0900625

8/2/2019 0900625

1/10

Influence of diet on growth yields of rumen micro-organismsin vitro and in vivo: influence on growth yield of variable carbonfluxes to fermentation products

M. Blummel1*, A. Karsli2 and J. R. Russell2

1Institute for Animal Production in the Tropics and Subtropics (480), University of Hohenheim, 70599 Stuttgart,Germany2Department of Animal Science, Iowa State University, Ames, IA 50011, USA

(Received 29 November 2001 Revised 18 March 2003 Accepted 10 May 2003)

The efficiency of rumen microbial production (EMP) in vitro and in vivo was examined for three roughages (lucerne (Medicago sativa L.)hay, oat ( Avenia sativa L.)berseem clover (Trifolium alexandrinum cultivar BigBee) hay and maize ( Zea mays L.) crop residue (MCR))and for five isonitrogenous (106 g crude protein (N 625)/kg) diets formulated from lucerne hay, oatberseem clover hay, MCR, soya-bean meal and maize grain to provide degradable intake protein for the production of 130 g microbial protein/kg total digestible nutrients.EMP in vivo was determined by intestinal purine recovery in sheep and ranged from 240 to 360 g microbial biomass/kg organic mattertruly degraded in MCR and in one of the diets respectively ( P,005). EMP in vitro was estimated by the substrate degraded : gasvolume produced thereby (termed partitioning factor, PF (mg/ml)) at times of estimated peak microbial production and after 160 and240h of incubation. For the diets, PF values were significantly related to EMP in vivo at peak microbial production (P004), but notafter 160 (P008) and 24.0h (P066). For roughages, PF values were significantly related to EMP in vivo only when measuredafter 160 h (P004). For MCR and diets, a close non-linear relationship was found between PF values at peak microbial productionand EMP in vivo (R 2 099, P,00001) suggesting a maximum EMP in vivo of 039. Low gas production per unit substrate degraded

(high PF) was associated with high EMP in vivo. The in vitro study of the products of fermentation, short-chain fatty acids, gases andmicrobial biomass (by purine analysis) after 160 h of incubation showed very strong relationships (R 2$ 089, P,00001) betweenshort-chain fatty acids, gases and gravimetrically measured apparent degradability. Except for maize grain, the true degradability oforganic matter estimated by neutral-detergent solution treatment agreed with the sum of the products of fermentation (R 2 081,P00004). After 160 h of incubation, the synergistic effects of diet ingredient on diets were greater for microbial biomass (18 %)than for short-chain fatty acids and gas production (7 %). It is concluded that measurement of gas production only gives incomplete infor-mation about fodder quality; complementation of gas measurements by true degradability measurements is recommended.

Fermentation products: In vitro gas production: Microbial efficiency

Proportionally high conversion of rumen-degraded feedinto microbial biomass, i.e. a high efficiency of microbial

production (EMP), is desired in ruminant animal nutritionbecause it leads to efficient feed N and C utilization(Beever, 1993; Leng, 1993; Van Soest, 1994). It was, forexample, demonstrated that the nature and fermentationcharacteristics of feed protein (N) and carbohydrates canaffect EMP (Brown & Pittman, 1991; Clark et al. 1992;Sinclair et al. 1995), thus rejecting the assumption of aconstant EMP still prevalent in some feeding systems(Ausschuss fur Bedarfsnormen der Gesellschaft furErnahrungsphysiologie der Haustiere, 1986; Agricultureand Food Research Committee, 1993). As an exception,the Cornell net carbohydrate and protein system (Fox

et al. 1992; Russell et al. 1992; Sniffen et al. 1992) con-siders feed- or diet-specific differences in EMP, suggesting

their prediction by the analysis of numerous feed carbo-hydrate and N fractions with associated degradation rates(National Research Council, 1996).

A simpler in vitro technique for the estimation of EMP ofroughages was suggested by Blummel et al. (1997) andconsists of a combination of two in vitro measurements inone 240 h incubation using rumen inoculum. Gas volumeproduced during the incubation is recorded as describedby Menke et al. (1979) and the substrate truly degradedis gravimetrically quantified by the modification of thetechnique of Tilley & Terry (1963) suggested by Goering& Van Soest (1970). The degradability measurement

* Corresponding author: Dr M. Blummel, present address, ILRI-South Asia Project, c/o ICRISAT, Patancheru 502324, Andhra Pradesh, India,

fax +91 40 241238, email [email protected]

Abbreviations: EMP, efficiency of microbial production; LH, lucerne hay; MCR, maize crop residue; MG, maize grain; OBH, oatberseem clover hay;

OM, organic matter; PF, partitioning factor; SBM, soyabean meal; SCFA, short-chain fatty acid; t1/2, time of half asymptotic gas production.

British Journal of Nutrition (2003), 90, 625634 DOI: 10.1079/BJN2003934q The Authors 2003

8/2/2019 0900625

2/10

reflects how much substrate was used for the formation ofall products of fermentation, namely short-chain fattyacids (SCFA), gases and microbial biomass. The gasmeasurement indicates how much substrate was used forthe formation of SCFA and gases, since SCFA and gas pro-duction are very closely associated stoichiometrically(Wolin, 1960; Beuvink & Spoelstra, 1992; Blummel &rskov, 1993; Blummel et al. 1999a). The substrate trulydegraded (mg) : gas thereby produced (ml) was termedthe partitioning factor (PF) and, proportional to theamount of substrate degraded, low gas production (highPF) was indicative of a high EMP in vitro (Blummel et al.1997). For roughages, PF values from 275 to 445 mg/mlapproximately reflected YATP ranges from 10 to 32 (forreview, see Blummel et al. 1997).

Microbial growth, however, has several stages, notably

lag, growth, stationary and decline phases (Van Soest,1994). Microbial growth phases relative to incubationtime will vary between substrates and comparisons ofEMP in batch systems at different microbial growthphases might result in false conclusions. This might bethe case with PF analysis fixedly conducted after 240 hincubations without regard for substrate-specific differ-ences in microbial growth kinetics.

In the present study, eight roughages and mixed dietswere investigated with two objectives in mind. The firstwas to examine if PF comparisons at similar microbialgrowth phases would result in greater agreement betweenin vitro and in vivo estimates of EMP. Second, by studyingthe products of microbial degradation, the study addressessome fundamental considerations about in vitro techniques,namely the comparative advantages and disadvantages ofmeasuring substrate disappearance (Goering & VanSoest, 1970) or appearance of fermentation products suchas gases (Menke et al. 1979; Steingass & Menke, 1986).

