楊天輝,常生華,莫本田,侯扶江
(1.草地農(nóng)業(yè)生態(tài)系統(tǒng)國家重點實驗室 蘭州大學(xué)草地農(nóng)業(yè)科技學(xué)院,甘肅 蘭州730020;2.寧夏農(nóng)林科學(xué)院動物科學(xué)研究所,寧夏 銀川 750002; 3.貴州省草業(yè)研究所,貴州 貴陽 550006)
黃土高原13種栽培牧草營養(yǎng)成分NIRS模型分析
楊天輝1,2,常生華1,莫本田3,侯扶江1
(1.草地農(nóng)業(yè)生態(tài)系統(tǒng)國家重點實驗室 蘭州大學(xué)草地農(nóng)業(yè)科技學(xué)院,甘肅 蘭州730020;2.寧夏農(nóng)林科學(xué)院動物科學(xué)研究所,寧夏 銀川 750002; 3.貴州省草業(yè)研究所,貴州 貴陽 550006)
對2012-2013年黃土高原種植的13個牧草品種、780份干草樣品的營養(yǎng)成分建立了近紅外光譜(near infrared reflectance spectroscopy,NIRS)的檢測模型。豆科牧草的粗脂肪(EE)、酸性洗滌纖維(ADF)和粗灰分(Ash)建模結(jié)果最好,其定標決定系數(shù)(RSQ)>0.94,交叉驗證相關(guān)系數(shù)(1-VR)>0.7最高,定標標準分析誤差(SEC)在0.071~0.713,交叉校驗定標標準分析誤差(SECV)在0.160~2.751。禾本科牧草的EE和可溶性糖(WSC)建模結(jié)果最好,RSQ分別達0.916和0.859,1-VR分別為0.609和0.810,SEC和SECV分別是0.250、1.488和0.505、3.172。菊科和車前科牧草的模型,除ADF外,其它指標預(yù)測的穩(wěn)定性和準確性較為理想,RSQ在0.85以上,1-VR在0.70以上,SEC和SECV分別在0.361~3.557和0.495~4.602。NIRS對豆科粗蛋白(CP)和WSC的數(shù)值預(yù)測較差,RSQ僅>0.55,對禾本科CP、ADF、中性洗滌纖維(NDF)、Ash及菊科和車前科的ADF的預(yù)測稍差,RSQ>0.7。
粗蛋白;粗脂肪;酸性洗滌纖維;中性洗滌纖維;粗灰分;菊苣;車前
隨著計算機技術(shù)和化學(xué)計量學(xué)的進步,近紅外檢測技術(shù)(near infrared reflectance spectroscopy,NIRS)因?qū)悠窡o損傷、用量少、速度快、精度高而應(yīng)用范圍不斷擴大[1],目前已成功應(yīng)用于動物、植物、微生物、土壤等生物和非生物樣品的成分檢測[2]。近紅外光譜數(shù)據(jù)分析的主旨是定量光譜數(shù)據(jù)與樣品成分含量之間的關(guān)系,核心是模型的建立與評價[3]。最小二乘法(PLS)是NIRS數(shù)學(xué)建模中應(yīng)用最多的方法[4];模型建立后,其校正和評價的主要參數(shù)包括定標決定系數(shù)、交叉驗證決定系數(shù)、外部驗證決定系數(shù)、交叉驗證標準差等,而且決定系數(shù)越大,標準差越小,模型的定標及預(yù)測效果越好[3]。
NIRS 技術(shù)在牧草常規(guī)營養(yǎng)成分分析的應(yīng)用雖然起步較晚,但是已經(jīng)成為一個重要分支[5]。1976年,Norris等[6]利用NIRS技術(shù)測定了牧草的粗蛋白(CP)。此后,該技術(shù)相繼用于分析紫花苜蓿(Medicagosativa)[7]、紅三葉(Trifoliumpretense)[8]、多年生黑麥草(Loliumperenne)[9]和紫羊茅(Festucarubra)[10]等牧草的營養(yǎng)成分檢測上,NIRS技術(shù)分析樣品粗蛋白含量的相關(guān)系數(shù)均在0.90以上。而且,青貯飼料樣品連續(xù)4年營養(yǎng)成分建模的決定系數(shù)均大于0.90[11]。