in rice plants, and signifcantly increase the selenium content in rice. According to the selenium content and yield of rice, it is suggested that the selenium-rich rice variety Meixiangzhan 2 was the choice and the amount of seleniumrich rape returned to the feld is 15 t/hm 2 [44].

Data Availability
Te data that support the fndings of this study are available from the corresponding author upon reasonable request.

Ethical Approval
Tis paper carried out a feld experiment with a conventional planting method to improve soil quality with green manure, and it does not involve ethical issues.