Abstract Quantitative evaluation and analysis was made to the soil nutrients， rice yield， rice and fertilizer utilization rate， and economic and ecological benefits of the slow controlled release fertilizer in typical rice fields with middle-low yield by comparing the one-time application and split applications of slow- or controlled-release fertilizer （CRF） with farmers fertilizer practice as the control， with the aim to explore the effect of CRF. Results showed that compared with the control， the application of CRF could meet balance nutrients required for rice growth， and the effective panicles and seed-setting rate were higher in the treatment groups of FVOL， SVOL， LADVOL which applied CRF. As for soil nutrients， soil pH maintained balance in FVOL， FCAI， SVOL， LADVOL. In addition， differences in soil nutrient reduction amount reached the extremely significant level between FVOL and SXNK. SVOL which applied CRF provided reasonable regulation of N， P and K release rate， which showed advantages over blending fertilization and farmers fertilizer practice. Comprehensive analysis found that the application of CRF could make rice tillering growth and decline leveled off， which effectively improved the quality of rice population， and it also had higher input-output ratio than that of the control and good economic benefits. In general， CRF suits to be used widely on rice crop in the regions with medium-low yield.
　　Key words Rice； Slow or controlled release fertilizer （CRF）； Population quality； Yield components
　　At present， the rice fields in Enping City are typical middle-low yield fields. The fertilization techniques in rice fields are still based on artificial application by farmers. Because farmers have inaccurately grasped the characteristics of fertilizers， fertilizers are generally applied twice， resulting in serious loss of fertilizers and low use efficiency. Moreover， large-scale growers tend to give up the principles of multi-application in low amount because of the high costs of artificial inputs， and prefer the application of base fertilizers and twice applications in crop-greening stage， or only add 1 artificial fertilization of granulate fertilizer， which is not conducive to improve of the use efficiency of fertilizer. In this study， tests were conducted to introduce the proper slow- or controlled-release fertilizers （CRF） together with scientific fertilization plans to further explore the agronomic traits and the demand for N， P， and K nutrients of rice at different growth stages， find out the effectiveness of CRF application for rice and fertilizer use efficiency. It was aimed to strive for the reduction of non-point source pollution and construction of environmental-friendly rural area through fertilization comparison test， field demonstration and promotion of new-type fertilizers. 　　Materials and Methods
　　Test area overview
　　The tests were conducted in the Encheng Sub-district Office of Enping City， Guangdong Province， which is located on the bank of Jinjiang River in the central part of Enping. It borders Dongcheng Town to the east， Hengbei Town and Dahuai Town to the southwest， Datian Town to the south， and Liangxi Town and Shengtang Town to the north. Located at 22.18°N， 112.3°W， it has a total area of 162.5 km2. The analysis on the 276 sampling sites in the rice fields of Encheng Sub-district Office showed that the soil had an average soil pH of 5.13， average organic matter content of 21.17 g/kg， average total nitrogen content of 1.1 g/kg， average available nitrogen content of 99.34 mg/kg[1]， available phosphorus content of 21.80 mg/kg， and available potassium content of 52.99 mg/kg. From the frequency of grading， the soil texture of Encheng Sub-district Office was mostly loamy soil with strong acidity， lack of organic matter content， moderate level of total nitrogen content， middle-low level of available nitrogen， insufficient available phosphorus and extremely insufficient available potassium.
　　The rice fields with middle-low yield were selected to conduct the test in Encheng Sub-district Office， and the test was conducted in the double cropping rice demonstration fields with an area of 0.67 hm2 of Hongyan Rice-Specilized Cooperative in Encheng， dividing into the early rice croppingfields and the late rice croppingfields.
　　Test materials
　　Rice variety was Guangdong Nongsu seedlings. Fertilizers included the 42% Volfertile Controlled Release Fertilizer （VOL for short） produced by Jinzhengda Company， a foreign variety from other provinces； Caihongpai （24-7-19） blended fertilizer （CAI for short）； 50% Xinnongker （23-7-20） Controlled Release Fertilizer （XNK for short） from the Institute of Soil and Fertilizer of Guangdong Academy of Agricultural Sciences， a local variety； straight fertilizers of farmers fertilizer practice （FFP）.
