| Agricultural non-point sources pollution is becoming one of the causes of eutrophication ofwater body in river, reservoir and lakes, and deterioration of rural eco-environment in our country,and seviously restricted sustainable development of agriculture economic and rural environment. Inorder to prove up the present situation and their parttern of nitrogen and phosphorus loss frompaddy field-based two cropping rice, and to seek for available fertilizer management practice forcontrolling nitrogen and phosphorus non-point sources pollution, this study simulated three type ofpaddy soils in Dongtinghu Lake area using lysismeter pool following the characteristics of originalsoil profile. Three type of soils were Alluvial Sandy Loamy Paddy Soil(ASL), Purple CalcareousClayey Paddy Soil(PCC) and Reddish Yellow Loamy Paddy Soil(RYL), respectively. There werefive treatments for each soil:ⅰ) no fertilizer (CK);ⅱ) ordinary chemical fertilizer (urea,superphosphate and potassium chloride, CF);ⅲ) equal N amount of controlled release nitrogenfertilizer (CRNF);ⅳ) combined-application of equal P2O5 of pig manure (no chemical phosphorusand complement of N and K by chemical fertilizer, OM); andⅴ) 70％N of CRNF applied only(70％CRNF). Nutrient loss amount due to leaching and runoff and its regulations, rice yields andnutrient use efficiency, nutrient balance and change of nutrient content in soil profile were studiedsystemically, starting with dynamics of pH, electronic conductivity (EC), total nitrogen (TN) andammonium nitrogen (NH4+-N) and nitrate nitrogen (NO--N), total phosphorus (TP) and totalpotassium (TK) in floodwater by modelling rice cultivation according to management practice oftwo-cropping rice cultivation in Dongting Lake area. Then, output load of the nitrogen andphosphorus by runoff and erossion from typical agricultural watershed, which is XiangyingGaoshan village watershed, were simulated using AnnAGNPS model, and source strength ofnitrogen and phosphorus in different land-use types and in different period was studied.Peaks of TN, TP, TK, and NH4+-N concentrations in floodwater were reached at 1st day and at3rd after applying ordinary chemical fertilizer in paddy field-based two cropping rice, respectively,and then all the nutrients were decreased over time rapidly. Attenuated range of the floodwater TPconcentrations was the largest, because phosphorus was easy to immobilize in soil. Speed of thefloodwater TK attenuation was slower than that of TN, because the chemical properties of K+ issimilar to NH4+, but K+ produces not as volatilization as NH4+. The concentrations of each nutrientin floodwater in 15～30 d after the CF application were distinctly higher than CK. The change of theTN (1～30 d), NH4+-N (1～30d) and TK (1～50 d) in floodwater followed the exponental attenuatingmodel as time progressed. In general, NO3--N concentrations in the floodwater were very lowduring two- cropping rice growth season. The floodwater pH rose gradually from early rice in 15 d (late rice in 3 d) after applying urea, and EC kept consistent with dynamics of NH4+-N, whichshowed significantly linear relationship, and also was simulated well using the exponentalattenuating model. CRNF, especially applying 70％CRNF reduced significantly TN, NH4+-Nconcentrations, pH and EC values in the floodwater compared with CF treatment in 15 d afterapplication because of controlling release of nitrogen in CRNF. Floodwater TN concentrations atpeak were decreased by about 30％for combined-application with pig manure and chemicalfertilizer, which was consistent with the ratio of urea-N was replaced by organic N in pig manure,and floodwater pH was higher than that in CF after application because of substitution of acidphosphorus fertilizer, but was lower than that in CF at peak resulting from slow release of NH4+.Floodwater TP concentrations at peak in OM treatment were lower than that for single CF in thecase of early rice, but were higher than that for CF in the case of late rice, and floodwater TPconcentrations during mid-stage of rice growth in the OM treatment were higher than that in the CF;Effect of combined-application with pig manure on floodwater TK was not obvious. In a word,CRNF application resulted in significant decrease of nitrogen concentrations in floodwater of paddyfield-based two cropping rice, and there also