基于叶绿素叶面分布特征的黄瓜氮镁元素亏缺快速诊断

为了快速、无损诊断作物氮(N)、镁(Mg)营养亏缺,该研究提出一种以叶绿素叶面分布特征诊断黄瓜N、Mg元素亏缺的方法。在设施栽培模式下精确控制N、Mg营养元素的供给,培养黄瓜缺N、缺Mg及对照植株(营养元素正常植株),然后采集对应的高光谱图像并结合化学计量学方法快速、无损检测叶绿素分布。与对照组叶片叶绿素分布相比,缺N叶片的叶绿素含量在整个叶面区域偏低,缺Mg叶片叶绿素在叶脉之间区域含量偏低。鉴于此,提取叶绿素叶面分布特征(叶片所有像素点对应的叶绿素含量均值及标准差)对N、Mg营养元素亏缺进行诊断,对预测集N、Mg元素亏缺正确诊断率达90%。研究结果表明叶绿素叶面分布特征可作为一种黄瓜N、Mg元素亏缺诊断依据。     

缺镁对菜豆幼苗膜脂过氧化及体内活性氧清除酶系统的影响

营养液中培养菜豆幼苗,缺镁菜豆植株生长受抑,生物量显著低于正常供应镁的植株,体内镁浓度和镁的积累总量均较低。严重缺镁黄化叶片中叶绿素含量下降。完整叶绿体的希尔反应值在缺镁与正常植株间无差别。缺镁菜豆植株的根和叶膜造性（以相对电导率表示）增加,叶片中丙二醛含量升高。活性氧清除酶系统:SOD、POD、CAT活性在缺镁菜豆真叶中高于正常植株,老叶中抗坏血醒过氧化物酶（AsA-POD）活性增加程度也较大。     

基于最小二乘支持向量回归的鹅肉弹性的可见-近红外光谱测定

为简化鹅肉弹性的可见-近红外光谱模型和提高预测精度,采用联合区间偏最小二乘法(synergyinterval partial least square algorithm,siPLS)结合遗传算法(Genetic algorithm,GA)提取可见-近红外光谱特征波长,用最小二乘支持向量回归(least square support vector for regression,LSSVR)建立鹅肉弹性的预测模型。试验以万能试验机获取恢复距离S作为鹅肉弹性实际值。在模型建立过程中,先利用sym8小波的2层分解对原始的可见-近红外光谱进行光谱预处理;然后用siPLS优选出4个特征光谱子区间(分别为第3、5、9、13子区间),在这4个特征光谱子区间内继续用GA优选出74个特征波长,并建立基于LSSVR的鹅肉弹性的预测模型。模型预测集的决定系数(R2)和预测均方根误差(root mean squarederror of prediction,RMSEP)分别为0.9096和0.0588。试验结果表明,siPLS结合GA法能够有效提取光谱中的鹅肉弹性对应的特征波长,有利于提高LSSVR模型预测鹅肉弹性的精度。     

特征波长筛选在近红外光谱测定梨硬度中的应用

为了提高应用近红外光谱分析技术快速测定梨硬度的精度和稳定性,该研究采用联合区间偏最小二乘和遗传算法（siPLS-GA）在校正模型中用来筛选特征光谱区域和波长,通过交互验证法确定模型的主成分因子数和筛选的波长,并以预测均方根误差（RMSEP）和相关系数（Rp）作为模型的评价标准。基于siPLS-GA的最优模型包含4个光谱区、96个变量和10个主成分因子。该模型结果显示：最佳预测模型相关系数（Rp）和RMSEP分别为0.9083和0.5573。研究结果表明,近红外光谱技术结合siPLS-GA建模用于无损、快速测定梨的硬度是可行的。     