Materials and methods

Feeds and feeding

The forages used for in vivo experimentation have pre-viously been described in detail by Karsli (1998). In brief,a first harvest oat (Avenia sativa L.)berseem clover (Trifo-

lium alexandrinum cultivar BigBee) was mowed at latevegetative stage and third harvest lucerne (Medicagosativa L.) was mowed at first flowering stage. Maize (Zeamays L.) crop residues (MCR) were obtained from maizefields and all three roughages were round-baled andground through a 50 mm screen of a tub-grinder shortlybefore the start of the feeding trial. The three roughageswere offered in restricted amounts and ad libitum (about115 % of intake; only results on ad libitum intake arereported in the present paper) to six rumen- and duode-num-cannulated wether sheep in a 6 6 Latin squaredesign. Five diets designed from the three roughages wereadjusted to approximately isonitrogenous levels by maizegrain (MG) and soyabean meal (SBM) incorporation and

were offered in a consecutive experiment to five sheep ina 5 5 Latin square design. The diets had degradableintake protein levels to provide 130 g microbial protein/kgtotal digestible nutrients, consumed according to level 1 of

the Nutrient Requirements of Beef Cattle computer program(National Research Council, 1996). In the calculation ofdietary crude protein, degradable protein intake and totaltract digestible organic matter (OM; see later) of MCR,oatberseem hay (OBH) and lucerne hay (LH) determinedin a previous experiment (Karsli, 1998) were used. Eachperiod lasted for 22 d, consisting of a 12 d adaptation anda 10 d collection period. The animals were housed in indi-vidual pens under controlled (258C) temperature conditionsand had free access to water and a trace minerals salt block.

Digesta flow, digestibility measurements and estimates ofin vivo microbial biomass production

Cr-mordanted fibre (15 g) containing approximately 20 gCr/kg was added to the rumen daily at 08.00 and 20.00

hours to determine digesta flow as described by Rojas-Bourillon et al. (1987). Feed and ort samples were col-lected on days 12 to 16 of each period and composited.On days 14, 15 and 16 of each period, 200 ml duodenaldigesta and 15 g fresh faecal matter were collected fourtimes per d. Sampling time was advanced 20 h per d sothat the samples collected over the 3 d represented each20 h of a 240 h cycle. After collection, samples werefrozen and stored. Approximately 1500 ml rumen contentswere collected from the ventral sac of the rumen for theisolation of rumen microbes by differential centrifugationaccording to Adamu et al. (1989).

DM concentrations of feeds, orts and faeces were deter-mined by drying in a forced-air oven at 608C for 480h.Duodenal digesta samples and the bacterial pellets werefreeze-dried. Dried feed, duodenal digesta and faecalsamples were ground through a 1 mm screen. Dry duodenaldigesta and faecal samples were composited on an equal dryweight basis for each animal in each period. OM concen-trations of dried feeds, orts, duodenal digesta and faecalsamples were determined as the weight loss during combus-tion at 6008C for 20 h in a muffle furnace. Concentrations ofneutral-detergent fibre and acid-detergent fibre in dry feeds,orts, duodenal digesta and faecal samples were determinedby the sequential procedure of Van Soest & Robertson(1985) and N concentrations were analysed by the Kjeldahlmethod. Purine concentrations of duodenal digesta and

rumen microbial pellet samples were determined by the pro-cedure of Zinn & Owens (1986). The amounts of N and trulydegraded OM in the rumen were calculated by correctingthe amounts of apparently degraded N and OM formicrobial N and OM as determined by the N:purines andN:OM ratios in the rumen bacterial pellet. Similarly,microbial biomass produced was calculated by dividingduodenal purine flow by the concentration of purines inthe DM of the bacterial pellet.

In vitro incubation procedures

Substrates were incubated in 100 ml calibrated glass syr-inges commonly used for the Hohenheim gas-production

test (Menke et al. 1979; Steingass & Menke, 1986), butthe incubation protocol followed was that of Blummel &Becker (1997). Briefly, four replicates of 500mg air-drysubstrate were weighed into the syringes and these were

M. Blummel et al.626

8/2/2019 0900625

3/10

incubated with 40 ml mixed suspension of rumen digesta.This suspension consisted of 10 ml rumen contents, 10 mlbicarbonate buffer, 5 ml macro- and micro-mineralssolutions (0002 ml of which was micro-mineral solution)and 15 ml distilled water. As a modification to themethod of Blummel & Becker (1997), no NH4HCO3 wasused in the preparation of the bicarbonate buffer, whichconsisted entirely of NaHCO3 (0467 M). Rumen inoculumwas collected from a ruminally fistulated GermanHinterwalder cow kept on a medium- to good-qualitygrass hay diet before morning feeding. The rumen inocu-lum consisted of about 600 ml rumen contents and400 ml rumen particulate matter/l. For details of inoculumhandling and preparation, see Blummel & Becker (1997).

In vitro gas volume and apparent DM and true organicmatter degradability measurements

Five different incubations were conducted. The rate andextent of in vitro gas production from roughages, concen-trates and diets were obtained during a 960 h incubationwith gas volume recordings after 2, 4, 6, 8, 10, 12, 24, 30,36, 48, 54, 60, 72 and 960 h of incubation. The rate andextent of gas production were calculated using an exponen-tial model yV(1e2ct), in which y is the gas volume attime t, V is the asymptotic value of gas production and cis the the fractional rate of gas production.

In vitro true OM degradability was determined afterrecording the gas volume and terminating the incubation

at the desired time by transferring the entire syringe con-tents through the outlet of the syringe into 600 ml spoutlessBerzelius beakers. The syringes were rinsed twice withneutral-detergent solution (prepared without sodium sulfiteand amylase) by dispensing 25 ml neutral-detergent sol-ution through the outlet into the syringe each time. The syr-inge was shaken each time to remove residual particles,after which the contents were added to the respectiveBerzelius beaker. The procedure of Goering & Van Soest(1970) for the determination of true digestibility was thenapplied by refluxing the incubation residue for 10 h andrecovering the undigested matter on filter crucibles (poros-ity 2). The PF value at a given time of incubation wascalculated as OM truly degraded (mg):gas volume pro-duced thereby (ml). The PF values were determined after240 and 160 h of incubation as well as at the time atwhich half-maximum gas production was achieved (t1/2).Preliminary data about kinetics of in vitro gas productionand microbial biomass yield had suggested that t1/2 couldserve as a common time denominator to facilitate across-substrate comparisons of PF values (Blummel et al.1999b). t1/2 in the model yV (12e

2ct) was calculated as:

t1=2 ln 2=c:

In vitro apparent DM degradability was determined after160 h of incubation by high-speed centrifugation. Theentire syringe content was transferred into centrifuge tubes

and centrifuged at 20 000g for 30min at 48C. The super-natant fraction was carefully removed with a Pasteur pipetteand stored for SCFA and NH3 analysis. The syringes werewashed three times with NaCl solution (4 g/l), each time

dispensing 15 ml of solution through the spike into thesyringe. The syringe was shaken each time to removeresidual particles and the contents were added to therespective centrifuge tube. After completing rinsing of thesyringes, centrifugation was repeated once and the super-natant fraction was discarded. The pellets were lyophilizedovernight and residual moisture was removed by drying in aforced-air oven at 1058C for 30h. In vitro apparent DMdegradability equals DM substrate incubated (mg) (DMpellet (mg) DM blank pellet (mg)), where blank pelletweight was determined by centrifugation of 4 40mlrumen suspension sampled at the 0 h of incubation.

