秸秆还田的利与弊

秸秆还田具有促进土壤有机质及氮、磷、钾等含量的增加；提高土壤水分的保蓄能力；改善植株性状，提高作物产量；改善土壤性状，增加团粒结构等优点。但是要达到这样的效果，并非易事。目前秸秆还田有多种形式，可分为5大类：秸秆粉碎翻压还田、秸秆覆盖还田、堆沤还田、焚烧还田、过腹还田。这5种形式的利弊如何?措施又怎样?笔者作如下简述。     

秸秆腐解剂在秸秆还田中的效果研究初报

大大田、微区和盆栽条件下，研究了秸杆腐解剂对小麦、水稻生长及产量的影响，同时研究了秸杆腐解剂对小麦、水稻秸秆腐解速率及对土壤肥力的影响。结果表明，稻、麦秸秆还田时施用秸秆腐解剂对提高稻、麦产量具有明显的增产效果。增产的原因是穗数和粒数增加；稻、麦秸秆还田量不同时，还田量大且配施秸秆腐解剂的效果较还田量小好；麦秸秆还田方式不同时，麦秸以栽稻前耕翻还田且配施秸秆腐解剂的效果较好，上水沤制的效果较差；秸秆腐解剂能促进稻、麦秸秆较快腐解，减轻和防止多量秸秆还田给作物生长带来不利影响，并可稳定和提高土壤养分含量。     

曲周试区秸秆还田配施氮磷肥的效应研究

长期进行秸秆还田并配合施用氮肥和磷肥,不仅减少了焚烧秸秆对环境的污染,而且对培肥土壤和提高作物产量有积极的作用。结果表明;秸秆还田配合施用氮肥和磷肥能增加土壤有机质含量,其中以播种前翻地深３０ｃｍ,施Ｐ２Ｏ５７５ｋｇ／ｈｍ＾２,秸秆４５００ｋｇ／ｈｍ６２,不施氮肥的处理为最高,达１．１９％。增加氮肥和秸秆还田量并采用免耕措施可以增加土壤碱解氮的含量,如免耕,施纯氮２２５ｋｇ／ｈｍ＾２秸秆２２５０ｋ     

秸秆配施氮肥还田对水稻土酶活性的影响

通过田间试验研究秸秆还田时间及配施氮肥比例对水稻土酶活性的影响,以期为培育水稻土肥力和稳定稻田生态系统功能提供理论依据。试验设置2个秸秆还田时间(WS,冬季还田;SS,春季还田)和4个氮肥配施量(N0,秸秆还田,试验期内全程不添加矿质氮;NB,常规施肥,还田时不添加矿质氮;N30B,秸秆还田时添加早稻基肥用量的30%矿质氮;N60B,秸秆还田时添加早稻基肥用量的60%矿质氮)。研究结果表明:1冬季秸秆翻耕还田能增加冬闲期6种与土壤碳周转相关酶(β-葡萄糖苷酶、β-纤维二糖苷酶、β-木糖苷酶、多酚氧化酶、过氧化物酶和蔗糖酶)的活性,冬闲期冬季秸秆还田条件下土壤酶活性均高于春季还田,生育期内冬季秸秆翻耕还田措施对土壤β-葡糖苷酶和过氧化物酶有增加作用;2秸秆还田并配施氮肥措施显著地增加冬闲期和生育期β-纤维二糖苷酶的活性,但配施氮肥的3个比例间土壤酶活性并无显著差异;3除多酚氧化酶外,其他5种酶均与其有机碳投入量呈显著正相关。因此,冬季秸秆还田及配施氮肥能在一定程度上调控与碳周转相关的土壤酶活性,对推广冬闲期秸秆翻耕还田及保障作物的产量具有重要的生态学意义。     