國外學(xué)者[3]對牧草、秸稈類粗飼料中CP、酸性洗滌纖維(ADF)、中性洗滌纖維(NDF)和其它成分均成功建立了NIRS 預(yù)測模型。20 世紀末,我國先后對紫花苜蓿[12]、燕麥(Avenasativa)[13]、黑麥草[14]、羊草(Leymuschinense)[15]等牧草的常規(guī)營養(yǎng)指標,建立了效果良好的NIRS定標模型。以往研究多對天然草原牧草混合樣品的營養(yǎng)成分進行實驗室檢測,結(jié)合近紅外光譜技術(shù)建立定標模型,或者對單一的栽培牧草品種進行測定和定標[2]。不同生態(tài)區(qū)域和生長季節(jié),各種牧草的養(yǎng)分含量可能大不相同,而在不同生育期,對多種利用方式的栽培牧草和混播草地的樣品建立近紅外定標檢查模型的研究鮮有報道。為此,本研究選取13種優(yōu)質(zhì)牧草,通過定期刈割模擬輪牧,采集牧草單一樣品和混合樣品參與定標,研究混合定標的可行性,以期為NIRS技術(shù)的改進、應(yīng)用范圍的擴展提供基礎(chǔ)數(shù)據(jù)。
1.1 栽培管理與取樣
2012年4月29日,12種多年生牧草品種和燕麥在蘭州大學(xué)榆中草地農(nóng)業(yè)實驗站播種(表1),完全隨機區(qū)組設(shè)計,每小區(qū)面積4 m×10 m,每個品種4次重復(fù)。每組小區(qū)間距1.5 m,組內(nèi)小區(qū)間隔0.5 m。播種前翻耕,耕深30~40 cm,施二胺150 kg·hm-2、尿素300 kg·hm-2作為底肥。每小區(qū)平均分為兩個裂區(qū),設(shè)置兩個處理,用定期刈割來模擬輪牧,生長季末收獲一次干草。2012年6月29日開始首次輪牧,牧草株高30 cm左右,此后每20 d刈割一次,留茬高度8 cm,末次齊地面刈割,全年共計刈割6次。每次模擬輪牧后,追肥7.5 g·m-2尿素,并灌水,灌溉量約為42.24 m3·hm-2。收獲干草的裂區(qū)刈割時間與模擬輪牧裂區(qū)末次刈割時間相同,齊地面刈割,施肥、灌溉的時間和量同模擬刈割。2013年,燕麥品種Makura復(fù)播,方法均與2012年相同;根據(jù)其它他牧草長勢,模擬輪牧的首次刈割時間為6月1日,末次為10月1日,共7次,其它管理均與2012年相同。所有草樣經(jīng)過60 ℃烘干,粉碎后過0.37 mm篩,用于測定營養(yǎng)成分和掃描近紅外光。
1.2 測定方法
1.2.1 化學(xué)成分測定方法 CP含量采用PROXIMA流動分析儀測定,粗脂肪(EE)采用ANKOMAXT15i型全自動脂肪分析儀濾袋法測定,NDF和ADF采用ANKOMA220型半自動纖維素分析儀進行測定,粗灰分(Ash)采用TM-O91OP型馬弗爐進行測定[16-18];可溶性糖(WSC)采用FOSS流動分析儀測定[19]。
表1 牧草品種試驗材料
1.2.2 樣品光譜測定 用Foss公司XDS-RCA可見/近紅外光譜儀掃描牧草樣品[2],光譜波長范圍400-2 498 nm,光譜分辨率2 nm,波長準確度<0.05 nm,數(shù)據(jù)采樣頻率2 f·s-1。每樣品重復(fù)裝樣掃描3次,計算其平均值,用軟件自動合并各處理重復(fù)和相似數(shù)值以增加精確度。雙檢測器系統(tǒng):硅檢測器(400-1 100 nm),硫化鉛檢測器(110-2 500 nm)。WinISⅢ光譜處理軟件建立牧草成分的標準曲線并驗證。
1.2.3 定標和校驗 把全部樣品分成兩份,其中585份作為定標集,195份作為預(yù)測集,在WinISⅢ定標軟件中對585份樣品先進行聚類分析,篩選出具有代表性的樣品514份,測定營養(yǎng)成分后輸入。采用軟件自動合并各處理重復(fù)值,將牧草干草分為豆科、禾本科和其它科,先用定標樣品集建立預(yù)測模型,再做交叉檢驗,根據(jù)交叉驗證相關(guān)系數(shù)(1-VR)、交叉校驗定標標準分析誤差(SECV)和定標標準分析誤差(SEC)等指標對模型進行優(yōu)化,并最終確定最佳定標方式和模型。然后,利用195份預(yù)測樣品集對模型進行外部驗證,檢測其預(yù)測效果,評價模型的外部驗證能力。