In vitro short-chain fatty acid production, ammoniaconcentration, stoichiometrical calculations and purine

analysis after 160 h of incubationSCFA in the supernatant fraction were analysed by GC(Hewlett Packard 5880 A (Hewlett Packard, Palo Alto,CA, USA) with flame ionization detection) as describedby Aiple (1993). Gas production was calculated fromSCFA according to Blummel et al. (1999a) and measuredgas volumes were corrected for pressure as described inthe same paper. The amounts of C, H and O required forthe production of SCFA (including the isoacids) andassociated fermentative CO2, CH4 and H2O were summedaccording to Blummel et al. (1997). ATP production fromthe acids was calculated using standard values from theliterature as described by Blummel et al. (1997). NH3-Nin the supernatant fraction was determined by the Kjeldahlmethod using direct steam distillation. Purines (adenineand guanine) were analysed by HPLC as described byMakkar & Becker (1999). Purines were analysed afterball-milling the pellets obtained by high-speed centrifu-gation in the determination of in vitro apparent DM degrad-ability and net purine production (mmol) was calculatedbased on pellet weights and purine analysis. The purine con-centration in the microbial biomass was estimated by purineanalysis of the lyophilized blank microbial pellet.

Statistical analysis

Statistical differences in the variables were analysed using

ANOVA procedure (1988; Statistical Analysis Systems,Cary, NC, USA) and mean values were separated by apply-ing least squares difference procedures with the probabilitylevel set to P005. Significance statements refer to thisprobability level unless otherwise stated. The computerprogram GraphPad Prism (1994; GraphPad Inc., SanDiego, CA, USA) was used to calculate the rate andextent of gas production by non-linear regression andalso to calculate the simple linear regression relationshipsused in the present study.

Results

Roughages, concentrates, diets and in vivoexperimentation

The chemical compositions of LH, OBH, MCR, SBM,MG and diets (diets 1 to 5) are summarized in Table 1.

Growth yields of rumen micro-organisms 627

8/2/2019 0900625

4/10

The mean crude protein (N 625) content of diets was 106(range 103 109) g/kg with little variation between thediets. For roughages, the crude protein content was lowestin MCR and highest in LH, while the reverse was true forfibre constituents. All diets included two roughages andat least one concentrate. MCR and MG were included inall diets.

Daily intakes of roughages and diets, OM truly degradedin the rumen and EMP are summarized in Table 2. DM andOM, crude protein and fibre intakes were substantiallyhigher in LH and OBH than in MCR and diets. Forroughages, OM truly degraded in the rumen was substan-

tially higher for LH than OBH and MCR, but only smalldifferences in OM truly degraded in the rumen wereobserved for the diets. The EMP in vivo, expressed as gmicrobial biomass production/kg OM truly degraded inthe rumen, was lowest for MCR and highest for diet 5.The diets had higher EMP than roughages, but the differ-ence was small for LH and OBH.

In vitro gas volume, substrate degradability measurementsand partitioning factor values: their relationships with theefficiency of microbial production in vivo

Asymptote, c and t1/2 of in vitro gas production, in vitrotrue OM degradability and PF of concentrates, roughages

and diets are summarized in Table 3. There were signifi-cant differences in V, c and t1/2 between components and

between diets. It can be calculated that the mean V valueof the diets was slightly (15 %), but significantly, higher(P,005) than expected from the summation of therespective values of the feed components according totheir proportion in the diets. The rate of gas productionof the diets was about 13 % higher than calculated fromthe incubations of the dietary components and theirrespective proportions in the diets. Similarly, t1/2 wasabout 13 % less in the diets than calculated from the dietarycomponents.

Gas volumes, in vitro true OM degradability and PFmeasured at t1/2 and after 160 h of incubation differed sig-

nificantly between components and diets. t1/2 varied widelyfrom 76 and 85 h for SBM and LH to 215 and 227 h forMG and MCR respectively. The difference between theshortest and longest t1/2 in diets was about 20 h. All t1/2were ,24 h and gas volumes and in vitro true OM degrad-ability after 240 h were consequently higher than theirrespective t1/2 and 160 h values (Table 3), but the reversewas true for PF values after 240 h (results not shown). Gasvolumes measured at t1/2 were not exactly 50 % of theasymptotic volumes, reflecting some deviation in therepeatability of gas production kinetics in differentincubations.

PF at t1/2 was significantly related to EMP in vivo in thediets, but not in the roughages (Fig. 1). In vivo EMP

across diets and roughages were not significantly relatedto gas volumes but were related to the in vitro true OM

Table 1. DM, organic matter, crude protein (N 625), neutral-detergent fibre and acid-detergent fibre in soyabeanmeal, maize (Zea mays L.) grain, lucerne (Medicago sativa L.) hay, oat (Avenia sativa L.)berseem clover and (Trifolium

alexandrinum cultivar BigBee) hay and maize crop residue and of diets made from these components*

SBM MG LH OBH MCR Diet 1 Diet 2 Diet 3 Diet 4k Diet 5{

DM (g/kg) 944 914 930 931 933 934 933 930 932 928OM (g/kg DM) 932 987 918 892 936 938 935 928 939 940CP (g/kg DM) 436 79 181 139 41 107 106 109 103 104NDF (g/kg DM) 206 169 420 622 757 659 645 610 628 554ADF (g/kg DM) 136 30 284 374 433 369 359 339 356 315

SBM, soyabean meal; MG, maize grain; LH, lucerne hay; OBH, oatberseem clover hay; MCR, maize crop residue; OM, organic matter; CP,crude protein; NDF, neutral-detergent fibre; ADF, acid-detergent fibre.

* Composition of diets. Diet 1: 0790 (MCR)005 (MG)016 SBM. Diet 2: 0690 MCR0103 OBH0083 MG0124 SBM. Diet 3: 0450 MCR035 OBH015 MG005 SBM.kDiet 4: 0691 MCR0103 LH0103 MG0103 SBM.{Diet 5: 0450 MCR0350 LH020 MG.

