Methane production and substrate degradation by rumen microbial communities containing single protozoal species in vitro


D. Morgavi, INRA, UR1213 Herbivores, Site de Theix, F-63122 Saint-Genès-Champanelle, France.


Aims:  To assess the effect of protozoal species on rumen fermentation characteristics in vitro.

Methods and Results: Entodinium caudatum, Isotricha intestinalis, Metadinium medium, and Eudiplodinium maggii from monofaunated wethers and mixed protozoa from conventional wethers were obtained by centrifugation, re-suspended at their normal densities in rumen fluid supernatants from defaunated or conventional wethers and incubated in vitro. The presence of protozoa increased the concentration of ammonia and altered the volatile fatty acids balance with more acetate and butyrate produced at the expense of propionate. Differences among species were observed, notably in the production of methane, which increased with E. caudatum as compared to other ciliates and to defaunated and mixed protozoa treatments (< 0·05). The increased methanogenesis was not correlated to protozoal biomass indicating that the metabolism of this protozoan and/or its influence on the microbial ecosystem was responsible for this effect.

Conclusions: Entodinium caudatum stimulated the production of methane, a negative effect that was reinforced by a concomitant increase in protein degradation.

Significance and Impact of the Study:  Comparison of individual species of protozoa highlighted the particular influence of E. caudatum on rumen fermentation. Its elimination (targeted defaunation) from the rumen could reduce methane production without affecting feed degradation.


Rumen ciliate protozoa are a metabolically important group of micro-organisms that influences digestion and the fermentation process in various ways. Mixed protozoa, considered as a group, efficiently degrade starch, soluble carbohydrates, and, to a lesser extent, plant structural polysaccharides. The provision of these activities to the rumen environment is the reason why mixed rumen protozoa are generally positively associated with increases in feed degradation (Williams and Coleman 1992). However, this positive effect on digestion is counterbalanced by an increase in methane production (Finlay et al. 1994; van Nevel and Demeyer 1996) and protein degradation (Ushida et al. 1986). Methane production is detrimental, because it represents a loss in energy for the animal and also due to the increased attention being paid to the accumulation of greenhouse gases associated with global warming. Ruminants are the main agricultural source of methane (Johnson and Johnson 1995) and the contribution of protozoa to the production of this greenhouse gas in the rumen has been estimated to be up to 37% (Finlay et al. 1994). Protozoa produce H2 as one of the main metabolic end products and they are intimately associated with methanogenic archaea (Ushida and Jouany 1996). The increase in ruminal protein degradation in the presence of protozoa has a negative impact on the supply of amino acids to the host and, at the same time, stimulates the elimination of nitrogen in urine also contributing to environmental pollution (Ushida et al. 1986; Williams and Coleman 1992). The global effect of protozoa on rumen fermentation characteristics have been mainly studied using mixed ciliate communities. However, the contribution of individual protozoal species is less known. The objective of this work was to evaluate in vitro the fermentation characteristics, in particular methane production, of four metabolically distinct protozoal species: Entodinium caudatum, Isotricha intestinalis, Metadinium medium and Eudiplodinium maggii incubated in association with two different microbial inocula isolated from defaunated and faunated animals.

Materials and methods

Donor animals and diets

Eight wethers fitted with rumen cannulae were used as donors of rumen fluid. Two animals had a regular, conventional mixed protozoal microbiota. The other six animals had previously been defaunated following the method of Jouany and Senaud (1979); two of them were kept free of fauna and the remaining four animals were each inoculated with E. caudatum, I. intestinalis, M. medium or E. maggii. These animals were kept defaunated or monofaunated for more than 1 year before the experimentation. Defaunated and monofaunated animals were fed a maintenance diet consisting of 700 g alfalfa pellet, 300 g cracked corn grain, and 200 g prairie hay and faunated animals were fed a hay-based diet. Feeds were given twice daily at 08:00 and 16:00 h, and access to water and mineral salt block supplement was unrestricted.

