近红外光谱法测定DDGS中粗蛋白含量的试验研究

试验建立DDGS粗蛋白含量测定的近红外光谱分析定标模型。采用化学分析法测定72个DDGS样品中的粗蛋白含量,利用FOSS InfraXact型近红外光谱分析仪采集样品光谱,光谱经2,4,4,1导数和标准正常化+散射处理(SNV+Detrend),用改进最小二乘法(MPLS)回归,获得了较好的定标模型,校正决定系数(RSQ)、交叉验证决定系数(1-VR)、校正标准误差(SEC)、交叉验证标准误差(SECV)分别为0.982 5、0.932 8、0.266 2、0.389 5。利用30个验证集的DDGS样品进行外部检验,预测值与真实值之间差异不显著(P>0.05)。结果表明,定标模型的预测性能较好,可以替代化学分析法快速测定DDGS中的粗蛋白含量。     

凯氏定氮法与杜马斯燃烧法测定大豆粗蛋白的比较研究

豆粕是现代畜禽养殖业基础日粮的主要蛋白来源。为比较目前最常用的粗蛋白含量检测技术的优劣,实验采自中国不同地区的162种大豆样品为实验材料,分别用凯氏定氮法和杜马斯燃烧法测定其粗蛋白含量,分析2种方法的差异,确定一种快速、安全的测定饲料原料粗蛋白含量的方法。结果表明:实验样品大豆粗蛋白均值在37%左右,最高可达49.04%(杜马斯燃烧法)。杜马斯燃烧法与凯氏定氮法在测定大部分大豆样品粗蛋白时无显著性差异(P>0.05),变异系数(CV)<1%,结果的相关系数为0.9789;燃烧法的变异系数更小(CV<0.04%),测定结果更加稳定。因此,在本试验中所选定的样品范围内,作为测定饲料原料粗蛋白含量的方法,两种方法可以相互替代,其中杜马斯燃烧法的稳定性优于凯式定氮法;至于2种方法的准确性比较需要进一步的研究。     

应用锦羊粪便的近红外光谱方程评定其日粮的粗蛋白

利用近红外光谱法 (NIRS)对绵羊粪便的扫描值和日粮粗蛋白的化学测定值来建立定标方程式。试验以绵羊为试验动物 ,日粮主要由各种牧草、作物秸秆和棉花籽壳组成 ,试验动物日粮设计了 78个蛋白水平。在 2 0 0 2年和 2 0 0 3年分别用 15只和 2 0只成熟母羊 (体重为 5 5± 2 .4kg)进行了为期 7周的试验。用凯式定氮法测定日粮的粗蛋白 (CP)水平是从 4 .3%到 2 3.5 % ,日粮粗蛋白的定标方程式决定系数R2 =0 .95 ,定标标准误差 (SEC) =1.0 8。用 12只饲喂美国北部大平原饲草的成年母羊的粪便光谱扫描值和与粪便对应的日粮粗蛋白的化学分析数据来校验粗蛋白预测方程式的有效性 ,结果显示 ,决定系数 (R2 ) =0 .81,预测标准误差 (SEP) =1.5 1,斜率 =0 .89,表明利用近红外光谱法 (NIRS)发展的粪便近红外光谱方程可以有效预测绵羊日粮的粗蛋白含量。     

应用绵羊粪便的近红外光谱方程评定其日粮的粗蛋白

利用近红外光谱法(NIRS)对绵羊粪便的扫描值和日粮粗蛋白的化学测定值来建立定标方程式.试验以绵羊为试验动物,日粮主要由各种牧草、作物秸秆和棉花籽壳组成,试验动物日粮设计了78个蛋白水平.在2002年和2003年分别用15只和20只成熟母羊(体重为55±2.4 kg)进行了为期7周的试验.用凯式定氮法测定日粮的粗蛋白(CP)水平是从4.3 %到23.5 %,日粮粗蛋白的定标方程式决定系数R2=0.95,定标标准误差(SEC)=1.08.用12只饲喂美国北部大平原饲草的成年母羊的粪便光谱扫描值和与粪便对应的日粮粗蛋白的化学分析数据来校验粗蛋白预测方程式的有效性,结果显示,决定系数(R2)=0.81,预测标准误差(SEP)=1.51,斜率=0.89,表明利用近红外光谱法(NIRS)发展的粪便近红外光谱方程可以有效预测绵羊日粮的粗蛋白含量.     

