grain and artificial stimulation of the rumen change the...

Grain and artificial stimulation of the rumen change

the abundance and diversity of methanogens and their

association with ciliates

by

Claus Thagaard Christophersen

Candidatus Agronomiae (M.Sc.)

This thesis is presented for the degree of Doctor of Philosophy

of The University of Western Australia

Animal Science

School of Animal Biology

Faculty of Natural and Agricultural Sciences

December, 2007

ii

“We create the future by citing the past”

(Unknown)

Declaration

iii

Declaration

The work presented in this thesis is my own work except where stated below. This work

was carried out in the School of Animal Biology at the University of Western Australia

and at CSIRO Livestock Industries at Floreat Park. The material presented in this thesis

has not been presented for any other degree.

The volatile fatty acids analysed were by the Department of Agriculture, WA.

Claus Thagaard Christophersen, December, 2007.

Publications

iv

Publications arising from this thesis

Refereed publications:

- Christophersen C.T., Wright, A-D.G. and Vercoe, P.E. (2007) Methane emission and

acetate/propionate ratio decreased when artificial stimulation of the rumen wall are

combined with increasing grain diets. J. Anim. Sci. in press

Chapter 4 in this thesis is identical to the paper published in Journal of Animal Science

except that the abstract has been removed.

- Christophersen, C. T., Wright, A-D.G. and Vercoe, P. E. (2004). Examining

diversity of free-living methanogens and those associated with protozoa in the rumen. J.

Anim. Feed Sci. 13: 51 – 54.

Conference abstracts

- Christophersen, C. T., Wright, A-D. G. and Vercoe, P. E. (2004). Does dietary

manipulation change the diversity of methanogens and protozoa that interact within the

rumen? 4th joint INRA-RRI Symposium, Gut Microbiology – Concerns and Responses

to food safety, Health and Environmental issues, 21-23 June 2004, Clermont-Ferrand,

France. J. Repro. Nutri. Develop. 44: 52.

Acknowledgements

v

Acknowledgements

Here at the very end of my PhD I would like to thank several people, who have helped

me through and been there when times were tough. Without the financial support from

The University of Western Australia (IPRS) and the Danish Research Agency, it would

never have been possible for me to undertake my PhD candidature. I am very grateful

for being granted such a possibility. I was also awarded a University of Western

Australia travel grant to attend the INRA-RRI 2004 gut microbes conference. For this I

am very grateful.

Huge thanks to my supervisors, Dr. Philip E. Vercoe (UWA) and Dr. Andre-Denis G.

Wright, for not just being supervisors but also becoming great friends. They have both

strongly supported my scientific research intellectually, and they are excellent

supervisors and have wonderful personalities. Their encouragement and understanding

have been invaluable.

I also would like to thank Mr. John Beesley for his great support and guidance. John

was always happy to share his 30+ years of experience working with animals, and he

would also be the first to put his hand up when samples had to be collected. The

measurements of liquid and particulate matters retention time had not been possible

without him.

Dr. Clare Engelke, thank you for being such a good friend and for taking samples and

looking after the animals while I was away, you are a true “life saver”.

My unofficial co-supervisors Dr. Lucy C. Skillman and Dr. Richard Cookson I also owe

great thanks for taking the time to listen, discuss and answer all my questions.

Numerous people have provided help, advice, encouragement and friendship during my

PhD and it is impossible to mention all of them. However, I would like in particular to

Acknowledgements

vi

thank: Mr. Andrew Toovey, Mr. Peter Hutton, Dr. Ian Williams, Dr. Suzanne Rea, Mrs.

Carolyn Pimm, Mr. Andrew Williams, Dr. Zoey Durmic and Mrs. Margaret Blackberry.

Thanks to my parents, Birthe and Ove Thagaard Nielsen, for supporting me in achieving

my goals and for visiting us here “down-under”. I would also like to thank my parents-

in-law, especially my mother-in-law for visiting us in Perth so many times, and for all

the washing and cleaning and baby sitting, making life with two PhDs and two boys

much easier for us.