Table 2. Daily intakes (g/kg live weight) of DM, organic matter and crude protein (N 625), organic matter truly degraded in the rumen andefficiency of microbial production*

(Mean values for six sheep)

LH OBH MCR Diet 1 Diet 2 Diet 3 Diet 4 Diet 5

DM intake 451a 311b 169c 171c 177c 220c 188c 217c

OM intake 414a 286b 159c 160c 165c 205c 183c 204c

CP intake 84a 44b 086d 20c 21c 25c 22c 23c

OM truly degraded (g/kg) 704a 612c 582c 655ab 647b 670a 645b 657ab

EMP (g microbial biomass/kgOM truly degraded)

295c 303c 241d 312bc 333b 344ab 326b 364a

LH, lucerne (Medicago sativa L.) hay; OBH, Oat (Avenia sativa L.)berseem clover (Trifolium alexandrinum cultivar BigBee) hay; MCR, maize (Zea maysL.) cropresidue; OM, organic matter; CP, crude protein; EMP, efficiency of microbial production.

a,b,cMean values within a row with unlike superscript letters were significantly different (P,005).* For details of diets and procedures, see Table 1 and p. 626.


8/2/2019 0900625

5/10

degradabilities (R 2 060, P002). Within diets, thisrelationship was again insignificant for gas volumes, butwas significant for in vitro true OM degradabilities(R 2 078, P0049). Within roughages, gas volumes andEMP in vivo were inversely related (R 2 099, P0016),but this relationship was insignificant for in vitro trueOM degradabilities. The PF and EMP in vivo of the fivediets and MCR could be fitted onto the same regressionline, which was better described by a curvilinear than alinear function (Fig. 1). The curvilinear function projectedan asymptote of 386 g microbial biomass/kg OM truly

degraded in the rumen. Combined gas volumes of thediets and MCR were linearly related to EMP in vivo (R 2

065, P,005). This relationship was closer between invitro true OM degradabilities and EMP in vivo (R 2 094,P,0001). No curvilinear regression of the kind reportedin Fig. 1 could be fitted to these two latter relationships.

The relationship between the PF after 160 h of incu-bation and the EMP in vivo was significant for theroughages (P004), but insignificant for the diets(Fig. 2). The PF after 160 h and EMP in vivo of the fivediets and MCR could not be fitted onto the same regressionline. The relationship between the PF after 240 h and EMPin vivo tended to be significant for roughages (P007),but was insignificant for diets (P066; results not shown).

In vitro fermentation products and their relationships

SCFA production, NH3 concentration, total andproportional purine recovery and in vitro apparent DMdegradabilities were measured after 160 h of incubation(Table 4). Acetate proportion was lowest in theconcentrates and highest in the roughages. Isoacids andvalerate were formed from SBM and were generally

Fig. 1. Relationship between in vitro efficiency of microbial pro-duction (EMP) estimated by combined substrate degradability andgas volume measurements at time of half asymptotic gas pro-duction (t1/2) and EMP measured in sheep. OM, organic matter. O,Maize (Zea mays L.) crop residue; K, oat (Avenia sativa L.)ber-

seem clover (Trifolium alexandrinum cultivar BigBee) hay; A,lucerne (Medicago sativa L.) hay (P033 for roughages); X, diet 1;O, diet 2; P, diet 3; B, diet 4; V, diet 5 (P004 for diets 1 to 5);. . ., curvilinear (R2 099); , linear (R2 093). For details of dietsand procedures, see Tables 1 and 2 and p. 626.

Table 3. Gas volumes, substrate degradabilities and partitioning factors (PF) of soyabean meal, maize (Zea mays L.) grain, lucerne(Medicago sativa L.) hay, oat (Avenia sativa L.)berseem clover (Trifolium alexandrinum cultivar BigBee) hay, maize crop residue and of diets

(diets 1 to 5) measured at time of half asymptotic gas production (t1/2) and after 160 h of incubation*

Variable SBM MG LH OBH MCR Diet 1 Diet 2 Diet 3 Diet 4 Diet 5

Gas asymptote (V, ml) 1496ef 2340g 1236b 1129a 1384c 1409c 1463de 1442d 1442d 1512f

Gas rate (c, % per h ) 901g 323b 816f 461d 305a 422c 419c 425c 449d 485e

Half-time (t1/2, h) 769a 215g 850b 151d 227h 164f 165f 163f 154e 143c

Gas at t1/2 (ml) 719f 1068g 552a 535a 633b 668c 711ef 678cd 708e 695de

ivTDOM at t1/2 (mg) 3745e 4259f 2511d 2029b 1676a 2166c 2268c 2280c 2301c 2634d

PF at t1/2 (mg/ml) 521f 399d 455e 379c 264a 310b 319b 336b 325b 370c

Gas 16 h(ml) 1086h 905g 855f 595b 531a 726c 721c 706c 760d 800e

ivTDOM 16 h (mg) 4147g 4453h 2925f 2053b 1589a 2178c 2203c 2351d 2356d 2624e

PF 16 h (mg/ml) 381d 492e 342bc 345c 299a 300a 305a 333bc 309a 328b

SBM, soyabean meal; MG, maize grain; LH, lucerne hay; OBH, oatberseem clover hay; MCR, maize crop residue; ivTDOM, in vitro truly degraded organicmatter.

a,b,c,d,e,f,g,hMean values within a row with unlike superscript letters were significantly different (P,005).* For details of diets and procedures, see Tables 1 and 2 and p. 626.

Variables are related to the incubation of 500 mg dry substrate.

Fig. 2. Relationship between in vitro efficiency of microbial pro-duction (EMP) estimated by combined substrate degradability andgas volume measurements after 16.0 h of incubation and EMPmeasured in sheep. OM, organic matter. O, Maize (Zea mays L.)crop residue; K, oat (Avenia sativa L.)berseem clover (Trifolium

alexandrinum cultivar BigBee) hay; A, lucerne (Medicago sativaL.)hay (roughages P004); X, diet 1; O, diet 2; P, diet 3; B, diet 4;V, diet 5 (P008 for diets 1 to 5); . . ., curvilinear (R2 056); , lin-ear (R2 051). For details of diets and procedures, see Table 1 andp. 626.


8/2/2019 0900625

6/10

,2 mmol/100 mmol total fatty acids in the other substrates.In the diet components, NH3 concentration was lowest inMG and highest in SBM. Significant differences in NH3concentrations were also observed in the diets, and it canbe calculated that their NH3 concentrations and EMP invivo were strongly inversely related (R 2 093, P0007).The highest purine concentration was found with MGand lowest with OBH. Except for SBM, guanine concen-tration was higher than adenine concentration. It shouldbe noted here that the purine contents in the non-incubatedsubstrates were 57, 57, 62, 63 and 61mmol/500 mg inSBM, MG, LH, OBH and MCR respectively. Purineyield per unit ATP produced (mmol/mmol) from the incu-bation of 500 mg substrate ranged from 128 in soyabeanmeal to 256 in MG. There was no significant relationshipbetween purine yield per unit ATP produced and EMP in

vivo either across diets and roughages or within diets androughages. The relationship tended to be inverse for theroughages (P022).

