1、Original articleEffect of maturity stage and chopping lengthof maize silage on particle size reduction in dairy cowsIsabelle F ERNANDEZ , Brigitte M ICHALET -D OREAU *Unité de Recherche sur les Herbivores, INRA-Theix, 63122 St-Genès-Champanelle, France(Received 25 February 2002; accepted 2

2、7 November 2002Abstract The effects of maturity stage and chopping length of maize silage on chewing behaviour,particle size reduction and rumen retention time were studied in four lactating cows fitted with a ruminal cannula,according to a 4×4Latin square design.Maize silage made up 80%of the

3、diet and the cows were fed at 90%of the ad libitum intake level.Chopping length had only a significant effect on the particle size of the maize harvested at the late maturity stage.For the latter,the proportion of particles larger than 2 mm (LP2from the feed to the rumen content decreased from 73to

4、38%for the coarse maize and from 58to 37%for the fine maize,mainly by ingestive mastication.Eating time expressed per unit of feed intake did not vary significantly between these two silages (21.3vs.19.8min·kg 1DMI.Ingestive mastication in dairy cows thus reduces long particles more efficiently

5、 than short particles.particle size / chewing behaviour / maize silage / dairy cowsRésuméEffet du stade de maturitéet de la finesse de hachage de lensilage de maïs sur le comportement alimentaire et la réduction de la taille des particules chez les vaches laitières.Quat

6、re vaches laitières canulées du rumen ont reçu en quantités limitées et dans un dispositif en carrélatin 4×4,quatre ensilages de maïs qui différaient par leur finesse de hachage et leur stade de maturi-té.La ration était composée de 80 %den

7、silage et 20 %de concentré.Les différences de taille de par-ticules des ensilages ont étéplus importantes au stade tardif que précoce.La mastication ingestive a réduit fortement la proportion de grosses particules (>2 mm(LP2par rapport àla comminution to-tale.Ce

8、tte diminution a étéplus importante avec lensilage grossièrement hachéquavec lensilage fi-nement haché,respectivement de 73à38 %vs.de 58à37 %.Parallèlement,la durée unitaire dingestion par gramme de fourrage ingéréna pas variésignificativem

9、ent avec les deux ensilages (21.3vs.19.8min·kg 1MSI.La mastication était donc plus efficace pour lensilage hachégrossière-ment.taille des particules / comportement alimentaire / ensilage de maïs / vaches laitières 445* Correspondence and reprintsTel.: 33 (04 62 40 93; f

10、ax: 33 (04 73 62 42 73; e-mail: bdoreausancy.clermont.inra.frAnim. Res. 51 (2002 445454 INRA, EDP Sciences, 2002DOI: 10.1051/animres:20020401. INTRODUCTIONParticle size plays a key role in the diges-tion and passage of feed through the gastro-intestinal tract of ruminants and therefore in feed intak

11、e.When studying the passage of food from the rumen,digesta particles are usually divided into two pools:a small particle pool(<2 mmthat leaves the ru-men,and a large particle pool(>2 mmthat has difficulty escaping23.The mean re-tention time of digesta depends on the sizes of the two particle p

12、ools and their outflow from the rumen.Most feed particle comminution to sizes small enough to leave the rumen is accomplished by chewing,and interrelationships exist between chewing activity,particle comminution,ruminal re-tention time and voluntary intake9.Only a few authors have studied particle c

13、omminution mechanisms with cattle in re-lation to control of feed intake and most re-sults have been obtained with hay or haylage 4, 20, 26.In this experiment,particle size reduc-tion in maize silages differing in maturity stage and chopping length was studied in cows fed restricted amounts.2. MATER

14、IALS AND METHODS2.1. Animals, experimental designand dietsMaize(Safrane variety,Limagrain Ge-neticswas grown in April1999in Limagne(Franceat a density of95 000 seeds per ha.Whole plants were harvested with a precision chop harvester(Claas SF80model,no roller millat two theoreti-cal chop lengths,fine

15、(4.2 mmand coarse (12.0 mm,and at two maturity stages, early stage(24%DMand late stage(31% DM.The different types of maize were ensiled in small experimental silos with a capacity of 4 m3.Four multiparous Holstein cows(aver-age initial body weight602±72 kg,average days in milk92±57dwere fi

16、tted with a permanent ruminal cannula made of polyamide and polyvinyl chloride (Synthesia,Nogent-sur-Marne,France. Surgery was performed more than12months before the initiation of the experiment,under general anaesthesia(Isoflurane,ICIU Pharma-vétérinaire,Paris,France.The cows were housed