不同产量水平下稻茬小麦的氮素营养指标特征

  【目的】  明确长江中下游地区不同产量水平稻茬小麦氮营养指标变化规律,为小麦氮素营养状况实时诊断提供理论依据。  【方法】  本研究通过江苏省多年多点田间不同氮肥、播期、密度和品种试验,构建不同产量水平小麦大数据,分析不同产量水平小麦在不同生育时期的干物质积累量、植株氮积累量和植株氮浓度的变化规律,并通过计算小麦临界氮浓度,得到累积氮亏缺和氮营养指数的动态变化规律,进而明确高产稻茬小麦氮素营养指标特征。  【结果】  干物质积累量和植株氮积累量的变化趋势一致,随着小麦生育进程的推进均逐渐增加,植株氮浓度逐渐减小,累积氮亏缺和氮营养指数会出现波动。整个生育期内,干物质积累量和植株氮积累量在高产小麦和中产小麦之间的差异不显著,但二者植株氮积累量均显著高于中低产小麦,中低产小麦又显著高于低产小麦。在抽穗期、开花期和灌浆期,高产和中产小麦干物质积累量显著高于中低产小麦,中低产小麦又显著高于低产小麦。依据累积氮亏缺值判断氮素营养状况,高产和中产小麦的累积氮亏缺变化趋势一致,在起身—孕穗期,高产小麦的累积氮亏缺值由0.3 kg/hm2降低为?23.0 kg/hm2,中产小麦由7.0 kg/hm2降低为?14.6 kg/hm2,孕穗—抽穗期又呈升高趋势,高产小麦由?23.0 kg/hm2升高为?11.4 kg/hm2,中产小麦由?14.6 kg/hm2升高为2.4 kg/hm2,开花—灌浆期的波动较小。表明高产小麦氮营养除起身期之外均为过剩,中产小麦在拔节—孕穗期的累积氮亏缺小于0,其余时期累积氮亏缺均大于0,但该水平的累积氮亏缺值一直在适宜范围内波动。中低产小麦在起身—拔节期,累积氮亏缺值由14.2 kg/hm2降低为9.5 kg/hm2,之后逐渐升高,灌浆期达到最大为43.9 kg/hm2;低产小麦在起身期到灌浆期,累积氮亏缺值由17.3 kg/hm2升高为71.1 kg/hm2。表明中低产和低产的小麦氮营养水平在整个生育期内逐渐降低,且整个生育期均处于亏缺状态 (累积氮亏缺值 > 0)。在拔节期、孕穗期、抽穗期和灌浆期,高产水平的小麦植株实际氮浓度高于植株临界氮浓度,中产小麦在孕穗期的植株实际氮浓度高于临界氮浓度,而中低产和低产的小麦在整个生育期植株实际氮浓度低于植株临界氮浓度。高产和中产的小麦氮营养指数在1附近波动,其中高产小麦的氮营养指数在起身—孕穗期由0.9升高为1.1,在抽穗—灌浆期,氮营养指数呈现先降低后升高趋势,其值分别为1.0、0.9和1.0,中产小麦与高产小麦的变化趋势一致,起身—孕穗期的氮营养指数由0.8升高为1.0,之后逐渐下降,其值均小于1,抽穗—灌浆期分别为0.9、0.9和0.9。中低产和低产的小麦氮营养指数始终低于1,中低产小麦在起身—拔节期氮营养指数由0.7升高为0.8,之后则逐渐下降,低产小麦从起身—开花期均逐渐下降,而这两个产量水平的氮营养指数在灌浆期会呈现略微上升趋势。  【结论】  随着产量水平的提高,小麦植株干物质和氮积累量、植株氮浓度、氮营养指数等都相应增加,累积氮亏缺相应下降。较高的干物质积累量和植株氮积累量是提高小麦产量的主要原因,在小麦生长过程中氮营养指数和累积氮亏缺均能准确诊断小麦氮营养状况,可为小麦氮肥精准管理提供理论支持。     

基于dbiPLS-SPA变量筛选的固态发酵湿度近红外光谱检测

为了提高基于近红外光谱技术的固态发酵关键过程参数——湿度快速检测的精度和稳定性,研究采用动态反向区间偏最小二乘（dbiPLS）法结合连续投影算法（SPA）进行最佳光谱子区间和特征组合变量的筛选,通过交互验证法确定偏最小二乘（PLS）模型的主成分因子数,并以预测均方根误差（RMSEP）和相关系数（Rp）作为模型的评价标准。试验结果显示,最佳dbiPLS-SPA模型筛选的组合变量个数为8,其RMSEP和Rp分别为1.1795%（质量分数）和0.9430。试验结果表明,dbiPLS-SPA是一个有效的波长组合变量筛选方法,可简化模型结构、增强模型精度和稳健性。     