芝麻、花生和田菁秸秆还田的化感效应研究

采用盆栽模拟实验研究了具化感作用的芝麻（Sesamum indicum）、花生（Arachis hypogaea）和田菁（Sesbania cannabina）秸秆全株还田对萝卜（Raphanus sativus）、黑麦草（Lolium multiforum）和黄瓜（Cucumis sativus）生长及土壤养分状况的影响,旨在评价秸秆还田释放养分与化感作用对后茬作物幼苗生长效应的综合影响。结果表明：秸秆还田能改善土壤养分状况,对后茬作物幼苗生长具有较好的促进作用。3种作物秸秆还田促进幼苗生长的大小顺序为花生秸秆〉田菁秸秆〉芝麻秸秆〉对照,其中芝麻秸秆对土壤速效磷、速效钾含量提高最大,而花生和田菁秸秆对土壤有机质、全氮的贡献值远远高于另两种处理。3种秸秆还田均对后茬作物幼苗生长和土壤养分具有促进作用,其释放的养分因子占主导作用,而化感作用并未明显表现出阻碍后茬作物幼苗生长。可能原因在于秸秆还田后导致较高pH和土壤养分利于植物生长,并影响了秸秆化感物质的活性和化感作用的发挥。     

稻麦秸秆直接还田技术的研究

通过４年多的研究,表明稻麦秸秆还田能有效地改善土壤理化性能和促进作物产量的提高。留高桩还田能有效地改善土壤物理性能,墒沟埋草能有效地提高土壤有机质的含量,麦田覆草能提高表层土壤的湿度和温度。几种秸秆还田方式中,以留高桩还田方法简便,效果最好;秸秆还田量以２２５０ｋｇ／ｈｍ＾２最为适宜,配合施用标准氮肥９５ｋｇ／ｈｍ＾２。     

不同作物秸秆还田对潮土结构的改良效果

为探究作物秸秆还田对土壤结构的改良效果,采用二因素裂区设计,主因素施肥(F):F1(施NPK复合肥)和F0(不施肥);副因素作物种类(B):B1(油菜)、B2(玉米)、B3(马铃薯)、B4(燕麦)和B5(荞麦),以不施肥不种作物的空白处理为对照CK。采集样品后测定土壤容重、孔隙度、毛管孔隙度和田间持水量,以及团聚体粒径和土壤MWD、GMD和D值。结果表明:5种作物秸秆还田降低了土壤0—60 cm土层的土壤容重,增加了毛管孔隙度和田间持水量,20—60 cm土层为主要改善土层;同时,增加了土壤0.25 mm粒级团聚体含量,且0—20 cm土层团聚体MWD、GMD值增加,D值降低,稳定性增加;秸秆还田配施化肥下,0—60 cm土层土壤0.25 mm粒级团聚体含量呈减少趋势,且团聚体MWD、GMD值降低,D值增加,稳定性下降;各处理下油菜秸秆还田后对土壤容重、孔隙度、田间持水量和团聚体稳定性改善效果较好,荞麦秸秆还田更有利于提高土壤田间持水量,燕麦秸秆还田更有利于提高土壤团聚体稳定性。由此可见,作物秸秆还田对土壤结构具有一定的改良效果,不同作物所起到的效果不同。整体而言,油菜秸秆还田效果更好。     

秸秆还田与施钾对华北低产潮土区作物产量及土壤钾素的影响

在华北平原的低产田上,研究了3年秸秆还田和施钾肥对小麦-玉米作物产量和耕层土壤钾素状况的影响。结果表明,相对于NP处理,NPK和NPK+秸秆还田(St)处理对小麦和玉米均有显著的增产效果;NPK+St处理增产最为明显,而且降低了年度间的产量变异系数。秸秆还田可以显著增加土壤速效形态钾含量。同季作物种植下,NPK+St处理的钾肥利用率、钾肥农学效率和钾肥偏生产力都高于NPK处理。NPK+St能提高0～20 cm耕层土壤速效钾、非交换性钾、矿物态钾和全钾含量。因此,秸秆还田有利于土壤不同形态钾素的稳定和增加,而NPK+St处理不仅能保证作物高产稳产,还能提高土壤各形态钾素含量。     