was obvious effect on decrease of nitrogenconcentrations but increase of the floodwater TP by pig manure application.The results of two-year's monitoring leaching water in the three type of paddy soils showedthat daily average quantity of leaching water from ASP, PCC and RYL were 1.85, 0.79 and 1.28ram, respectively; total amount of nitrogen leaching loss in the three soils accounted for 2.28％,0.66％and 1.50％of N applied in CF treatment of ASL, PCC and RYL, respectively. TNconcentrations in leaching water in ASL was near to RYL, but amount, of leachate was higher thanRYL, resulting in distinctly increase of TN leaching losses. Both the amount of leaching water andTN concentrations were significantly lower in PCC than in the other two soils. For nitrogen formsin leachate, the leaching losses of NH4+-N, organic nitrogen and NOB-N occupied 39.7％, 56.8％and 3.5％of TN leaching losses, respectively. Leaching losses of total potassium in the three soilsaccounted for 14.0％, 4.68％and 11.5％of K applied, respectively. Magnitude of TK concentrationsin leachate were: RYL＞ASL＞PCC. Leaching losses of total phosphorus in all soils were less;there were no difference among treatments with all fertilization and treatment with on fertilizer.This may be due to easy immobilization of P in soil. Comparing with nutrient leaching losses intreatment with application of different fertilizer in all the three soils, TN concentrations in leachatein treatment with CRNF were highest, and significantly higher than CK treatment. And also TNleaching losses in treatments with OM and CF were significantly higher than CK treatment in ASLand RYL. TK concentrations in leachate in treatment with combined-application with pig manureand KCI, and single application of potassium chloride were distinctly higher than CK in all soils.But because nutrient leaching loss was a long-term and accumulating process, and this study hadcontinually been conducted only for two years, so the difference of N and K leaching losses amongdifferent N and K fertilizer sources applied was not significant. In the three type of soils for paddy field-based two cropping rice, average losses of TN, TP andTK by runoff in CF treatment were 8.01,484.1 and 29.4 kg·ha-1, respectively, and increased by280％, 255％and 153％compared with treatment with no fertilizer, respectively. The runoff lossesfor TN, TP and TK were 2.67％, 0.74％, 11.8％of fertilizer- N, P and K applied, respectively. Forthe all runoff events, the losses of TP (20 d), TN (20 d) and TK (30 d) resulting from runoff eventsin 20～30 d after application were the most important. TN losses (two events in 20 d), TP losses(two events in 20 d) and TK losses (three events in 30 d) were accounted for 72.4％, 65.6％and87.8％of TN, TP and TK losses by runoff in whole early rice growth season. TN losses by runoff inrice growth season in treatments with CRNF and 70％CRNF application were decrased by 25.1％and 28.5％compared with urea treatments, respectively. It is important to decrease TN loss in the1st runoff event in 15 d. While the difference of TN losses by runoff between urea and CRNFtreatment followed was not distinct, which was consistent with the floodwater TN dynamics thatCRNF application caused to significant decrease of floodwater TN concentrations compared withurea. TN losses due to runoff for combined-application with pig manure were decreased by 24.4％,but TP increased by 82.9％compared with urea and superphosphate application, respectively. Thedifference of TK losses between pig manure and KCl treatment was not significant.The grain yields of two-cropping rice from ordinary chemical fertilizer application were15228.6～16529.4 kg·ha-1, and N, P and K fertilizer use efficiency was 27.3～34.7％, 22.4～30.0％,41.2～60.7％, respectively, agronomic efficiency of total nutrient was 5.58～7.72 kg·kg-1, and two-cropping rice yield was the highest in RYL and fertilizer use efficiency was higher than the othertwo soils. Yield increase from treatment with application of 300 kg N·ha-1 of CRNF presented themost dominance, which average yields from three type of soils were increased by 3.63％comparedwith CF treatment from paddy field-based two cropping rice. But it resulted in the distinct increaseof rice straw yield, and promoted the nitrogen uptake of rice plants and grains. Therefore, CRNFcaused significant increase of