Effect of plant secondary compounds on in vitro methane, ammonia production and ruminal protozoa population

Authors

  • R. Bhatta,

    Corresponding author
    • Energy Metabolism Laboratory, Division of Bioenergetics and Environmental Sciences, National Institute of Animal Nutrition and Physiology, Bangalore, India
    Search for more papers by this author
  • M. Saravanan,

    1. Energy Metabolism Laboratory, Division of Bioenergetics and Environmental Sciences, National Institute of Animal Nutrition and Physiology, Bangalore, India
    Search for more papers by this author
  • L. Baruah,

    1. Energy Metabolism Laboratory, Division of Bioenergetics and Environmental Sciences, National Institute of Animal Nutrition and Physiology, Bangalore, India
    Search for more papers by this author
  • K.T. Sampath,

    1. Energy Metabolism Laboratory, Division of Bioenergetics and Environmental Sciences, National Institute of Animal Nutrition and Physiology, Bangalore, India
    Search for more papers by this author
  • C.S. Prasad

    1. Energy Metabolism Laboratory, Division of Bioenergetics and Environmental Sciences, National Institute of Animal Nutrition and Physiology, Bangalore, India
    Search for more papers by this author

Correspondence

Raghavendra Bhatta, Energy Metabolism Laboratory, National Institute of Animal Nutrition and Physiology, Bangalore 560 030, India. E-mail: ragha0209@yahoo.com

Abstract

Aims

The objective of this study was to evaluate the potential of secondary plant metabolites from 38 sources to serve as antimethanogenic additives in ruminant diets. The effect of leaf tannins from these different plant sources on rumen fermentation, protozoal populations and methanogenesis was also studied.

Methods and Results

Samples (200 mg dry matter, DM) were incubated without and with polyethylene glycol (PEG)-6000 (400 mg DM) as a tannin binder during 24-h incubation in the in vitro Hohenheim gas system. In the leaf samples, total phenol (g kg−1 DM) was maximum in Pimenta officinalis (312) followed by Oenothera lamarckiana (185) and Lawsonia inermis (105). Of the 38 samples, condensed tannins exceeded 4·0 g kg−1 in only Alpinia galanga (7·50), Cinnamomum verum (4·58), Pelargonium graveolens (18·7) and Pimenta officinalis (23·2) and were not detected in seven samples. When the bioactivity of the leaf samples was assessed using the tannin bioassay, the percentage increase in the amount of gas produced during incubation of samples with the tannin-binding agent PEG-6000 over the amount produced during incubation without the tannin binder ranged from nil (zero) to 367%, with the highest being recorded with A. galanga leaves. The ratio of methane reduction per ml of total gas reduction was maximum with Rauvolfia serpentina (131·8) leaves, followed by Indigofera tinctoria (16·8) and Withania somnifera (10·2) leaves. Total and differential protozoal counts increased with added PEG in twenty-two samples, maximum being in Pimenta officinalis. Increased accumulation of total volatile fatty acids during incubation with added PEG-6000 was recorded, and the values ranged from zero to 61%. However, the increase was significant in only 11 of the 38 tannin sources tested indicating noninterference of tannin on in vitro fermentation of carbohydrates by the majority of samples tested. Conversely, in 26 of 38 plant sources, the leaf tannins reduced N-digestibility as evidenced by increased accumulation of NH3-N with added PEG.

Conclusions

Our study unequivocally demonstrated that plants containing secondary metabolites such as Rauvolfia serpentine, Indigofera tinctoria and Withania somnifera have great potential to suppress methanogenesis with minimal adverse effect of feedstuff fermentation.

Significance and Impact of the Study

It was established that methanogenesis was not essentially related to the density of protozoa population in vitro. The tannins contained in these plants could be of interest in the development of new additives in ruminant nutrition.

Introduction

Reduction in enteric methane emissions from livestock production is a high priority, because this biological process accounts for 2–12% loss of dietary gross energy in ruminants (Johnson and Johnson 1995). Moreover, methane is a potent greenhouse gas with a global warming potential 23 times higher than that of carbon dioxide (IPCC 2001). Ruminal methanogenic archaea utilize hydrogen produced during microbial fermentation of feeds to reduce carbon dioxide to methane, thereby maintaining low partial pressure of hydrogen, which allows the re-oxidation of reduced nucleotides produced during fermentation (Morgavi et al. 2010). Optimizing the rumen fermentation for efficient nitrogen utilization is critical to minimize the release of nitrogen into the atmosphere. Manipulating the rumen microbial ecosystem to reduce methane emission and nitrogen (N) excretion by ruminants to improve their performance are important issues for animal nutritionists. There is a need to identify feed additives with potential to modify rumen fermentation, thereby enhancing the efficiency of feed utilization while decreasing the methane emission and nitrogen excretion. A number of mitigation strategies that can reduce enteric methane and ammonia production have been identified and reported in the literature (Martin et al. 2010). Plant secondary metabolites such as tannins are particularly attractive as rumen modifiers as these compounds are natural products, which are generally accepted to be environmental friendly and safe in food production systems. Due to their potential to adversely affect feed intake and nutrient utilization, tannins are sometimes regarded as antinutritional; however, when administered at low concentrations, certain tannins can beneficially alter ruminal fermentation (Bhatta et al. 2002), improve microbial protein synthesis (Bhatta et al. 2001) and suppress methanogenesis (Bhatta et al. 2012, 2013).