秸秆还田对稻麦轮作农田活性有机碳组分含量、酶活性及产量的短期效应

【目的】秸秆还田作为一种有效的秸秆处理方式不仅能够提高土壤肥力,增加作物产量,还可以缓解农田生态压力。研究稻麦轮作系统下不同秸秆还田量对土壤活性有机碳库、酶活性和作物产量的短期影响,可为提出适宜当地生产的秸秆还田量提供理论依据。【方法】利用稻麦轮作农田定位试验进行了研究。采用随机区组设计,设7个处理,以稻麦季秸秆均不还田为对照处理(CK),6个不同秸秆还田量处理。测定了秸秆还田后土壤活性有机碳库和土壤酶活性的变化,稻麦产量以及三者之间的相关关系。【结果】1)与秸秆不还田处理相比,试验范围内的秸秆还田量能在一定程度上提高土壤活性碳组分的含量和土壤酶活性,并能增加水稻和小麦的产量及其构成因素;2)土壤总有机碳和微生物生物量碳的含量随着秸秆还田量增加,增幅呈先增大后减小的趋势,以连续两季50%秸秆还田量处理下显著较高,而水溶性有机碳、活性有机碳含量和碳库管理指数在连续两季25%秸秆还田量处理下最高;3)相比秸秆不还田处理,连续两季25%秸秆还田量对土壤脲酶、过氧化氢酶和蔗糖酶活性的影响均最显著;4)水稻和小麦的产量均为在连续两季25%和50%秸秆还田量处理下增产较显著,与秸秆不还田相比,水稻增产达9.0%,小麦增产达11.45%;5)土壤碳库、土壤酶活性以及水稻和小麦产量之间均存在显著相关。【结论】在本试验条件下,连续两季25%和50%秸秆还田量表现出显著提高土壤碳汇能力和增加作物产量的优势。     

紫色土丘陵区秸秆还田的腐解特征及对土壤肥力的影响

在分析水稻、小麦、油菜秸秆的养分含量和化学组成的基础上，通过田间两年四作的定位试验，探讨了紫色丘陵区秸秆还田的腐解变化及对土壤肥力的影响。结果表明，秸秆含有丰富的养分，经腐解后给土壤提供了大量的碳、氮、磷、钾。在腐解过程中，还田后的前3个月分解快，后期分解缓慢；3种秸秆在田间的分解速率（K）：麦秸〉稻草〉油菜秸，秸秆翻埋〉秸秆面施；养分释放速度钾〉磷〉氮。测定了不同处理的土壤水分特征曲线．并将实验结果拟合为经验关系式。秸秆覆盖还田促进土壤团粒结构形成，提高了土壤水稳性团聚体含量，从而改善了土壤通透性和保水保肥性。秸秆覆盖还田还降低了土壤容重，增加了土壤总孔隙度，使土壤有机质、速效氮、磷、钾得到一定提高，从而达到培肥地力的目的。     

Impact of tillage, stubble management and crop rotation on nematode populations in a long-term field experiment