For diets and roughages, a strong relationship was foundbetween stoichiometrically calculated and actuallymeasured gas volumes (Fig. 3). In MG, measured gasvolume was about 20 % higher than the calculated value,while for SBM measured gas volume was less than thestoichiometrically calculated value by approximately16 %. The relationship was described by the regressionequation y248068x (R 2 089, P,00001).

In vitro apparently degraded substrate was calculated bysummation of the mass of C, H and O recovered in SCFAand the mass of C, O and H stoichiometrically calculated

for CO2, CH4 and H2O produced. There was a very goodagreement between in vitro apparent DM degradability asdetermined by high-speed centrifugation and in vitro appar-ent DM degradability calculated by C, O and H balance

(Fig. 4). The relationship was described by the regressionequation y22010x (R 2 098, P,00001).

In vitro truly degraded substrate was calculated as thesum of C, O and H recovered in SCFA and fermentativeCO2, CH4 and H2O plus the amount of microbial biomassproduced (the latter calculated from purine production).The relationship between in vitro true OM degradabilityas measured by neutral-detergent solution treatment andcalculated true degradability was strong, except for MG,where the measured degradability was about 130 mg

Fig. 3. Relationship between in vitro gas volumes stoichiometricallycalculated from short-chain fatty acid recovery after 160 h of incu-bation and measured gas volumes. O, Maize (Zea mays L.) crop

residue; K, oat (Avenia sativa L.)berseem clover (Trifolium alex-andrinum cultivar BigBee) hay; A, lucerne (Medicago sativa L.) hay;L, maize grain; S, soyabean meal; X, diet 1; O, diet 2; P, diet 3; B,diet 4; V, diet 5; , y x. For details of diets and procedures, seeTables 1 and 2 and p. 626.

Table 4. Short-chain fatty acid (SCFA) production, ammonia-nitrogen, total purine bases (PB), PB relative to ATP and in vitro apparentdegradabilities of DM obtained from the incubation of diet components and of diets*

Variable SBM MG LH OBH MCR Diet 1 Diet 2 Diet 3 Diet 4 Diet 5

SCFA (mmol) 270i 147 g 169h 112b 102a 138ef 133cd 131c 142fg 146g

Acetate(mmol/100 mmol total fatty acids)

575a 583b 697e 714f 693e 661c 663c 672d 664c 674d

Propionate(mmol/100 mmol total fatty acids)

252 h 235g 211c 201a 238g 229f 223e 205b 218d 198a

Butyrate(mmol/100 mmol total fatty acids)

92c 156h 60a 68b 68b 90c 97d 108f 99e 112 g

Isobutyrate(mmol/100 mmol total fatty acids)

14e 04c 05d 04c 0a 05d 04c 03b 04c 03b

Valerate(mmol/100 mmol total fatty acids)

30 g 10e 20f 10e 05a 08cd 07c 06b 08cd 10e

Isovalerate (mmol/100 mmoltotal fatty acids)

37f 12e 07c 03a 07c 05b 03a 05b 03a

NH3 (mg) 152a 003g 30b 22c 06f 25c 19d 12e 17d 08f

PB (mmol) 848e 999f 681b 494a 502a 714c 673b 681b 732c 782d

Adenine (mmol) 457e 473e 303c 236a 248b 349cd 324c 329c 351cd 376dGuanine (mmol) 391b 526c 378b 258a 254a 365b 349b 352b 381b 406b

PB/ATP (mmol/mmol) 128a 256 g 159b 172c 187d 200ef 194de 198ef 198ef 205f

ivADDM (mg) 3115f 1782d 1974e 1494b 1251a 1551bc 1581c 1506b 1591c 1714d

SBM, soyabean meal; MG, maize (Zea maysL.) grain; LH, lucerne (Medicago sativa L.) hay; OBH, oat (Avenia sativa L.)berseem clover (Trifolium alexandrinumcultivar BigBee) hay; MCR, maize crop residue; ivADDM, in vitro apparently degraded DM.

a,b,c,d,e,f,g,h,iMean values within a row with unlike superscript letters were significantly different (P,005).* For details of diets and procedures, see Tables 1 and 2 and p. 626. Variables are related to the incubation of 500 mg dry substrate. Negative value calculated.


8/2/2019 0900625

7/10

higher than the calculated value (Fig. 5). The relationshipwas described by the regression equation y846062x(R 2 081, P00004).

Discussion

Relationships between efficiency of microbial productionin vitro and in vivo

As a high conversion of degraded feed into microbialbiomass, i.e. high EMP, decreases the need for supplement-

ing rumen undegradable feed protein and decreases thefeed C flow into fermentative CO2 and CH4, high EMPis preferred over high SCFA production (Beever, 1993;Leng, 1993; Van Soest, 1994). The prediction of possible

feed-related differences in EMP is, therefore, of consider-able interest in feed analysis. In the present work, signifi-cant differences in EMP in vivo were found for bothroughages and diets, even though the latter were designedby Karsli (1998) to provide uniformly the degradableintake of protein required for the synthesis of 130 gmicrobial protein/kg total digestible nutrients consumed(National Research Council, 1996; level 1). These findingssupport the concept of varying EMP implicit in the Cornellnet carbohydrate and protein system (National ResearchCouncil, 1996; level 2). Unfortunately, the analyticaltools suggested by the Cornell net carbohydrate and proteinsystem for the prediction of EMP are quite laborious andsimpler techniques are required.

The results presented in Fig. 1 suggest that combinedtrue degradability of substrate and gas volume measure-

ments in vitro, i.e. PF analysis (Blummel et al. 1997), isa promising technique with regards to the detection of vari-ations in EMP. For diets, there was a significant relation-ship between PF values measured at t1/2 and EMP invivo, which accounted for 94 % (non-linear relationship)of the variation in EMP. However, despite the relativetime proximity between PF measurements at t1/2 andafter 160 h of incubation (Table 3), this relationship wasonly close to significant (P008) when measured after160 h and was insignificant (P066) when determinedafter 240 h. The 240 h findings can be explained by a dis-tortion of PF measurements through secondary fermenta-tion of lysed microbial cells into SCFA and,consequently, of the gases after microbial peak yield(Blummel & rskov, 1993; Cone et al. 1997). Still, thePF value in diet 5, for example, decreased by about11 %, from 370 (PF at t1/2) to 328 (PF after 160 h) mg/ml within the relative short incubation period from 143to 160h (Table 3). These findings suggest that for thekind of diets designed by Karsli (1998) detection of vari-ations in EMP might require more laboratory input thanPF analysis at only one fixed incubation time. Theadditional laboratory input consists of the incubationrequired for estimating asymptote and rate of gas pro-duction for the calculation of t1/2.