In the rumen ecosystem, the ubiquitous protozoa are large producers of hydrogen. In addition, a physical association between protozoal cells and methanogens exists in the rumen ecosystem, which favours hydrogen transfer (Morgavi et al. 2010). Rumen protozoa contribute to ruminal methane and N emissions by synergistically providing hydrogen as a reducing substrate to methanogens and by their predatorial and digestion of large proportions of rumen bacteria. Tropical plants containing tannins/saponins have been found to suppress or eliminate protozoa from the rumen and reduce methane and ammonia production (Lila et al. 2003). Nevertheless, plants produce many different types and concentrations of secondary metabolites, which are likely to affect the nutritional value of these plants.

The screening of these plants is an important step for the discovery of new compounds and their development as feed additives to mitigate rumen methanogenesis and N turnover. Whereas the effects of some medicinal plants, their extracts or essential oils on rumen fermentation have already been reported, these studies have only dealt with a small number of plant species, and only a few have dealt specifically with the possibility of decreasing methane production using phytogenic additives. Undoubtedly, the richness of the plant kingdom contains many promising plants that have not yet been investigated for such purposes. Moreover, plants screened for biological activity may be limited in their geographical distribution or the type, and concentration of tannin they contain may vary regionally, thus having marked impacts on their activity. This justifies the development of a database on various plants, their nutritive value and methane reduction potential. In the present study, 38 not-well-researched plants containing contrasting type and concentration of tannins were screened to determine their effect on protozoa population, methanogenesis and ammonia nitrogen in vitro. There are no previous reports on the phenolic composition, rumen fermentation and methane reduction properties of these plants. Other in vitro fermentation variables were studied as well in the gas production system so that a selection could be made for the most ideal antimethanogenic source.

Materials and methods

Experimental plants

All samples used in this study were plant leaves. About 500 g of the leaves (both mature and immature) was harvested from individual plants, and pooled samples of the leaves were taken for this study. All samples were air-dried for 3–4 days and ground to pass a 1-mm sieve. They were preserved in tightly closed plastic jars stored in a dry, dark and cool place to prevent phenolic degradation.

Nutrient and tannin analyses

The tree leaves samples were analysed in triplicate for crude protein (CP) (AOAC 1990), neutral detergent fibre (NDF) and acid detergent fibre (ADF) (Van Soest et al. 1991). The NDF was analysed in leaf samples without sodium sulfite and amylase. Both NDF and ADF were expressed with residual ash.

For the tannin assays, samples were ground to a fine powder in a Cyclotec mill. Samples (0·2 g) were extracted in 10 ml aqueous acetone (acetone/water, 7 : 3) twice for 20 min in an ultrasonic water bath. The extracted samples were centrifuged (6000 g, 10 min, 4°C), and the supernatants were combined and used for tannin analysis on the same day. Determinations of total phenols (TP), total tannins (TT) and condensed tannins (CT) in the samples were carried out based on the methods described by Makkar (2003). Accordingly, TP and TT were assessed by a modified Folin–Ciocalteu method using polyvinylpolypyrrolidone to separate non-tannin phenols (NTP) from tannin phenols. The CT was analysed by the butanol–HCl–iron method. While both TP and TT were expressed as gallic acid equivalents, CT was given as leucocyanidin equivalents. Hydrolysable tannins (HT) were calculated as the difference between TT and CT (Singh et al. 2005). Analyses were carried out in duplicate except for detergent fractions that were measured in triplicate.

For the in vitro gas production test, rumen liquor was collected after morning feeding from two cannulated Holstein Friesian crossbred bulls fed a total mixed ration containing finger millet (Eleusine coracana) straw and a commercial concentrate mixture in a 1 : 1 ratio. The ration contained 16% CP and 9·0 MJ kg−1 dry matter of metabolizable energy. The rumen liquor, strained through a muslin cloth, was pooled and used as the source of inoculum. A total of 200 mg air-equilibrated sample was each incubated with 30 ml buffered rumen inoculum (Menke et al. 1979) in 100-ml calibrated syringes placed in a water bath maintained at 39°C. The incubations were conducted in triplicate for each sample on two successive days, and these incubations were repeated three times within an interval of 1 week. Incubations without leaf sample served as the blanks with every set. The difference in the composition and activity of the rumen inoculum among incubations was controlled by parallel incubation of reference standard feedstuffs as suggested by Menke et al. (1979). Incubations were run for 24 h with recording of gas production at 8 and 24 h. The interference of tannins on fermentations in vitro was assessed using polyethylene glycol (PEG-6000). The magnitude of the increase in gas production on PEG addition to the tannin sources at a ratio of 2 : 1 was taken as an index of tannin's interference on rumen fermentation (Makkar et al. 1995).