The population abundance of free-living and plant-parasitic nematodes was investigated in a long-term rotation/tillage/stubble management experiment at Wagga Wagga Agricultural Institute, New South Wales (NSW), Australia. The treatments were a combination of two crop rotations: wheat (Triticum aestivum)–wheat and wheat–lupin (Lupinus angustifolius); two tillage systems: conventional cultivation (CC) and direct drill (DD); and two stubble management practices: stubble retention (SR) and stubble burnt (SB). Plots of one of the wheat–wheat treatments received urea at 100 kg N ha⁻¹ during the cropping season. Soil samples from 0–5 and 5–10 cm depths were collected in September (maximum tillering), October (flowering) and December (after harvest), 2001, to analyse nematode abundance. Soil collected in September was also analysed for concentrations of total and labile C, and pH levels.Three nematode trophic groups, namely bacteria-feeders (primarily Rhabditidae), omnivores (primarily Dorylaimidae excluding plant-parasites and predators) and plant-parasites (Pratylenchus spp. and Paratylenchus spp.) were recorded in each soil sample. Of them, bacteria-feeders (53–99%, population range 933–2750 kg⁻¹ soil) dominated in all soil samples. There was no difference in nematode abundance and community composition between the 0–5 cm and 5–10 cm layers of soil. The mean population of free-living and plant-parasitic nematodes varied significantly between the treatments in all sampling months. In most cases, total free-living nematode densities (Rhabditidae and Dorylaimidae) were significantly (P < 0.001) greater in wheat–lupin rotation than the wheat–wheat rotation irrespective of tillage and stubble management practices. In contrast, a greater population of plant-parasitic nematodes was recorded from plots with wheat–wheat than the wheat–lupin rotation. For treatments with wheat–wheat, total plant-parasitic nematode (Pratylenchus spp. and Paratylenchus spp.) densities were greater in plots without N-fertiliser (295–741 kg⁻¹ soil) than the plots with N-fertiliser (14–158 kg⁻¹ soil).Tillage practices had significant (P < 0.05) effects mostly on the population densities of plant-parasitic nematodes while stubble management had significant effects (P < 0.05) on free-living nematodes. However, interaction effects of tillage and stubble were significant (P < 0.01) for the population densities of free-living nematodes only. Population of Rhabditidae was significantly higher in conventional cultivated plots (7244 kg⁻¹ soil) than the direct drilled (3981 kg⁻¹ soil) plots under stubble retention. In contrast, plots with direct drill and stubble burnt had significantly higher populations of Dorylaimidae than the conventional cultivation with similar stubble management practice. No correlations between abundance of free-living nematodes, and concentration of total C and labile C in soil were observed in this study. These results showed that stubble retention contributed for enormous population density of free-living (beneficial) nematodes while conventional cultivation, irrespective of stubble management, contributed for suppressing plant-parasitic nematodes.     

Temperature and stubble management influence microbial CO2-C evolution and gross N transformation rates

Few studies have examined the kinetics of gross nitrogen (N) mineralization, immobilization, and nitrification rates in soil at temperatures above 15 °C. In this study, ¹⁵N isotopic pool dilution was used to evaluate the influence of retaining standing crop residues after harvest versus burning crop residues on short-term gross N transformation rates at constant temperatures of 5, 10, 15, 20, 30, and 40 °C. Gross N mineralization rates calculated per unit soil organic carbon were between 1 and 7 times lower in stubble burnt treatments than in stubble retained treatments. In addition, significant declines in soil microbial biomass (P=0.05) and CO₂-C evolution (P<0.001) were associated with stubble burning. Immobilization rates were of similar magnitude to gross N mineralization rates in stubble retained and burnt treatments incubated between 5 and 20 °C, but demonstrated significant divergence from gross N mineralization rates at temperatures between 20 and 40 °C. Separation in the mineralization immobilization turnover (MIT) in soil at high temperatures was not due to a lack of available C substrate, as glucose-C was added to one treatment to test this assumption. Nitrification increased linearly with temperature (P<0.001) and dominated over immobilization for available ammonium in soil incubated at 5 °C, and above 20 °C indicating that nitrification is often the principal process controlling consumption in a semi-arid soil. These findings illustrate that the MIT at soil temperatures above 20 °C is not tightly coupled, and consequently that the potential for loss of N (as nitrate) is considerably greater due to increased nitrification.     