For the roughages, time-uniform PF analysis after 160 h(and 240 h) of incubation was better related to EMP in vivo

than PF analysis at t1/2. This was caused by the high PFvalue of LH (455 mg/ml) when determined at t1/2, whichwas only 85h. It is possible that this PF value of455 mg/ml, which according to Blummel et al. (1997)would indicate a YATP of approximately 32, presents anartifact caused by removal of still unfermented cellularcomponents of LH soluble by neutral-detergent solutiontreatment, resulting in an overestimation of in vitro trueOM degradability and consequently the PF value. This pro-blem will be addressed in greater detail later on. Despitethese reservations, theoretical PF values for diets calcu-lated from PF values of dietary components at t1/2and the proportion of each component in a diet werestrongly related to the in vivo EMP of diets (R 2 093).

These findings suggest that the in vivo EMP may be pre-dicted from PF values of possible feed components.

Interestingly, the regression of the EMP in vivo on PF ofdiets and MCR at t1/2 could be fitted onto one regression

Fig. 4. Relationship between in vitro apparent degradabilitymeasured by high-speed centrifugation after 160h of incubationand recovery of C, H and O in short-chain fatty acids, (SCFA) andgases. O, Maize (Zea mays L.) crop residue; K, oat (Avenia sativaL.)berseem clover (Trifolium alexandrinum cultivar BigBee) hay;A, lucerne (Medicago sativa L.) hay; L, maize grain; S, soyabeanmeal; X, diet 1; O, diet 2; P, diet 3; B, diet 4; V, diet 5; , y x.For details of diets and procedures, see Tables 1 and 2 and p. 626.

Fig. 5. Relationship between in vitro true organic matter degradabil-ity estimated by neutral detergent solution after 160 h of incubationand recovery of short-chain fatty acids, gases and microbial bio-mass. O, Maize (Zea mays L.) crop residue; K, oat (Avenia sativaL.)berseem clover (Trifolium alexandrinum cultivar BigBee) hay;A, lucerne (Medicago sativa L.) hay; L, maize grain; S, soyabeanmeal; X, diet 1; O, diet 2; P, diet 3; B, diet 4; V, diet 5. For details

of diets and procedures, see Tables 1 and 2 and p. 626.


8/2/2019 0900625

8/10

line, which was better (R 2 099 v. 093) described by anon-linear than a linear function (Fig. 1). Both curvilinear-ity, as well as projected asymptote of EMP (386gmicrobial biomass/kg OM truly degraded in the rumen),agree well with the findings on maximum EMP (in thepresence of protozoa) reported by Russell et al. (1992)and Van Soest (1994). For LH and OBH, estimates ofEMP in vitro by PF analysis were consistently higherthan the respective EMP in vivo (Figs 1 and 2). Crude pro-tein content was higher in these two roughages than in dietsand MCR, and this crude protein might have been usedmore efficiently for microbial synthesis in vitro, whereno absorption and passage of NH3 occurs, than in vivo.It can be calculated from in situ results on N and OMdegradation of the LH and OBH (Karsli, 1998) thatrumen N availability in LH was much higher than the

OM availability, which might have resulted in a less thanoptimal EMP in vivo (Sinclair et al. 1995; Witt et al.1999). Similarly, Blummel et al. (2001) calculated thesynchronization indices of N:OM fermentation of thefive diets according to Sinclair et al. (1995) and Wittet al. (1999) and concluded that both overall N and syn-chronization of N:OM fermentation was insufficient formaximum EMP. However, in the current in vitro work,NH3 was still present in the diet incubation (Table 4),which does not support the assumption of N as the primar-ily limiting factor for EMP. Interestingly, the NH3 concen-tration of the diets in vitro was highly negatively related tothe EMP in vivo (R 2 093) and differences in the EMP aretherefore probably less due to lack of total N and more tomicrobial inefficiencies in utilizing it.

In the roughage incubations, the PF analysis after 160and 240 h ranked the roughages in the same order as theEMP in in vivo determinations, the relationship being sig-nificant for the PF after 16.0 h (Fig. 2), but not for the PFafter 240h (P007). The 240 h incubation time wasapparently too long for LH, considering the short t1/2(85 h) of this roughage, resulting in a low PF value after240 h because of microbial lysis. However, only threeroughages have been investigated in the present workand caution is, therefore, required to not over-interpretthese results.

Relationships between in vitro fermentation products

The concept of PF analysis demands a close stoichiometri-cal relationship between SCFA and gas production and areliable determination of true degradability of the substrate.In the present work, gas volumes of diets and roughageswere well predicted by SCFA analysis and the applicationof the stoichiometrical relationships outlined by Wolin(1960). In MG, however, more gas was measured thanwas stoichiometrically calculated, while the reverse wastrue for SBM (Fig. 3). The latter findings agree with resultsreported by Blummel et al. (1999a) obtained with SBM ofa different provenance. These authors concluded thatthe stoichiometrical relationship between SCFA and gas

production is not good for feeds with a crude proteincontent .400 g/kg.

A possible explanation for this observation resides withthe high production of NH3 in these feeds, ultimately

impairing the bicarbonate buffering system in the in vitrogas test (Cone 1998; Blummel et al. 1999a). Thus, eventhough SCFA are produced, the buffering effect does notnecessarily allow the expulsion of CO2 into the gasphase. In the bicarbonate-buffered gas test with an intactbuffering mechanism, approximately 50 % of the totalgas volume consists of CO2 originating from bufferingthe SCFA produced (Blummel & rskov, 1993). Moregas was measured in the MG incubation than wasaccounted for by SCFA analysis and stoichiometricalcalculations. Lactate was not analysed in the presentwork because several authors (Steingass & Menke, 1986;Getachew et al. 1998) have shown the buffering capacityof the employed in vitro test to be adequate for maintainingpH .6 in concentrate and starch incubations. These pHconditions do not favour lactate production.

Nevertheless, buffering of lactate would lead to gas pro-duction, but, besides the unfavourable pH conditions men-tioned earlier, it appears unlikely that lactate was producedto any significant extent for another reason. There was avery good agreement between the amount of substrateapparently degraded in vitro, as determined by high-speed centrifugation and the amount of C, O and H recov-ered in SCFA, and calculated for fermentative CO2, CH4and H2O (Fig. 4). Were lactate in the MG incubationresponsible for the stoichiometrical underestimation ofgas production, more substrate should have been appar-ently degraded than recovered in SCFA and fermentativeCO2, CH4 and H2O, but this was not the case. Morework is required to understand why more gas was recov-ered in MG than was accounted for by SCFA, and itappears worthwhile investigating if the complete uptakeof NH3 (very little NH3 was recovered after terminatingthe MG; Table 4) was responsible, possibly due to somemechanism converse to the one observed in SBM (excessNH3). However, lactate analysis should be included inthis work.

