Original article

Effect of ruminal inoculation of Isotricha aloneor a mixed B-type fauna in a defaunated rumen

on the digestion of a hay-maizediet (70:30) in sheep

JP Jouany J Sénaud S Toillon M Ben Salah

J Bohatier G Prensien

1 INRA-Theix, SNRH, unité de la digestion microbienne, 63122 Saint-Genès-Champanelle;2 Université Biaise-Pascal, laboratoire de protistologie, UA 138,complexe scientifique des Cézeaux, 63177Aubière cedex;

3 Université dAuvergne, faculté de pharmacie, laboratoire de biologie cellulaire,63001 Clermont-Ferrand, France

(Received 19 October 1993; accepted 22 August 1994)

Summary ― Two adult sheep (75 kg live weight) fitted with rumen cannulas were defaunated by theemptying method during the first period of the experiment. They were inoculated with the genusIsotricha alone during the second period, and with a mixed ciliate population (Entodinium, Eudiplodinium,Epidinium) during the third. They were fed a diet of grass hay (840 g) and pelleted maize grains (360 g)in 8 meals per day, every 3 h. Defaunation was successful and no accidental contamination occurredduring the experiment. The protozoa had no significant effect on the volume of rumen digesta, noron the turnover of the particulate phase. The addition of Isotricha and of the mixed fauna increased theADF digestibility of the diet but, in the same animals, lowered the in sacco degradation of wheat straw.The ruminal pool sizes of dry matter (DM), organic matter (OM), nitrogen (N), neutral detergent fibre(NDF), acid detergent fibre (ADF) and acid lignin detergent (ADL) remained unchanged after protozoainoculations. The concentration of total volatile fatty acids (VFA) was not altered by faunation withIsotricha or a mixed fauna. The molar proportion of acetate increased at the expense of all the otherVFAs (mainly propionate with the mixed fauna). Correspondingly, the proportion of methane in therumen gases increased and that of C02 decreased in inoculated animals. The ammonia concentrationwas highest in animals with a mixed fauna and lowest in those inoculated with Isotricha alone. This trendis explained in terms of the specific effect of the different genera of protozoa on nitrogen metabolism.

rumen / protozoa / digestion / kinetics of digesta

Résumé ― Effet de l’inoculation d’Isotricha ou d’une faune conventionnelle de type B dansun rumen de mouton défauné sur la digestion d’un régime mixte foin-maïs (70/30). Deux mou-tons adultes (75 kg de poids vif) porteurs de fistule permanente du rumen ont été défaunés par la

méthode de vidage au cours de la première période. Leur rumen a été ensuite inoculé avec le seul genreIsotricha pendant la deuxième période et avec une faune mixte de type B au cours de la troisièmepériode. Ils ont reçu une ration composée de foin de Graminées (840 g) et de grains de mai’s agglo-mérés (360 g) distribuée en 8 repas par jour, toutes les 3 h. La refaunation du rumen n’a pas modifiéle volume total des digesta dans le rumen, ni la vitesse de renouvellement de la phase solide. Lamasse totale de MS, MO, N, NDF, ADF, ADL présente dans le rumen n pas non plus été affectée parla présence des protozoaires ciliés. La digestibilité de la lignocellulose de la ration (fraction ADF) a étéaccrue (P < 0,05) par l’addition dlsotricha et d’une faune mixte dans le rumen bien que la dégradationin sacco d’une paille de blé ait été plus faible chez ces mêmes animaux. La concentration totale desAGV n a pas été modifiée par l’introduction d1sotricha ou d’une faune mixte dans le rumen ; la part del’acétate dans le mélange des AG V a augmenté aux dépens de celle des autres acides. Corrélative-ment, la proportion de méthane dans les gaz a augmenté. La concentration en N-NH3 a été la plus éle-vée chez les animaux à faune mixte et la plus faible chez les animaux mono-inoculés avec Isotricha.Des explications sur l’évolution de ce dernier paramètre sont données en relation avec l’action spéci-fique du genre Isotricha sur le métabolisme azoté dans le rumen.

rumen / protozoaires / digestion / cinétique des digesta

INTRODUCTION

It is now generally agreed that ciliate rumenprotozoa have a negative effect on nitrogenuse by ruminants. In contrast, the results con-cerning their action on cellulolysis are con-flicting (Bird and Leng, 1984; Romulo et al,1986). This is an important issue since forageis the basis of ruminant diet. It can be of vital

importance in countries where the nutritionalvalue of roughage is low and where energyand nitrogen enrichment of the basic dietcannot be readily achieved on a wide scale.