2.1 牧草營養(yǎng)成分分析結(jié)果
對定標樣品各營養(yǎng)成分NIRS模型進行隨機樣品驗證(表2),豆科牧草、禾本科和其它科牧草驗證集的預(yù)測值與化學(xué)測定值間的平均偏差(Bias)分別為-0.10~0.21、-0.16~0.24和-0.06~0.05,牧草定標模型預(yù)測效果較為理想。
2.2 可見/近紅外校正模型的建立與優(yōu)化
用WinISⅢ定標軟件對光譜數(shù)據(jù)進行預(yù)處理,選擇全譜范圍(400―2 500 nm),確定主因子數(shù)。采用NIRS處理軟件自帶的共12種回歸、光譜和數(shù)學(xué)處理方法組合對定標集樣品光譜進行定標[2](表3)。
2.3 牧草養(yǎng)分的最佳光譜處理和數(shù)學(xué)處理方法
豆科牧草定標集樣品進行NIRS掃描后,分別采用改進偏最小二乘法、最小二乘法建立模型,選擇最優(yōu)模型和光譜處理模式(表4);其中EE、ADF和Ash的定標決定系數(shù)(RSQ)均在0.9以上,交叉驗證相關(guān)系數(shù)(1-VR)在0.7以上,定標標準分析誤差(SEC)在0.071~0.713,交叉校驗定標標準分析誤差(SECV)在0.160~2.751,說明豆科牧草所建模型在EE、ADF和Ash數(shù)值的預(yù)測具有較高的穩(wěn)定性和準確性;所建定標模型在NDF數(shù)值預(yù)測方面稍差,RSQ達0.831,1-VR為0.631,SEC和SECV分別是3.331和4.624,需對定標集進一步擴充和完善以進一步提高預(yù)測準確度;模型對CP和WSC的預(yù)測能力最差,RSQ分別僅為0.557和0.559,1-VR分別為0.543和0.404,SEC和SECV分別為1.889、2.017和2.461、2.224。
表2 牧草定標和預(yù)測樣品集牧草營養(yǎng)成分分析結(jié)果
注:Lab Mean為實驗室分析結(jié)果平均值;NIR Mean為近紅外預(yù)測結(jié)果平均值; Bias為驗證集預(yù)測值與化學(xué)分析值的平均偏差。樣品為隨機選取,用“刈割年份+品種代號+茬次”表示,品種代號同表1所示。
Note: Lab Mean,mean value of calibration sample lab date; NIR Mean,mean value of forecast sample lab date; Bais, average deviation between calibration sample lab date and forecast sample lab date. Sample were randomly selected, and represent as sampling year+variety abbreviation+the cutting times, the varieties abbreviation are the same as Table 1.
表3 定標集樣品的回歸、光譜和數(shù)學(xué)處理方法
注:Modified PLS為改進最小二乘法;PLS為最小二乘法;SNV+Detrend為標準正常化+去散射處理;2,4,4,1為二階導(dǎo)數(shù)處理,數(shù)據(jù)間隔點為4,平滑處理間隔點為4,二次平滑處理間隔點為1。下表同。
Note: Modified PLS, Modified Partial Least Squares; PLS, Partial Least Squares; SNV + Detrend, standard normal variant+Detred; 2, 4, 4, 1, 2 derivative processing, 4 Gap, 4 Smooth, 1 Smooth 2. similarly for Table 4.