样品集的数据分布特征对建立玉米粗蛋白近红外预测模型的影响

本实验旨在研究当建模样品集的数据分布分别呈正态分布与均匀分布时对构建玉米粗蛋白的傅立叶近红外预测模型的影响,探讨建立近红外光谱预测模型的快速方法。本试验组建3个不同定标集,且其粗蛋白含量的数据分布分别呈现均匀分布(10.00,0.85)、正态分布1(10.02,0.692)、正态分布2(10.01,0.692)特征,建立粗蛋白的近红外预测模型。结果表明:均匀分布、正态分布1和正态分布2所对应的模型的R2分别为0.9879、0.9858、0.9862,RMSECV分别为0.1055、0.1079、0.1069,RSD%分别为1.06、1.08、1.07;均匀分布模型在预测各个范围的粗蛋白时其误差均在0.04以内,而正态分布1模型的误差依次为0.09、0.06、0.02、0.01、0.07、0.10。结果显示,在相同定标样品数下,定标集呈均匀分布时所建预测模型的预测误差变异小,并且在预测含量偏离平均数较大的样品时效果好于正态分布,而正态分布则是在预测含量在接近平均数的样品时有优势;同时在减少一定数量的定标样品后,使用均匀分布的定标集仍然可以保持所建预测模型的准确性。     

饲料真蛋白的定量检测

目前，在饲料中掺杂使假的现象屡见不鲜，其中较常见的是粗蛋白掺假。粗蛋白含量是饲料质量评定的关键指标之一，也是一般化验室检测的重要项目。检测方法为凯氏定氮法。有人利用凯氏法只检测饲料中含氮量，再根据凯氏常数推算粗蛋白含量的原理，于饲料中混入尿素、硫酸铵等含氮量高的非蛋白氮类物质，以提高表观粗蛋白含量。而一般饲料厂采用普通凯氏法测粗蛋白，无法鉴别真伪。现提供一种用凯氏法测真蛋白的方法。将样品充分混匀后分成两份，准确称取。一份用于常规凯氏定氮测粗蛋白含量。将另一份加5倍的三氯乙酸溶液（10％浓度），搅动…     

饲料中硝态氮对燃烧法与凯氏法总氮含量测定结果的影响

为了确定硝态氮对饲料含氮量的杜马斯燃烧法测定值(Cn)和凯氏法测定值(Kn)的影响,对14种硝态氮含量较高的饲料原料,采用杜马斯燃烧法和凯氏法测定各样本中的含氮量,并以化学法测定硝态氮含量为对照进行分析。结果表明:14种样本含氮量的燃烧法测定值显著高于凯氏法测定值(P<0.01),虽然2种方法测定结果间高度相关(r=0.9960),但拟合曲线与Y=X存在显著差异(P<0.01),燃烧法与凯氏法测定值间差值(Cn-Kn)与饲料中硝态氮含量相关性低(R2=0.6036)。由此可得出,当饲料中含有大量硝态氮时,燃烧法测定值显著高于凯氏法测定值,但2种方法测定值差与硝态氮含量相关性低,因此硝态氮是造成饲料含氮量的燃烧法与凯氏法测定值间差异的原因之一,但不是惟一原因。     

饲料粗蛋白含量分析方法比较

本文利用3种仪器设备,PE2400N元素分析仪、凯氏氮/蛋白质测定仪、凯氏定氮仪(国标法)对13种饲料原料和2种配合饲料样品的粗蛋白含量进行了分析测试,旨在探讨3种分析测试方法的测量误差,以检验各设备在饲料生产质量控制中的作用。经统计分析,发现这3种方法的测定结果无显著差异。另外,与国标法相比,PE2400N元素分析仪自动化程度较高,具有快速、精确、省时省力等特点,值得提倡;凯氏氮/蛋白质测定仪体积小,重量轻,省工省力,样品不转移,测试数据准确,价格低廉,适合我国国情,具有推广价值。     

应用近红外光谱分析技术检测毒死蜱的含量

运用近红外光谱技术结合偏最小二乘法进行了溶液中微量的毒死蜱含量的测定试验。配制39个质量浓度为0.005 0、0.007 5、0.010 0……0.100 0 mg/kg的毒死蜱溶液样品,依2∶1分为校正集和预测集,其中26个毒死蜱溶液样品作为校正集,13个毒死蜱溶液样品作为预测集,校正集样品浓度为0.005 0、0.007 5、0.012 5、0.015 0……0.095 0、0.100 0 mg/kg,预测集样品浓度为0.010 0、0.017 5、0.025 0……0.097 5 mg/kg。选取1 100~1 500 nm波长范围的光谱,用二阶导数(2nd-der)结合标准正态变量变换(SNV)方法进行预处理,采用主因子数8进行建模,在波长为1 100~1 500 nm时得到了校正集(留一交叉验证法)相关系数R2为0.996 1,预测集相关系数R2为0.993 9,校正标准差为0.001 76 mg/kg,预测标准差为0.002 40 mg/kg的结果。在检测模型中预测值与实际值之间具有显著的线性相关性。研究结果有助于利用近红外光谱技术快速测定溶液中毒死蜱的含量。     

