Multifactorial models to assess responses to sorghum proportion, molasses and bacterial inoculant on in vitro quality of sorghum–soybean silages

Multifactorial models to assess responses to sorghum proportion, molasses and bacterial inoculant on in vitro quality of sorghum–soybean silages

Please download to get full document.

View again

of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.

Press Releases

Publish on:

Views: 0 | Pages: 13

Extension: PDF | Download: 0

  Animal Feed Science and Technology 164 (2011) 161–173 Contents lists available at ScienceDirect AnimalFeedScienceandTechnology  journal homepage: Multifactorial models to assess responses to sorghum proportion,molasses and bacterial inoculant on  in vitro  quality of sorghum–soybean silages R. Lima a , d , R.F. Díaz a , b , A. Castro b , S. Hoedtke c , V. Fievez d , ∗ a Central University of Las Villas, Department of Veterinary Medicine and Zootechny, Carretera a Camajuaní km 5  ½ , 54830 Santa Clara, Cuba b Central University of Las Villas, Agriculture Science Institute (CIAP), Carretera a Camajuaní km 5  ½ , 54830 Santa Clara, Cuba c University of Rostock, Institute of Farm Animal Sciences and Technology, Justus von Liebig Weg 8, 18059 Rostock, Germany d Ghent University, Department of Animal Production, LANUPRO, Proefhoevestraat 10, 9090 Melle, Belgium a r t i c l e i n f o  Article history: Received 5 August 2010Received in revised form 7 January 2011Accepted 16 January 2011 Keywords:In vitro  fermentationLaboratory silagesSilage qualitySorghum ( Sorghum bicolor  )Soybean ( Glycine max )Silage inoculantSilage additives a b s t r a c t Cuban climatological conditions allow culture of both soybean and sorghum. Combinedensiling of these crops might provide an excellent feed for ruminants during the dry sea-son of forage shortage. This study aimed to assess the optimal proportions of both crops toensure good quality silage using two approaches, being on  in vitro  fermentation test (Ros-tock fermentation test; RFT) and lab scale silages. Effects in the RFT of sorghum proportion(SGP), molasses and bacterial inoculant on silage quality characteristics were assessed bymultifactorialresponsemodels.Twosorghumvarieties( i.e. ,CIAP2E-95,SG 1  andCIAP49V-95, SG 2 ) and one soybean (SB) variety ( i.e. , INCASOY-35) were sown, harvested, choppedand ensiled. RFT was run with 72 treatments, including the pure forages (SB, SG 1  and SG 2 )oreitherSG 1  orSG 2  incombinationwithSBinthreeproportions(SGP;0.40,0.60and0.80),withorwithoutwatersolublecarbohydrates(WSC;10,20and30g/kgfreshmaterial)frommolasses (17, 35 and 52g/kg of fresh material) and with or without  Lactobacillus plan-tarum  as the bacterial inoculant. Quadratic response models were fitted and are presentedas contour plots. Responses of pH, lactic acid and NH 3 –N/N to addition of WSC and SGPwhen bacterial inoculant was added or not were quantified. Models allowed predictionof minimum SGP or WSC contents to reach target values of desired parameters. More-over, quantification of equivalent responses to WSC and SGP allowed assessment at theirexchange rates. RFT results ( i.e. , pH and lactic acid) were used to select acceptable qualitytreatments for laboratory silages. Additionally, pure sorghum or soybean lab scale silageswere prepared to enlarge the distribution of the silage quality characteristics. In both pro-cedures ( i.e. , RFT and lab scale silos), fermentation of good quality was produced with bothsorghumvarietiesalone,butcombinedsilageshadimprovedqualitycomparedtosoybeansilage (P<0.05). Addition of WSC and bacterial inoculant further improved silage quality(P<0.05). High correlations between RFT and laboratory silage parameters suggest RFT tobe an acceptable alternative to laboratory silage calculation before upscaling treatments. © 2011 Elsevier B.V. All rights reserved.  