The present study investigates the actionof rumen protozoa in cellulolysis and in pro-teolysis; it is novel in that the main physio-logical, digestive and metabolic parameterswere measured in the rumen of animals thatwere successively defaunated, monoino-culated with Isotricha sp and inoculated witha conventional mixed fauna. The overall

effect of the protozoa was assessed by thedifferences recorded between the faunatedand defaunated states. The intermediatemonoinoculated state provided a means ofstudying the effects of lsotricha sp, whichdevelop in the molasses-rich feed used byBird and Leng (1978, 1984) and Bird (1989,1991 It was with such a feed that theseauthors observed a negative effect of pro-

tozoa on cellulolysis, while most of the testscarried out in Europe, North America andJapan have yielded opposite results on ani-mals harbouring a population rich in

Ophryoscolecidae (Jouany et al, 1988).Furthermore, Jouany et al (1992) observedin vitro that addition of Isotricha alone in adefaunated rumen decreased the proteindegradation while the degradation wasincreased with a mixed-fauna inoculation.It seems therefore that Isotricha, which hasrarely been studied as the only protozoa invivo, have a specific action in rumen diges-tion. In addition, Jouany et al (1981 ) showedthat the introduction of Isotricha alone intodefaunated sheep rumens decreased lig-nocellulose digestion in a mixed feed basedon dehydrated alfalfa. It was thus decided to

perform measurements of the dynamics ofdigestion as a complement to the standardmeasurements to know how Isotrichabehaves in a grass-hay-based diet.

MATERIALS AND METHODS

Animals

Two castrated adult male Texel sheep weighing73 kg (sheep A) and 76 kg (sheep B), fitted with

large rumen fistula (75 mm in diameter), wereused for 3 measurement periods. Their rumenswere defaunated according to the procedure ofJouany and Sénaud (1979a) during the firstperiod, and measurements were started 60 dlater. They were then inoculated with the singlegenus Isotricha as described by Jouany et al(1981) at the beginning of the second period.Measurements were started 30 d after stabiliza-tion of the lsotricha population. Finally they wereinoculated with a mixed inoculum comprising thegenera Entodinium, Epidinium, and Eudiplo-dinium, in addition to Isotricha. This inoculumwas characterized as a B-type according to Eadie(1962). Measurements were started 30 d afterstabilization of the ciliate population.

The sheep were housed in a building set apartfrom other animal facilities and specially designedso that was no inter-animal contact. In order toavoid contamination the staff were trained to han-dle the sheep that had been defaunated and sub-sequently refaunated with a single protozoalspecies or with a defined mixed protozoal popu-lation (see Jouany, 1978).

The welfare of the animals was strictlyrespected throughout the experiment, which wasperformed in accordance with the regulations laiddown by French law.

Feed

The sheep were fed daily 840 g of chopped nat-ural grass hay and 360 g of pelleted maize grainsdistributed semi-continuously by means of anautomatic feeder (8 equal meals per day, every3 h). Dietary concentrations of organic matter(OM), nitrogen (N), neutral detergent fibre (NDF),acid detergent fibre (ADF) and starch were 91.8,1.3, 49.9, 29.6 and 21.2% of the dry matter (DM)respectively.

Measurements

Characterization of rumen digesta

The volume and weight of the rumen contentswere measured after all rumen digesta were col-lected by completely emptying the rumen at14.00 h, between 2 meals. Pools of DM, OM andcell-wall components were assayed twice per

period for each sheep on a representative sampleof digesta.

Breakdown of cell-wall components

The digestibility of the diet and cell-wall compo-nents was calculated by measuring feed intakeand collecting faeces for 7 d.

In sacco degradation of wheat straw

Wheat straw is considered to be a substrate com-

posed almost exclusively of pure cell walls. Toavoid physical losses from bags it was ground ina mill fitted with 2 mm grating and then screenedso as to retain only particles larger than 200microns. Nylon bags 5.5 x 12 cm in size (AnkomCo, Fairport, NY, USA), made of a woven fabric95 microns in mesh size and containing 3 g ofstraw, were left for 6, 12, 18, 24, 48, 72 and 96 hin the rumen. Two kinetic runs with 2 bags pertime and per run were carried out in each sheep.The rates and the levels of straw degradation(parameters a, b and c) were estimated by plottingthe curve of DM disappearance according to themodel of PJrskov and Mc Donald (1979).

Measurement of feed retention timein the digestive tract

A fraction of the hay from the diet was washedin a domestic washing machine with a commercialwashing soap to eliminate cytoplasmic contentsand to obtain a residue containing about 90%NDF. It was then labelled with ytterbium (Yb) bysoaking for 20 h in a solution of Yb acetate (50 mgYb g-1 hay); it was then thoroughly washed withrunning water to eliminate unbound Yb. The haythus treated contained about 25 mg Yb g-1 DM.The exact content of Yb was then determined.

Exactly 20 g of labelled hay was introducedvia the cannula into the rumen and an aliquot ofexcreted faeces was taken every 4 h for 48 h,then every 8 h for the next 36 h, and finally every12 h up to 168 h. The faecal Yb excretion curvewas plotted and modelized according to the modelof Grovum and Williams (1973), Faichney (1986),and Thielemans et al (1978) to determine theretention time of solid particles in the differentdigestive compartments.

The liquid phase was labelled by addition tothe rumen of a 100 ml dose of a 20% (w/v) solu-

tion of polyethylene glycol (PEG 4000). A 200 mlsample of rumen content was taken with a 12 mmdiameter tube connected to a rubber syringe 2, 4,6, 8, 12, 16, 20, 24, 32, and 48 h afterwards.Twenty millilitres were used for PEG assay andthe remainder was returned to the rumen. Theretention time (Th) and turnover rate (k 1/Th)of the liquid phase was calculated by the equatjionC = Co e -kr.