fertilizer N use efficiency by rice, and fertilizer- P and K useefficiency was increased also because of rice biomass promoted by CRNF. Two-cropping riceyields from treatment with 210 kg·ha-1 of CRNF application was the most obvious in PCC, it wasnear to the highest yield's treatment in the other two soils, and it was showed the highest agronomicefficiency. N uptake by rice was only lower in 70％CRNF treatment than in CRNF treatment, N useefficiency was higher than CRNF treatment, especially in ASL, it was increased by 39.1％compared with urea treatment. There was some effect of OM treatment on yield increase of twocropping rice, especially in RYL, the yield increase was higher than CRNF and CF treatments; boththe fertilizer use efficiency and agronomic efficiency in OM treatment was higher than in CFtreatment, and phosphorus use efficiency in OM treatment was increased.At present, nutrient balance of fertilizer management in paddy field-based two cropping rice inDongting Lake area was: N surplus slightly in ASL and PCC, but deficit slight in RYL. N deficit inRYL mainly showed in two aspects of promoting nutrient uptake and increasing rice grain yield, wherese N surplus in PCC were mainly due to weaker nutrient release capacity and lower Ncontents in rice abovepart plants and grains; Phosphorus in paddy field-based cropping rice was instate of surplus on the whole, quantity of P surplus in PCC was the largest; potassium in paddy fieldwas in deficit state, especially K in RYL and ASL. Compared with CF treatment,combined-application with pig manure and chemical fertilizer promoted rice yield increaseresulting from activating soil phosphorus, and P surplus in soil was decreased. It should be anexcellent fertilizing measure for raising fertilizer use efficiency. N deficit in treatment with CFapplication in paddy field reflected that soil available N contents in ploughed layer were decreasedby year after year, however, though N was in deficit stage for application of 300 kg·ha-1 CRNF,increasing available N in ploughed layer soil may be due to soil N mineralization and N release inCRNF at late growth stage of rice. P and K balance state in paddy field was consistent with thechange of available P and K in ploughed layer soil. The available P in soil was increased, butavailable K in soil was decreased by year after year. The change of nutrient in soil profile showedthat available N and P in ploughed layer soil indicated the trend of cumulation, which was droppedoff as profile deepened, wherese available K in soil was increased as profile deepened, indicated thetrend of decreasing.By means of GIS technique, AnnAGNPS model parameters were ascertained based oncombining observation in-site and experience parameter in reference, and the model was debuggedand verified through data observed in-site in several rain events of different intensity and differentstage. The results showed that output load of TN and TP, amount of runoff water in small watershedof Dongtinghu Lake area were simulated well by AnnAGNPS. The N and P output load per area inwatershed were 12.2 kg·ha-1 of TN and 0.65 kg·ha-1 of TE Rice growth season from May to Octoberwas the main period of N and P loss, TN and TP load were accounted for 92.2％and 87.5％of totaloutput load at this stage, respectively. Among the source strength of agricultural non-point sourcepollution from different land-use types, N and P output load from arable land with fertilizerapplication were the largest because of ratio of land-use, its output load of N and P were occupied69.4％and 61.9％of total load, respectively. Output load of N and P from forest land was thesecond because of water and soil loss, which accounted for 16.2％and 18.6％of the total.Contribution from ratio of N and P in village and vegetable land to non-point source pollution inwatershed was less, because the watershed belongs to a typical natural village of agriculture andsmaller area of land-use in the both, as N and P loss load per area was relative high. In conclusion,AnnAGNPS model could estimate TN and TP output load in watershed well, and forecast thetime-distribution characteristic of N and P output in watershed, therefore, it could provide scientificgist for establishing field water and fertilizer management scheme, planning and management ofenvironment in watershed.
|
PDF Full Text Request