Suppression of wheat early growth in standing stubble

Early growth and development are often lower when wheat is sown into standing stubble. A study was conducted to determine whether this difference in early growth could be explained by the effects of stubble on soil temperature in the vicinity of the young plant. The roles of nitrogen nutrition and soil strength were also assessed. Three crops were monitored (1990–1992), with the wheat being sown into either standing wheat stubble after a no-till fallow (NT), or into no-tilled plots from which the stubble had been removed by burning (NB). Measurements were made of wheat growth and development, soil and plant N, soil temperature and penetration resistance. The site was on a black earth near Warialda in the northern wheatbelt of New South Wales, Australia. In 1992 wheat was also grown under simulated stubble to isolate the shading and soil temperature effects of stubble from other factors. A significant (P<0.05) relationship was found between average soil temperature and above ground dry matter (DM) at 65 days after sowing (DAS) but not at 107 DAS. This relationship accounted for differences in DM production at 65 DAS between NT and NB treatments in 1991 and 1992, but not in 1990. In that year the lower DM production in NT plots was associated with poorer N nutrition, and possibly disease. Laboratory incubations indicate that immobilisation of N as stubble decomposed could have contributed to this. Burning stubble produced no immediate increase in soil N availability, so that it is unlikely that N contained in stubble contributed to the difference. Soil strength differences between treatments and phytotoxic effects are unlikely to have contributed to growth differences in this soil.     

Effects of tillage and traffic on crop production in dryland farming systems: I. Evaluation of PERFECT soil-crop simulation model

Agricultural production systems are complex involving variability in climate, soil, crop, tillage management and interactions between these components. The traditional experimental approach has played an important role in studying crop production systems, but isolation of these factors in experimental studies is difficult and time consuming. Computer simulation models are useful in exploring these interactions and provide a valuable tool to test and further our understanding of the behavior of soil–crop systems without repeating experimentation.Productivity erosion and runoff functions to evaluate conservation techniques (PERFECT) is one of the soil–crop models that integrate the dynamics of soil, tillage and crop processes at a daily resolution. This study had two major objectives. The first was to calibrate the use of the PERFECT soil–crop simulation model to simulate soil and crop responses to changes of traffic and tillage management. The second was to explore the interactions between traffic, tillage, soil and crop, and provide insight to the long-term effects of improved soil management and crop rotation options. This contribution covers only the first objective, and the second will be covered in a subsequent contribution.Data were obtained from field experiments on a vertisol in Southeast Queensland, Australia which had controlled traffic and tillage treatments for the previous 5 years. Input data for the simulation model included daily weather, runoff, plant available water capacity, and soil hydraulic properties, cropping systems, and traffic and tillage management. After model calibration, predicted and measured total runoffs for the 5-year period were similar. Values of root mean square error (RMSE) for daily runoff ranged from 5.7 to 9.2 mm, which were similar to those reported in literature. The model explained 75–95% of variations of daily, monthly and annual runoff, 70–84% of the variation in total available soil water, and 85% of the variation in yield. The results showed that the PERFECT daily soil–crop simulation model could be used to generate meaningful predictions of the interactions between crop, soil and water under different tillage and traffic systems.Ranking of management systems in order of decreasing merit for runoff, available soil water and crop yield was (1) controlled traffic zero tillage, (2) controlled traffic stubble mulch, (3) wheeled zero tillage, and (4) wheeled stubble mulch.     

Effects of tillage and traffic on crop production in dryland farming systems: II. Long-term simulation of crop production using the PERFECT model

Soil water conservation is critical to long-term crop production in dryland cropping areas in Northeast Australia. Many field studies have shown the benefits of controlled traffic and zero tillage in terms of runoff and soil erosion reduction, soil moisture retention and crop yield improvement. However, there is lack of understanding of the long-term effect of the combination of controlled traffic and zero tillage practices, as compared with other tillage and traffic management practices.In this study, a modeling approach was used to estimate the long-term effect of tillage, traffic, crop rotation and type, and soil management practices in a heavy clay soil. The PERFECT soil–crop simulation model was calibrated with data from a 5-year field experiment in Northeast Australia in terms of runoff, available soil water and crop yield; the procedure and outcomes of this calibration were given in a previous contribution. Three cropping systems with different tillage and traffic treatments were simulated with the model over a 44-year-period using archived weather data.Results showed higher runoff, and lower soil moisture and crop production with conventional tillage and accompanying field traffic than with controlled traffic and zero tillage. The effect of traffic is greater than the effect of tillage over the long-term. The best traffic, tillage and crop management system was controlled traffic zero tillage in a high crop intensity rotation, and the worst was conventional traffic and stubble mulch with continuous wheat. Increased water infiltration and reduced runoff under controlled traffic resulted in more available soil water and higher crop yield under opportunity cropping systems.     