Gas and methane estimation

After 24 h of incubation, the volume of fermentation gas produced was recorded from visual assessment of the calibrated scale on the syringe. The gas produced due to fermentation of substrate was calculated by subtracting gas produced in blank syringe (containing inoculum and buffer but no substrate) from total gas produced in the syringe containing substrate and inoculum. The gas produced in standard syringe (containing concentrate and hay standard from Hohenheim University) was used to check day-to-day variation in the quality of inoculum. For methane estimation, 1·0 ml of gas was sampled with an airtight syringe (Hamilton Company, Reno, Nevada) from the head space of each incubation syringe (having one outlet) using a adapter fitted to the silicon tubing and injected into a Thermo-fisher gas chromatograph equipped with a thermal conductivity detector and stainless steel column packed with Porapak-Q. Temperatures of injector oven, column oven and detector were 60, 100 and 110°C, respectively (Kajikawa et al. 2007). Based on the methane percentage estimated in the gas produced, methane production in ml was calculated in each sample (methane volume (ml) = methane% × total gas produced (ml) in 24 h).

Rumen fluid analysis

The incubation was terminated after 24 h by removing the syringes from the water bath, and the supernatant fluids from each syringe were stored at −20°C until further analysis. The rumen fluid samples were analysed for ammonia N (Conway 1957). In brief, 1·0 ml of incubation fluid was kept in the outer chamber of the Conway diffusion dish and 1·0 ml of boric acid indicator in the central chamber; after that, 1·0 ml of potassium carbonate solution is placed in the outer chamber opposite to that of rumen fluid. After mixing the contents of the outer chamber, it was kept for 20 min for ammonia diffusion and titrated against standard sulfuric acid.

The total volatile fatty acid (TVFA) concentration was estimated according to Barnett and Reid (1957). Two millilitre of rumen fluid and 2·0 ml of buffer (potassium oxalate + oxalic acid) were taken in Markham apparatus, and 100 ml of distilled VFA is collected in a flask kept in ice bath. After adding few drops of phenolphthalein, it was titrated against standard alkali.

Enumeration of ciliated protozoa

Rumen ciliates were identified according to Hungate (1966). The counting chamber had a depth of 0·1 mm. Spirotrichs not identified to generic level were classified into small Spirotrichs (mainly Entodinia with an average size 42 μm × 23 μm) and large Spirotrichs (mainly Diplodinia with an average size of 132 μm × 66 μm). The protozoa numbers were calculated according to Kamra et al. (1991).

Observations and calculations

Net methane and gas productions were calculated as per Jayanegara et al. (2009) from the differences of the methane and gas in the test syringe and the corresponding blank, and the methane concentration was determined as:

display math

The percentage increase in methane production after PEG addition, which was calculated as:

display math

Statistical analysis

The statistical analysis was carried out using PROC GLM (General Linear Model) of SAS 9.2 (SAS/STAT User's Guide, SAS Institute Inc., Cary, NC, USA). The specific ancova model was Yij = μ + αi + βjXij+∈ij, where Yij is the dependent variable, μ is the least squares mean, αI is the effect of PEG, βj is the jth effect of covariate, and Xij are the covariates, viz. CP, NDF, ADF, tannin fractions and protozoa, and was used for evaluating the effect of PEG.

Further, a second model reading Yij = μ + αi+ αβij + ∈ij, where Yij is the dependent variable, μ is the least squares mean, αI is the effect of PEG, βj is the effect of plant, and αβij is the interaction effect, was used for evaluating the effect of plant, PEG and plant × PEG interaction.

Results

The nutrient composition of the leaves is presented in Table 1. Energy (ME, MJ kg−1 DM) content varied from 3·40 to 9·1. Highest CP (g kg−1 DM) was recorded in Indigofera tinctoria (260) followed by Artemisia absinthium (245), Ruta graveolens (245) and lowest in Cymbopogon martinii (53·6). NDF and ADF contents ranged from 218 to 711 and 116 to 401, respectively. Most of the samples contained >5·0 g kg−1 ether extract.

Table 1. Energy value (MJ kg−1 DM) and nutrient contents of leaves (g kg−1 DM)
Name of the plantME (MJ kg−1 DM)OMCPNDFADFADLEEAsh
  1. All were leaf samples; ME, metabolizable energy; OM, organic matter; CP, crude protein; NDF, neutral detergent fibre; ADF, acid detergent fibre; ADL, acid detergent lignin; EE, ether extract.