17、in individual tie stalls, and were randomly assigned to the4exper-imental diets in a4×4Latin square design. The dietary treatments were based on the experimental maize silages(79.3%sup-plemented with19.4%concentrate and 1.3%minerals.The concentrate was com-posed of53.3%soybean meal,10.1% wheat,

18、7.5%rapeseed meal,10.1%barley, 15.1%beet pulp,2%beet molasses,1% calcium carbonate,0.5%dicalcium phos-phate,0.3%magnesium oxide and0.3%so-dium chloride and the mineral composition was6%P,24%Ca and5%Mg,respec-tively.The diets were formulated to contain 12.3%crude protein(on a DM basisand to meet the

19、cowsmineral requirements16. They were given twice daily in equal pro-portions at09.00and16.00 h,and in re-stricted amounts(at90%of the individual ad libitum DM intake.Each period of the Latin square lasted29days,a14-day adap-tation period followed by a15-day experi-mental period.2.2. MeasurementsFee

20、d intake was recorded daily.Feed samples were collected daily from d 15to d 29.Silage samples were analysed daily for DM,48h in an80 °C forced-air oven. Samples of the other ingredients were col-lected once weekly and analysed for DM. Dried composite samples of feeds were pooled within each per

21、iod and ground to pass a1-mm hammer mill screen before chemical analysis.A representative sample of maize silage was collected weekly446I. Fernandez, B. Michalet-Doreauduring each period and frozen at20 °C and pooled within each silage at the end of the trial, for sieve and silage analysis.Chew

22、ing activity was recorded for three consecutive days(from d 24to d 26,using a small balloon filled with a foam rubber placed in the sub-mandibular space and connected to a pressure gauge transducer. Signals from the transducers were recorded to determine ingestion and ruminating times.Feed intake du

23、ring the first morning meal was measured by weighing refusals 2 h after feeding.Ruminal mean retention time was deter-mined from a single dose of Europium (20 g of Eu oxide·kg1DM feedlabelled on maize stover11given just before the morning meal on d22.After washing(1h, 40 °Cwith a commercia

24、l detergent with-out EDTA,followed by four rinses with wa-ter,maize stover was macerated in a Eu acetate solution(15mL·g1stover DM;0.02Mfor24h.Labelled particles of stover were rinsed under tap water for2min and immersed for1.5h in a water bath (15mL·g1stover DMto remove loosely or unbound

25、 rare earth.Labelled particles were dried at80 °C for48 h.Faecal sam-ples were collected from the rectum of each cow from d21to d26every3 h for12 h, then every4 h for36 h,then every6 h for 36 h,and finally every24 h until168 h after marker distribution.Faecal samples were dried(80 °C for48

26、 hand ground and passed through a 1-mm screen.At the end of each period(on d29,the rumen was manually emptied5 h after the morning feeding and the contents were weighed.After homogenisation,dry matter content was determined and a representa-tive sample(5 kgwas frozen at20 °C for sieve analysis.

27、Another sample was strained through a250-µm nylon filter to separate the liquid and solid phases and weighed.When the rumen was empty,the cows were fed maize silage alone.Boluses of maize silage,chewed,and delivered through the cardia,were collected by hand through the rumen cannula.Eight bolus

28、es were pooled for each cow per period and frozen at 20 °C for sieve analysis.2.3. Chemical and physical analysesFeed samples were analysed for DM, ash,crude protein2,starch12and NDF and ADF using-amylase 29.Silage fermentation characteristics were measured on maize silage juice obtained with a

29、 grape press.pH was immediately de-termined with a digital pH-meter(CG840, Ag/AgCl electrode,Schott Gerate,Hofheim, Germany.Acetic acid and ethanol con-tents were determined by gas liquid chro-matography17,lactic acid content by Nolls method22and NH3content by Conways method7.The DM content of maize

30、 silage was corrected for losses of vol-atile components during oven-drying10.Particle size distribution in maize silage, boluses and ruminal content were deter-mined by wet sieving13.Samples(50 g of fresh matterwere sieved with a system (Retsch AS200,Germanythat provided water spray and vibrations,

31、equipped with six sieves of8.00,4.00,2.00,0.40,0.10and 0.05 mm aperture size,and a bottom pan. Three5-min sievings were carried out us-ing on average12L of water per sample. The material retained on each sieve was dried for24 h in a forced-air oven at60 °C, then for48 h at80 °C and weighed

32、.Total filtration water,containing the finest parti-cles(<0.05 mm,was weighed,and five fractions(100mLwere weighed and dried at80 °C.Two replicates were completed for each sample.The particle size distribution was ex-pressed as DM percentage of particles col-lected on the different screens.M