Abbreviations:  ADFom, acid detergent fiber; CIAP, agriculture science institute; CP, crude protein; DM, dry matter; FM, fresh material; Hcell, hemi-cellulose; INOC − , without addition of bacterial inoculant; INOC+, with addition of bacterial inoculant; L, lactic acid; NDFom, neutral detergent fiber; OM,organic matter; RFT, Rostock fermentation test; SB, soybean; SCFA, short chain fatty acid; SG 1 , sorghum variety CIAP 2E-95 containing more tannins ascompared to SG 2 ; SG 2 , sorghum variety CIAP 49V-95; SGP, sorghum proportion; TFa, total fermentation acids ratio; WSC, water soluble carbohydrates. ∗ Corresponding author. Tel.: +329 264 90 02; fax: +329 264 90 99. E-mail address: (V. Fievez).0377-8401/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.anifeedsci.2011.01.008  162  R. Lima et al. / Animal Feed Science and Technology 164 (2011) 161–173 1. Introduction Forage conservation is a key element in livestock production, especially for ruminants as it prevents nutrient deficitsduring periods of feed shortage, such as during the dry seasons. Conservation stabilizes production among environmentalconditions (Ojeda et al., 1991; Abdelhadi, 2007), typically by sun or artificial drying as hay, or addition of acids or natural fermentation to produce silage. In tropical regions, forage is of acceptable quality for conservation early in the wet season.However, the weather at that time is often unreliable for sun drying which makes hay making difficult, and artificial dryingisexpensiveandfacilitiesarenotwidelyavailable.Fermentationthroughensilingoffreshmaterialremainsafeasibleoption(Mannetje, 2000), and natural fermentation is preferred as addition of acids may be beyond the resources of smallholders, arecorrosive,canbehazardoustofarmerhealthandcancontributetoenvironmentalpollution(Mannetje,2000;Mühlbach,2000).However, tropical grasses and legumes are not an ideal material for ensiling, mainly because of their low level of watersoluble carbohydrates (WSC) which are essential for successful ensiling (Titterton and Bareeba, 2000). In addition, legumes have high buffering capacity, which increases pH and susceptibility of their proteins to proteolysis (McDonald et al., 1991).Several management practices can improve fermentable carbohydrate levels, reduce buffering capacity and limit prote-olysis, and can therefore contribute to production of good quality silage. These options include: mixing cereal crops withlegumes, wilting, using silage additives and conservation in small scale silos to achieve and maintain adequate anaerobicconditions (Titterton and Bareeba, 2000). Both soybean and sorghum can be grown under Cuban climatic conditions. Thus their combined ensiling is of interest, but their optimal proportion, as well as the potential to further improve silage qualitythrough additives such as molasses and/or lactobacilli should be assessed. Obviously, large scale silos are inappropriatefor this purpose and small, or laboratory-scale, silos in which conditions can be better controlled (Perkins and Pratt, 1951)should be used. Laboratory silos are considered a practical method to compare a number of treatments and are necessarywhen evaluating numerous experimental variables and their interactions (Cherney et al., 2006), as their entire contents can be weighed, processed and analyzed accurately. Obviously, such experiments are only of value under the assumptionthat the fermentation process is similar to that taking place in field-scale silos (Cherney et al., 2004). Still, laboratory silos might be too laborious when