Adhatoda zeylanica 7·0479614732420566·18·70104
Alpinia galanga 3·4383110963537164·141·369·5
Andrographis paniculata 6·6581682·045934181·416·584·4
Artemisia annua 9·0581021632415534·119·390·5
Artemisia absinthium 7·4881524532619969·124·484·5
Cichorium intybus 6·9184956·824117691·320·651·2
Cinnamomum verum 4·9786785·271134595·316·932·9
Cymbopogon flexuosus 7·0382195·067738157·326·679·4
Cymbopogon martinii 6·6480053·666644340·216·4100
Cymbopogon winterianus 5·8182668·965232349·232·375·3
Foeniculum vulgare 6·9374815727820586·120·1152
Hemigraphis colorata 7·5472610130920742·56·10174
Indigofera tinctoria 7·9780026026315556·613·8100
Lavandula stoechas 6·9984576·557950180·433·055·0
Lawsonia inermis 8·6485111824517964·833·649·4
Leucas aspera 8·3876421434217383·822·4136
Lippia citriodora 6·7979076·535227972·19·60110
Medicago sativa 8·8581625931918256·024·784·2
Melissa officinalis 5·0478517647641146·70115
Mentha citrata 6·6381012748144375·129·289·6
Matricaria chamomilla 6·5678711646638918·522·3113
Murraya koenigii 7·0180111433418856·429·198·9
Myristica fragrans 6·2292065·644927497·822·380·0
Oenothera lamarckiana 5·3281015121813434·611·990·4
Origanum vulgare 5·6881689·637933437·025·684·3
Pandanus amaryllifolius 8·2582710742833768·839·472·9
Pelargonium graveolens 4·6983412950436120·614·866·3
Pimenta officinalis 3·4083744·840035771·47·5562·8
Pogostemon cablin 6·7378013841831248·352·6120
Rauvolfia serpentina 8·1188020524215541·023·6120
Rauvolfia tetraphylla 6·9778317327217569·931·5117
Rosmarinus officinalis 4·1883710149844932·669·962·9
Ruta graveolens 8·2080424123212730·424·595·7
Salvia officinalis 6·5479313741432118·645·6107
Strobilanthes flaccidifolius 6·2576818415711654·52·40132
Teraxacum officinale 8·8179713123317843·534·3103
Thymus vulgaris 8·4982511341921314·838·974·9
Withania somnifera 6·1890015331218856·429·198·9

The phenolic compositions of the leaves are presented in Table 2. Total phenol (g kg−1 DM) was maximum in Pimenta officinalis (312) followed by Oenothera lamarckiana (185) and Lawsonia inermis (105). Of the 38 samples, only Alpinia galanga (7·50), Cinnamomum verum (4·58), Pelargonium graveolens (18·7) and Pimenta officinalis (23·2) contained >4·0 g kg−1 CT, and it was not detected in seven samples. Most of the phenolics were present in the form of HT.

Table 2. Total phenols (TP), total tannins (TT) and condensed tannin (CT) of leaves (g kg−1 DM)
Name of the plantTPaTTaCTbHT
  1. All were leaf samples; TP, total phenol; TT, total tannin; CT, condensed tannin; HT, hydrolysable tannin; ND, not detected.

  2. a

    As tannic acid equivalent.

  3. b

    As leucocyanidin equivalent.

Adhatoda zeylanica 34·02·040·0032·04
Alpinia galanga 81·47·017·50
Andrographis paniculata 40·32·500·0232·47
Artemisia annua 34·92·090·0122·07
Artemisia absinthium 24·01·450·131·32
Cichorium intybus 34·01·820·1021·72
Cinnamomum verum 42·73·124·58ND
Cymbopogon flexuosus 34·62·480·0202·46
Cymbopogon martinii 26·01·58ND1·58
Cymbopogon winterianus 27·02·13ND2·13
Foeniculum vulgare 34·42·500·0072·49
Hemigraphis colorata 71·15·05ND5·05
Indigofera tinctoria 71·05·321·523·80
Lavandula stoechas 12·30·86ND0·86
Lawsonia inermis 1058·170·1248·05
Leucas aspera 57·24·450·0204·43
Lippia citriodora 74·25·450·0335·42
Medicago sativa 10·80·600·0060·594
Melissa officinalis 11·17·600·0100·748
Mentha citrata 16·41·0701·07
Matricaria chamomilla 16·01·160·0061·15
Murraya koenigii 47·93·230·0673·16
Myristica fragrans 26·11·511·220·294
Oenothera lamarckiana 18516·910·00716·9
Origanum vulgare 41·83·340·0053·33
Pandanus amaryllifolius 34·42·130·0242·12
Pelargonium graveolens 88·47·8418·7ND
Pimenta officinalis 31229·223·26·02
Pogostemon cablin 43·63·060·0043·06
Rauvolfia serpentina 51·63·270·0183·25
Rauvolfia tetraphylla 37·62·280·0202·26
Rosmarinus officinalis 36·22·340·2582·08
Ruta graveolens 27·41·560·0241·54
Salvia officinalis 39·92·81ND2·81
Strobilanthes flaccidifolius 84·75·850·0795·77
Teraxacum officinale 39·82·96ND2·96
Thymus vulgaris 69·84·74ND4·74
Withania somnifera 22·31·180·0151·16

Activity of tannins represented by increment in gas volume with PEG (tannin bioassay) ranged from nil (zero) to 367% with highest being recorded in A. galanga (Table 3). The ratio of methane reduction per ml of total gas reduction is presented in Table 3. The maximum ratio was recorded in Rauvolfia serpentina (131·8) followed by I. tinctoria (16·8) and Withania somnifera (10·2); nine tannin sources did not affect methanogenesis.