33、ean particle size was calculated by plotting the cumulative percentages of retained parti-cles against the logarithm of sieve size30. The d75-d25inter-fractile gap,which reflects heterogeneity of particle size distributionParticle size reduction in dairy cows44721,was calculated.The proportion of la

34、rge particles larger than2 mm(LP2was determined by the sum of the particle pro-portions retained on8,4,and2 mm aper-ture sieves,this being considered as the critical size for leaving the reticulo-rumen in cattle23,28.As performed by Bailey et al.4,the differences in particle size from feed to bolus

35、and from bolus to rumen were expressed using differences in mean particle size and LP2 proportions.Europium content in faecal content was determined by atomic absorption spectro-photometry(Model2380Spectrophotometer, Perkin-Elmer,Bois-dArcy,Franceafter extraction of the marker from the dried samples

36、 by the method of Hart and Polan 15.The ruminal mean retention time was calculated by plotting the natural logarithm of marker concentration against time,with the linear descending part of the curve cor-responding to the ruminal particulate out-flow rate 27.2.4. Statistical analysisData were analyse

37、d as a4×4Latin square design using the GLM procedure of SAS24with maize silage,period,and cow as the sources of variation.The means were considered significant at P<0.05.The model was:Y ijkl=µ+ S i+ P j+ C k+ E ijkl where Y ijkl=observation,µ=overall mean, S i=effect of silage trea

38、tment(i=1to4, P j=effect of period(j=1to4,C k=effect of cow(k=1to4and E ijkl=residual error.3. RESULTS AND DISCUSSION3.1. Maize silage characteristicsThe chemical composition and fermen-tation characteristics of the four maize448I. Fernandez, B. Michalet-DoreauTable I.Chemical composition and fermen

39、tation characteristics of the maize silages.Maturity stage Early LateChopping length Fine Coarse Fine Coarse Chemical compositionDM (g·kg1244240308328 OM (g·kg1DM952949955956 Crude protein (g·kg1DM75817574 Starch (g·kg1DM255246326304 NDF (g·kg1DM471477418432 ADF (g·kg1D

40、M272283238230 Fermentation characteristicspH37.137.838.538.1 Lactic acid (g·kg1DM47.046.337.152.5 Acetic acid (g·kg1DM11.27.77.210.2 Ethanol (g·kg1DM7.68.3 1.5 4.7 N-NH3(g·kg1DM0.420.440.320.53 DM:dry matter,OM:organic matter,NDF:neutral detergent fibre,ADF:acid detergent fibre,N

41、-NH3:ammonia nitrogen.silages are presented in Table I.The chemi-cal composition did not vary with the chop-ping length of the maize silage.The increase of DM content from24to32%was accompanied by an increase of starch con-tent with6%units and a decrease of NDF and ADF content with respectively5and

42、4%units.These variations were due to a relative increase in grain content with an advancing maturity stage.The fermentation characteristics of the four maize silages indicated a normal pres-ervation.At the early maturity stage,chop-ping length had no effect on fermentative parameters.At the late mat

43、urity stage, coarsely chopped silage was more fer-mented than the fine silage,resulting in higher lactic acid and ethanol concentra-tions,despite the slightly higher DM con-tent.At the early maturity stage,the coarse si-lage contained more particles larger than 8 mm compared with the fine silage(22v

44、s. 16%,but less particles between2and 4 mm in size(from11to19%(Tab. II.Consequently,the LP2proportion was not affected(61%on a DM basisby chopping length.At the late maturity stage,coarse chopping induced a marked increase in the proportion of particles larger than8 mm (from9to45%that was only partl

45、y offset by a lower proportion of particles between 2and8 mm in size(from51to30%. Coarse chopping thus resulted in an in-crease in the LP2proportion(from58to 73%and in the extent of dispersion,which rose from3to14 mm,showing the high heterogeneity of the silage.3.2. Maturity stage and particle sizer

46、eductionBecause the cows were limited-fed,DM intake per day or per meal did not vary with the maturity stage.The intake during the morning meal accounted for nearly half of the daily intake,on average7 kg DM (Tab. III.Maturity stage did not affect the time spent eating per day or per unit of feed. T

47、he time spent chewing was higher for theParticle size reduction in dairy cows449 Table II.Particle size distribution of the maize silages.Maturity stage Early LateChopping length Fine Coarse Fine Coarse (% DM retained on sieve8.0 mm15.521.99.345.44.0 mm31.130.531.122.22.0 mm18.710.919.58.00.4 mm7.7