a large number of variables and their interactions are to be evaluated. For this reason, othermethods allowing evaluation of numerous treatments within a short time period prior to selection for further upscaling,are required. The Rostock fermentation test (RFT) is an example of such a method and has been used to rapidly assess theensiling potential of forages through pH kinetics during 46 (Pieper et al., 1989; Zierenberg, 2000) or 58h (Martens et al., 2008) of incubation. Generally the test is an incubation of forage in water with or without additives. During fermentation,the pH is measured and incubation characteristics, which typically indicate silage quality such as lactic, acetic and butyricacids, ammonia and alcohol, are analyzed in the extract (Pieper et al., 1989; Zierenberg, 2000; Martens et al., 2008).Hence, quality of soybean silages either or not in combination with sorghum and either or not supplemented withmolasses and/or lactobacilli was assessed, both by an  in vitro  fermentation test (RFT) as well as using laboratory silos.Further, comparison of silage characteristics as obtained using both techniques was used to assess whether parameters of the rapid RFT reliably predict the nutritional quality of soybean–sorghum silage mixtures. 2. Materials and methods  2.1. Plant material Two sorghum ( Sorghum bicolor   (L.) Moench) varieties (SG 1 =CIAP 2E-95 and SG 2 =CIAP 49V-95, with the former havinghighergrainandtotalDMyield,butalsoincreasedtannincontentscomparedtothelatter)andonesoybean( Glycinemax (L.)Merr.)variety(INCASOY-35)weresownataresearchfarm(22 ◦ 43  N,79 ◦ 90  W)fromtheagricultureresearchinstitute(CIAP)of the Central University of Las Villas, Santa Clara, in the centre of Cuba. Soybean and sorghum varieties were intercropped(two rows of soybean and one row of sorghum with 45cm row spacing) at a rate of 22.4plants/m 2 at 18/05/2007 and22/07/2007 for the soybean–SG 1  combination and soybean–SG 2  combination, respectively. Each intercropped pair wassown in separate fields. Average ± SD precipitation, temperature and humidity during the cropping period of soybean–SG 1 mixtureandsoybean–SG 2  mixturewere221 ± 44mmand218 ± 29mm,26 ± 0.6 ◦ Cand26 ± 0.4 ◦ Cand81 ± 2%and83 ± 1%,respectively. Crops were not fertilized nor irrigated. The pasty grain state of sorghum determined the harvesting date(23/08/2007 and 27/10/2007 for soybean–SG 1  combination and soybean–SG 2  combination, respectively). Bean formationwas completed at the time of harvest for the soybean crop. For experimental purposes, soybean and sorghum plants wereharvestedseparatelyaboutnoonduringtherainyseason.Sorghumwasharvestedbelowthelastleafwithyellowcolorationand soybean was harvested about 20cm above the soil. Whole sorghum and soybean plants were chopped separately to aparticle size of 2cm and frozen at − 20 ◦ C immediately, allowing use of material of the same srcin in both the RFT and labscale silage.  2.2. Rostock fermentation test (RFT) Materialusedwasstoredfrozenat − 20 ◦ Cfor3days.Afterthawing,plantmaterial(50goffreshmaterial(FM))and200mlofdistilledwaterwerepackedintosterileplastic400mlrecipientsintriplicate.Testsusedpuresorghumandsoybeanaswell  R. Lima et al. / Animal Feed Science and Technology 164 (2011) 161–173 163 as their combinations (sorghum proportion (SGP): 0.40, 0.60 and 0.80 on FM basis) with or without additives. The additivesused in this study were 10, 20 or 30g of WSC per kg of fresh material as molasses ( i.e. , 17, 35 and 52g of molasses/kgFM) and of a bacterial inoculant of   Lactobacillus plantarum  (DSMZ 8862 and DSMZ 8866, BIO-SIL  ® , Dr. Pieper Technologie-und Produktentwicklung GmbH, Wuthenow, Germany). In each of the inoculated treatments, 3 × 10 5 colony forming units(CFU)/gofFMwasappliedaccordingtoPieperetal.