　　Test methods
　　The tests were mainly designed to plot tests and large area demonstrative tests. The early rice croppingand late rice croppingwere planted in the plot tests， while only the early rice croppingwas planted in the large area demonstrative tests （LAD）. The rice in the plot test groups was artificially transplanted， and mechanical transplanting was performed in LAD. In the plot test groups， the early rice croppingfields were set 3 fertilization treatments of FVOL， FCAI， FFFP， the late rice croppingfields with 3 treatments of SVOL， SCAI， SFFP， and the early rice croppingin LAD with 3 treatments with LADVOL， LADCAI， LADFFP （Table 1 & 2）. 　　Agricultural Biotechnology2018
　　Survey and measuring items
　　Determination of soil properties in the cropping fields and LAD
　　The measured soil properties in the test plots and LAD included the changes of soil nutrients before rice transplanting and harvesting. The soil nutrients of plot tests were measured according to the standardized test of soil testing and formula fertilization.
　　Survey of agronomic traits and comprehensive economic benefits of cropping fields and LAD
　　For each treatment plot， 75 rice seedlings were selected for the survey using 5-poing sampling method. The growth conditions of the selected rice seedlings were observed and recorded， as well as the mean values of major agronomic traits. The yield of each plot was measured after harvesting， and the per unit yield was calculated to compare the rate of growth. In the meantime， all expenses during the production were recorded for the analysis of comprehensive economic benefits.
　　Statistical analysis
　　Calculation method
　　a. Nutrients absorbed by the rice kernels to produce 100 kg of economic yield[5]： N 2.25 kg； K2O 1.1 kg； K2O 2.7 kg. b. Soil nutrient coefficient was 0.16； Fertilizing amount （kg/hm2） = （Nutrient absorption amount per unit yield of crop× Target yield-Soil measured value × 0.16）/（Fertilizer nutrient content × Fertilizer utilization rate）.
　　Data processing
　　SPSS 17.0 was used to process the clustering， analysis of variance， and F-test was used to proceed with multiple comparisons using Duncans new multiple range method.
　　Results and Analysis
　　Effect of different treatments on soil nutrient reduction amount
　　As shown in Table 3， after harvesting， the soil nutrient reduction amounts were significant in both the early and late rice cropping fields and LAD. According to the analysis on soil nutrients （Table 4）， in the treatments of FVOL， SVOL， LADVOL， the application of CRF had no negative effects on soil environment， and the soil pH remained balanced with no acidification. However， the difference in pH values with other treatments was extremely significant. After variance analysis of F-test of multiple comparisons using Duncans new multiple range method， the differences were extremely significant （P<0.05） in the contents of organic matter， available N， available P， available K of all treatments[6]， and high nitrogen absorption and utilization rate indicated less nitrogen fertilizer loss[7]. It not only saved fertilizer， helped increase the economic benefits of planting rice， increase the income of farmers， but also reduced the pollution of water bodies and the eutrophication of rivers and lakes， which was of great significance for protecting the ecological environment and human health[2-4， 8]. The test results of soil testing and formula testing showed that the available N content in the second rice cropping fields were of fields with significant low-yield. The N content of treatment SVOL was （-39.533 3±8.025 1） kg/hm2， of treatment SXNK was （-33.733 3±2.020 7） kg/hm2， and of treatment SFFP was （-31.400 0±12.124 3） kg/hm2， and the amounts of soil nutrient reduction reached an extremely significant level[9]. 　　As shown in Fig. 1， SPSS was used to analyze the cluster dendrogram of nutrient reduction amount in the early and late rice cropping fields and LAD. The analysis on the squared Euclidean distance clustering indexes of between-groups linkage were significant and clustered in the contents of organic matter， available N， available P， available K of the rice in all treatments in the test plots. Compared with all other treatments， the differences of the 3 treatments of LADFFP， SFFP， FFFP were the biggest， and the squared Euclidean distance of between-groups linkage showed that the agronomic traits were significantly clustered in all the test treatments with the biggest different distance clustering index of 1.825. The squared Euclidean distance clustering index of between-groups linkage was 5.625-10.425 in the treatment groups of FVOL， FCAI， SVOL， LADVOL， SXNK， LADCAI. The comprehensive analysis found that， compared with the amounts of fertilizer applied in different treatments， the reduction amounts of soil nutrients clustered at different fertilization amounts[10].