The love, encouragement and support from my lovely wife, Helle Martha

Christophersen, have been outstanding. Words just cannot express how grateful I am for

that.

Claus Thagaard Christophersen, December, 2007.

Abstract

vii

Abstract

In Australia, there is pressure to reduce the amount of methane produced by ruminant

livestock because they are the single largest source of methane emitted from

anthropogenic sources, accounting for 70.7% of agricultural methane emissions. In

addition, methane production represents a loss of gross energy intake to the animal. The

organisms that are responsible for methane production in the animal gut are a distinct

group of Archaea called methanogens. Methanogens occupy three different niches

within the rumen. Some live freely in the rumen digesta (planktonic), others are

attached to the outer surface of the rumen ciliates (ectosymbiotic), and some reside

within the ciliates (endosymbiotic). The types and number of methanogens, as well as

rumen ciliates and their symbiotic interactions, influence the amount of methane

produced from the rumen. These factors in turn are affected by many factors, including

diet and ruminal retention time. In this thesis, I tested the general hypothesis that

increasing the amount of grain in the diet and reducing the retention time would affect

the abundance and diversity of methanogens in their different niches, including their

association with ruminal ciliates.

Twenty-four fistulated sheep were used in a complete factorial design with the

sheep randomly divided into four groups. The sheep had a 2-wk acclimatization period

on an oaten-chaff diet, followed by three, 3-wk diet phases. In diet phase 1 all sheep

were given the same oaten-chaff diet. Two of the four groups were maintained on the

oaten-chaff diet for the duration of the experiment with pot scrubbers added to the

rumen of one of the two groups. The remaining two groups were offered a low grain

diet (35% grain) in the second diet phase followed by a high grain diet (70% grain) in

the third diet phase. Pot scrubbers were also added to the rumen of one of these two

groups of grain-fed sheep. The pot scrubbers were inserted with the intention of

Abstract

viii

increasing rumen stimulation without changing the diet composition. Ruminal pH,

volatile fatty acids was measured to monitor rumen fermentation. In addition, methane

production (in vitro) and ruminal retention time (in vivo) were measured in the 4

treatment groups at the end of each diet phase. A robust real-time PCR assay was

developed to quantify methanogens and a denaturing gradient gel electrophoresis

(DGGE) assay was developed to examine the diversity of methanogens. Rumen ciliates

were also examined using the same methods but with already published protocols. The

DGGE gels were analysed using Gelcompar II and the resulting DGGE banding

patterns were analysed using a multivariate statistical software program (PRIMER). The

Shannon index for each treatment group was also calculated based on the banding

patterns to indicate the direction of diversity changes.

The addition of grain and pot scrubbers changed rumen parameters. Methane

production in vitro was reduced on a low grain diet in sheep with pot scrubbers.

Methane production was reduced in sheep fed the high grain diet, with or without pot

scrubbers. Acetate/propionate ratios were also lower in sheep fed the high grain diet.

The total abundance of methanogens and ciliates in diet phase 3 was not different

between treatments. However, the abundance of methanogens associated

endosymbiotically with rumen ciliates was significantly higher in high grain-fed sheep

with and without pot scrubbers.

Fifteen of the most dominant methanogen DGGE bands were sequenced and

identified using a BLAST search. Ten of the 15 different bands had >98% identity to

Methanobrevibacter spp., whereas the other five bands were found to be between 82%

and 96% similar to Methanobrevibacter spp. The diversity of methanogens and ciliates

were also found to vary between treatments. The change in DGGE banding patterns and

Shannon indices when sheep were fed grain indicated that the types of methanogens

Abstract

ix

changed when sheep were fed low and high grain diets, but their diversity did not. In

contrast, the diversity of rumen ciliates decreased when sheep were fed a high grain

diet. A total of 18 bands from the DGGE analysis of the ciliates were sequenced. All

except one, which was 98% similar to Cycloposthium sp. not found previously in the

rumen, matched the sequences for previously identified rumen ciliates. Some of the

rumen ciliates identified were not present in sheep fed the high grain diet.