(1989)andZierenberg(2000).Bacterialinoculantwasacivatedinsaline physiological solution 3h before utilisation in order to obtain the refered dose within one ml of solution. Samples weremixed thoroughly prior to the start of the RFT, before each pH measurement and again at the end. Each plastic recipientwas covered and placed in an incubator at 31 ◦ C for 46h. pH was measured directly in the recipient at the beginning of theexperiment and after 14, 18, 22, 26, 38 and 46h of incubation. At the end of incubation, the contents were filtered throughcheese cloth and the extract used for chemical analysis.  2.3. Ensiling procedure for laboratory silages Selection of treatments for laboratory silages was based on RFT pH ( ≤ 4.5) and ratio of lactic acid to total fermentationacids( ≥ 0.7and<0.9).Additionally,combinationswithaCPcontentoflessthan100g/kgDMwerediscarded.Irrespectiveof the RFT pH and lactic acid proportion, treatments without molasses and inoculant were retained for comparative proposeswhen their supplemented counterparts were selected. Finally, silages were prepared of all pure forages ( i.e. , SB, SG 1 , SG 2 ).Mixtures meeting the formerly mentioned criteria consisted of a SGP of 0.40 and 0.60. Material used for laboratory silageswas stored frozen for 10 days at − 20 ◦ C. After thawing, plant material (520g of FM) was packed into laboratory silo flaskswith volume of 800ml in triplicate. The additive used in this study was of 10, 20 or 30g of WSC/kg FM as molasses and of abacterial inoculant of   L. plantarum . The treatment of SG 1 /SB (0.60/0.40) combined with 10g of WSC/kg FM was discarded aslactic acid (L) in total fermentation acids (TFa) did not meet the thresholds as outlined before. Indeed, fermentation acids of this treatment almost exclusively consisted of lactic acid (L/TFa=0.94) and some impairment of aerobic stability has beenassociatedwithsuchlowamountsofaceticacid.Asaneffectofthebacterialinoculantsoccurredforlacticacidconcentrationand L/TFa, particularly in combination with WSC, it was decided to only consider treatments with addition of lactobacilliin the combined silages with molasses. In each of the inoculated laboratory scale silages, 3 × 10 5 CFU/g of FM, was appliedaccordingtoPieperetal.(1989)andZierenberg(2000),andcompactiondensitywas650kgFM/m 3 .Sampleswerethoroughlymixed (by hand) before packing the laboratory silos and compacted by hand, with the aid of a rod, assuring removal of airand the desired compaction density. The laboratory silos were covered and placed in dark at room temperature (28 ± 3 ◦ C)until their opening 30 days later.  2.4. Sampling and chemical analysis of fresh and ensiled forage 2.4.1. Fresh forage After chopping and thorough mixing, 350g of fresh material was dried at 65 ◦ C for 72h in triplicate. Both pure plantmaterial as well as their combinations were dried separately, after which the material was ground to pass a 1mm screenand stored in glass bottles at room temperature (28 ± 3 ◦ C) for chemical analysis.  2.4.2. Ensiled forage After 30 days of ensiling, and immediately at the opening, a subsample was collected for extraction to determine pH andanalyze for ammonia N/total N ratio (NH 3 –N/N), lactic acid, short chain fatty acids (SCFA) and alcohol. Another subsamplewas dried (65 ◦ C during 72h) and stored for chemical analysis.The extract was prepared and stored as described by Lima et al. (2010). In all extracts from RFT and laboratory silage, pH wasmeasuredbeforeacidification(210pH/ion,Hanna,UK),lacticacidwasoxidizedtoacetaldehydeusingConwaymicrodif-fusionchambersandmeasuredspectrophotometrically(224nm)accordingtoConway(1957a).Ammoniaandalcoholwere determined before acidification according to Voigt and Steger (1976) and Conway (1957b), respectively. For SCFA analysis, acidified extracts were centrifuged (10min at 22,000 ×  g  , Beckman J2-HS, Palo Alto, CA, USA) before determination by gaschromatography according to Van Ranst et al. (2010).  