　　Effect of different fertilization treatments on main agronomic traits of rice
　　As shown in Table 5， comprehensive analysis was made to the factors composed rice yield [11]， namely effective panicles， seed setting rate， 1 000-grain weight， and panicle setting rate. pike-forming rate. The results showed that for treatments FVOL， SVOL， LADVOL which were applied with CRF， the effective panicles were （265.600 5±22.047 0） × 104/hm2（Aa）， （243.000±237.85） × 104/hm2 （Aa）， and （240.300 0±212.529） × 104/hm2 （Ab）. The variance analysis of F-test with multiple comparisons using Duncans new multiple range method showed that only treatment LADVOL had a significant difference （0.05）， while the other treatments showed extremely significant differences （P<0.01）. The seed setting rates of treatments FVOL， SVOL， LADVOL[8] were （75.603 3±4.136 9）% （Aa）， （71.293 3±5.319 9）% （Bb）， （75.328 7±2.547 4）% （Aa）， respectively. Significant difference existed in treatment SVOL， while there was no significant difference in treatments FVOL and LADVOL. On the other hand， all other treatments showed extremely significant differences （P<0.01）. There was no difference in the main agronomic traits of treatments SVOL and SXNK， but the difference was extremely significant with SFFP and FFFP （P<0.01）.
　　As shown in Fig.2， SPSS was used to analyze the cluster dendrogram of major agronomic traits of rice in the early and late rice cropping fields and LAD. The analysis on the squared Euclidean distance clustering indexes of between-groups linkage were significant and clustered in the highest seedlings， effective panicles， panicle lengths， panicle setting rates， grains per panicle， seed setting rate， 1 000-grain weights and theoretical yield[12]. The squared Euclidean distance of between-groups linkage showed that the agronomic traits were significantly clustered in the treatments of SVOL and SXNK with the clustering index of 1.000. The clustering index of squared Euclidean distance of between-groups linkage was 2.125 in the treatments of FCAI and LADCAI. The comprehensive analysis found that the agronomic traits were significantly clustered in the different treatments applied with CRF. 　　Effect of different fertilization treatments on the fertilizer use efficiency of rice
　　As shown in Table 6， in terms of yield， the plot test groups applied with CRF showed yield advantages， and there was no significant difference in the actual yields （kg/hm2） between 5 CRF-applied treatments of FVOL， FCAI， FFFP， SVOL and SFFP， but the difference in actual yield of treatment SXNK， which was 3 955.35 ± 225.790 4 kg/hm2， reached the significant level. For the treatments applied with CRF， the N utilization rates were SVOL （36.169 8±3.172 8）>SFFP （32.992 8±2.503 0）>SXNK （32.674 4±1.868 1）>LADFFP （23.953 4±1.591 6）>FFFP（21.941 6±1.280 0）>FVOL（21.178 3±1.904 2）>LADVOL（18.763 9±1.101 8）>FCAI（18.630 7±1.487 1）>LADCAI（16.060 2±1.780 7）， and SVOL and SXNK showed significant differences with other treatments （P<0.01）. The P2O5utilization rates of the treatments applied with CRF were LADFFP （42.972 8±1.056 7）>FVOL（36.974 6±1.587 1）>LADVOL（35.746 4±2.095 8）>FFFP（33.660 9±1.960 0）>LADCAI（32.366 9±1.581 5）>FCAI（30.651 5± 2.443 5）>SFFP（28.079 0±2.130 9）>SVOL（26.581 9±2.331 6）>SXNK（24.475 1±1.396 5）. The K2O utilization rates of treatments applied with CRF were FFFP（59.565 2±3.470 0）>SVOL（57.882 1±5.077 9）>FVOL （54.529 2±2.336 4）>LADFFP（49.330 7±1.221 1）>SFFP（49.069 1±3.721 7）>FCAI（48.525 6±3.869 6）>LADVOL（43.622 9±2.560 7）>SXNK（43.467 5±2.478 4）>LADCAI（38.542 9±1.881 7）. The variance analysis of F-test with multiple comparisons using Duncans new multiple range method showed each treatment reached a very significant level （P<0.01）[13].