Measurement of fermentation end

products and the number of protozoaand bacteria in the rumen

Samples of 200 ml rumen fluid were taken on 3successive days in each of the sheep halfwaybetween 2 successive meals, 5 times in thecourse of the day. About 15 ml was used to countprotozoa and to estimate the bacterial popula-tion, to assay microbial metabolites, such as VFA,ammonia-N (NH3-N), and to record pH; theremainder was returned to the rumen. Gaseswere sampled, at the same times as above, witha rumen cannula specially designed for this typeof measurement (Jouany and S6naud, 1979b)that was fitted the day before to avoid air con-tamination of the rumen gas phase.

Experimental design

Each sampling period started after rumen con-ditions were stabilized (60 d after defaunationwas achieved for period 1; 30 d after the Isotrichasp population had become stable for period 2;30 d after the mixed fauna had established for

period 3). The periods lasted at least 4 weeks.The 2 in sacco kinetic runs were carried out inthe first week. The distribution of transit markersand collection of faeces for the measurement ofretention time and digestibility were made duringthe second week. pH measurements and micro-bial metabolite assays in the rumen were per-formed on 3 successive days during the thirdweek. The rumen was emptied twice during thefourth week, on Tuesday and Friday.

Assays

DM was determined by oven drying at 80°C for 48h. OM was measured after ashing in a muffle fur-

nace (550°C for 6 h). Nitrogen was assayed bythe method of Kjeldahl (AOAC, 1965) automatedwith the Kjeltec (Tecator, Paris, France). Cell-wall components were analyzed by the method ofVan Soest and Wine (1967) automated with theFibertec (Tecator, Paris, France) after starch hadbeen eliminated by a glucoamylase treatment.The method of Thivend et al (1972) was used forstarch assay. Ytterbium was analysed by flamespectrophotometry according to the method ofSiddons et al (1985). PEG was determined byturbidimetry (Hyden, 1955).

Volatile fatty acids were assayed by themethod of Jouany (1982). Samples of rumen fluid(5 ml) containing 5% (w/v) solution of orthophos-phoric acid were kept at -15°C. Two milliliters ofrumen fluid were mixed with 8 ml of a 12.5% (w/v)NaCl solution and stored at -15°C for determi-nation of NH3-N by a modified version of themethod of Berthelot (Na salicylate - dichloroiso-cyanurate reactants were used instead of phenol- Na hypochlorite) with a Technicon autoanalyser(Van Eenaeme et al, 1969). Gases were anal-ysed by gas-solid chromatography (Jouany andS£naud, 1 978).

The protozoa were counted under a binocularmagnifier (x 80) in Dollfuss cells according to themethod described by Jouany (1978) in rumenfluid obtained by filtration through metal gauze(1 mm mesh size). Total protozoa volume wascalculated from an assumed volume of about0.5 x 106 )im3 for each cell of Isotricha, Epidiniumand Eudiplodinium and 2.0 x 104 pm3 for smallentodinia (Warner, 1962). The correspondingmass of protozoa was established on the basisthat 1.2 x 105 large ciliates (Isotricha, Epidinium,Eudiplodinium) contained approximately 1 mg N(Leng et al, 1981) and that the mean content of Nin their DM was close to 6.0% (Jouany andThivend, 1972). We also assumed that the vol-ume, and therefore the weight, of small entodiniawas 100 times less that of the large ciliates.According to these estimations, we calculatedthat 9 mg DM of ciliates was reached with 105

large ciliates and 10! small ciliates.The total numbers of bacteria were determined

by direct counts after staining with orange acridineunder epifluorescent microscopy (Prevot et al,1988).

Statistical analysis

As it is not possible to remove ciliates by mildtreatment, the experimental design had to follow

the sequential order of defaunation, refaunationwith only one species of protozoa, and faunationwith a mixed fauna. A latin-square treatment wastherefore inappropriate. Because of the difficultyof carrying out such an experiment with differentfauna (including a single species inoculation),only 2 animals were used. Each was consideredas its own control. Comparisons between the 3different periods were made for each animal. In naddition, samplings were several times to obtainmore precise values of the means.

Data were treated by an Anova procedure(SAS, 1985). The model was as follows: Yi = p +Ai+ T lij, where Yij = trait observed, A; = animaleffect, Ti= protozoa effect and Lij = error,assumed to be the interaction between animaland protozoa effects.

RESULTS

No accidental contamination of the rumenfauna occurred during any of the 3 experi-mental periods.

There was no significant ’animal x pro-tozoa’ interaction (p > 0.05) in any of the

digestive parameters tested. Hence, onlyone mean value is given for each periodand comparisons were made between the 3periods.