On a high grain diet, methanogens associate endosymbiotically with rumen

ciliates to get better access to hydrogen. It appears that the association between

methanogens and rumen ciliates is dictated by the availability of hydrogen in the rumen

and not the generic composition of the ciliate population. Furthermore, endosymbiotic

methanogens appear to produce less methane than methanogens in other niches. The pot

scrubbers did not change ruminal retention time but they did reduce the

acetate/propionate measurements observed in sheep on the high grain treatment. The

reason why pot scrubbers had this effect remains unknown, but it is interesting to

consider that some physical interaction has occurred between the pot scrubbers, the

grain and the sheep that has improved the fermentation parameters in sheep fed a high

grain diet. The results from this study have advanced our understanding of the

interaction between methanogens and ruminal ciliates, and methanogenesis in the rumen

in response to dietary changes and mechanical challenges. Extending this work to look

more specifically at the species of methanogens that are most closely linked to high

methane production and how they interact with the ruminal ciliates will be critical for

manipulating enteric greenhouse gas emissions.

Table of content

x

Table of content

Chapter 1: General introduction ..............................................................................................................1

Chapter 2: Literature review ....................................................................................................................4

2.1 Introduction .......................................................................................................................................5

2.2 Taxonomy and characteristics of rumen methanogens......................................................................6

Taxonomy ......................................................................................................................................7

Pathways of methanogenesis .........................................................................................................8

Substrate range...............................................................................................................................9

Ecology ........................................................................................................................................11

2.3 Taxonomy and characteristics of rumen ciliates..............................................................................12

2.3.1 The entodiniomorphid ciliates .................................................................................................13

Digestion and metabolism of dietary components .......................................................................13

2.3.2 The vestibuliferid ciliates ........................................................................................................14

Digestion and metabolism of dietary components .......................................................................14

2.3.3 Interrelationships between species of rumen ciliates ...............................................................15

2.4 Interaction between methanogens and rumen ciliates .....................................................................17

Niches occupied by methanogens in the rumen...........................................................................17

Why is there an interaction between ciliates and methanogens? .................................................17

Methanogens and ciliate species that associate............................................................................19

Time after feeding and the interaction between ciliates and methanogens ..................................20

2.5 Grain diets and ruminal retention times influence hydrogen availability, methanogens, rumen

ciliates and their association..................................................................................................................22

Hydrogen availability...................................................................................................................22

Methanogens, rumen ciliates and their interaction.......................................................................23

2.6 Value and limitations of two molecular methods for examining methanogens and ciliates in the

rumen.....................................................................................................................................................27

2.6.1 Denaturing gradient gel electrophoresis (DGGE)....................................................................27

2.6.2 Quantitative real-time PCR (real-time PCR) ...........................................................................32

2.7 Summary .........................................................................................................................................36

Table of content

xi

Chapter 3: Materials and methods .........................................................................................................38

3.1 Introduction .....................................................................................................................................39

3.2 Experimental design ........................................................................................................................39

Chapter 4: In vitro methane emission and acetate/propionate ratio are decreased when artificial

stimulation of the rumen wall is combined with increasing grain diets in sheep................................41

4.1 Introduction .....................................................................................................................................42

4.2 Materials and methods.....................................................................................................................43

Experimental design ....................................................................................................................43

Estimation of mean retention time of liquid and particulate matter .............................................43

Volatile fatty acids, pH and in vitro methane production ............................................................45

Statistical analyses .......................................................................................................................46

4.3 Results .............................................................................................................................................46

Effect of diet ................................................................................................................................46

Effect of pot scrubbers .................................................................................................................49

Combined effect of diet and pot scrubbers ..................................................................................49

4.4 Discussion .......................................................................................................................................50


Effect of diet ................................................................................................................................52

Effect of pot scrubbers .................................................................................................................53