Relationship between soil structure and runoff/soil loss after 24 years of conservation tillage

A better understanding of tillage and stubble management effects on surface soil structure is vital for the development of effective soil conservation practices for the long-term. Relationships between aspects of soil structure and runoff/soil loss were investigated in 24 year old field experiment on an Oxic Paleustalf, in NSW, Australia. Two tillage/stubble management systems were compared, namely direct drilled/stubble retained (DD/SR) versus conventional tillage/stubble burnt (CC/SB). Tillage and stubble burning significantly increased bulk density and decreased the macro-aggregate stability, mean weight diameter (MWD), geometric mean diameter (GMD) and total porosity, particularly macroporosity (>60 μm). For the 0–5 cm layer, DD/SR had significantly higher water stability of macro-aggregates >2 mm than CC/SB (165 g/kg versus 78 g/kg), and the volume of pore space of diameter >60 μm at 0–5 cm depth was significantly greater (more than 11%) for DD/SR than for CC/SB. Under simulated rainfall (100 mm/h) and the removal of surface stubble, both runoff and soil loss were significantly higher under CC/SB compared to DD/SR. The infiltration rate at the end of the experiment under DD/SR was 3.7 times that of CC/SB (85 mm/h versus 23 mm/h). There were significant correlations between the proportion of soil particles >0.25 mm measured after wetting by rain and both final infiltration rate (P < 0.001) and soil loss (P < 0.001). It was concluded that 24 years of direct drilling and stubble retained practices significantly reduced runoff and soil erosion hazards, due to a fundamental change in soil structure, viz. higher soil aggregate stability and higher macroporosity of the surface soil.     

农田休闲期作物留茬对近地表风场的影响

茬是防治农田休闲期土壤风蚀的有效措施。该文利用野外观测数据,通过比较不同高度的4种作物留茬及冰草覆盖地表的风速廓线,分析各种处理下近地面风速和空气动力学特征,揭示作物留茬防止土壤风蚀的机制。结果表明：作物留茬降低了地面以上2 m高度内的风速,降低幅度与留茬高度成正比,与观测高度成反比。高度相同时不同类型的留茬比较,莜麦和小麦茬下风速降低幅度较大,植株稀疏的油菜茬和秸秆较细的胡麻茬的防风效果较差。同类作物不同高度的留茬覆盖下,地表空气动力学粗糙度随留茬高度的增加而增大,在莜麦、油菜和小麦茬高度达到15、30和15 cm时显著增大,是3种留茬防风的有效高度。可见,高度适宜的作物留茬能降低近地面风速,削弱侵蚀力,有效防治土壤风蚀。     

带田残茬带宽度及高度对土壤风蚀模数影响的风洞试验

为了探究带宽和残茬高度对土壤风蚀模数的影响,利用野外移动式风洞对阴山北麓地区带田土壤风蚀模数进行了原位测试。结果表明：作物残茬带对间作翻耕地具有保护作用,土壤风蚀模数在全部留茬时最小,全部裸露时最大,残茬与耕翻裸露地间作时居中。不同残茬降低风蚀模数程度不同,莜麦茬可以降低48.92%～67.39%,油菜茬可以降低5.22%～34.36%。不同带宽和残茬高度的间作组合,土壤风蚀模数都随风速的增加而增大,呈指数曲线变化规律。在一定风速吹蚀下,残茬带宽从0.5～3.0?m变化,土壤风蚀模数随带宽的增加而减小,符合幂函数曲线变化规律;残茬高度从5～25?cm变化,土壤风蚀模数随残茬高度增加而线性降低。研究说明减少耕翻和作物留茬是控制风蚀的有效措施。     