Establishment of lsotricha spand mixed fauna in sheep rumen

The presence of Isotricha in the rumen of

defaunated sheep was detected (minimalconcentration = 50 cells ml-1) 2 d after theintroduction of 250 ml of rumen contents

containing 4 x 103 ciliates ml-!, sampledfrom a monoinoculated sheep previouslyprepared according to Jouany (1978). Theconcentration peaked by day 14 (3 700Isotricha sp ml-! remained at this level forabout 20 d, and then settled down at a valueclose to 2 000 ml-1 (table I).

At the beginning of the third period, ani-mals were inoculated with 250 ml of mixed

rumen contents sampled from sheepmonoinoculated with Entodinium sp (75 000ml-1), Epidinium sp (3 500 ml-1), andEudiplodinium sp (30 000 ml-1). The intro-duction of this mixed inoculum into therumen of sheep monoinoculated withIsotricha spp resulted in a complex faunawhose population could be followed imme-diately. Peak values were reached betweenday 7 and day 14 for all the ciliates: 3 000Isotricha, 2 000 Epidinium, 40 000 Eudiplo-dinium, 32 000 Entodinium ml-1 (fig 1 TheEntodinia reached the highest number atabout day 50, attaining a level of more than105 ml-1, which accounted for about 75%of the total population, while the numbersof Epidinium dropped from day 35 (50 - 400Epidinium ml-1). The population of Eudiplo-dinium fluctuated between 104 and 4 x 104ml-!. The population of Isotricha spp, whenmixed with the other ciliates, stabilized atthe same level (about 3 000 ml-1) as itreached when alone in the monoinoculated

sheep (figs 1, 2; table I).

Variations in protozoa populationsduring the day

Although the sheep were fed every 3 h(pseudo-steady state of the rumen), the con-centrations of the protozoa varied over the24 h cycle. The Isotricha sp populationincreased 1.5 h after each feeding anddecreased just before feed was distributed.

Besides the variations linked to feeding,we observed nycthemeral variations inIsotricha concentration that may have beendue to sequences of cell division. There wasno significant variation between 9.00 h and17.00 h, but a decrease occurred at 21.00 hand there was a significant increase at1.00 h.

The concentrations of entodiniomorphs inthe rumen of sheep inoculated with themixed fauna slightly decreased immediatelyafter feeding and reverted to their initial

values just before the following feeding. Thegeneral time course of Isotricha in animalsinoculated with the mixed fauna was thesame as that observed during the periodwhen the sheep were monoinoculated withthis genus.

Biomass of protozoa and concentrationof bacteria according to the natureof the fauna

Both the volume and the mass of the cili-

ates in the rumen of the sheep with mixedfauna were about 10-fold greater than in therumen of the monoinoculated sheep (tableI). In the latter, the biomass of the ciliatesranged between 3.1 and 4.0 g 1-1.

The number of total bacteria was close to3.4 x 109 ml-1 during the defaunated period.It stabilized at 1.7 x 109 ml-1 duringmonoinoculation with Isotricha or inocula-tion with a mixed conventional fauna. Giventhe large variations within each period, thedifferences between periods were not sta-tistically significant.

Effect of protozoa on physiological andphysical characteristics of rumen digesta

The volume and weight of total rumen con-tents were not significantly influenced bythe rumen fauna (table II). Consequently,there was no difference in the specific weightof rumen content between periods. The

rumen pool sizes of DM, OM, N, NDF, ADF,acid detergent lignin (ADL) were unchangedafter the inoculation of ciliates into the defau-nated rumens.

End products of rumen fermentation

The pH of the rumen contents and total VFAconcentrations were not significantly modifiedby the presence of either of the fauna typesstudied (table 111). The increase in the pro-portion of acetate in the VFA mixture afterthe introduction of the mixed fauna (+9.4%)in the defaunated rumen was lower after

inoculation of Isotricha (+4.5%). This ten-dency towards an increase is consistent withthe greater amount of methane (+6%) pre-sent in the rumen gases. The proportion ofpropionate decreased in sheep inoculatedwith the mixed fauna (-40%). Addition ofIsotricha to the defaunated rumens had noeffect on propionate molar proportion. Cor-respondingly, the proportion of C02 in the

rumen gases was slightly lower in the mixedfaunated sheep (-4%). The butyrate pro-portion remained unchanged after the ino-culation of protozoa in defaunated animals.The sum of branched VFA and valerate and

caproate was lowest in Isotricha-inoculatedanimals.

Ammonia N concentration in the rumen

fluid increased (+49%) after the addition ofmixed fauna to the defaunated rumens anddecreased (-29%) after inoculation ofIsotricha alone.

In sacco degradation of wheat straw

The addition of Isotricha or a mixed fauna tothe defaunated rumens did not modify therapidly degradable fraction (fraction ’a’) ofthe DM of wheat straw (table IV). The poten-tially digestible fraction (fraction ’b’) wasgreatest in the defaunated sheep. It fell inthe mixed-fauna animals while it reached

an intermediate value in the presence of

Isotricha. Compared to that of defaunatedanimals, the degradation rate (fraction ’c’)was higher in the presence of a mixed fauna(+54%), and lower in the Isotricha-monoinoculated animals (-25%).

Point-by-point analysis of the kineticsshows that the positive effect of defauna-tion on DM breakdown of wheat straw was

only significant after 12 h retention in therumen.