Experimental setup

Rumen contents from defaunated, monofaunated and faunated animals were collected before the morning feeding and strained through a polyester monofilament fabric (250 μm mesh aperture) under a stream of CO2 to remove solids. The strained rumen fluids from the two defaunated and the two faunated animals were combined to make one defaunated and one faunated rumen fluid sample, respectively. A total of 200 ml of monofaunated and faunated strained rumen fluid were transferred to centrifuge tubes (×2) and centrifuged at low speed (500 g; 5 min) to collect the protozoal cells. The same treatment was applied to the defaunated strained rumen fluid. Supernatants were decanted and protozoal pellets were re-suspended in an equivalent volume of defaunated or faunated rumen fluid supernatants obtained as above to produce, for each protozoon, a defaunated- and a faunated-based inoculum. Pellets from defaunated and faunated rumen fluids were re-suspended in their own supernatants. During this manipulation, care was exerted to keep anaerobiosis and to avoid changes in temperature that could affect the viability of the inocula. These reconstituted rumen fluids were mixed in a 1 : 3 ratio with an anaerobic buffer solution (Goering and VanSoest 1970) kept at 39°C under O2-free CO2 gas and used immediately to inoculate fermentation vials. Forty millilitres of the rumen fluid-buffer mixture was added to vials containing 300 mg of an alfalfa hay : maize [70 : 30; 470 g neutral detergent fibre (NDF), 156 g crude protein (CP) and 171 g NDF, 78 g CP per kg dry matter (DM) for alfalfa hay and maize, respectively] mixed feed ration ground to pass a 1 mm sieve as substrate and incubated anaerobically at 39°C for up to 24 h. Vials without substrate were used as control blanks. At the end of the incubation period, gas production was measured with the aid of a pressure transducer and a sample was collected for analysis of constituents by gas chromatography. Vial contents were centrifuged at 4000 g, 10 min, 4°C. Supernatants were processed for analysis of soluble fermentation end products and pellets used for estimation of DM degradation (DMD). For volatile fatty acids (VFA) and ammonia determination, 2 ml supernatant were mixed with 0·2 ml 5% (v/v) metaphosphoric acid in duplicate tubes and stored at −20°C until analysis. Pellets were dried at 60°C for 48 h for DMD. All treatments were done in triplicate and the experiment was repeated twice.

Analytical procedures

Strained rumen fluid samples were mixed with methylgreen-formalin solution (MFS) (Ogimoto and Imai 1981) in a 1 : 1 ratio and stored at room temperature in the dark until used for protozoal counting. Samples were further diluted in MFS, if necessary and enumeration was done using a Jessen counting chamber.

VFA and fermentation gases were analysed by gas chromatography as described before (Broudiscou et al. 1999; Morgavi et al. 2003). The amounts of VFA produced were obtained by subtracting the amounts present initially in the incubation medium from those determined at the end of the incubation period. Ammonia was measured by colorimetry following the method of Weatherburn (1967) and using a technicon auto-analyser II system (Technicon Instruments Co., Tarrytown, NY, USA).

Data were statistically analysed by one-way analysis of variance using the Mixed procedure of sas (SAS Institute Inc., Cary, NC, USA). The fermentation parameters were analysed with a model that included protozoa as fixed effect and experimental run (replication in time) as random effect. Data for protozoal counts were converted using a log transformation and analysed for inocula differences between experiments by t-test. Correlation between protozoal biomass and fermentation parameters was measured using the non-parametric Spearman’s rank test analysis. All tests were declared significant at < 0·05.


Incubation mixtures containing protozoa were prepared by re-suspending the protozoal pellet to the original volume of rumen fluid. Protozoal densities varied widely among inocula depending on the species (Table 1), a factor that certainly affected protozoal influence on rumen fermentation characteristics. However, they represented the naturally occurring densities of protozoa in the monofaunated and faunated animals used. Donor animals used in this study had been kept in a defaunated, monofaunated or faunated state for more than a year and the microbiota can therefore be considered as stable. Protozoal volumes in the inocula were calculated for each species in order to further discriminate whether the effects on fermentation were related to a species particular metabolism or to a more general, species-independent, ‘protozoal effect’ correlated to biomass. To take into account the large differences in cell size among species, protozoal biomass was estimated using the following cell volumes: 2·0 × 104 μm3 for each small entodiniomorph, e.g. E. caudatum, and 0·5 × 106 μm3 per cell of Isotricha sp. and for large entodiniomorphs, e.g. E. maggii and M. medium (Williams and Coleman 1992). Dasytricha ruminantium found in the mixed fauna was considered a small-sized ciliate (2·0 × 104 μm3).

Table 1.   Protozoal densities and volumes contained in inocula used for experiments*
InoculumCells 103 ml−1mm3 ml−1
  1. *Values are means (standard errors) of two repetitions except for Metadinium where only one data point was available. Volume was calculated according to Williams and Coleman (1992).

  2. †Composed of 88% small entodiniomorphids, 9% large entodiniomorphids, 2·6%Dasytricha, and less than 1%Isotricha sp.

Entodinium caudatum1404·1 (105·0)28·08
Eudiplodinium maggii19·2 (4·0) 9·58
Isotricha intestinalis10·8 (2·7) 5·38
Metadinium medium0·7 (–) 0·34
Mixed fauna†1893·3 (170·2)43·21

Table 2 shows the fermentation data for the individual protozoal species tested. The H2 recovery, estimated by stoichiometry (Demeyer 1991) averaged 104% and 96% for treatments containing defaunated and faunated supernatants, respectively, indicating that fermentations proceeded properly. The fauna-free and mixed fauna treatments had similar DMD and fermentation characteristics, except for the production of gas, butyrate and the acetate to propionate ratio that were higher when protozoa were present (< 0·05). In contrast, propionate production decreased in the presence of protozoa (P < 0.01).

Table 2.   Fermentation characteristics of alfalfa hay : corn grain (70 : 30) diet by individual protozoa incubated in rumen fluid supernatant from defaunated or faunated wethers
 DMD (%)MethaneGas mlVolatile fatty acidsNH3 mg l−1
μmol§μmol/ 100 mg DMDTotal μmolAcetate (A), (mol per 100 mol)Propionate (P), (mol per 100 mol)Butyrate, (mol per 100 mol)Iso-acids, (mol per 100 mol) A : P ratio
  1. Mean values within a treatment in a column with unlike superscript letters were significantly different (< 0·05). Statistical analysis on individual volatile fatty acids was done on actual production data.

  2. *P < 0·05, **P < 0·01 difference between mixed fauna and fauna-free inocula.

  3. †Rumen fluid supernatant (500 g, 5 min) from defaunated or faunated wethers was used to resuspend protozoal pellets from monofaunated and mixed fauna animals and pellets from defaunated animals (fauna-free) (see ‘Materials and methods’ for details).

  4. ‡Standard error of the mean.

  5. § μmol produced per fermentation vial.

Defaunated RFS†
Faunated RFS
 Mixed fauna58·2b532bc305b52·4ac**1740b61·3b19·7ab**14·1a*3·3a3·2**215b
SEM5·6629·9 26·12·60243·82·822·891·042·20·5630·8

The addition of individual protozoal species to the defaunated supernatant increased the production of methane for E. caudatum, E. maggii and I. intestinalis (< 0·05). However, when adjusted per unit of substrate disappearance only the presence of E. caudatum boosted methanogenesis by 13% (< 0·05) as compared to the fauna-free treatment. When protozoa were incubated with the faunated supernatant, methane values were also higher for E. caudatum (+16%, < 0·05). In contrast, the presence of the other protozoal species did not increase production of methane – either absolute or relative to the amount of substrate degraded – in these fermentors with the faunated supernatant. Although methane production was positively correlated to protozoal biomass in incubations with defaunated supernatant (Spearman: r = 0·9, < 0·05), there was no correlation in the amount of methane produced per unit of DMD. This indicates that stimulation of CH4 production by E. caudatum was not the consequence of an increased utilisation of substrate but rather the effect of the particular metabolism of this protozoon and/or its distinctive influence on the microbial ecosystem.

In fermentors with defaunated supernatant, addition of protozoa induced a numerical increase in the amount of DMD as compared to the fauna-free treatment with the exception of E. caudatum inocula. In contrast, when protozoa were added to rumen fluid supernatant prepared from faunated animals, the presence of I. intestinalis improved substrate degradation (< 0·05).

It was observed that the fauna-free fermentors generally produced more propionate and less butyrate than the protozoa-containing fermentors. NH3 concentrations increased in the presence of E. caudatum (< 0·05) with both types of supernatants. This rise in NH3 was accompanied by a numerical increase in the amount of branched-chain VFA produced. The productions of both butyrate and NH3 were positively correlated to protozoal biomass when incubated with defaunated supernatant (Spearman: r = 0·9, < 0·05) being therefore highest with fermentors inoculated with E. caudatum, lowest with M. medium, and had intermediate values with E. maggii and I. intestinalis. Although the trend was similar with the faunated supernatant, this correlation was not significant.


The main objective of this work was to evaluate the contribution of individual protozoa to methane production and other end products of fermentation at densities naturally present in animals. The protozoal numbers found for individual species were similar to published data on mixed faunated and monofaunated animals (Jouany et al. 1992; Ivan et al. 2000). The difference observed between experimental repetitions (shown as standard errors in Table 1) were within normal, day-to-day animal variations reported for ciliate protozoal numbers (Jouany 1989).

Methane production was the same in the mixed fauna (532 μmol) and the fauna-free fermentors (531 μmol), although it is generally accepted that defaunation decreases methane emissions (Hegarty 1999; Boadi et al. 2004). Due to experimental and practical constraints, faunated animals were kept separated from monofaunated and defaunated ones and were fed a hay-based diet instead of the mixed hay-concentrate ration. Difference in diet composition could be responsible for this apparent discrepancy.

In the rumen, protozoa are an important source of H2 that is formed during the fermentation of organic matter and used to produce methane. E. caudatum, which actively ingest and ferment starch granules (Williams and Coleman 1992), grow faster in concentrate rich diets than in forage diets. Consequently, its contribution to methane emissions is more important in starch containing diets. In agreement with this, incubations using as substrate the components of the diet separately showed that E. caudatum produced more methane per unit of DMD from corn grain than from alfalfa hay (unpublished data). In contrast, E. maggii, I. intestinalis, M. medium and fauna-free inocula produced similar amounts of methane with both substrates (unpublished data). Newbold et al. (1995) using rumen fluid from animals harbouring different species of protozoa also found that combinations containing Entodinium sp. gave the highest methanogenesis. Methane in cultures of E. caudatum is produced almost exclusively by intracellular methanogens (Varadyova et al. 2001) and could partly explain why the presence of this protozoon increased methane in the in vitro incubations tested in this work.

Because of the large difference in protozoal numbers it can be argued that variation observed among species, in particular for E. caudatum, just reflect the dissimilar biomasses. I. intestinalis, a large protozoal species, was present at a modest concentration as compared to mixed fauna or E. caudatum (Table 1), even if adjusted for their biomass. However, inocula containing I. intestinalis were positively associated with increases in substrate DMD in faunated-based inocula. In addition, the fermentation profiles of I. intestinalis differed in a number of parameters from those of defaunated or mixed fauna controls, indicating subtle differences induced by the presence of this protozoon belonging to the family Isotrichidae, whereas the other ciliates tested belong to the family Ophryoscolecidae. Methane produced per unit of feed degraded was not correlated to protozoal biomass, which also supports the idea that individual protozoal species influence rumen functioning differently depending on their metabolism. In contrast, ciliates’ biomass was positively correlated to NH3. However, this could be explained because E. maggii and M. medium, the two species with the lowest biomass in our experiments have also a low proteolytic activity (Williams and Coleman 1992).

Of the four protozoan species tested, only E. caudatum had an effect on rumen fermentation characteristics that could be considered as negative for the host. It not only increased the production of methane, but also enhanced protein degradation as evidenced by the increased level of NH3 and tendency to decrease substrate degradation. This effect on protein metabolism is in agreement with in vivo data (Ivan et al. 2000).

The differences observed in the fermentation characteristics of individual protozoa incubated with rumen fluid supernatants from defaunated or faunated animals suggest that important interactions (synergistic or antagonistic) exist among protozoa and other rumen microbes. The presence of distinct protozoal communities certainly induce changes in the ruminal microbial ecosystem that help to explain the differences observed in this work among inocula, differences that cannot be attributed to protozoa alone. More precise information on the microbial community shifts induced by the presence of protozoa seems necessary to better understand their role in the rumen ecosystem.


The authors wish to acknowledge the financial support received from the MCYT (Acción Integrada, grant no. HF2004-0026) and the French Ministry of Foreign Affairs (09219UK). M.J. Ranilla was the recipient of a fellowship under the OECD Co-operative Research Programme: Biological Resource Management for Sustainable Agriculture Systems.