可见-近红外与中红外光谱预测土壤养分的比较研究

对土壤养分的快速和准确测定有助于适时指导施肥。为进一步研究可见-近红外（350～2500 nm）与中红外光谱（4000～650 cm-1）对土壤养分的预测能力,以贵州省500个土样为例,对光谱进行Savitzky-Golay（SG）平滑去噪处理,再用标准正态化（SNV）方法进行基线校正,然后分别应用偏最小二乘回归（PLSR）和支持向量机（SVM）两种方法进行建模,探讨了可见-近红外和中红外光谱对土壤全氮（TN）、全磷（TP）、全钾（TK）和碱解氮（AN）、有效磷（AP）、速效钾（AK）共六种土壤养分的预测效果。结果表明：（1）无论基于可见-近红外光谱还是中红外光谱,PLSR模型的预测精度整体均优于SVM模型。（2）中红外光谱对TN、TK和AN的预测精度均显著高于可见-近红外光谱,可见-近红外和中红外光谱均可以可靠地预测TN和TK（性能与四分位间隔距离的比率（RPIQ）大于2.10）,中红外光谱可相对较可靠地预测AN（RPIQ=1.87）;但两类光谱对TP、AP和AK的预测效果均较差（RPIQ<1.34）。（3）当变量投影重要性得分（VIP）大于1.5时,PLSR模型在中红外光谱区域预测TN和TK的重要波段多于可见-近红外光谱区域,TN的重要波段主要集中于可见-近红外光谱区域的1910和2207 nm附近,中红外光谱区域的1 120、1 000、960、910、770和668 cm-1附近;TK的重要波段主要集中于可见-近红外光谱区域的540、2176、2225和2268 nm附近,中红外光谱区域的1 040、960、910、776、720和668 cm-1附近。因此,中红外光谱技术结合PLSR模型对土壤养分预测效果较好,可快速准确预测土壤TN和TK,可为指导适时施肥提供技术支撑。     

土壤全氮的无人机高光谱响应特征及估测模型构建

为更好地体现出光谱与土壤全氮（soil totalnitrogen,STN）含量之间的响应关系,实现以高光谱快速估测土壤全氮含量,该研究以无人机搭载高光谱传感器获取农田土壤高光谱影像,提取光谱反射率并进行数学变换,基于灰色关联度和皮尔逊相关系数提取各光谱中土壤全氮含量的敏感波段,基于敏感波段采用偏最小二乘回归（partialleastsquares regression,PLSR）、岭回归（ridge regression,RR）和随机森林（random forest,RF）构建土壤全氮的高光谱反演模型,筛选出最优模型并对研究区土壤全氮含量进行反演制图。结果表明：1）反射率的倒数光谱中的敏感波段（996～1 003 nm）集中在近红外长波范围内,反射率的一阶微分（first derivative of reflectance,FDR）光谱中的敏感波段（398～459、469和472～1 003 nm）和反射率对数的一阶微分光谱中的敏感波段（398～459、463～973和978～1 003 nm）在可见光和近红外范围内都有分布,反射率的一阶微分光谱中的敏感波段（615～625、632和...     

缺素培养对大豆营养生长和根系形态的影响

通过营养液栽培试验,探讨了巴西10号和本地2号两个大豆基因型在氮、磷、钾、钙、镁、硫、铁缺乏条件下的缺素症状和植株生物量变化,以及根系形态指标的改变。结果表明,缺素处理对大豆植株生长影响显著,在各种缺素条件下大豆植株生物量均明显降低,地上部与根部均表现出各种营养元素缺乏时的特有症状;缺氮、缺钾、缺钙、缺镁和缺铁处理时大豆总根长和根表面积也均显著降低。大豆对不同缺素处理的反应存在基因型差异。     