Chapter 5: Grain and artificial stimulation of the rumen wall changes the association between

methanogens and rumen ciliates .............................................................................................................55

5.1 Introduction .....................................................................................................................................56

5.2 Materials and Methods ....................................................................................................................58


Rumen sampling ..........................................................................................................................58

DNA extraction and quantification ..............................................................................................59

Denaturing Gradient Gel Electrophoresis (DGGE) .....................................................................60

Phylogenetic analysis...................................................................................................................62

Quantitative real-time PCR..........................................................................................................63

Table of content

xii

Statistical analysis and diversity index ........................................................................................66

Nucleotide sequence accession number .......................................................................................66

5.3 Results .............................................................................................................................................66

Effect of diet ................................................................................................................................66


Identification of DGGE bands and their phylogenetic relationship .............................................71

Validation of real-time PCR assay...............................................................................................74

5.4 Discussion .......................................................................................................................................75

Validation of real-time PCR and DGGE......................................................................................81

Chapter 6: Grain changes the diversity of rumen ciliates but not their abundance ..........................84

6.1 Introduction .....................................................................................................................................85

6.2 Materials and methods.....................................................................................................................86


Rumen sampling, extraction and quantification of DNA.............................................................86

Denaturing gradient gel electrophoresis (DGGE)........................................................................87

Real-time PCR .............................................................................................................................88

Statistical analysis and diversity index ........................................................................................89

Nucleotide sequence accession number .......................................................................................90

6.3 Results .............................................................................................................................................90

Effect of treatments......................................................................................................................90

Identification of DGGE bands .....................................................................................................93

6.4 Discussion .......................................................................................................................................94

Chapter 7: General discussion ................................................................................................................99

Future studies.............................................................................................................................107

Conclusion .................................................................................................................................108

References ...............................................................................................................................................110

Chapter 1: General introduction

1

CHAPTER 1

General introduction


2

Methane is the second most important greenhouse gas emitted from anthropogenic

sources and has a global warming potential 23 times more potent than carbon dioxide

(Wuebbles and Hayhoe, 2002). Globally, methane production from ruminants accounts

for about 28% of all methane produced from anthropogenic sources (Food and

Agriculture Organization of the United Nations, 2000). In Australia, ruminant livestock

are the single largest source of methane emissions, accounting for 70.7% of agricultural

methane emissions (Australian Greenhouse Office, 2007). Methane is formed in the

rumen during fermentation of feed by methanogenic Archaea (methanogens), expired

via the lungs and exhaled at the nose and mouth. The production of methane represents

an energy loss to the animal, which has been estimated to be between 2 – 15% of the

animal’s gross energy intake (Johnson and Johnson, 1995; McAllister et al., 1996; Van

Nevel and Demeyer, 1996).

The rumen methanogens occupy three different niches within the rumen. Some

live freely in the rumen (free-living), others are attached to the outer surface of the

rumen ciliates (ectosymbiotic), and some reside within the rumen ciliates

(endosymbiotic) (Vogels et al., 1980; Stumm et al., 1982). The percentage of methane

produced by methanogens living in or on the rumen ciliates has been estimated to be

between 9 - 37% (Finlay et al., 1994; Newbold et al., 1995). Therefore, rumen ciliates

have a significant role in methane production from ruminants (Krumholz et al., 1983;

Finlay et al., 1994; Newbold et al., 1995). There is also evidence that changes in the

abundance of rumen ciliates can affect methanogenesis, as Krumholz et al. (1983)

found that the methanogenic activity in rumen fluid was highest in fractions containing

high numbers of protozoa. Furthermore, a change in the generic composition of the

rumen ciliates can also lead to a change in methane production (Itabashi et al., 1994).