Table 3. Gas production (ml per 200 mg DM), methane concentration (%) and methane increase with and without polyethylene glycol (PEG) among plant species
Name of the plantGas production (ml per 200 mg DM)TBA (% increase in gas)*LSDP-valueMethane (ml)Methane reduction per ml of total gas reductionLSDP-value
−PEG+PEG−PEG+PEG
Adhatoda zeylanica 29·329·71·370·470·5236·216·921·780·031<0·001
Alpinia galanga 0·94·23671·56<0·0010·311·000·210·0031·026
Andrographis paniculata 28·829·52·430·070·5235·746·370·900·0020·954
Artemisia annua 41·442·21·930·650·6246·867·901·300·001<0·001
Artemisia absinthium 27·427·40·000·000·6355·576·510·000
Cichorium intybus 31·436·415·91·09<0·0018·055·050·000
Cinnamomum verum 16·218·212·30·96<0·0012·874·310·720·0150·826
Cymbopogon flexuosus 30·130·82·330·080·5327·657·870·310·0040·725
Cymbopogon martinii 30·430·50·330·010·9633·263·714·500·042<0·001
Cymbopogon winterianus 21·522·23·260·56<0·0013·272·770·000
Foeniculum vulgare 27·428·12·550·57<0·0014·525·030·730·0030·649
Hemigraphis colorata 35·0375·711·65<0·0015·884·880·000
Indigofera tinctoria 31·231·30·320·010·6343·425·1016·800·024<0·001
Lavandula stoechas 29·830·21·340·010·5063·964·140·450·0030·934
Lawsonia inermis 40·142·14·990·43<0·0013·366·401·520·003<0·001
Leucas aspera 36·538·45·210·86<0·0015·415·640·120·0021·003
Lippia citriodora 30·430·50·330·010·9046·316·895·800·017<0·001
Medicago sativa 38·038·51·320·050·5795·387·734·700·023<0·001
Melissa officinalis 13·514·910·40·840·0063·063·590·380·0010·695
Mentha citrata 25·525·81·180·360·7131·502·964·870·025<0·001
Matricaria chamomilla 26·227·86·110·96<0·0013·583·940·230·0010·945
Murraya koenigii 28·930·24·500·23<0·0012·182·880·540·0040·834
Myristica fragrans 25·830·819·40·84<0·0014·226·080·370·0040·789
Oenothera lamarckiana 16·427·467·11·45<0·0017·037·030·000
Origanum vulgare 20·521·23·410·51<0·0012·072·680·870·010<0·001
Pandanus amaryllifolius 36·837·21·090·030·6257·968·671·770·062<0·001
Pelargonium graveolens 12·525·110112·3<0·0012·125·310·250·0030·895
Pimenta officinalis 6·920·21931·26<0·0014·091·850·000
Pogostemon cablin 21·822·11·380·230·3254·324·661·130·015<0·001
Rauvolfia serpentina 33·8340·590·020·8457·2533·6131·751·230<0·001
Rauvolfia tetraphylla 25·826·83·880·48<0·0015·165·360·200·0011·036
Rosmarinus officinalis 10·212·219·60·42<0·0011·662·880·610·0120·648
Ruta graveolens 32·8330·610·010·6347·847·920·400·0010·694
Salvia officinalis 26·127·13·830·54<0·0016·316·060·000
Strobilanthes flaccidifolius 22·122·51·810·410·5985·223·700·000
Teraxacum officinale 40·742·74·911·95<0·0016·736·100·000
Thymus vulgaris 38·439·42·600·120·0898·368·420·060·0010·682
Withania somnifera 21·121·41·420·120·6424·527·5910·230·892<0·001
SourceGas production (ml per 200 mg DM)Methane (%)
LSDP-valueLSDP-value
  1. TBA, tannin bioassay; LSD, least square difference.

  2. All the samples were leaf samples.

  3. Samples were incubated at 1 : 2 ratio with PEG-6000 (w/w basis).

  4. a

    SAS programme gives LSD for main effect only, not for interaction.

  5. For calculation of methane (%) and methane increase with PEG addition (%), see Section Materials and Methods.

Plant0·3216<0·0010·2609<0·001
PEG0·0531<0·0010·0864<0·001
Plant × PEGa<0·001<0·001

Total and differential protozoal counts are depicted in Table 4. Twenty-two samples showed increment in total protozoal count with added PEG, maximum being in Pimenta officinalis.