冬枣氮素含量预测模型中特征波长选择方法的应用

为了提高近红外光谱法快速测定枣叶氮含量的准确性和鲁棒性。采用偏最小二乘法建立了冬枣叶片氮含量近红外光谱模型。模型的相关系数为0.799,均方根误差为0.055。整个光谱区域包含了许多与冬枣氮含量无关的光谱变量。冗余信息的存在降低了模型的预测性能。所以采用间隔偏最小二乘(IPLS)结合遗传算法和模拟退火算法来选择冬枣叶片氮含量的特征波长。用凯氏定氮法测定冬枣叶样品的氮含量。试验选用15棵枣树,每棵树5个叶片作为试验对象。用于光谱测量的仪器是ASD光谱仪,测试仪在350~2 500 nm波长范围内,光谱分辨率为1 nm。在数据采集前使用了白板进行校正(标准白板反射系数为1),每个样品测量了5次,取平均值作为样品的相对反射率。遗传算法结合间隔偏最小二乘法选取的4个特征波长为685,689,781,783 nm。根据这4个波长,建立了冬枣叶片氮含量近红外光谱模型。模型预测相关系数为0.9175,预测均方根误差为0.063。利用模拟退火算法,建立了7个波长的冬枣叶片氮含量的近红外光谱模型。模型的相关系数为0.9301,均方根误差为0.052。因此,近红外光谱结合光谱选择方法的特点,可以有效地提高模型的精度,使模型更实用。但光谱选择方法的特点并不普遍。基于单波长变量选择的模型更为敏感,更适用于均匀采样。基于波长间隔选择的模型抗干扰能力相对较强,但更适合于不均匀采样。因此,基于状态与模型相结合的特征选择可以更好地应用于模型。     

论水土保持生态环境监督管理发展新趋势及其对策

通过深入贯彻落实《水土保持法》,水土保持生态环境监督管理呈现出了新的发展趋势 :社会大环境明显趋好 ;水土流失防治呈社会化、市场化、效益化发展 ;水土保持监督执法向全面性的监督管理转化 ;“预防为主”方针正在得到加强和落实 ;规范化建设成效代表了向高层次发展的趋势 ;城市水土保持监督管理工作加强等。水保监督管理好的发展趋势与问题并存 ,应采取以下对策 :依法行使归口管理职责 ;提高认识 ,加强管理 ;强化“预防为主”首要方针 ;综合发展人为水土流失防治技术 ;开展水土保持监测工作 ,提高监督管理整体水平等 ,以促进水土保持生态环境监督管理事业的顺利发展     

Study on Soil and Pine-Seedling Zn and Mn and Predicting-Model Design

Soils were collected from 2-year (2-y) and 3-year (3-y) old red-pine seedling plots in two tree nurseries, Hayward in the north and Wilson in the southwestern part of Wisconsin State respectively, and equilibrated with 0.01 M Ca(NO₃)₂ for soil solution Zn and Mn (solu-Zn and Mn), and with 0.01 M Ca(NO₃)2+0.005 M EDTA for soil adsorbed Zn and Mn (ad-Zn and Mn). Buffering capacity of soil Zn and Mn (b-Zn and Mn) was obtained from the ratio of ad-Zn and Mn to the solu-Zn and Mn. The concerned traces in pine seedling needles (ndls), stems(sts) and roots (rts) were simultaneously measured.     

Carbon and nitrogen natural stable isotopes in Slovene honey: adulteration and botanical and geographical aspects

Isotope parameters (δ(13)C(honey), δ(13)C(protein), δ(15)N) were determined for 271 honey samples of 7 types (black locust, multifloral, lime, chestnut, forest, spruce, and fir honeys) from 4 natural geographical regions of Slovenia. Carbon and nitrogen stable isotope ratios were measured to elucidate the applicability of this method in the identification of the botanical and geographical origin of honey and in honey adulteration. Only 2.2% of the samples were adulterated according to the internal standard carbon isotope ratio analysis method. Botanical origin did not have any major influence on the honey isotope profiles; only black locust honey showed higher δ(13)C values. Some differences were seen across different production years, indicating that the influence of season should be further tested. Statistical and multivariate analyses demonstrated differences among honeys of various geographical origins. Those from the Alpine region had low δ(13)C (-26.0‰) and δ(15)N values (1.1‰); those from the Mediterranean region, high δ(13)C (-24.6‰) and medium δ(15)N values (2.2‰); those from the Pannonian region, medium δ(13)C (-25.6‰) and high δ(15)N value (3.0‰); and those from the Dinaric region, medium δ(13)C (-25.7‰) and low δ(15)N values (1.4‰).     

Procyanidin and catechin contents and antioxidant capacity of cocoa and chocolate products

Cocoa and chocolate products from major brands were analyzed blind for total antioxidant capacity (AOC) (lipophilic and hydrophilic ORAC(FL)), catechins, and procyanidins (monomer through polymers). Accuracy of analyses was ascertained by comparing analyses on a NIST standard reference chocolate with NIST certified values. Procyanidin (PC) content was related to the nonfat cocoa solid (NFCS) content. The natural cocoa powders (average 87% of NFCS) contained the highest levels of AOC (826 +/- 103 micromol of TE/g) and PCs (40.8 +/- 8.3 mg/g). Alkalized cocoa (Dutched powders, average 80% NFCS) contained lower AOC (402 +/- 6 micromol of TE /g) and PCs (8.9 +/- 2.7 mg/g). Unsweetened chocolates or chocolate liquor (50% NFCS) contained 496 +/- 40 micromol of TE /g of AOC and 22.3 +/- 2.9 mg/g of PCs. Milk chocolates, which contain the least amount of NFCS (7.1%), had the lowest concentrations of AOC (80 +/- 10 micromol of TE /g) and PCs (2.7 +/- 0.5 mg/g). One serving of cocoa (5 g) or chocolate (15 or 40 g, depending upon the type of chocolate) provides 2000-9100 micromol of TE of AOC and 45-517 mg of PCs, amounts that exceed the amount in a serving of the majority of foods consumed in America. The monomers through trimers, which are thought to be directly bioavailable, contributed 30% of the total PCs in chocolates. Hydrophilic antioxidant capacity contributed >90% of AOC in all products. The correlation coefficient between AOC and PCs in chocolates was 0.92, suggesting that PCs are the dominant antioxidants in cocoa and chocolates. These results indicate that NFCS is correlated with AOC and PC in cocoa and chocolate products. Alkalizing dramatically decreased both the procyanidin content and antioxidant capacity, although not to the same extent.     

NMR and molecular mechanics study of pyrethrins I and II.

Bioassay-directed fractionation of the organic extract of the Kenyan pyrethrum flowers (Chrysanthemum cinerariaefolium Vissiani) resulted in the isolation of two natural pyrethrin esters, pyrethrin I (PI) and pyrethrin II (PII) as the major constituents. These esters elicited inhibition of the multiple drug resistant (MDR) Mycobacterium tuberculosis. The high-field (1)H and (13)C nuclear magnetic resonance (NMR) chemical shifts of PI and PII were unequivocally assigned using modern two-dimensional (2D) proton-detected heteronuclear multiple-quantum coherence (HMQC) and heteronuclear multiple-bond correlation (HMBC) experiments. The conformations of both esters were deduced from (1)H-(1)H vicinal coupling constants and confirmed by 2D nuclear Overhauser effect spectroscopy (NOESY). Computer molecular modeling (MM) studies revealed that PI and PII molecules adopt a "love-seat" conformation in chloroform (CDCl(3)) solution.     

A-type proanthocyanidins from lychee seeds and their antioxidant and antiviral activities

Two new A-type trimeric proanthocyanidins with two doubly bonded interflavanoid linkages, litchitannin A1 [epicatechin-(2β→O→7,4β→6)-epicatechin-(2β→O→7,4β→8)-catechin] (1) and litchitannin A2 [epicatechin-(2β→O→7,4β→6)-epicatechin-(2β→O→7,4β→6)-epicatechin] (2), were isolated from lychee (Litchi chinensis Sonn. cv. Heiye) seeds together with aesculitannin A (3), epicatechin-(2β→O→7,4β→8)-epiafzelechin-(4α→8)-epicatechin (4), proanthocyanidin A1 (5), proanthocyanidin A2 (6), proanthocyanidin A6 (7), epicatechin-(7,8-bc)-4β-(4-hydroxyphenyl)-dihydro-2(3H)-pyranone (8), and epicatechin (9). Their structures were elucidated on the basis of spectroscopic and chemical evidence. It is the first time that compounds 1-4, 7, and 8 have been reported in this species. Compounds 1-9 showed more potent antioxidant activity than L-ascorbic acid with ferric reducing antioxidant power (FRAP) values of 3.71-24.18 mmol/g and IC50 values of 5.25-20.07 μM toward DPPH radicals. Moreover, litchitannin A2 (2) was found to exhibit in vitro antiviral activity against coxsackie virus B3 (CVB3) and compounds 3 and 6 displayed antiherpes simplex virus 1 (HSV-1) activity.     