Kinetics of soluble and particulatemarkers

The volume of liquid in the rumen, as eval-uated from the marker concentration at ’timezero’ (ie the time the dose of marker wasintroduced) obtained by calculating the inter-section of the decay curve of the markerconcentration with the y-axis, was unaf-fected by the introduction of Isotricha indefaunated rumen, as confirmed by theemptying method (table V). In contrast, it

was lowered (-15%) by the addition of themixed fauna while with the emptyingmethod this type of fauna had no discernibleeffect.

The retention time of the particulatephase in the different parts of the digestivetract calculated from the 2-mixing-compart-ment-one-delay model of Grovum andWilliams (1973) indicated that the time delay,

which reflects retention in the tubular partof the tract (omasum, abomasum, smallintestine), was not influenced by the additionof the mixed fauna to the rumen of defau-nated or monoinoculated sheep. There wasno difference in the mean retention time ofsolid particles (34.2 ± 5.3 h) in compartment1 (taken to be the rumen) between the 3experimental periods, nor in that (16.6 ± 0.4h) of compartment 2 (taken to be the cae-cum and proximal colon).

Digestibility of the diet throughoutthe digestive tract

Protozoa had no influence on the digestibil-ities of DM, OM, or N (table VI). Both typesof protozoal inoculations (Isotricha alone ormixed fauna) increased the digestibility ofthe lignocellulose fraction (ADF) of the feed

(+14% respectively). The positive effect ofprotozoa on the digestibility of the total cell- l-

wall components (NDF) was only significantin the mixed faunated animals.

DISCUSSION

In agreement with previous studies ofS6naud et al (1980), Groliere et al (1980)and Grain et al (1980), our results showedthat ciliates become rapidly established in arumen that has been defaunated. The pop-ulations of Ophryoscolecidae and Isotrichareached their maximum size after 7 and14 d, respectively. As observed by the pre-vious authors, a stable plateau was obtainedsome 20 d after this peak, but at a lowerlevel.

After the inoculation of the mixed fauna,the numbers of Isotricha did not decreasefrom their previous level, which proves thatthere was no interaction between the 2 typesof ciliates. An absence of competitionbetween Isotricha and Entodinium was pre-

viously observed by Groli6re et al (1980) ina mixed diet similar to that used in our study.The only interaction observed by theseauthors was slight competition betweenIsotricha and a type-A fauna (Polyplasfron),

which was unfavourable to the former. This

suggests that Isotricha behaves differentlytowards type-B and type-A ciliates.

Although the animals were fed 8 equalmeals during the day, the populations ofprotozoa were not in steady-state condi-tions. The decrease in Isotricha numbers

noted before each meal is consistent withthe results of Abe et al (1981 ) and Murphy etal (1985) obtained in cows and steersrespectively, fed 2 meals per day. Theauthors explained that Isotricha rapidlymigrate from the reticulum towards therumen in response to a chemical stimulus

originating from dietary soluble compounds.The decrease in entodiniomorphid ciliatesafter feed intake owing to their dilution in a ahigher volume of digesta is in good agree-ment with the previous results of Jouany etal (1973). The increase in these ciliates thatoccurred at 1.00 h could have been due tothe high ruminating activity of animals from21.00 h and during the beginning of the nightwhich made more substrates available torumen microbes even when animals were

continuously fed (S6naud et al, 1977).The biomasses of protozoa in the rumen

of mixed-faunated sheep (3.1 and 4.0 g 1-1)were close to those recorded by Jouany(1978) who used a method based on cen-trifugations to isolate protozoa: 1.5 g I-1

for a diet based exclusively on grass hay;3.6 g I-1 for a hay-barley diet; 5.0 gl -1 for a

hay + Jerusalem artichoke diet, and 3.7 g 1-1for a diet of hay and sugar beet.

Although the statistical threshold was notreached, the tendency for the number ofbacteria to decrease after ruminal inocula-tion of protozoa, whatever the type and levelof the protozoa, confirms the negative cor-relation between bacterial and protozoaldensities previously reported by Teather etal (1984). The variations in bacteria con-centration we observed from one day toanother and between animals were proba-bly due to the method of direct counts used.

The variance analysis of kinetics of fae-cal output of markers used here to evalu-ate the dynamics of particles in the digestivetract indicated that animal effect was higherthan protozoa effect. Differences betweenthe models of Faichney and Thielemanswere slight while the Grovum and Williammodel gave higher values for digestiveretention times of particles (+10 to +16%).The kinetics of liquid phase estimated byPEG dilution rate were influenced by pro-tozoa, the highest rate being obtained withfaunated animals which had also the low-est rumen volume of liquid. It must be notedthat the emptying method showed no effectof protozoa on rumen volume.

The absence of any difference in themeasured pool sizes of rumen digesta fol-lowing the inoculation of protozoa is anindication that protozoa had no significanteffect on quantitative aspects of ruminaldigestion, since the retention time of solidparticles was not changed by the inocula-tion of protozoa. According to Demeyer(1987), there are great variations in thedynamics of digesta in the rumen betweenanimals in the response to defaunation;sometimes the outflow rate of digestaincreases and the rumen volume

decreases; sometimes opposite results arerecorded. As in our experiment, Faichneyand Griffiths (1978), Orpin and Letcher

(1983-84), and Punia et al (1987) notedno effect of protozoa on the volume of therumen nor on the mean retention time of

particulate phase in the rumen.There is still a wide debate about the real

impact of refaunating a defaunated rumenon the digestion of plant cell walls. Moststudies made during the last 15 yearsshowed and increase in cell-wall carbohy-drate digestion after inoculation of protozoainto defaunated rumens (see Jouany, 1989,1991 ). A few inverse results were obtainedin Australia (Bird and Leng, 1984; Romulo etal, 1986).

Our results clearly show that the inocu-lation of a single species or a mixed speciesof protozoa into defaunated rumens

improved the digestion of the lignocellulosicfraction (ADF) of the diet more than thedigestion of the total cell-wall fractions(NDF). This indicates that the role of proto-zoa was more important for the digestion ofcellulose than hemicellulose fractions, whichis in disagreement with the proposal ofUshida et al (1991 Consequently, we sug-gest that rumen ciliates have an effect onall the plant cell-wall fractions (hemicellu-lose or lignocellulose) and that the effects onthe different components of cell wallsobserved depend on the chemical organi-zation of the complex matrix of the cell-wallcarbohydrates.

The positive effect of Isotricha or a mixedfauna on the digestion of the lignocellulosein the whole digestive tract we observedhere is difficult to relate to the decrease inthe potentially digestible fraction of wheatstraw estimated in sacco in the rumen ofthe refaunated sheep. This indicates thatthe same substrate has to be used for insitu and in vivo digestion trials. Jouany(1989) discussed the ’substrate effect’ oncell-wall digestion in relation to the responseof protozoa. He observed that wheat straw,unlike other roughages, is not a suitablesubstrate to study the effect of protozoa oncellulolysis. Also, as suggested by Trabalza

Marinucci et al (1992), in sacco degrada-tion may not reflect the real in situ degra-dation. However, if we suppose that in saccotrials were not biased, it follows that the

lower digestion of cell-wall carbohydratesin the rumen of faunated animals was morethan compensated for by a higher digestionin the hindgut of the same animals. How-ever the reasons for such a shift in digestionwhich goes beyond mere compensation aredifficult to explain.

The absence of any significant differencein VFA concentration and pH values afterinoculation of protozoa in defaunatedrumens indicates that protozoa had no greateffect on the amount of fermented organicmatter. It is noteworthy that no increase inthe molar proportion of butyrate wasobserved after refaunation which is in con-trast to findings reported in most previousreports (see Jouany, 1991 This trend ispartly explained by the increase in theacetate proportion in faunated sheep sinceacetate is considered to be a major precur-sor of butyrate during rumen fermentations(Russell and Wallace, 1988).

The higher NH3-N concentration in the,rumen of animals faunated with the mixed

B-type fauna can be accounted for by theinvolvement of the protozoa in feed andmicrobial protein breakdown, if we considerthat the use of NH3-N for microbial synthesisis not affected by the presence of ciliates asindicated by ltabashi and Kandatsu (1975).Interestingly, the presence of the genusIsotricha alone caused a fall (-30%) in rumenNH3-N concentration. These results are inagreement with a previous in vitro experi-ment (Jouany et al, 1992) in which defau-nation decreased the degradation of spar-ingly soluble protein in the rumen, and theaddition of Isotricha sp to defaunated rumensresulted in an even greater decrease in pro-tein degradation up to the NH3-N stage, irre-spective of protein solubility. This effect ofIsotricha was confirmed by the decrease weobserved in branched-chain VFA and

caproate and valerate concentrations in therumen of /sotricha-inoculated animals com-

pared to the same concentrations in defau-nated and mixed-faunated animals.

This study shows that when protozoa areadded to the rumen, cellulolysis and prote-olysis are affected. It confirms the greaterdigestibility of cell-wall carbohydrates in thepresence of protozoa along the whole diges-tive tract and explains why the degradationin sacco of wheat straw is lower in the sameanimals. Isotricha had no specific effect oncellulolysis, which is at variance with ahypothesis we previously made from resultsof Bird and Leng (1984) and Romulo et al(1986). It induced changes similar to thoseobserved with mixed fauna but to a lesserextent because it had a smaller biomass.The specific role of Isotricha in preservingprotein degradation in the rumen must beunderlined.

REFERENCES

Abe M, Iriki T, Tobe N, Shibui H (1981) Sequestration ofholotrich protozoa in the reticulo-rumen of cattle.Appl Environ Microbiol4l , 758-765

AOAC (1965) Official Methods of Analysis of the Asso-ciation of Official Agricultural Chemists, l0th EditionAssociation of Official Analytical Chemists. (W Hor-/<ssoc/ahon of 0!/c/a/ !natyf/caf Cnam/srs. (W Hor-witz, ed). Washington, USA, 15-17 7

Bird SH (1989) Production from ciliate-free ruminants. In:The Roles of Protozoa and Fungi in Ruminant Diges-tion, (JV Nolan, RA Leng, DI Demeyer, ed). Penam-bul Books, Armidale, NSW, Australia, 233-246

Bird SH (1991) Role of protozoa in relation to the nutri-tion of the host animal. In: Recent Advances on theNutrition of Herbivores, (YW Ho, MK Wong, N Abdul-lah, ZA Tajuddin ed), Malaysian Society of AnimalProduction, Vinlin Press, Kuala Lumpur, Malaysia,171-180

Bird SH, Leng RA (1978) The effects of defaunation ofthe rumen on the growth of lambs on low-proteinhigh-energy diets. Br J Nutr 42, 81-87

Bird SH, Leng RA (1984) Further studies on the effectsof the presence or absence of protozoa in the rumenon live-weight gain and wool growth of sheep. BrJ JNutr52, 607-611 1

Demeyer DI (1987) Interd6pendence des effets de ladefaunation sur I’activit6 muralytique, le volume et la

cin6tique du contenu de rumen. R6sultats pr6limi-naires et hypotheses. Reprod NutrDev27, 161-162

Eadie JM (1962) Inter-relationships between certainrumen ciliate protozoa. J Gen Microbiol 29, 579-588

Faichney GJ (1986) The kinetics of particulate matterin the rumen. In: Control of Digestion and Metabolismin Ruminants (LP Milligan, WL Grovum, A Dobson,ed), Reston Books, Prentice-Hall, Englewood Cliffs,NJ, USA 173-195

Faichney GJ, Griffiths DA (1978) Behaviour of soluteand particle markers in the stomach of sheep givena concentrate diet. BrJ Nutr40, 71-82

Grain J, S6naud J, Grolibre CA, Jouany FP (1980)Implantation et d6veloppement des populations deprotozaires cili6s (Polyplastron multivesiculatum,Entodinium sp, Isotricha prostoma) dans le rumen demoutons recevant diff6rents regimes alimentaires.III Considerations g6n6rales sur 1’6cosyst6me rumen.Protistologica 16, 501-506

Grolibre CA, S6naud J, Jouany JP, Grain J, De Puy-torac P (1980) Implantation et d6veloppement despopulations de protozoaires ciii6s (Polyplastron mul-tivesiculatum, Entodinium sp Isotricha prostoma)dans le rumen de moutons recevant diff6rents

regimes alimentaires. II Regimes a base de foinenrichis en c6r6ales. Protistologica 16, 385-394

Grovum WL, Williams VJ (1973) Rate of passage ofdigesta in sheep. 3. Rate of passage of water and drymatter from the reticulo-rumen, abomasum and cae-cum and proximal colon BrJNutr30, 313-329

Hyden S (1955) A turbidimetric method for the determi-nation of higher polyethylene glycols in biologicalmaterials. Ann K Lantlrh6ghs 22, 139-145

Itabashi H, Kandatsu M (1975) Influence of rumen ciliateprotozoa on the concentration of NH3 and VFA inconnection with the utilization of NH3 in the rumen.

Jpn J Zootech Sci 46, 409-416 6

Jouany JP (1978) Contribution a I’dtude des protozoairescili6s du rumen : leur dynamique, leur r6le dans ladigestion et leur int6r6t pour le ruminant. Doctoralthesis, Université Clermont II, France

Jouany JP (1982) Volatile fatty acid and alcohol deter-mination in digestive contents, silage juices, bacte-rial cultures and anaerobic fermentor contents. Sci

Alim 2, 131-144

Jouany JP (1989) Effects of diet on populations of rumenprotozoa in relation to fibre digestion. In: The Rolesof Protozoa and Fungi in Ruminant Digestion (JVNolan, RA Leng, DI Demeyer, eds), PenambulBooks, Armidale, NSW, Australia, 57-79

Jouany JP (1991) Rumen Microbial Metabolism andRuminant Digestion (JP Jouany, ed), INRA Editions,Science Update, Versailles, France, pp 374

Jouany JP, Thivend P (1972) Evolution postprandialede la composition glucidique des corps microbiens durumen en fonction de la nature des glucides du

regime. I. Les protozoaires. Ann Biol Anim BiochimBiophys 12, 673-677

Jouany JP, S6naud J (1978) Utilisation du monensindans la ration des ruminants. II. Effets sur les fer-mentations et la population microbienne du rumen.Ann Zootech 27, 61-74

Jouany JP, S6naud J (1979a) Defaunation du rumen demouton. Ann Biot Anim Biochim Biophys 19, 619-624

Jouany JP, S6naud J (1979b) Description d’une tech-nique permettant d’effectuer des prelevementsrepetes de gaz dans le rumen. Ann Biol AnimBiochim Biophys 19, 1007-1010 0

Jouany JP, Senaud J, Grain J, De Puytorac P (1973)Observations sur la dynamique des populations decil!6s oligotriches dans le rumen du mouton. AnnBiol Anim Biochim Biophys 13, 741-745