48、6.09.2 3.20.1 mm 1.3 1.7 2.20.90.05 mm0.40.50.80.3< 0.05 mm25.428.528.119.9 Mean particle size (mm0.800.760.55 2.64d75-d25 (mm* 4.07 5.20 3.0813.76LP2 (% on DM basis61.061.058.073.0*Extent of dispersion.DM: dry matter; LP2 = sum of the particle proportions retained on 8, 4, and 2 mm sieve apertur

49、e.early than for the late maturity stage,836 vs.751min.This difference persisted when the results were expressed per unit of feed intake,but was no longer significant.In lac-tating cows,De Boever et al.8reported a higher chewing time per unit of feed intake for the less mature maize,which they attri

50、b-uted to the lower grain content of this silage.The LP2proportions,being on average comparable for the early and late stage(re-spectively61and65%on DM basis (Tab. IV,were similar in the correspond-ing boluses,respectively44.2and46.1%on a DM basis.By mastication overall,mean particle size and LP2pro

51、portion for the4si-lages decreased from1.19to0.25 mm and from63to45.1%on a DM basis.Ingestive mastication strongly reduced mean particle size by79%.With hay,mastication during consumption reduced the large particle pro-portion on average by44%in sheep6and by51%in cattle19.Ingestive mastication incre

52、ased the proportion of small particles, particularly those smaller than0.05 mm (on average with 11% units.Rumen fill expressed as fresh matter was significantly higher with the early than with the late maturity stage,and the rumen fill expressed in dry matter followed the same trend(Tab. V.Neither t

53、he solid phase pro-portion,nor DM content,nor DM particle size distribution varied with the maturity stage.The higher rumen fill with the early silage was not accompanied by any modifi-cation to the structure of the rumen con-tents.A higher NDF content as in the early stage silage,may induce a highe

54、r rumen fill 1.This did not depress DMI,because the cows were limited-fed.Ruminal mean re-tention was not affected by the maturity stage.3.3. Chopping length and particle sizereductionAt the early maturity stage there was lit-tle difference in particle size characteristics between fine and coarse ch

55、opped silages, so we focused on late maize silages.Table III.Influence of maturity stage and chopping length of maize silage on DM intake and chew-ing activity of dairy cows.Maturity stage Early Late SE Chopping length Fine Coarse Fine CoarseDry matter intake (DMITotal DMI (kg·d114.57a14.89a14.

56、56a14.45a0.99a.m. meal DMI (kg 6.86a 6.98a 6.97a 6.81a0.46 Time spent eatingmin·d1 min·kg1DMI297a21.70a273a19.51a261a19.84a304a21.28a152.19a.m. meal (mina.m. meal (min·kg1DMI124a19.28a114ab17.08a103b16.74a121a18.30a41.87Time spent ruminatingmin·d1 min·kg1DMI538a37.50a550a37.

57、83a491a34.48a444b31.72a292.90Time spent chewingmin·d1 min·kg1DMI835a59.17a836a57.41a752b54.25a749b52.86a244.45a,b Within the same line, means with different letters are significantly different (P< 0.05; SE: standard error.Table IV.Influence of maturity stage and chopping length of maize

58、 silage on particle size in digesta. Maturity stage Early Late SE Chopping length Fine Coarse Fine CoarseParticle size distribution in bolus (% DM retained on sieve8.0 mm9.8a12.2ab7.7a16.5b 1.8 4.0 mm17.0a22.7b20.3ab19.7ab 1.8 2.0 mm14.8a11.9a17.7a10.3b 1.1 0.4 mm14.6a10.9a14.6a11.8a 1.3 0.1 mm 5.5a

59、 4.4a 4.4a 4.9a0.6 0.05 mm 1.6a 1.2a 1.3a 1.5a0.2< 0.05 mm36.6a38.6a33.8a35.6a 2.6 Mean particle size (mm0.21a0.21a0.27a0.30a0.06 d75-d25 (mm* 1.67a 2.09a 1.96a 2.54a0.25 LP2 (% on DM basis41.6a46.7a45.7a46.5a 2.9 Change inMean particle size (mm LP2 (% on DM basis0.59a21.0a0.55a14.3b0.28a12.3b2.3

60、5b26.6a0.062.4Particle size distribution in rumen (% DM retained on sieve8.0 mm 5.1a10.7b 6.5a13.5b 1.9 4.0 mm13.3a14.5a11.1a12.8a 1.3 2.0 mm19.3a14.3b19.5a11.5b 1.4 0.4 mm27.3a24.6a26.7a28.1a 2.7 0.1 mm14.4a14.2a12.6a13.7a 1.6 0.05 mm 2.9a 2.5a 2.5a 2.6a0.4 < 0.05 mm18.2a19.2a21.2a18.7a 5.6 Mean particle size (mm0.57a0.55a0.46a0.62a0.20 d