Table 4. Total and differential protozoa population (per ml) with and without polyethylene glycol (PEG)
PlantEntodinia (105)Holotricha (105)Total% increaseaLSDP-value
−PEG+PEG−PEG+PEG−PEG+PEG
Adhatoda zeylanica 0·1040·1130·00100·1050·1137·620·036<0·001
Alpinia galanga 0·1330·135000·1330·1351·500·0420·759
Andrographis paniculata 0·1480·1290·0010·0010·1500·13000
Artemisia annua 0·1720·1670·0020·0040·1740·17100
Artemisia absinthium 0·0710·108000·0710·10953·521·023<0·001
Cichorium intybus 0·2730·26500·0020·2730·26700
Cinnamomum verum 0·2380·306000·2380·30728·90·005<0·001
Cymbopogon flexuosus 0·1510·1980·0240·0190·1750·21724·00·002<0·001
Cymbopogon martinii 0·1560·1880·0080·0080·1640·19518·90·002<0·001
Cymbopogon winterianus 0·1140·1210·0140·0080·1280·1290·780·0100·854
Foeniculum vulgare 0·1430·1270·0080·0110·1500·13800
Hemigraphis colorata 0·2240·2400·0030·0020·2270·2426·610·045<0·001
Indigofera tinctoria 0·1470·1580·00200·1490·1596·710·032<0·001
Lavandula stoechas 0·2370·2050·0130·0130·2500·21800
Lawsonia inermis 0·1940·15000·0010·1950·15200
Leucas aspera 0·1560·121000·1560·12100
Lippia citriodora 0·1420·1330·0010·0020·1430·13600
Medicago sativa 0·1900·2310·0020·0040·1920·23522·400·053<0·001
Melissa officinalis 0·2470·2390·0060·0090·2530·24800
Mentha citrata 0·0980·1320·0040·0100·1020·14239·20·016<0·001
Matricaria chamomilla 0·0140·2020·0110·0210·1510·22347·70·012<0·001
Murraya koenigii 0·0930·1050·0040·0020·0970·10710·30·036<0·001
Myristica fragrans 0·1010·1310·0010·0010·1020·13128·430·063<0·001
Oenothera lamarckiana 0·2560·2430·0060·0030·2620·24500
Origanum vulgare 0·1140·1240·0080·0060·1230·1316·500·018<0·001
Pandanus amaryllifolius 0·1760·1750·0070·0140·1830·1882·730·0010·639
Pelargonium graveolens 0·1400·2110·0050·0110·1450·22253·10·002<0·001
Pimenta officinalis 0·1380·1760·0030·0010·01410·17611480·042<0·001
Pogostemon cablin 0·1410·1380·0020·0050·1430·14200
Rauvolfia serpentina 0·1770·0950·01200·1890·09600
Rauvolfia tetraphylla 0·0840·079000·0840·07800
Rosmarinus officinalis 0·0920·0780·00700·0990·07800
Ruta graveolens 0·1610·1850·0030·0010·1640·18512·800·036<0·001
Salvia officinalis 0·1520·1420·0050·0080·1560·15100
Strobilanthes flaccidifolius 0·2530·2260·0140·0020·2660·22800
Teraxacum officinale 0·1530·2160·0030·0060·1560·22342·950·428<0·001
Thymus vulgaris 0·1870·2610·0100·0050·1970·26635·00·0030·856
Withania somnifera 0·2060·2860·0020·0050·2090·29139·230·036<0·001
SourceEntodinia (105)Holotricha (105)Total (105)
LSDP-valueLSDP-valueLSDP-value
  1. LSD, least square difference

  2. All samples were leaf samples.

  3. a

    % Increase represents increase with PEG addition compared with without PEG.

  4. b

    SAS programme gives LSD for main effect only not for interaction.

  5. Spirotrichs not identified to generic level were classified into small Spirotrichs (mainly Entodinia with an average size 42 μm × 23 μm) and large Spirotrichs (mainly Diplodinia with an average size of 132 μm × 66 μm).

Plant0·0006<0·0010·0003<0·0010·0018<0·001
PEG0·0004<0·0010·0001<0·0010·0005<0·001
Plant × PEGb<0·001<0·001<0·001

The TVFA concentration and ammonia N values are presented in Table 5. The TVFA concentration also increased when incubated with PEG, and the values ranged from zero to 61%. The increment was significant in only 11 of 38 samples, thereby indicating noninterference of tannin sources on the in vitro fermentation. However, in 26 of 38 samples, tannin has affected the N-digestibility, reflected in the NH3-N values. The maximum increase was recorded in M. sativa (35·7) and lowest in Pandanus amaryllifolius.

Table 5. Total volatile fatty acid and ammonia nitrogen concentration with polyethylene glycol (PEG) among plant species
PlantTVFA (m Mol dl−1)% increaseaLSDP-valueNH3-N (mg dl−1)% increaseaLSDP-value
−PEG+PEG−PEG+PEG
Adhatoda zeylanica 11·811·10·0000·52612·114·918·80·854<0·001
Alpinia galanga 12·312·30·0001·36516·819·513·80·965<0·001
Andrographis paniculata 12·312·30·0000·96511·214·422·21·106<0·001
Artemisia annua 12·813·33·910·6181·03523·627·112·90·985<0·001
Artemisia absinthium 12·813·33·910·2480·71810·714·425·71·002<0·001
Cichorium intybus 13·615·715·40·5261·00620·522·69·290·006<0·001
Cinnamomum verum 11·913·816·00·321<0·00114·416·311·70·214<0·001
Cymbopogon flexuosus 16·717·22·990·5120·74526·428·05·710·098<0·001
Cymbopogon martinii 12·017·243·30·365<0·00116·818·27·690·209<0·001
Cymbopogon winterianus 13·513·60·740·0590·90011·214·422·21·026<0·001
Foeniculum vulgare 16·718·18·380·9650·75920·522·69·290·624<0·001
Hemigraphis colorata 13·818·231·91·035<0·00111·214·422·20·985<0·001
Indigofera tinctoria 10·717·260·70·682<0·00114·915·64·490·2300·965
Lavandula stoechas 14·015·07·140·8530·64813·514·03·570·0450·853
Lawsonia inermis 12·513·68·800·6321·02613·619·630·61·004<0·001
Leucas aspera 11·811·10·00019·621·06·670·0640·429
Lippia citriodora 9·210·413·00·6950·6249·5114·434·01·023<0·001
Medicago sativa 11·514·828·70·654<0·00112·619·635·71·035<0·001
Melissa officinalis 10·910·30·00013·013·53·700·2350·965
Mentha citrata 15·716·44·461·0260·86312·114·918·80·354<0·001
Matricaria chamomilla 12·113·07·440·6240·95114·916·811·30·563<0·001
Murraya koenigii 19·022·116·30·984<0·00110·914·926·81·035<0·001
Myristica fragrans 13·915·511·50·8640·95316·317·56·860·5140·965
Oenothera lamarckiana 13·515·615·60·9651·05010·715·229·61·026<0·001
Origanum vulgare 10·813·525·00·635<0·00110·714·425·70·954<0·001
Pandanus amaryllifolius 14·017·625·70·634<0·00129·830·83·250·3411·002
Pelargonium graveolens 12·912·60·00012·613·24·550·1250·856
Pimenta officinalis 15·116·710·60·7510·75610·111·411·40·4160·845
Pogostemon cablin 12·813·87·810·9850·86513·514·03·570·2050·645
Rauvolfia serpentina 10·411·49·620·6240·79612·614·211·30·954<0·001
Rauvolfia tetraphylla 10·717·260·71·235<0·00120·522·69·290·4150·756
Rosmarinus officinalis 17·518·13·430·6521·60423·627·112·90·900<0·001
Ruta graveolens 11·612·57·760·6421·02614·717·214·51·002<0·001
Salvia officinalis 8·7311·026·00·536<0·00127·528·43·170·6210·652
Strobilanthes flaccidifolius 12·515·624·80·726<0·00112·114·918·80·964<0·001
Teraxacum officinale 15·015·74·670·4871·03614·916·28·020·3250·965
Thymus vulgaris 14·715·55·440·8620·69827·028·03·570·6350·751
Withania somnifera 17·318·67·510·9541·82314·416·311·70·045<0·001
SourceTVFA (m Mol dl−1)NH3-N (mg dl−1)
LSDP-valueLSDP-value
  1. All samples were leaf samples.