禾本科牧草定標集樣品處理參數(shù)及模型選擇方法與豆科相同(表4);其中EE和WSC的RSQ分別達到0.916和0.859,1-VR分別為0.609和0.810,SEC和SECV分別是0.250、1.488和0.505、3.172;說明所建模型在EE和WSC數(shù)值的預(yù)測較為穩(wěn)定和準確;定標模型在CP、ADF、NDF和Ash的數(shù)值預(yù)測方面稍差,RSQ在0.75~0.78,1-VR在0.62~0.65,SEC和SECV分別在0.744~3.472和0.909~5.378。
其它科牧草定標集樣品所建模型的處理參數(shù)和模型選擇方式與上述兩類相同(表4);除ADF數(shù)值預(yù)測方面稍差RSQ為0.708,1-VR為0.801,SEC和SECV分別是2.799和4.709;其它指標數(shù)值預(yù)測穩(wěn)定性和準確性較為理想,RSQ在0.85以上,1-VR在0.70以上,SEC和SECV分別在0.361~3.557和0.495~4.602。
表4 牧草各營養(yǎng)成分的最佳光譜處理和數(shù)學(xué)處理
注:RSQ,定標決定系數(shù);SEC,定標標準分析誤差;SECV,交叉校驗定標標準分析誤差;1-VR,交叉驗證相關(guān)系數(shù)。
Note: RSQ,determination coefficient of Standard curve; SEC,Standard Error of Calibration; SEVC,Standard Error of cross validation; 1-VR,1 minus the variance ratio.
牧草和飼料的常規(guī)養(yǎng)分分析是NIRS技術(shù)的傳統(tǒng)應(yīng)用領(lǐng)域[3],國際上建立了系統(tǒng)的NIRS模型庫[20]。NIRS技術(shù)測定反芻家畜粗飼料、多年生黑麥草和紫羊茅(F.rubra)的CP含量,模型的相關(guān)系數(shù)在0.97~0.98[8,21]。NIRS技術(shù)還可以測定多花黑麥草的CF、NDF、ADF含量[22],玉米秸稈的NDF和ADF含量[23],紫花苜蓿的Ash等[7,24]。但是,由于牧草種類構(gòu)成、營養(yǎng)成分含量以及環(huán)境等諸多因素的差異,目前還沒有一個較為完善的NIRS模型可以適應(yīng)于各種牧草的生產(chǎn)和多種草地類型[25]。本研究建立了2012-2013年模擬輪牧條件下,13種優(yōu)質(zhì)牧草品種干草營養(yǎng)成分的NIRS模型;各品種牧草營養(yǎng)成分的NIRS模型,除豆科CP和WSC外,定標決定系數(shù)均在0.7以上,交叉驗證相關(guān)系數(shù)均在0.63以上(表4),均達實用標準,這與Marten等的結(jié)果較為一致[10],他們雖然建立了4種豆科牧草常規(guī)營養(yǎng)成分的NIRS 模型,并證實NIRS 可分析混合牧草營養(yǎng)成分,但未涉及不同科牧草NIRS模型的比較。本研究發(fā)現(xiàn),將NIRS模型運用于不同科牧草品種營養(yǎng)成分的預(yù)測效果較為理想,也表明定標模型對樣品集的變異描述能力較強,定標模型建立較為成功。
然而,單一牧草樣品建模較混合樣品更準確[26]。本研究中,豆科牧草NIRS模型,CP和WSC的決定系數(shù)較其它模型小,RSQ僅>0.55,禾本科CP、ADF、NDF和Ash及其它牧草的ADF的預(yù)測稍差,RSQ>0.7(表4),既可能與混合樣品有關(guān),也可能因為實驗室中樣品測定值存在誤差,需要擴充樣本量、降低實驗室化學(xué)分析的誤差等,以校正定標方程。
References:
[1] Givens D I,de Boever J L,Deaville E R.The principles,practices and some future applications of near infrared spectroscopy for predicting the nutritive value of foods for animals and humans.Nutrition Research Reviews,1997,10(1):83-114.