Root development and growth of oilseed, wheat and pea crops on tilled and non-tilled soil

Field experiments were done on red brown earth soils at two sites in South Australia between 1979 and 1981. The effects of different amounts and types of tillage on root growth and plant development were compared for linseed, pea, rape, safflower, soybean, sunflower and wheat. Crops were grown in 2 × 1.5 m² plots with the following treatments: (a) tine-tilled to 120 mm; (b) non-tilled with seeds sown in 30 mm slots, (c) non-tilled with seeds sown in narrow, 30 mm deep holes, and (d) and (e) as for (b) but with vertical cracks extending from the base of the slots to depths of 120 and 300 mm, respectivley.

In the naturally-compacted non-tilled soil, penetrometer resistance was highly dependent on water content. Treatments (b) and (c) restricted seminal root growth relative to treatment (a), and the degree of restriction was dependent on penetrometer resistance in the post-planting period. Growth of lateral roots was also less in (b) and (c) than in (a). The reductions in early root growth for (b) and (c) were associated with reduced dry matter production and yield of the crops. However, the presence of cracks in (d) and (e) provided a zone for unrestricted root growth and the dry matter production and yields were comparable to those in (a).

The magnitude of the reductions in root and dry matter productions for (b) and (c) relative to (a) ranged from 0% to > 50%, depending on the crop and the planting time. On the basis of reductions in dry matter production from (a) to (b), it appears that the suitability of these crops for growth in non-tilled soil decreases in the order: wheat > pea > rape > linseed > safflower and sunflower. In separate experiments, soybean and sunflower suffered similar yield reductions in non-tilled soil.     

Influence of Plant Residues after Steam-Treatment with High Temperature and Pressure on Soil Microbial Biomass C and N, Water-Soluble C and N, and pH

Steam-treated grass trimmings (GrT), and wooden chips of pine (PnT) and walnut branches (WnT) or non-treated grass trimmings (GrNoT) and wooden chips of pine (PnNoT) and walnut branches (WnNoT) were mixed with soil (S) and then incubated for 10 weeks to measure microbial biomass C (MBC) and N (MBN), water-soluble C (WSC) and N (WSN), and soil pH. The results showed that there were a larger amount of MBC and WSC in steam-treated samples, principally in early period. It was probably due to the steam-treatment facilitating solubilization of organic C that served with substrate for microbial biomass. In grass trimming, the MBN and WSN had differences between treated and non-treated. No such differences in wooden chips was probably because the low total N content in these plants. The higher value of WSN showed in S+GnNoT showed more potential of nitrate pollution. The steam-treatment did not influence on soil pH after 1 week of incubation.     

Tillage and mulch effects on productivity and water use of pea and soil carbon stocks

Soil management can notably influence crop production under rainfed farming; however, improper soil management is one of the key factors threatening sustainability. The objective of this 3 years’ study was to evaluate the performance of two types of tillage: conventional tillage (CT) and zero tillage (ZT) systems with four mulches; paddy straw mulch (PSM), maize stubble (MS), Imperata cylendrica (thatch grass, TG) and no mulch (NM) on yield and water use of pea (Pisum sativum) and physico-chemical properties of soil. CT and PSM have registered 11.1% and 36.1% higher pod yield, 10.0% and 40% improvement of water use efficiency (WUE) and 9.7% and 49.2% better harvest monetary benefit (HMB) over ZT and NM, respectively. Soil moisture content (SMC) was higher with ZT than CT. Placement of mulch has considerably higher SMC on upper soil layer over NM, but at deeper depth, there was no such evidence. The soil organic carbon (SOC) was considerably improved by 4.1% with ZT than CT, whereas, MS improved SOC by 1.9% than NM. The findings clearly suggested that CT along with PSM registered improvement in pod yield, WUE and HMB, but ZT and MS improved SOC at different soil depths and responded to carbon stock management.     

干旱胁迫与复水条件下高粱、玉米苗期渗透调节及抗氧化比较研究

盆栽试验结果表明,高粱和玉米2种作物均存在渗透调节和抗氧化耐旱途径。高粱耐旱性较强,干旱条件下叶片相对含水量较高,水势亦较大。高粱以可溶性糖和蛋白质为主要渗透调节物质,以过氧化氢酶(CAT)为主要抗氧化酶。玉米以K 和脯氨酸为主要渗透调节物质,以超氧化物歧化酶(SOD)和过氧化物酶(POD)为主要抗氧化酶。