Gravimetrically determined true degradability andrecovery of fermentation products

As mentioned earlier, reliable determination of true sub-strate degradability is a second condition for meaningful

PF analysis. The relationship between in vitro true OMdegradability, as determined by neutral-detergent solutiontreatment, and true OM degradability calculated as thesum of C, H and O recovered as SCFA, CO2, CH4 andH2O plus the microbial biomass produced, is presented inFig. 5. There was a good agreement between measuredand calculated true degradability for the diets and mostof the diet components, confirming that neutral-detergentsolution treatment achieves an effective separation ofmicrobes and undegraded feed, thereby yielding a reliableestimate of true feed degradability (Van Soest, 1994).However, neutral-detergent solution-derived in vitrodegradability in MG was about 130 mg higher than wasaccounted for by the fermentation products (Fig. 5).

The low rate of gas production from MG (323 % per h,Table 3) strongly suggests that not all the starch was fer-mented after 160 h of incubation. Starch can be solublein neutral-detergent solution (Van Soest, 1994) and


8/2/2019 0900625

9/10

removal of unfermented starch is probably responsible forthe high discrepancy between the measured and calculatedtrue degradability evident in MG. It appears that determi-nations of in vitro true degradability of pure, high-starchconcentrates by neutral-detergent solution treatment arepotentially prone to substantial analytical errors, resultingin an overestimation of degradability.

In the balance of fermentation products, microbialbiomass was estimated by purine analysis. It should bementioned here that the relationship between 160 hpurine production per unit ATP and EMP in vivo was insig-nificant for diets and roughages, the relationship tending tobe inverse for the roughages. The purine contents of non-incubated diet components was on average 600mmol/500 mg, which appears high when compared with thepurine yield after incubating the five substrates for 160 h

(mean value 705mmol/500 mg; see Table 4). Eventhough McAllan & Smith (1973a,b) suggested that feedpurines are completely degradable in the rumen, Djouvinovet al. (1998) have recently shown the effective feed purinedegradability in the rumen to be dependent on incubationtime. This could affect in vitro measurements where theentire apparently undegraded pellet was analysed for pur-ines after a short (160 h) incubation period relative torumen retention times. Studies of the kind presented inFig. 5 should therefore be confirmed using differentmicrobial markers.

Complementarity of measurements of substrate

disappearance and of fermentation productsGravimetric methods measuring substrate disappearancehave frequently been criticized on the grounds thatsubstrate may be lost, for example by solubilization orfiltration, without having actually being fermented(Menke et al. 1979; Blummel & rskov, 1993; Grootet al. 1998; Pell et al. 1998) and these concerns can be justified (see Fig. 5). These reservations provided astrong rationale for the development of feed evaluationsystems based on in vitro gas production (Menke et al.1979). Blummel et al. (1997), on the other hand, haveargued that measurement of only one fermentation product,particularly of gases that are waste products, is question-

able unless the proportionality of fermentation products,including SCFA, gases and microbial biomass, is constant.If this relationship is not constant, gravimetric measure-ment of true substrate degradability (Goering & VanSoest, 1970) provides a more convincing concept.

Yet more information about the fermentation can beobtained by combining true degradability and gas volumemeasurements. The results presented in the present studysupport this hypothesis, despite the occasional problemsencountered with neutral-detergent solution treatment inthe in vitro determination of true OM degradability dis-cussed earlier. For example, in vitro true OM degradabilityof diets measured at t1/2 was significantly related to theirEMP in vivo, while gas volume at t1/2 was not. In vitro

OM degradability measured at t1/2 is probably a goodindicator of the rate of substrate degradation. As high sub-strate degradability rates decrease microbial maintenancerequirements, EMP is potentially increased (Pirt, 1982;

Russell et al. 1992). The problem associated with measur-ing in vitro gas production only is also manifest in theroughage incubation, where gas volumes at t1/2 and EMPin vivo were inversely related (P002). MCR had ahigher (P,005) gas volume at t1/2 than LH and OBH,reflecting a proportionally higher conversion of degradedsubstrate into SCFA and gases. In a related context, after160 h of incubation the mean purine production from500 mg of the mixed diets was 716mmol. It can be calcu-lated from the purine content of components and theirproportion in the diet that only 606mmol purines wereto be expected, assuming additive effects, i.e. the associat-ive effect of supplementation on purine yield was about18 %. On the other hand, mean SCFA yield from the incu-bation of 500 mg mixed diets was 138 mmol, while129 mmol SCFA were to be expected from additive effects

of SCFA production from the incubation of the dietarycomponents. i.e. the associative effect on SCFA was onlyabout 7 % (calculated from Table 4).

In conclusion, nutritionally significant variations werefound in the proportions of rumen fermentation products,including microbial biomass yield. In vitro degradabilitymeasurements were related to EMP in vivo in a moremeaningful manner than gas volume measurements, butmore comprehensive information about substrate degra-dation was obtained by combining both measurements.Therefore, to achieve a high EMP it appears sensible tosuggest the selection of feeds with a high true degradabilityand low gas production in proportion to the amount ofsubstrate degraded.

Acknowledgement

The authors would like to thank Dr H. Steingass fromthe Institute for Animal Nutrition (450), University ofHohenheim, 70 593 Stuttgart, Germany, for assistancewith the in vitro and GC analysis.

References

Adamu AM, Russell JR, McGillard AD & Trenkle A (1989)Effects of added dietary urea on the utilization of maizestover silage by growing beef cattle. Anim Feed Sci Technol22, 227 236.

Agriculture and Food Research Committee (1993) Energyand Protein Requirement of Ruminants. An Advisory ManualPrepared by the AFRC Technical Committee on Response toNutrients. Wallingford, Oxon: CAB International.

Aiple KH (1993) Vergleichende Untersuchungen mit Pansensaftund Kot als Inoculum im Hohenheimer Futterwerttest(Comparisons of rumen and faecal inoculum in the HohenheimFeed Evaluation Test). PhD Thesis, Universitat Hohenheim.

Ausschuss fur Bedarfsnormen der Gesellschaft fur Ernahrung-sphysiologie der Haustiere (1986) Energie- und Nahrstoffbe-darf landwirschaftlicher Nutztiere. Nr 3. Milchkuhe undAufzuchtrinder (Energy and nutrient requirement of farm ani-mals). Frankfurt (Main): DLG-Verlag.

Beever DE (1993) Ruminant animal production from forages present position and future opportunities. In Grassland forOur World [M Baker, editor]. Wellington: SIR Publishing.

Beuvink JMW & Spoelstra SF (1992) Interaction betweensubstrate, fermentation end-products, buffering systems and


8/2/2019 0900625

10/10

gas production upon fermentation of different carbohydrates bymixed rumen microorganism in vitro. Appl Microbiol Biotech-nol 37, 505509.

Blummel M, Aiple K-H, Steingass H & Becker K (1999a) A noteon the stoichiometrical relationship of short chain fatty acidproduction and gas formation in feedstuffs of widely differingquality. J Anim Physiol Anim Nutr 81, 157167.