[2] 李潔,王建福,吳建平,張昌吉,王欣榮.可見/近紅外漫反射光譜分析法測定牧草營養(yǎng)成分研究.家畜生態(tài)學(xué)報,2014,35(4):54-58. LI J,Wang J F,Wu J P,Zhang C J,Wang X R.Study on the determination of nutritional ingredient in native grass with visible/near infrared reflectance spectroscopy.Journal of Domestic Animal Ecology,2014,35(4):54-58.(in Chinese)
[3] 杜雪燕,王迅,柴沙駝,劉書杰.近紅外光譜技術(shù)在天然草地牧草營養(yǎng)價值評價中的應(yīng)用及展望.中國農(nóng)學(xué)通報,2014,30(35):1-6. Du X Y,Wang X,Chai S T,Liu S J.Application of near infrared reflectance spectroscopy (NIRS) in assessing nutritional value of natural grassland pasture and its prospect.Chinese Agricultural Science Bulletin,2014,30(35):1-6.(in Chinese)
[4] 張琳,張黎明,李燕,劉丙萍,王曉斐,王俊德.偏最小二乘法在傅里葉變換紅外光譜中的應(yīng)用及進展.光譜學(xué)與光譜分析,2005,25(10):1610-1613. Zhang L,Zhang L M,Li Y,Liu B P,Wang X F,Wang J D.Application and improvement of partial-least-squares in fourier transform infrared spectroscopy.Spectroscopy and Spectral Analysis,2005,25(10):1610-1613.(in Chinese)
[5] 胡超,白史且,張玉,鄢家俊,游明鴻,李達旭,白玲,張勁.近紅外光譜技術(shù)(NIRS)及其在牧草品質(zhì)檢測中的應(yīng)用.草業(yè)與畜牧,2013(4):40-46. Hu C,Bai S Q,Zhang Y,Yan J J,You M H,Li D X,Bai L,Zhang J.Near-infrared reflectance spectroscopy analytic model established for the IVDMD ofCichoriumintybusL.Prataculture & Animal Husbandry,2013(4):40-46.(in Chinese)
[6] Norris K H,Barnes R F,Moore J E,Shenk J S.Predicting forage quality by infrared replectance spectroscopy.Journal of Animal Science,1976,43(4):889-897.
[7] Alexandrov A N,Atanassova S L,Peltekova V D,Broderick G A,Todorov N A.The use of near infrared spectroscopy for estimation of microbial contamination of in sacco forage residues and protein and dry matter degradability in the rumen.Publication-European Association for Animal Production,1995,81:133-140.
[8] Gislum R,Micklander E,Nielsen J P.Quantification of nitrogen concentration in perennial ryegrass and red fescue using near-infrared reflectance spectroscopy (NIRS) and chemometrics.Field Crops Research,2004,88(2):269-277.
[9] Hansen J L,Viands D R,Steffens J C,Sniffen C J.Heritability and improvement of protein and nitrogen concentrations in wilted alfalfa forage.Crop Science,1992,32(4):879-883.
[10] Marten G C,Brink G E,Buxton D R,Halgerson J L,Hornstein J S.Near infrared reflectance spectroscopy(NIRS):Analysis of forage quality.Crop Science,1984,24(6):1179-1182.
[11] Villamarín B,Fernández E,Mendéz J.Analysis of grass silage from northwestern spain by near-infrared reflectance spectroscopy.Journal of AOAC International,2002,85(3):541-545.