  2. TVFA, total volatile fatty acids; NH3-N, ammonia nitrogen; LSD, least square difference

  3. a

    % Increase represents increase with PEG addition compared with without PEG.

  4. b

    SAS programme gives LSD for main effect only not for interaction.

Plant0·1157<0·0010·1656<0·001
PEG0·0357<0·0010·0511<0·001
Plant × PEGb<0·001<0·001

Discussion

The aim of the present study was to screen the activity of 38 leaves containing different levels and proportions of phenolics for suppressing in vitro methanogenesis. Most of the leaves were a good source of energy except Pimenta officinalis and Alpinia galanga wherein the ME content was <4·0 MJ kg−1 DM. There was large variation in the protein content among the samples probably due to their different geographical origin and growth conditions.

Eight of 38 samples screened did not contain CT, but had appreciable HT. Most of the phenols were present as nontannins. Tannin bioassay (which measures the percentage increase in gas volume in cultures incubated with added PEG compared with cultures incubated without PEG addition) that reflects the effect of tannin on gas production was highest in A. galanga (367) followed by Pelargonium graveolens (101) and Oenothera lamarckiana (67·1), their TP contents being 81·4, 88·4 and 185 g kg−1, respectively. However, A. galanga and P. graveolens contained only CT (7·50, 18·7), whereas O. lamarckiana contained only HT (16·9). In an earlier study, Jayanegara et al. (2012) established that TBA had higher correlation with total phenol rather than TT content signifying the role of non-tannin phenol in suppressing methanogenesis. Because non-tannin phenols are not likely to decrease utilization of proteins and other nutrients in ruminants, our finding can have practical application in mitigating rumen methanogenesis.

In an earlier study, we demonstrated that in vitro incubations of tamarind (Tamarindus indica) seed husk (containing about 140 g CT kg−1 DM) produced less gas than incubations of same substrate containing added PEG (Bhatta et al. 2001). Recently, it was also shown that samples containing both HT and CT (HT + CT) were more effective in reducing in vitro total gas and methane production than samples containing only HT (Bhatta et al. 2009). Not unexpectedly, however, this phenomenon is likely dependent on concentrations and availability of the phenolic fractions as not all leaves containing both HT + CT were able to effectively reduce total gas and methane production in the in vitro incubations. In support of this conclusion, an earlier study had shown that phenolic fractions present in tannin extracts were more effective than leaves containing the tannins (Bhatta et al. 2009).

Comparing the ratios of the reductions in the amount of methane produced to the reductions in total gas produced during the incubations can be used as a scale to screen large number of samples, and this comparison confirmed that nine leaf sources were not effective candidates for methane mitigation. In seventeen samples, the ratio was <1·0, and in seven samples, the methane reduction ratio was between 1 and 5. The highest ratio was recorded in Rauvolfia serpentine (131·2) and Lippia citriodora (5·80). These values did not correlate well with the TBA, establishing that methane reduction per ml of total gas reduction is a better indicator than TBA to identify methane-suppressing candidates because it reflects effects on both total gas and methane production.

Two samples, Medicago sativa and Mentha citrata, although containing very low TT (0·06 and 1·07 g kg−1 DM), significantly reduced methane production (4·87 and 4·70). This could be because of the presence of saponin and essential oils that are also known to suppress methanogenesis (Lila et al. 2003).