Blummel M & Becker K (1997) The degradability characteristicsof fifty-four roughages and roughage neutral-detergent fibre asdescribed by in vitro gas production and their relationship tovoluntary feed intake. Br J Nutr 77, 757786.

Blummel M, Krishna N & rskov ER (2001) Supplementationstrategies for optimizing ruminal carbon and nitrogen utiliz-ation: concepts and approaches. In Review Papers: 10thAnimal Nutrition Conference, Karnal, India, November 9th to11th 2001, pp. 1023. Karnal, Haryana, India: Animal Nutri-tion Society of India.

Blummel M, Makkar HPS & Becker K (1997) In vitro gas pro-

duction a technique revisited. J Anim Physiol Anim Nutr77, 2434.

Blummel M, Moss A, Givens I, Makkar HPS & Becker K (1999 b)Preliminary study on the relationship of microbial efficienciesof roughages in vitro and methane production in vivo. ProcSoc Nutr Physiol 8, 76 Abstr.

Blummel M & rskov ER (1993) Comparison ofin vitro gas pro-duction and nylon bag degradability of roughages in predictingfeed intake in cattle. Anim Feed Sci Technol 40, 109119.

Brown WF & Pittman WD (1991) Conservation and degradationof nitrogen and fiber fraction in selected tropical grasses andlegumes. Trop Grassl 25, 305.

Clark JH, Klusmeyer TH & Cameron MR (1992) Microbialprotein synthesis and flow of nitrogen fractions to the duode-num of dairy cows. J Dairy Sci 75, 2304.

Cone JW (1998) Influence of protein fermentation on gasproduction profiles. Proc Soc Nutr Physiol 7, 36 Abstr.

Cone JW, Van Geldern AH & Driehuis F (1997) Description ofgas production profiles with a three-phasic model. Anim FeedSci Technol 66, 3145.

Djouvinov DS, Nakashima Y, Todorov N & Pavlov D (1998) Insitu degradation of feed purines. Anim Feed Sci Technol 71,6777.

Fox DG, Sniffen CJ, OConnor JD, Van Soest PJ & Russell JB(1992) A net carbohydrate and protein system for evaluatingcattle diets. III Cattle requirement and diet adequacy. J AnimSci 70, 35783596.

Getachew G, Blummel M, Makkar HPS & Becker K (1998) Invitro gas measuring techniques for assessment of nutritionalquality of feeds: a review. Anim Feed Sci Technol 72,261281.

Goering HK & Van Soest PJ (1970) Forage Fiber Analyses(Apparatus, Reagents, Procedures and Some Applications). Agricutural Handbook no. 379. Washington, DC: USDA-ARS.

Groot JC, Williams BA, Oostam AJ, Boer H & Tamminga S(1998) The use of cumulative gas and volatile fatty acidproduction to predict in vitro fermentation kinetics of Italianryegrass leaf cell walls and contents at various time intervals.Br J Nutr 79, 519525.

Karsli MA (1998) Ruminal microbial protein synthesis in sheepfed forages of varying nutritive value. PhD Thesis, IowaState University.

Leng RA (1993) Quantitative ruminant nutrition a greenscience. Aust J Agric Sci 44, 363 380.

McAllan AB & Smith RH (1973a) Degradation of nucleic acidderivatives in the rumen. Br J Nutr 29, 331342.

McAllan AB & Smith RH (1973b) Degradation of nucleic acid

derivatives by rumen bacteria in vitro. Br J Nutr 29, 467474.Makkar HPS & Becker K (1999) Purine quantification in digesta

from ruminant animals by spectophotometric and HPLCmethods. Br J Nutr 81, 107 111.

Menke KH, Raab L, Salewski A, Steingass H, Fritz D &Schneider W (1979) The estimation of the digestibility andmetabolizable energy content of ruminant feedstuffs from thegas production when they are incubated with rumen liquor.J Agric Sci 93, 217 222.

National Research Council (1996) Nutrient Requirement of BeefCattle. Washington, DC: National Academy Press.

Pell AN, Pitt RE, Doane PH & Schofield P (1998) The develop-ment, use and application of the gas production technique atCornell University, USA. In In vitro Techniques for Measuring Nutrient Supply to Ruminants, pp. 4554 [ER Deaville,

E Owens, AT Adegosan, C Rymer, JA Huntington and TLJLawrence, editors]. Edinburgh: BSAS.

Pirt SJ (1982) Maintenance energy: a general model for energy-limited and energy-sufficient growth. Arch Microbiol 133,300302.

Rojas-Bourillon A, Russell JR, Trenle A & McGillard AD (1987)Effects of rolling on the composition and utilization by growingsteers of whole-plant corn silages. J Anim Sci 64, 303311.

Russell JB, OConnor JD, Fox DG, Van Soest PJ & Sniffen CJ(1992) A net carbohydrate and protein system for evaluatingcattle diets. I Ruminal fermentation. J Anim Sci 70,35513561.

Sinclair LA, Garnsworthy PC, Newbold JR & Buttery PJ (1995)Effects of synchronizing the rate of dietary energy and nitrogenin diets with similar carbohydrate composition on rumen fer-

mentation and microbial protein synthesis in sheep. J AgricSci 124, 463472.

Sniffen CJ, OConnor JD, Van Soest PJ, Fox DG & Russell JB(1992) A net carbohydrate and protein system for evaluatingcattle diets. II Carbohydrate and protein availability. J AnimSci 70, 35623577.

Steingass H & Menke KH (1986) Schatzung des energetischenFutterwertes aus der in vitro mit Pansensaft bestimmten Gasbil-dung und der chemischen Analyse (Estimation of the energycontent of feedstuffs from in vitro gas production and chemicalanalysis). Ubersicht Tierernahr 14, 251270.

Tilley JMA & Terry RA (1963) A two stage technique for thein vitro digestion of forage crops. J Br Grassl Soc 18, 104 111.

Van Soest PJ (1994) Nutritional Ecology of the Ruminant, 2nded., Ithaca, NY: Cornell University Press.

Van Soest PJ & Robertson JB (1985) A Laboratory Manual for Animal Science no. 612. Ithaca, NY: Cornell UniversityPress.

Witt MW, Sinclair LA, Wilkinson RG & Buttery PJ (1999)The effect of synchronizing the rate of dietary energy andnitrogen supply to the rumen on the production and metab-olism of sheep: food characterization and growth and metab-olism of ewe lambs given food ad libitum. Anim Sci 69,223236.

Wolin MJ (1960) A theoretical rumen fermentation balance. JDairy Sci 43, 14521459.

Zinn RA & Owens FN (1986) A rapid procedure for purinemeasurement and its use for estimating net ruminal protein syn-thesis. Can J Anim Sci 66, 157166.


Documents

0900625