Jouany JP, Zainab B, S6naud J, Groli6re CA, Grain J,Thivend P (1981) Role of the rumen ciliate protozoaPolyplastron multivesiculatum, Entodinium sp andIsotricha prostoma in the digestion of a mixed diet insheep. Reprod Nutr Dev 21, 871-884

Jouany JP, Demeyer DI, Grain J (1988) Effect of defau-nating the rumen. Anim Feed Sci Technol 21, 229-265

Jouany JP, Ivan M, Papon Y, Lassalas B (1992) Effectsof Isotricha, Eudiplodinium, Epidinium+ Entodiniumand a mixed population of rumen protozoa on thein vitro degradation of fishmeal, soybean meal andcasein. Can JAnim Sci 72, 871-880

Leng RA, Gill M, Kempton TJ et al (1981 ) Kinetics oflarge ciliate protozoa in the rumen of cattle givensugar cane diets. BrJNutr46, 371-384

Murphy MR, Drone PE, Woodford ST (1985) Factorsstimulating migration of holotrich protozoa into the.rumen. Appl Environ Microbiol49, 1329-1331

Orpin CG, Letcher AJ (1983-84) Effect of absence ofciliate protozoa on rumen fluid volume, flow rate,and bacterial population in sheep. Anim Sci Technol10, 145-153

Orskov ER, Mc Donald IW (1979) The estimation of pro-tein degradability in the rumen from incubation mea-surements weighted according to rate of passage.J Agric Sci (Camb) 92, 499-503

Prevot S, Senhaji M, Bohatier J, S6naud J (1988) Comp-tage par 6pifluorescence des bactéries du rumen,cultivdes in vitro. Estimation de leur 6tat physi-ologique. Reprod Nutr Dev 28, 137-138

Punia BS, Leibholz J, Faichney GJ (1987) The role ofrumen protozoa in the utilization of paspalum (pas-palum dilatum) hay by cattle. Br J Nutr 57, 395-406

Romulo B, Bird SM, Leng RA (1986) The effects ofdefaunation on digestibility and rumen fungi countsin sheep fed high-fibre diets. Proc Austr Soc AnimProd 16, 327-330

Russell JB, Wallace RJ (1988) Energy yielding and con-suming reactions. In: The Rumen Microbial Ecosys-

tem. (PN Hobson, ed), Elsevier Applied Science,London, UK, 185-215 5

SAS (1985) SAS User’s guide: Statistics. SAS Institute,Cary, NC, USA

S6naud J, Jouany JP, Grain J, De Puytorac P (1977)Influence de la frequence des repas sur le com-portement alimentaire du mouton et sur les varia-tions de densite des populations de cili6s au cours dunycth6mbre. Ann Biol Anim Biochim Biophys 17,567-572

Sdnaud J, Groli6re CA, Zainab B, Grain J, Jouany JP(1980) Implantation et d6veloppement des popula-tions de protozoaires cili6s (Polyplastron multi-vesiculatum, Entodinium sp, Isotricha prostoma)dans le rumen des moutons recevant diff6rents

regimes alimentaires. I. Regime luzeme-orge (40/50).Protistologica 16, 329-337

Siddons RC, Paradine J, Beever DE, Cornell PR (1985)Ytterbium acetate as a particulate-phase-digesta-flow marker. BrJNutr54, 509-519 9

Teather RM, Mahadevean S, Erfle JD, Sauer FD (1984)Negative correlation between protozoal and bact-erial levels in rumen samples and its relation to thedetermination of dietary effects on the rumen micro-bial population. Appl Environ Microbiol47, 566-570

Thielemans RF, Frangois E, Bodart C, Thewis A (1978)Mesure de transit gastro-intestinal chez le porc a

I’aide de radiolanthanides. Comparaison avec lemouton. Ann Biol Anim Biochim Biophys 18, 237-247

Thivend P, Mercier C, Guilbot A (1972) Determination ofstarch with glucoamylase. In: Methods in Carbohy-drate Chemistry, vol Vi (RL Whistler, JN Bc Millered), Academic Press, New York, USA 100-105

Trabalza Marinucci M, Dehority BA, Loerch SC (1992) Invitro and in vivo studies of factors affecting diges-tion of feeds in synthetic fiber bags. J Anim Sci 70,296-307

Ushida K, Jouany JP, Demeyer Di (1991) Effects ofpresence or absence of rumen protozoa on the effi-ciency of utilization of concentrate and fibrous feeds.In: Physiological Aspects of Digestion andMetabolism in Ruminants (T Tsuda, Y Sasaki, RKawashima, ed) Academic Press, Tokyo, Japan,625-654

Van Eenaeme C, Bienfait JM, Lambot 0, Pondant A(1969) Determination automatique de I’ammoniaquedans le liquide du rumen par la m6thode de Berthelotadapt6e 6 I’autoanalyzer. Ann Med Vet7, 419-425

Van Soest PJ, Wine RH (1967) Use of detergents in theanalysis of fibrous feeds, II. A rapid method for thedetermination of fibre and lignin. JA O A C 50, 50-55

Warner ACI (1962) Some factors influencing the rumenmicrobial population. J Gen Microbiol28, 129-146