紫花苜蓿青贮饲料粗灰分含量的近红外评定方法研究

为了探讨近红外光谱分析技术在青贮饲料粗灰分测定中的效果,按照不同苜蓿品种、刈割次数和青贮方法制备了160份紫花苜蓿青贮饲料样品,用液氮冷冻技术制备近红外测定样品,用偏最小二乘回归法建模,对傅里叶变换近红外光谱技术测定青贮饲料新鲜样品中粗灰分含量的可行性进行了分析。结果表明:紫花苜蓿青贮饲料样品粗灰分近红外光谱分析的最佳光谱范围为9 736. 26~4 123. 20 cm-1,交叉检验相关系数(Rcv)和交叉检验标准误(RMSECV)分别为0. 978和0. 177。用50个样品对模型进行外部检验,预测相关系数(r)为0. 978,预测标准误(RMSEP)为0. 207。说明近红外光谱技术可以测定紫花苜蓿青贮饲料鲜样中的粗灰分含量。     

Energy evaluation of extruded compound foods for dogs by near-infrared spectroscopy

Near-infrared reflectance spectroscopy (NIRS) was used to predict the chemical composition, apparent digestibility and digestible nutrients and energy content of commercial extruded compound foods for dogs. Fifty-six foods of known chemical composition and in vivo apparent digestibility were analysed overall and 51 foods were used to predict gross energy digestibility and digestible energy content. Modified partial least square calibration models were developed for organic matter (OM), crude protein (CP), ether extract (EE), crude fibre (CF), nitrogen free extracts (NFE) and gross energy (GE) content, the apparent digestibility (OMD, CPD, EED, NFED and GED) and the digestible nutrient and energy content (DOM, DCP, DEE, DNFE and DE) of foods. The calibration equations obtained were evaluated by the standard error and the determination coefficient of cross-validation. The cross-validation coefficients of determination (R) were 0.61, 0.99, 0.91, 0.96, 0.94 and 0.92 for OM, CP, EE, CF, NFE and GE, the corresponding standard error of cross-validation (SECV) being 5.80, 3.51, 13.35, 3.64 and 16.95 g/kg dry matter (DM) and 0.29 MJ/kg DM respectively. The prediction of apparent digestibility was slightly less accurate, but NIRS prediction of digestible nutrient (g/kg DM) and DE (MJ/kg DM) gave satisfactory results, with high R (0.93, 0.97, 0.93, 0.83 and 0.93 for DOM, DCP, DEE, DNFE and DE respectively) and relatively low SECV (11.55, 6.85, 12.14 and 22.98 g/kg DM and 0.47 MJ/kg DM). It is concluded that the precision of NIRS in predicting the energy value of compound extruded foods for dogs is similar or better than by proximate analysis, as well as being faster and more accurate.     

The use of near infra‐red reflectance spectroscopy for the determination of plant nitrogen

Abstract

A Near Infra‐Red Reflectance Spectrophotometer was calibrated to analyse Italian ryegrass for protein nitrogen (N). Rye‐grass samples having a wide range in N content were analysed by standard “wet” chemistry techniques and the resulting data used to calibrate the Near Infra‐Red Spectrophotometer for ryegrass N analysis. A correlation (r) of 0,99 and standard error of calibration (SEC) of 0,209 resulted from the initial regression analysis between the Near Infra‐Red Spectroscopy (NIRS) estimated and “wet” chemistry data. In order to further evaluate the accuracy of the NIRS calibration a separate set of ryegrass samples were analysed for N content, by both the “wet” chemistry and NIRS methods, resulting in a correlation (r) of 0,98 and standard error of prediction (SEP) of 0,235. The applicability of the NIRS ryegrass calibration to other species was briefly examined by estimating the N contents of kikuyu (Pennisetum clandestinum) (r = 0,97 and SEP = 0,277).     

近红外漫反射光谱定量分析天然牧草营养成分

为了快速测定内蒙古锡林郭勒盟草原天然牧草的营养成分,试验选用内蒙古锡林郭勒盟草原2016年5-11月份的主要牧草及混合牧草样品共407份,研究利用近红外漫反射全光谱扫描技术结合实验室检测数据,用修正偏最小二乘法（MPLS）,进行粗蛋白（CP）、中性洗涤纤维（NDF）、酸性洗涤纤维（ADF）、木质素（ADL）、粗灰分（Ash）、粗脂肪（EE）、钙（Ca）、磷（P）的定标和验证。结果表明：Ca、NDF、DM、CP、Ash的外部验证相对分析误差RPD（SD/SEP）均 > 3,NIRS预测值与化学值的相关系数RSQ均在0.9以上,说明这5个指标的定标效果较好, 建立的定标模型可以用于实际检测;ADF外部验证相对分析误差2.5     

黄土高原13种栽培牧草营养成分NIRS模型分析

对2012-2013年黄土高原种植的13个牧草品种、780份干草样品的营养成分建立了近红外光谱(near infrared reflectance spectroscopy,NIRS)的检测模型。豆科牧草的粗脂肪(EE)、酸性洗涤纤维(ADF)和粗灰分(Ash)建模结果最好,其定标决定系数(RSQ)0.94,交叉验证相关系数(1-VR)0.7最高,定标标准分析误差(SEC)在0.071~0.713,交叉校验定标标准分析误差(SECV)在0.160~2.751。禾本科牧草的EE和可溶性糖(WSC)建模结果最好,RSQ分别达0.916和0.859,1-VR分别为0.609和0.810,SEC和SECV分别是0.250、1.488和0.505、3.172。菊科和车前科牧草的模型,除ADF外,其它指标预测的稳定性和准确性较为理想,RSQ在0.85以上,1-VR在0.70以上,SEC和SECV分别在0.361~3.557和0.495~4.602。NIRS对豆科粗蛋白(CP)和WSC的数值预测较差,RSQ仅0.55,对禾本科CP、ADF、中性洗涤纤维(NDF)、Ash及菊科和车前科的ADF的预测稍差,RSQ0.7。     

饲料添加剂L-赖氨酸硫酸盐中L-赖氨酸含量近红外速测方法研究

为了探讨利用近红外漫反射光谱技术(NIDRS)快速定量分析饲料添加剂L-赖氨酸硫酸盐中L-赖氨酸含量的可行性,本试验在全国范围内收集了具有代表性的L-赖氨酸硫酸盐添加剂76个,采用国家标准方法对样品中的L-赖氨酸含量进行化学赋值;用光栅型近红外光谱仪扫描L-赖氨酸硫酸盐样品,获取了不同物理状态下样品的近红外光谱图。依据L-赖氨酸含量将样品分为定标集和验证集,运用适当的光谱预处理方法,采用竞争性自适应重加权(CARS)算法结合偏最小二乘法(PLS)建立了L-赖氨酸硫酸盐的近红外定标分析模型,并将该模型与全波长模型进行了比较。结果表明:用烘干、60目粉碎后的样品结合CARS算法建立的定标模型最优,定标集校正决定系数(R2C)为0.954,校正集标准偏差(SEC)为0.510,交互验证标准偏差(SECV)为0.659;验证集预测决定系数(R2P)为0.952,预测标准偏差(SEP)为0.554,相对分析误差(RPD)值为3.83。由此可见,NIDRS定量分析L-赖氨酸硫酸盐具有一定可行性,对于丰富我国氨基酸盐及其他氨基酸制品的快速检测方法具有实际的应用意义。     

检测燕麦干草主要营养成分含量的近红外光谱分析模型

为探索NIRS技术在测定燕麦(Avena sative)干草品质上的应用,试验于2020—2021年收集了249份不同品种、年限和生长时期的燕麦干草,通过WinISI III定标软件建立燕麦干草主要营养成分的近红外光谱模型。结果显示：粗蛋白(CP)、中性洗涤纤维(NDF)和粗脂肪(EE)预测模型的定标系数(RSQ)和外部验证决定系数(RSQv)均在0.83以上,校正标准误(SEC)、交叉验证误差(SECV)和预测标准误差(RMSEP)均小于0.02,相对标准误差(RPD)均大于3,预测值逼近化学分析的精度具有良好的预测效果。酸性洗涤纤维含量(ADF)建模效果较差,定标系数和外部验证决定系数分别为0.83和0.84,校正标准误(SEC)、交叉验证误差(SECV)和预测标准误差(RMSEP)均小于0.01,接近化学分析精度,且RPD大于2.50。因此,所建ADF模型也可用于近红外预测。     