Type and time of autumn tillage with and without herbicides at reduced rates in southern Sweden: 1. Yields and weed quantity

Environmental as well as economic incentives support the use of integrated weed management (IWM) systems for crop production. In order to reduce the input of agrochemicals for weed control, it may be possible to combine reduced doses of herbicides with appropriate tillage strategies and still maintain acceptable weed population levels. The purpose of this study was to compare the efficiency of different tillage strategies, with and without herbicides at 50% of normal rates, on weed populations and crop yields. The influences of type and time of stubble cultivation, harrowing and mouldboard ploughing were followed in eight long-term field trials in southern Sweden from 1988 until the spring of 1994. The crop rotations were dominated by spring-sown oats (Avena sativa L.) and barley (Hordeum distichon L.). Although significant site-by-year-by-treatment interactions were found, certain trends in influences of tillage on weeds and yields were identified. Weed density and weight often decreased in the order: stubble cultivation without ploughing > solely ploughing > stubble cultivation succeeded by ploughing. Grain yields of oats and barley varied 6–50% among tillage systems and the highest yields were often obtained in plots where ploughing succeeded stubble cultivation. Effects of tillage on weeds or yields were usually not influenced by herbicide application. Although only subnormal herbicide rates were used, weed density and weight were mostly reduced by 70–90%, while yields increased by 10–20%. In the spring of 1994, 1 year after the last herbicide application, densities of annual broad-leaved weeds were 40–65% lower in plots previously treated by herbicides than in non-treated plots, and differences among tillage systems were still significant. This study shows the advantage of combining herbicides at reduced rates with stubble cultivation and ploughing.     

稻茬高度影响轮作紫云英生长及还田效应

大田试验条件下,研究了水稻机械收割不同留茬高度(15cm 、30cm 、40cm,分别标注低茬、中茬、高茬)对轮作紫云英生长、养分积累以及翻压还田后土壤养分、化学性状和微生物量等的影响。结果表明：稻茬高度显著影响轮作紫云英生长、养分积累以及翻压还田效果。紫云英单位面积株数随留茬高度的降低而显著减少；除单株分枝数外,植株株高、茎粗、单株鲜、干重均表现为高茬处理＞中茬处理＞低茬处理,高、低茬处理间差异显著或极显著。产草量受稻茬高度影响明显,单位面积鲜、干草产量均随稻茬高度的降低而显著减少,高茬处理的紫云英鲜、干草产量均最高,分别达到19.22t.hm_^-2和3.27 t.hm_^-2,较低茬处理每公顷增加15.21 t和2.53 t。不同稻茬高度处理的紫云英植株碳及氮、磷、钾含量无显著差异,植株养分积累量则随稻茬高度增加而显著增加,各养分积累量变化趋势一致；不同处理紫云英翻压还田后土壤pH值变化差异不显著,但均较试验前略有上升；土壤电导率则表现为高茬处理大于中、低茬处理,处理间差异极显著; 翻压后土壤养分变化差异较大,各养分含量均表现为随水稻留茬高度降低而减少的趋势,高留茬处理土壤的有机质、全氮、碱解氮、有效磷、速效钾含量分别较低留茬处理增加41.51%、27.35%、34.96%、68.78%、82.12%。不同处理紫云英翻压还田后,土壤可培养微生物数量存在差异。三大类微生物细菌、真菌、放线菌的数量均以高茬处理最高,除放线菌外,差异均达显著水平,其中,对细菌数量的影响最为明显,高茬处理极显著大于中、低茬处理,中、低茬处理间差异显著。综上,水稻留高茬(茬高40cm)轮作紫云英可获得较高的生物量和养分积累,翻压还田后土壤培肥效果显著优于中、低茬(茬高≦30cm)收割的处理。