　　In this study， fertilizers were applied according to the levels of "3-control techniques"， which indicated that the application of CRF to low-yield fields[14] could achieve the levels of "3-control techniques" and could save the secondary labor work， greatly improving labor efficiency and reducing production costs.
　　Effects of different fertilization treatments on rice economic benefits
　　As shown in Table 7， in terms of yield， the application of CRF could show the advantages. The analysis on the effects of application on rice economic benefits of early cropping fields[7] showed that there was no significant difference between the treatments of FVOL， FCAI， FFFP. The analysis on the effects of application on rice economic benefits of late cropping fields showed that there was no significant difference in treatments SVOL and SFFP， but the difference was significant in treatment SXNK （P<0.01）. Comprehensive analysis showed that there was no significant difference between the 4 treatments of SVOL， SFFP， LADVOL， LADCAI， and treatments FVOL， SVOL， and LADVOL were all applied with CRF， which showed no significance with FFP treatments. Moreover， their advantages in saving labor and reducing pollution gradually transformed into ecological and economic benefits， thereby achieving the aim to control agricultural non-point source pollution. 　　Discussion and Conclusion
　　Application of CRF can stabilize soil nutrient balance in low yield fields
　　The test results showed that after harvesting， the soil nutrient reduction amounts were significant in both the early and late rice cropping fields and LAD. According to the analysis on soil nutrients， the application of CRF in low-yield fields could balance soil nutrients， but had no negative effects on soil environment， and the soil pH remained balanced with no acidification. In treatments FVOL and SXNK， the organic matter contents increased to a certain extent， and the reduction amounts of soil nutrients even reached the extremely significant level. In treatments SVOL， SXNK and SFFP， the contents of available N increased significantly， indicating that the application of CRF in low-yield fields can effectively control the release of fertility and improve soil fertility yield.
　　Application of CRF can increase rice yield components in low-yield fields
　　The application of CRF in low-yield fields could greatly improve the components of rice yields like effective panicles， panicle setting rate， grains per panicle and seed setting rates. The indicators of effective panicles and seed setting rates were in a high level in the treatments of FVOL， SVOL， SXNK and LADVOL， which applied with CRF. However， since the test parameters were measures in early and late cropping fields， there were difference in climate. In general， the application of CRF can increase rice yield components in low-yield fields.
　　Application of CRF can enhance fertilizer use efficiency of rice in low-yield fields
　　The test results showed that the N utilization rate of treatment SVOL， P2O5 utilization rate of treatment FVOL， K2O utilization rates of treatments SVOL and FVOL all reached high levels， and all showed advantages over the application of blended fertilizer and farmers fertilizer practice.
　　Application of CRF can maximize rice economic and ecological benefits in low-yield fields
　　In terms of yield， CRF showed the advantages. In the early rice cropping fields， the input-output ratio of 1∶2.08 of treatment FCAI was the lowest among treatments FVOL， FCAI and FFFP， and in the late rice cropping fields， the input-output ratio of 1∶91 of treatment SVOL was the highest among treatments SVOL， SXNK and SFFP. Moreover， their advantages in saving labor and reducing pollution gradually transformed into ecological and economic benefits， thereby achieving the aim to control agricultural non-point source pollution. 　　Outlook
　　Rice growers in Enping City have low organic fertilizer application levels. The application amounts of nitrogen and phosphorus fertilizers are large-oriented in rice production， while the application amounts of potash fertilizers are low. The irrational application of nitrogen， phosphorus and potash fertilizers results in increasing cost input. As for the application of CRF in small and medium sized plots of typical low-yield fields， the yield components are not determined by the application amounts of fertilizers only. This study was conducted according to the requirements of the Technical Specification for Test Fertilization by Soil Testing （Trial） combined with the actual situation since the implementation of the soil testing and fertilization project in the city. However， tests were conducted in plots and demonstrative fields， and there were differences in field fertility， which was not under control， so fertilization was not the only factors composing yield[15]. There were external factors like diseases， pests， weeds as well as human-induced operational errors affecting the yield components. Therefore， the agronomic traits of rice during the harvesting time should be investigated for different treatments to further clarify the effects of fertilization amounts on yield， which can better show， express and determine which fertilization mode can better achieve the aim of high-yield， high-efficient and quality production of rice. In the further， it is necessary to further strengthen the soil testing and formulated fertilization experiment in Enping City[16]， and determine the relationship between the fertilization amounts and yield， so as to play a greater positive role in improving agricultural efficiency and increasing farmers income in the whole city.
　　References
　　[1] DONG GC， WANG Y， YU XF， et al. Differences in nitrogen uptake and utilization of rice varieties at different growth stages[J]. Scientia Agriculutra Sinica， 2011， 44（22）： 4570-4582.
　　[2] MAI RJ， TAN YC， XIE SP， et al. Research on measuring the effect of nitrogen and phosphorus fertilizer potassium soil applying rice Yangdong County[J]. Guangdong Agricultural Sciences， 2011（04）： 66-67， 75.
　　[3] LI RM， ZHUANG DA. A preliminary experiment of soil testing and fertilizer fertilization of rice[J]. Guangdong Agricultural Sciences， 2010（12）： 73-74.
　　[4] PENG XL， LIU YYM LUO SG， et al. Effects of the site-specific nitrogen management on yield and dry matter accumulation of rice in cold areas of northeastern China[J]. Scientia Agriculutra Sinica， 2006（11）： 2286-2293. 　　[5] WANG Y， FU LD， LI X， et al. Effects of different nitrogen application rates on the growth and yield of rice[J]. North Rice， 2013（5）： 14-17， 21.
　　[6] National Agricultural Technology Promotion Center. Journal of organic fertilizers in China[G]. Beijing： China Agriculture Press， 1999.
　　[7] PENG SB， HUANG JL， ZHONG XH， et al. Strategies for improving nitrogen use efficiency in rice field in China[J]. Chinese Agricultural Science， 2002， 35（9）： 1095-1103.
　　[8] HARMUT K， SABINE SB. Development， growth and chemical composition of the potato crop （Solanum tuberosum L.）. II. Tuber and whole plant[J]. Potato Research， 1997b， 40： 135-153.
　　[9] ALLISON MF， FOWLER JH， ALLEN EJ. Responses of potato （Solanum tuberosum） to potassium fertilizers[J]. Journal of Agricultural Science， 2001， 136：407-426.
　　[10] WANG XB， LIU XD， BAO HP， et al. Study on formula fertilization of high quality and high yield in rice[J]. Journal of Jilin Agricultural University， 2000， 22（3）： 14-18.
　　[11] BAO HP， LIU SB， WANG XB， et al. Effects of three elements of N， P and K on rice yield[J]. Journal of Jilin Agricultural University， 2001， 23（2）： 5-8.
　　[12] LING QH， SU ZF， ZHANG HQ. Study on the relationship between rice yield and population quality and its influencing factors[J]. Journal of Crop Science， 1995， 21（4）： 464-469.
　　[13] ZHENG SX， NIE J， XIAO J， et al. Effects of film-degraded rice controlled release fertilizer on the nitrogen uptake and yield of the hybrid rice[J]. Agricultural Science & Technology， 2002， 3（2）： 4-10.
　　[14] TANG SH， CHEN JS， XU PZ， et al. Study on nitrogen release from controlled release fertilizer and dynamic absorption of rice[J]. Chinese Journal of Soil Science， 2004， 35（2）：186-190.
　　[15] CONG HL， WANG KJ， LI H， et al， Application test of nitrogen control technology[J]. North Rice， 2009， 39（5）：31-33.
　　[16] LIU SQ， ZHEN YX， HU CY， et al. Study on mathematical model of change formula fertilization for rice[J]. Journal of Hebei Normal University of Science & Technology， 1987， 1（4）：35-43.