[12] 趙環(huán)環(huán),胡躍高,趙其波,張鵬.近紅外光譜分析技術(shù)在黑麥草粉粗蛋白測定中的應(yīng)用.動物營養(yǎng)學(xué)報,2001,13(4):40-43. Zhao H H,Hu Y G,Zhao Q B,Zhang P.Application of near infrared reflectance spectroscopy (NIRS) to analysis crude protein in ground rye hay.Chinese Journal of Animal Nutrition,2001,13(4):40-43.(in Chinese)
[13] 趙秀芳,李衛(wèi)建,黃偉,曹喆,戎郁萍.燕麥干草品質(zhì)的近紅外光譜定量分析.光譜學(xué)與光譜分析,2008,28(9): 2094-2097. Zhao X F,Li W J,Huang W,Cao Z,Rong Y P.Near infrared determination of dry hay quality in oats.Spectroscopy and Spectral Analysis,2008,28(9):2094-2097.(in Chinese)
[14] 尚晨,張月學(xué),李炎,蒿若超,唐鳳蘭,韓微波,劉杰淋.紫花苜蓿粗蛋白和粗纖維近紅外分析模型的建立.光譜學(xué)與光譜分析,2009,12(29):3250-3253. Shang C,Zhang Y X,Li Y,Hao R C,Tang F L,Han W B,Liu J L.Near-infrared reflectance spectroscopy analytic models established for the crude protein and crude fiber of alfalfa property.Spectroscopy and Spectral Analysis,2009,12(29):3250-3253.(in Chinese)
[15] 石丹,張英俊.近紅外光譜法快速測定羊草干草品質(zhì)的研究.光譜學(xué)與光譜分析,2011,31(10):2730-2733. Shi D,Zhang Y J.Determination ofLeymuschinensisquality by near-infrared reflectance spectroscopy.Spectroscopy and Spectral Analysis,2011,31(10):2730-2733.(in Chinese)
[16] 段倩雯,成慧,侯扶江.景泰綠洲3種小谷物的產(chǎn)草量與牧草營養(yǎng)成分預(yù)測.草業(yè)科學(xué),2016,33(10):2041-2053. Duan Q W,Cheng H,Hou F J.Prediction off orage yield and nutrition of three small grains in Jingtai oasis.Pratacultural Science,2016,33(10):2041-2053.(in Chinese)
[17] 成慧,侯扶江,常生華,陳先江.黃土高原秋播時間對3種小谷物牧草生產(chǎn)性能的影響.草地學(xué)報,2013,21(6):1162-1168. Cheng H,Hou F J,Chang S H,Chen X J.Effects of fall sowing time on the forage productivity of small grain planted in the Loess Plateau.Acta Agrestia Sinica,2013,21(6):1162-1168.(in Chinese)
[18] 楊天輝,成慧,Eun J K,常生華,張瑜,莫本田,侯扶江.黃土高原模擬輪牧對高糖黑麥草生產(chǎn)性能的影響.草業(yè)科學(xué),2015,32(9):1473-1481. Yang T H,Cheng H,Eun J K,Chang S H,Zhang Y,Mo B T,Hou F J.Responses of stimulated grazing to high-sugar ryegrass productive qualities on the Loess Plateau.Pratacultural Science,2015,32(9):1473-1481.(in Chinese)
[19] 張麗英.飼料分析及飼料質(zhì)量檢測技術(shù).北京:中國農(nóng)業(yè)大學(xué)出版社,1993:46-74.
[20] Park R S,Agnew R E,Gordon F J,Steen R W J.The use of near infrared reflectance spectroscopy (NIRS) on undried samples of grass silage to predict chemical composition and digestibility parameters.Animal Feed Science and Technology,1998,72(1):155-167.
[21] Barton F E,Windham W R.Determination of acid-detergent fiber and crude protein in forages by near-infrared reflectance spectroscopy:Collaborative study.Association of Official Analytical Chemists,1988,71(6):1162-1169.
[22] Bruno-Soares A M,Murray I,Paterson R M,Abreu J M F.Use of near infrared reflectance spectroscopy (NIRS) for the prediction of the chemical composition and nutritional attributes of green crop cereals.Animal Feed Science and Technology,1998,75(1):15-25.
[23] Cozzolino D,Fassio A,Gimenez A.The use of near-infrared reflectance spectroscopy (NIRS) to predict the composition of whole maize plants.Science of Food and Agriculture,2001,81(1):142-146.
[24] de Boever J L,Cottyn B G,de Brabander D L,Boucqué C V.Prediction of the feeding value of maize silages by chemical parameters,in vitro digestibility and NIRS.Animal Feed Science and Technology,1997,66(1-4):211-222.
[25] 聶志東,韓建國,張錄達,李軍會.近紅外光譜技術(shù)(NIRS)在草地生態(tài)學(xué)研究中的應(yīng)用.光譜學(xué)與光譜分析,2007,27(4):691-696. Nie Z D,Han J G,Zhang L D,Li J H.Application of near infrared reflectance spectroscopy technique(NIRS) to grassland ecology research.Spectroscopy and Spectral Analysis,2007,27(4):691-696.(in Chinese)
[26] Berardo N.Prediction of the chemical composition of white clover by near-infrared reflectance spectroscopy.Grass and Forage Science,1997,52(1):27-32.
(責(zé)任編輯 王芳)
Analysis of nutritional content in 13 forage crop varieties in the Loess Plateau based on visible/near infrared reflectance spectroscopy
Yang Tian-hui1,2, Chang Sheng-hua1, Mo Ben-tian3, Hou Fu-jiang1
(1.State Key Laboratory of Grassland Agro-ecosystems, College of Pastoral Agriculture Science and Technology, Lanzhou University, Lanzhou 730020, China;2.Institute of Animal Science, Ningxia Academy of Agricultural and Forestry Sciences, Yinchuan 750002, China;3.Guizhou Institute of Prataculture, Guiyang 550006, China)
A visible/near-infrared reflectance spectroscopy (visible/NIRS) model was developed to determine the forage cultivar nutritional composition of 780 hay samples under simulated rotational grazing. Hay samples (n=780) from 13 forage crop varieties under simulated grazing in Loess Plateau during the 2012 to 2013 growing season were evaluated using calibration methods for prediction of nutrient contents using NIRS. The following results were obtained. The optimal calibrations in Leguminosae were ether extract (EE), acid detergent fiber (ADF), and crude ash(Ash). The multiple correlation coefficients(RSQ) and 1-variance ratio (1-VR) were > 0.94 and > 0.7, and standard error of calibration (SEC) and standard error of cross validation (SECV) were 0.071~0.713 and 0.160~2.751, respectively. Optimal calibrations in Gramineae were EE and water-soluble carbohydrate content(WSC). RSQ were 0.916 and 0.859 and 1-VR were 0.609 and 0.810 for EE and WSC, respectively, and SEC was 0.250 and 1.488 and SECV was 0.505 and 3.172, respectively. For the other species, the results for nutrient predication were reasonably good, with the exception of ADF. RSQ and 1-VR were >0.85 and >0.70, and SEC and SECV were 0.361~3.557 and 0.495~4.602, respectively. These results indicate that the accuracy of prediction using NIRS was acceptable for 13 forage crop nutrients, although the crude protein(CP), ADF, neutral detergent fibre(NDF), Ash of Gramineae and the ADF of others species (RSQ > 0.7) may require further calibration. The accuracies of the predictions for CP and WSC in Leguminosae (RSQ > 0.55) were not acceptable and thus more samples and greater precision during measurement will be required in further investigations.
crude protein; ether extract; ADF; NDF; Ash;Cichoriumintybus;Plantagolanceolata
Hou Fu-jiang E-mail:cyhoufj@lzu.edu.cn
10.11829/j.issn.1001-0629.2016-0137
2016-03-18 接受日期:2016-05-18
長江學(xué)者和創(chuàng)新團隊發(fā)展計劃(IRT13019);甘肅省2016年草牧業(yè)試驗試點和草業(yè)技術(shù)創(chuàng)新聯(lián)盟科技支撐(GCLM2016001)
楊天輝(1989-),男,寧夏同心人,助理研究員,碩士,研究方向為草地栽培。E-mali:1178981668@qq.com
侯扶江(1971-),男,河南扶溝人,教授,博導(dǎo),博士,研究方向為草地資源利用與管理。E-mail:cyhoufj@lzu.edu.cn
S816.11
A
1001-0629(2017)3-0575-07*
楊天輝,常生華,莫本田,侯扶江.黃土高原13種栽培牧草營養(yǎng)成分NIRS模型分析.草業(yè)科學(xué),2017,34(3):575-581.
Yang T H,Chang S H,Mo B T,Hou F J.Analysis of nutritional content in 13 forage crop varieties in the Loess Plateau based on visible/near infrared reflectance spectroscopy.Pratacultural Science,2017,34(3):575-581.