2.4.3. Chemical analysis Samples were assayed in duplicate for the chemical fractions dry matter (DM; ID 930.15), N (ID 954.01) of  AOAC (1995)and organic matter (OM) content by EEC (1971). Neutral detergent fiber was analyzed with sodium sulfite and not with a heat stable amylase and expressed exclusive of residual ash (NDFom; Van Soest et al., 1991). Acid detergent fiber was determined by sequential analysis of the residual NDFom and expressed exclusive of residual ash (ADFom; Van Soest et al.,1991) and hemicellulose (Hcell) was calculated as the difference between NDFom and ADFom. Chemical analysis of freshforage and silages are in Table 1. Dry matter was corrected for silages considering the pH, SCFA, lactic acid, alcohol and ammonia losses according to Beyer et al. (2003). Alcohol and lactic acid losses are considered to be independent of medium pH, whereas losses of SCFA (g/kg DM) and ammonia (g/100g DM) depend of medium pH. Hence, different formulas wereused depending on the pH range being:  pH   ≤ 4 . 0 :  DMcorr   = DMa + 0 . 94 ∗ SCFA + 0 . 08 ∗ L +  A + 0 . 16NH 3 ;  164  R. Lima et al. / Animal Feed Science and Technology 164 (2011) 161–173  Table 1 Dry matter (DM) content (g/kg FM), organic matter (OM), crude protein (CP), neutral detergent fiber (NDFom), acid detergent fiber (ADFom) and hemicel-lulose (Hcell) content (g/kg DM) of pure sorghum and soybean both fresh as well as ensiled on a laboratory scale for 30 days ( n =3).Treatments a DM OM CP NDFom ADFom HcellFresh forageSB 286 926 154 420 339 81SG 1  391 952 66 460 250 211SG 2  286 946 83 541 417 124SEM 17.9 4.0 13.6 20.2 25.0 20.0Ensiled forageSB 299 913 148 492 392 101SG 1  392 949 61 468 225 243SG 2  345 943 80 539 361 178SEM 14.7 5.7 13.3 11.0 25.8 20.7 a SB: soybean; SG 1 : CIAP 2E-95; SG 2 : CIAP 49V-95. pH  >  4 . 0 ≤ 4 . 5 :  DMcorr   = DMa + 0 . 80 ∗ SCFA + 0 . 08 ∗ L +  A + 0 . 32NH 3 ;pH  >  4 . 5 ≤ 5 . 0 :  DMcorr   = DMa + 0 . 68 ∗ SCFA + 0 . 08 ∗ L +  A + 0 . 48NH 3 ;pH  >  5 . 0 ≤ 5 . 5 :  DMcorr   = DMa + 0 . 58 ∗ SCFA + 0 . 08 ∗ L +  A + 0 . 64NH 3 ;pH  >  5 . 5 :  DMcorr   = DMa + 0 . 50 ∗ SCFA + 0 . 08 ∗ L +  A + 0 . 80NH 3 ;with  DMcorr  , the corrected  DM  ;  DMa , the analyzed DM pre-drying at 65 ◦ C, final drying at 105 ◦ C;  L , the lactic acid content(g/kg DM) and  A , the alcohol content (g/kg DM). Buffering capacity was determined according to Weissbach (1967) in pure and mixed fresh forage samples of soybean, sorghum and their mixtures.  2.5. Statistical analysis A general linear means (GLM) model was applied using SPSS (SPSS, 2007) and correlations, regressions and figures were calculated and prepared with STATISTICA (StatSoft, 2008).The model (GLM, univariate) used to assess effects of addition of molasses and lactic acid bacteria on RFT quality ( i.e.  pH,ammonia N/total N ratio, lactic, acetic and butyric acids, alcohol, lactic acid/total fermentation acids ratio (L/TFa)) was: Y  ijkl  =   + SGP i = 1 – 5 + SV  j = 1 – 2 + WSC k = 1 – 4 + INOC l = 1 – 2 + SGP i ∗ SV  j + SGP i ∗ WSC k + SGP i ∗ INOC l + SGP i ∗ SV  j ∗ WSC k + SGP i ∗ SV  j ∗ INOC l + SGP i ∗ WSC k ∗ INOC l + SGP i ∗ SV  j ∗ WSC k ∗ INOC l + SV  j ∗ WSC k + SV  j ∗ INOC l + SV  j ∗ WSC k ∗ INOC l + WSC k ∗ INOC l + ε ijkl , with:SGP i =1–5 ,thesorghumproportion(1.0 versus 0.80 versus 0.60 versus 0.40 versus 0.0),SV  j =1–2 ,thesorghumvariety(SG 1 versus SG 2 ),WSC k =1–4 ,thedoseofwatersolublecarbohydratessupplemented(0.0 versus 10 versus 20 versus 30gofWSC/kgFM),INOC l =1–2  (with versus withoutbacterialinoculant);SGP i *SV  j ,SGP i *WSC k ,SGP i *INOC l ,SGP i *SV  j *WSC k ,SGP i *SV  j *INOC l ,SGP i *WSC k *INOC l , SGP i *SV  j *WSC k *INOC l , SV  j *WSC k , SV i *INOC l , SV  j *WSC k *INOC l , WSC k *INOC l , the interaction between theindicated factors. Interactions were discarded if P>0.15.A nonlinear regression was completed to quantify responses of pH, NH 3 –N/N, lactic acid, acetic acid, alcohol and L/TFato SGP and WSC for the two sorghum varieties with or without bacterial inoculant after 46h of RFT using the model:  z   = a ∗ SGP + b ∗ WSC + c  ∗ SGP 2 + d ∗ SGP ∗ WSC + e ∗ WSC 2 +  f, with:  a ,  b ,  c  ,  d ,  e  and  f  , parameter estimates and  z  , the response. These relations are presented as contour plots.Linear regression analysis was used to correlate buffering capacity and sorghum proportion, as well as pH, alcohol,ammonia, lactic, acetic and butyric acid from RFT and from laboratory silages.The model (GLM, univariate) used to assess the effect of SGP and variety as well as addition of molasses combined withlactic acid bacteria on quality of laboratory silages ( i.e. , DM, OM, CP, NDFom, ADFom, Hcell, pH, ammonia N/total N ratio,lactic, acetic and butyric acid, alcohol and lactic acid/total fermentation acids ratio) was: Y  ijkl  =  + SGP i = 1 – 2 + SV  j = 1 – 2 + EA k = 1 – 4 + SGP i ∗ EA k + SV  j ∗ EA k + ε ijk , with: SGP i =1–2 , the sorghum proportion (0.60  versus  0.40), SV  j =1–2 , the sorghum variety (SG 1  versus  SG 2 ), EA k =1–4 , theensilingadditive(silagewithoutadditives versus incubatedwith10gofWSC/kgFMandbacterialinoculant versus incubated20g of WSC/kg FM and bacterial inoculant  versus  incubated with 30g of WSC/kg FM and bacterial inoculant), SGP i *EA k  R. Lima et al. / Animal Feed Science and Technology 164 (2011) 161–173 165 Fig. 1.  Regression of buffering capacity of pure sorghum and soybean forage and their combination 95 days post sowing according to sorghum proportion(SGP) [SG 1 : CIAP 2E-95; SG 2 : CIAP 49V-95]. Fig. 2.  Contour plot of the interaction between sorghum proportion (SGP) and water soluble carbohydrates (WSC) from molasses for pH after 46h of RFTrun either with (black lines) or without (grey lines) bacterial inoculant ( n =3) [SG 1 : CIAP 2E-95; SG 2 : CIAP 49V-95]. and SV  j *EA k , the interaction between the indicated factors. In addition, a Dunnett’s test was completed to assess whetherchemical characteristics of mixed silages differed from those of the pure silages.In addition, results of chemical analysis of sorghum and soybean combinations, both fresh and ensiled, were comparedwith the composition calculated from pure fresh forage or silage values according their proportion. A Student’s  t  -test wasused for this comparison. 3. Results  3.1. Buffering capacity The buffering capacity of sorghum and soybean alone or in combination is in Fig. 1. Buffering capacity decreased linearly withincreasingsorghumproportionirrespectiveofvariety.Puresoybeanhadthehighestbufferingcapacity,andthebuffer-ing capacity of the combinations of soybean and sorghum could be calculated from values of the pure forages as calculatedvalues did not differ from measured ones (P>0.05).  3.2. RFT  Figs.2–4showcontourplotsoftheinteractiveeffectonpH,lacticacidandNH 3 –N/NofSGPandadditionofWSCthroughmolasses both when lactobacilli were added or not. Effects after 46h of RFT are for models with  R 2 >0.65 ( i.e. , pH, NH 3 –N/Nandlacticacidcontent).CoefficientsofmultivariatemodelsareinTable2.Accordingtothemodel,theresponseofpH(Fig.2)
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks