【作者】 李松；

【导师】 陈英旭；

【作者基本信息】 浙江大学 ， 环境工程， 2009， 博士

【摘要】 农村生活污水的排放对地表水环境的恶化有着十分显著的贡献,富营养化现象的发生与农村生活污水氮磷的大量排放有着密切的关系。本研究运用田间试验、室内模拟、微观示踪等方法,考察了太湖地区稻田湿地处理农村生活污水氮素的降解行为和脱氮机理,研究了稻田湿地田面水磷素的浓度特征及去除过程,明确了稻田湿地氮素的迁移、转化、分配及平衡过程,分析了稻田湿地径流氮素及磷素的流失特征、产生条件及其影响因素,建立了径流氮磷浓度与施肥降雨间隔时间、淹水深度、降雨强度及降雨时间等影响因素之间的线性回归方程,并采用野外田间试验数据对模拟方程进行了验证。主要研究结论如下:1、稻田湿地处理农村生活污水田间试验表明,7月13日水稻移栽后,由于水稻土壤的吸附,稻田湿地田面水总氮（TN）含量迅速下降,但追肥后又快速上升,且TN含量顺序为:地表水（SW）＞灰水（GW）＞生活污水（DW）＞黑水（BW）＞对照（CK）。经过水稻的一个生长周期的处理,氮去除率顺序为:BW（96.8%）＞DW（96.2%）＞GW（95.6%）＞SW（94.1%）＞CK（93.8%）,同时,11月13日各处理出水化学需氧量（COD）均稳定在20 mg L^-1左右,可满足地表水环境质量Ⅳ类标准。计算发现,来自农村生活污水的氮素去除率（62.9%-69.3%）显著高于源自尿素的氮素去除率（27.5%-32.7%）,表明源自农村生活污水中的氮素相对化肥氮更易于被稻田湿地去除。田面水总氮负荷（TNL）具有与TN相似的变化特征,同时,COD与TN浓度及TNL具有正相关性。此外,GW、DW和BW的水稻产量要显著高于CK、SW以及当地平均水稻产量,说明稻田湿地处理农村生活污水不但能生态高效脱氮,而且可以增加水稻产量。2、与氮肥不同,7月13日磷肥施用后COD变化不大,但总磷（TP）迅速升高,随后逐渐下降,并在10月15日左右COD和TP浓度逐渐趋于稳定,CK、SW、GW、DW和BW出水中的COD分别为:13.54、20.98、20.87、21.09和17.86 mg L^-1,TP去除率顺序为:GW（98.17%）＞DW（97.28%）＞BW（97.04%）＞SW（96.78%）＞CK（75.20%）。同时,GW、DW和BW中来自农村生活污水中磷的去除率（60.3%-71.4%）显著性（P≤0.05）高于化肥磷的去除率（26.8%-36.7%）,表明相对于磷肥,农村生活污水中磷的形态更易于被稻田湿地去除。随着水稻的生长,磷素从根部分别向茎部、叶部迁移,最后在谷籽中富积。除CK外,其它处理稻田湿地总磷负荷（TPL）和TP浓度随时间非线性下降,直到10月1日达到稳定。此外,五个处理的水稻产量差异不显著（P≤0.05）,说明采用农村生活污水替代地表水灌溉稻田湿地可在保证水稻产量的同时高效除磷。3、¹⁵N示踪试验表明,对照处理（CK）、普通尿素处理（UR）与同位素尿素处理(¹⁵NUR)的氨挥发量在10月1日达到了一个稳定值10.3 mg pot^-1,最终UR与¹⁵NUR氨挥发量分别占施入肥料氮总量的43.54%和41.18%。追肥后UR和¹⁵NUR田面水TN浓度快速上升,分别达到了最大值19.52和18.64 mg L^-1,而CK仅为3.78 mg L^-1,并在10月1日后达到了稳定值,分别为0.32、1.05和0.93 mg L^-1。分蘖期水稻植株吸收的氮素浓度逐渐上升,并在孕穗期达到了较高的水平,CK、UR、¹⁵NUR的吸氮量分别为:45.3、165.4和173.6 mg pot^-1。从水稻各生长期吸氮量的变化来看,秧苗期吸氮量低,孕穗期吸氮量最高,成熟期下降,且水稻吸氮总量为:¹⁵NUR＞UR＞CK。CK中的氮残留量最低(0.064 g pot^-1),UR的残留量及残留率比¹⁵NUR分别高0.171 g pot^-1和1.4%,但氮肥利用率、氮素回收率却分别低3.95%、2.55%。同时,UR和¹⁵NUR的硝化反硝化损失率分别为5.81%和5.69%,其产量较CK分别增加了33.69%和29.90%,再次说明施氮对增加水稻产量是十分必要的。4、人工降雨模拟试验表明,稻田湿地径流是一种典型的机会径流。产流时间随降雨强度增大而减少,径流TN与降雨强度呈正相关性,与施肥降雨时间间隔和淹水深度呈明显的负相关性,与降雨时间基本无关。同时,时间间隔对径流TN的影响最大,淹水深度次之,降雨强度再次之,降雨时间最小。在各种试验条件下的径流TN都超过地表水环境质量标准中规定的TN标准限值2 mg L^-1,并可用线性回归方程y=-2.124x₁-1.147x₂+0.097x₃+0.001x₄+32.987来模拟四个影响因素作用下的机会径流TN流失浓度,这在田间试验得到了很好的验证。NO₃^-和NH₄⁺是稻田湿地径流氮素流失的主要形态,各处理条件下NO₃^-和NH₄⁺比重分别在32%-75%、24%-66%之间波动;TN流失负荷随时间间隔的延长明显降低,NO₃^-流失风险主要发生在前7 d内,其后流失风险将大大降低。5、稻田湿地是一个天然的弱碱性缓冲系统,早期径流中溶解态磷（DP）比重较大（83%左右）,后期颗粒态磷（PP）比重逐渐升高（38%左右）,但总体上DP还是主要流失形态。径流TP与降雨强度、时间间隔及降雨时间呈正相关性,与淹水深度负相关。与TN相似,时间间隔对径流TP浓度的影响最大,淹水深度次之,降雨强度再次之,降雨时间最小。在各种试验条件下,径流TP都超过了地表水Ⅴ类标准限值0.4 mg L^-1,并可采用线性回归方程,y=-0.548x₁-0.243x₂+0.014x₃-0.001x₄+7.386来模拟径流TP浓度。稻田湿地磷素的流失风险主要发生在施肥后前10 d内,其后将大大降低。因此,施肥后短期径流会导致氮磷等肥料养分的大量流失,但可通过淹水深度的合理调配来避免径流发生,可以实现稻田湿地从氮磷库的输出源转变为吸收氮磷的汇,从而生态高效去除农村生活污水的氮磷,减轻农村生活污水面源污染。更多还原

【Abstract】 Rural domestic wastewater discharge is the leading source of water quality impacts to rivers and lakes in many countries,and a lot of N and P discharged from rural domestic wastewater are closely related to eutrophication.The overall objectives of this research has been to investigate the degradation behavior of N and the N-removal mechanism during the treatment of rural domestic wastewater by the paddy wetlands,and study the floodwater P concentration character and P-removal process,and discover the features of N transference, transformation,distribution and balance,and analysis the runoff characteristics,production conditions and influencing factors of N and P in the paddy wetlands,and establish the linear regression equation among runoff concentration,rain-fertilization interval,floodwater depth, rainfall intensity and rainfall time,and verify it by the data from the field experiment in field scale and microscopic scale.The followings were the main results.1.The results from the field experiment of rural domestic wastewater treatment by the paddy wetlands showed that floodwater TN concentration decreased quickly after transplantation,due to the adsorption of the paddy soil,but increased greatly after topdressing, and the sequence was SW＞GW＞DW＞BW＞CK.After a growth period,N-removal efficiency was in the following sequence of BW（96.8%）＞DW（96.2%）＞GW（95.6%）＞SW（94.1%）＞CK（93.8%）,and COD of the effluent was stabilized at about 20 mg L^-1,which met gradeⅣof surface water discharge standard.N-removal rate from rural domestic wastewater （62.9%-69.3%） was higher significantly（P≤0.05） than that from urea（27.5%-32.7%）,which showed that N from rural domestic wastewater was removed preferentially compared with that from urea by the paddy wetlands.TN load in the paddy wetlands floodwater had the similar variation characteristics with TN,and COD had a positive relation with TN and TN load.Moreover,the rice yields of GW,DW and BW were higher significantly than that of CK, SW,which indicated the paddy wetlands could not only remove effectively N from rural domestic wastewater,but also could improve the rice yield.2.Different from N fertilizer,COD concentration in the paddy wetlands floodwater changed little after P fertilizer application,but TP concentration rapidly rose,and gradually decreased,and was at about Oct.15.COD of CK,SW,GW and BW treatments were 13.54, 20.98,20.87,21.09,17.86 mg L^-1 respectively,and dephosphorization rates were in the following order at:GW（98.17%）＞DW（97.28%）＞BW（97.04%）＞SW（96.78%）＞CK（75.20%）. Dephosphorization rate from rural domestic wastewater（60.3%-71.4%） was higher significantly（P≤0.05） than that from P fertilizer（27.5%-32.7%）,which showed that P from rural domestic wastewater was removed preferentially.P transferred from roots to stems and leaves,as the rice grown,and accumulated in the grain lastly.TP and TP load in the paddy wetlands floodwater linearly decreased with time except CK treatment,and was stabilized on Oct.1.In addition,the rice yields of five treatments had not significant difference,which suggested irrigating the paddy wetlands by rural domestic wastewater instead of surface water could remove P and keep the rice yield simultaneously.3.The results of ¹⁵N isotopic tracer experiment showed NH₃ volatilization losses of CK, UR and ¹⁵NUR treatments was stabilized at 10.3 mg pot-1 until Oct.1,finally those of UR and ¹⁵NUR treatments account for 43.54%and 41.18%of the input total N,respectively. After fertilizer application,TN concentration of UR and ¹⁵NUR treatments rose quickly and reached the max value of 19.52,18.64 mg L^-1,respectively,but only 3.78 mg L^-1 for CK,and was stabilized on Oct.1 at the value of 0.32,1.05,0.93 mg L^-1,respectively.N concentration of the rice plant rose gradually at tillering stage,and reached the maximum value at the booting stage,absorptive N concentration of CK,UR and ¹⁵NUR treatments were 45.3,165.4, 173.6 mg pot^-1,respectively.N content of the rice plant was the lowest at the seedling stage, the highest at the booting stage and decreased in the mature stage,and followed the sequence of ¹⁵NUR＞UR＞CK.Soil residual N content of CK was the lowest of 0.064 g pot^-1,and residual N content and rate of UR were higher by 0.171 g pot^-1,1.4%than those of ¹⁵NUR, respectively,but N utilization efficiency and recovery rate were lower by 3.95%,2.55%, respectively.At the same time,nitrification and denitrification loss rates of UR and ¹⁵NUR were up to 5.81%,5.69%,respectively,and their rice yields increased by 33.69%,29.90%, respectively,which showed it was necessary for urea application to increase rice yield again.4.The results of indoor artificial rainfall experiment showed that the runoff of the paddy wetlands is a kind of opportunity runoff,which happened occurred accidentally.Producing runoff time decreased,as rainfall intensity increased,and runoff TN concentration had a positive correlation with rainfall intensity,and had a negative correlation with rain-fertilization interval and floodwater depth,and had no correlation with rainfall time. Rain-fertilization interval had the most important influence on runoff TN concentration, floodwater depth the second,rainfall intensity the third,rainfall time the least,under all tested conditions,runoff TN concentration was above the limited value(2 mg L^-1) of surface water environment quality standard,and could be simulated by the linear regression equation of y=-2.124x₁-1.147x₂+0.097x₃+0.001x₄+32.987 under the effect of four factors,which was verified well by the field experiment.NO₃^- and NH₄⁺ were the major N forms in the runoff of the paddy wetlands,and NO₃^- and NH₄⁺ rates fluctuated in the range of 32%-75%,24%-66%, respectively.TN loss load in the paddy wetlands decreased obviously,as rain-fertilization interval increased,and there had been a great NO₃^- runoff risk within 7 days,then the N loss risk would decrease greatly.5.The paddy wetlands is a natural weak alkaline system,DP rate was high（about 83%） in the early runoff period,and PP rate increased gradually（about 38%）,and DP was the major P form in the runoff.TP had a positive correlation with rainfall intensity,rain-fertilization interval and rainfall time,and had a negative correlation with floodwater depth.Similar to TN concentration,rain-fertilization interval had the most important influence on runoff TP concentration,floodwater depth the second,rainfall intensity the third,rainfall time the least, and TP concentration was all above the limited value(0.4 mg L^-1) of surface water quality standard gradeⅤ,and could be simulated by the linear regression equation of y=-0.548x₁-0.243x-2+0.014x₃-0.001x₄+7.386.Runoff P loss risk increased,mainly when rain-fertilizer interval was less than 10 days,after that it would decrease greatly.So,it will lead to great losses of N and P in the paddy wetlands,when runoff occurred shortly after fertilizer application.Runoff can be avoided by the adjustment of floodwater depth,and the paddy wetlands can be changed from output sources of N and P pools to the absorpted sinks, and highly-efficient denifrification and dephosphorization of rural domestic wastewater can be realized,and non-point source pollution of rural domestic wastewater was decreased.更多还原