紫色土光谱特征及其碱解氮的预测研究

文章分析探索了应用可见近红外光谱技术快速、高效、便捷测定土壤营养参数的可能性。采集蓬莱镇组紫色土样本,比对分析了不同肥力水平、土壤厚度和土壤粒径条件下采集土壤光谱对可见近红外光谱特征的影响,筛选出不同厚度、粒径土壤条件下的碱解氮含量预测模型。研究结果表明,土壤样本厚度为30mm时具有最大的光谱反射率,建立的氮含量预测模型效果最佳,校正集和验证集的相关系数分别为0.84和0.83,均方根误差分别为1.79和1.87。土样粒径在0.25-0.85mm时氮含量的预测效果最佳,校正集和验证集的相关系数均超过0.8,且均方根误差较小;但当土样粒径<0.25mm时,氮含量预测模型效果明显下降。采用20目（<0.85mm）过筛、30mm厚度土壤样本采集可见近红外光谱和偏最小二乘法（PLS）模型预测,可以实现对蓬莱镇组紫色土碱解氮含量的较好光谱预测。     

Analysis of bovine udder, plate and viscera using near infrared reflectance spectroscopy

The objective of this study was to evaluate near infrared reflectance spectroscopy (NIRS) as an accurate and inexpensive alternative to conventional chemical analyses of nonconsumer bovine tissue. Udder, plate and visceral samples were collected from mature, Charolais-Angus and Hereford-Angus crossbred beef cows at slaughter, ground and analyzed for concentrations of lipid, protein and dry matter using standard AOAC chemical procedures. Samples were analyzed using NIRS. The collection of samples was randomly split into separate calibration and validation sets. Nine calibration equations representing each constituent and tissue combination were developed, using first- or second-order derivative mathematical transformations, and those calibration equations were validated. Correlation coefficients of calibration (R) and validation (r) ranged from .95 to .98 and from .87 to .97, respectively, for all analyses except plate dry matter (r = .77). Standard errors of calibration and prediction ranged from 1.89 to 5.81%. Results from this study are interpreted to indicate that bovine udder, plate and visceral tissue composition can be accurately, quickly and efficiently predicted using NIRS technology.     

Comparison of fecal crude protein and fecal near-infrared reflectance spectroscopy to predict digestibility of fresh grass consumed by sheep

Organic matter digestibility (OMD), an essential criterion for the evaluation of the nutrition of ruminants, cannot be measured easily at pasture. Therefore, the objective of this study was to test and compare 2 methods of OMD prediction based on the fecal CP content (CPf) or near infrared reflectance spectroscopy (NIRS) applied to feces. First, published equations derived from fecal N (Eq. 1(CP), n = 40) and from fecal NIRS (Eq. 1(NIRS), n = 84) were used to predict OMD of an independent validation data set from which in vivo OMD, ranging from 58 to 74%, was measured for 4 regrowth stages of Digitaria decumbens. Second, to establish equations usable in grazing situations and to improve the efficiency of the predictions, new equations were calculated from a large data set (n = 174) using CPf (Eq. 2(CP)) or fecal NIRS (Eq. 2(NIRS)). By applying the CPf method, Eq. 2(CPf) (OMD, % = 88.4 - 263.9/CPf, % of OM; residual SD = 2.92, r(2) = 0.63) showed similar statistical parameters (P < 0.01) when compared with Eq. 1(CP) (OMD, % = 86.6 - 266.2/CPf, % of OM; residual SD = 2.95, r(2) = 0.79). When using fecal NIRS, Eq. 2(NIRS) showed decreased SE of calibration (SEC = 1.48) and of cross-validation (SECV = 1.75) and greater coefficient of determination of cross-validation (R(2)(CV) = 0.85) than the previously published Eq. 1(NIRS) (SEC = 1.78, SECV = 2.02, R(2)(CV) = 0.77). The validation of the 4 equations on the validation data set was satisfactory overall with an average difference between the predicted and the observed OMD ranging from 0.98 to 2.79 percentage units. The Eq. 2(NIRS) was nevertheless the most precise with a decreased residual SD of 2.53 and also the most accurate, because the SD of the average difference between predicted and observed OMD was the lowest. Therefore, fecal NIRS provided the most reliable estimates of OMD and is thus a useful tool to predict OMD at pasture. However, an adequate number of reference data are required to establish good calibration. Indeed, better calibration statistics were obtained by increasing the data set from 84 (Eq. 1(NIRS)) to 174 (Eq. 2(NIRS)). In contrast, using fecal N on a set of 84 or 174 points did not improve the prediction. Both methods are useful for predicting OMD at pasture in certain circumstances, using fecal NIRS when a large data set (n = 84 and n = 174) is available and fecal CP with smaller data sets (n = 40).