When comparing samples incubated in the presence of PEG to those incubated without added PEG, the increase (%) in Entodinia population was greater than that in Holotricha indicating higher susceptibility of Entodinia to tannin (Table 4). Of 38 samples, 21 samples recorded a significant (P < 0·05) increase in Entodinia as compared to seven with increased Holotricha populations. The methane inhibition in these samples was accompanied with inhibition in protozoa population, indicating that methanogens associated symbiotically with the ciliate were adversely affected. Protozoa can synergistically provide hydrogen as a source of electrons to the methanogens, and hence, antiprotozoal effects of tannins would be expected to decrease methane production by methanogens attached to protozoa (Wang et al. 2009). It was also established that methane production is usually higher when protozoa are present or present in greater numbers in the rumen than when absent or low in numbers (Jouany and Lassalas 1997). However, Hess et al. (2003) reported that only a small portion of total methane production was due to the presence of methanogens attached with the ciliate protozoa, and Machmüller et al. (2003) observed an increased number of methanogens in defaunated sheep, thus suggesting that association between protozoa and methanogens may not always play an important role in rumen methanogenesis. The differential response of protozoal populations and methane production observed between our samples incubated with and without PEG were highly variable depending on the different tannin sources. For example, in samples such as Artemisia absinthium, Hemigraphis colorata, Pimenta officinalis and Teraxacum officinale, there was an increase (P < 0·05) in the protozoa population with PEG, but there was no effect (P > 0·05) on methanogenesis. In contrast, in incubations containing added Andrographis paniculata, Artemisia annua, Lippia citriodora and Rauvolfia tetraphylla, protozoa numbers were unaltered, but methane concentrations were increased (P < 0·05) with PEG. Probable reasons for these observations could be that some of the tannins may directly affect the methanogenic archaea not associated with the protozoa. The finding that tannins suppressed methanogenesis directly through their antimethanogenic activity and indirectly through their antiprotozoal property was observed in an earlier study (Bhatta et al. 2009). Moreover, others have reported that HT and CT may differentially effect on ruminal ciliated protozoa, with HT generally being less inhibitory against protozoa than CT (Leinmüller et al. 1991; Sliwinski et al. 2002). Overall, our results support the conclusions of the recent meta-analysis of Jayanegara et al. (2012), indicating that there is no direct relationship between dietary tannin and protozoa counts.

The effect of tannin sources on in vitro TVFA production varied among the samples. Tannins did not cause a substantial inhibition of the TVFA production that would generally indicate a nonspecific inhibition of fermentation activity. If the inhibition on methane was nonspecific rather than a targeted effect, then the proportion of methane in the total gas would be hardly affected. The slight depression in fermentative activity could be a consequence of an impaired methanogenesis, and it was observed with other additives causing similar effects on ruminal fermentation such as monensin (García-González et al. 2008). Jayanegara et al. (2009) reported that inhibition of fibre degradation will shift fatty acid composition away from acetate and hence less production of hydrogen and less methane fermentation. In the present study, the effect of the different tannin sources on TVFA accumulations appeared to be less than their effect on methane production, thus indicating that methane suppression was primarily due to antimethanogenic activity rather than lowered fibre digestibility.

Tannins decreased in vitro NH3-N concentrations in all incubations compared with PEG-supplemented incubations; however, the magnitude of decrease was greater (P < 0·05) in samples containing both HT + CT than in samples containing only HT (Table 5). This could be attributed to the formation of complexes between HT and CT with proteins. The HT and proteins usually form complexes at an optimal pH range of 3–4, but these complexes can also occur at pH 6–7 (Leinmüller and Menke 1990). Some of the ruminal bacteria can dissociate the protein–HT complexes. However, the dissociation of protein–CT complexes is not easy (McSweeny et al. 2001). The observed difference in the NH3-N concentration between samples containing only HT and those with HT + CT was attributable to the reversible nature of protein–HT complex. In the absence of tannin (or when bound with PEG), degradability of protein was higher, resulting in greater NH3-N concentration, possibly overcoming the inhibition of microbial deaminase by tannins (Leinmüller and Menke 1990). Similar findings were also observed by Min et al. (2003) who found that populations of ruminal proteolytic Butyrivibrio fibrisolvens were decreased when sheep were changed from a diet of perennial rye grass/white clover (which did not contain CT) to a diet of Lotus corniculatus containing CT (32 g CT kg−1 DM). Conversely, when PEG was infused into the rumen, populations of those proteolytic bacteria increased significantly (P < 0·05). Decrease (P < 0·05) in the NH3-N concentration was also attributed to the inhibition of the bacteria-degrading activity of protozoa (Jouany 1994). The overall results indicate that the candidate plants did not cause substantial modifications in fermentation parameters, apart from methane production and ammonia N, suggesting that these phenols did not affect substrate degradation and were not toxic to ruminal microbes.

Our results based on the effect of PEG on in vitro incubation established that plants containing secondary metabolites such as Rauvolfia serpentine, Indigofera tinctoria and Withania somnifera have great potential to suppress methanogenesis. Further, it was also established that methanogenesis was not essentially related to the density of protozoa population in vitro. The tannins contained in these plants could be of interest in the development of new additives in ruminant nutrition. The optimum dose to obtain maximum methane suppression without any adverse effect on digestibility needs to be determined.

Acknowledgements

The financial assistance provided to this work by the Indian Council of Agricultural Research (ICAR), New Delhi, under the ‘Outreach Project on Methane’ is gratefully acknowledged.

Ancillary