3

The amount of methane emitted from ruminants depends on the conditions in the

rumen, which are controlled by factors including diet and rumen retention time

(McAllister et al., 1996). For example, methane production decreases from ruminants

when fed high levels of grain (Russell, 1998). This is thought to occur because of an

increased competition for hydrogen between the methanogens and the hydrogen-

utilising propionate producing bacteria. One reason methanogens associate with ciliates

is to get access to hydrogen and it is likely that when sheep are fed a high grain diet the

competition for hydrogen would affect this association. However, how changing these

factors affect the numbers and diversity of methanogens, rumen ciliates, and especially

their association is largely unknown. It would be beneficial to explore this gap in

knowledge to help reduce methane emissions from ruminant livestock. Therefore, in

this thesis I have examined the effect that increasing the grain content of the diet and

decreasing ruminal retention time has on the number and diversity of rumen

methanogens and rumen ciliates, and how this affects their association. The general

hypothesis tested in this thesis was that increasing the amount of grain in the diet and

reducing the retention time would affect the abundance and diversity of methanogens in

their different niches, as well as their association with ruminal ciliates.

Chapter 2: Literature review

4

CHAPTER 2

Literature review


5

2.1 Introduction

The rumen contains a microbial population made up of Archaea (methanogens) (Woese

et al., 1990), bacteria, ciliates and fungi (Hungate, 1966). These microorganisms

ferment the food that a ruminant consumes to produce energy. The microorganisms in

the rumen also function via complex interactions with each other. Rumen microbes have

developed different strategies to survive in a highly competitive environment where a

change of feed source is likely to make significant changes to the structure of the

microbial ecosystem. Due to the complexity of the rumen environment this Chapter is

limited to a review of the methanogens and the ciliates living in the rumen.

Understanding how different rumen manipulations, in particular diet and retention time,

affect methanogens and their association with the rumen ciliates would be beneficial for

reducing methanogenesis from ruminant livestock. In order to understand why change

in diet and decreased retention time in the rumen may change the diversity and the

numbers of methanogens, ciliates and the close association between them, four specific

fields of literature need to be examined in more detail: The first two areas are a

description of the taxonomy and characteristics of rumen methanogens and ciliates.

Then the association between methanogens and rumen ciliates will be explained, and

finally the value and limitations of molecular techniques, denaturing gradient gel

electrophoresis (DGGE) and quantitative real-time PCR, used to study rumen microbes

will be covered.


6

2.2 Taxonomy and characteristics of rumen methanogens

The organisms that produce methane are a distinct group of Archaea (Woese et al.,

1990) called methanogens. Methanogens are a normal component of the microbial

population in the rumen, but are also found in a wide range of other environments

(Miller and Wolin, 1986).

Classification of methanogens was initially based on a wide range of

characteristics because they were considered to be bacteria. The minimal standards for

classifying methanogens was monoculture, morphology, Gram staining, electron

microscopy, susceptibility to lysis, motility, colony morphology, nutritional spectrum,

end products, growth rates, growth conditions, G + C content of the DNA, lipid

analysis, cell wall structure, protein analysis and antigenic fingerprinting (Boone and

Whitman, 1988). However, classification is now based almost solely on DNA (16S

rRNA) analysis and it has been determined that Archaea belong to their own

phylogenetic kingdom (Woese et al., 1990). Balch (1979) reorganized the taxonomy of

the methanogens based upon these phylogenetic relationships. The 16S rRNA gene has

been very useful because it is conserved in all known species of methanogens, but is

different from the 16S rRNA gene found in other Archaea and bacteria. In this section I

will concentrate on the taxonomy of methanogens, pathway of methanogenesis, their

substrate range and their ecology. These areas are central in the understanding of

methanogenesis from ruminants, as different methanogens have different affinity for

hydrogen (Zinder, 1993). Therefore, a change in the methanogens population can

potentially result in a changed methane production from the rumen.


7

Taxonomy

Based on the above characteristics, Archaea are classified into four phyla within the

domain Archaea: Crenarchaeota, Euryarchaeota, Korarchaeota and Nanoarchaeota

(Barns et al., 1994; 1996; Burggraf et al., 1997a; 1997b; Huber et al., 2002). To date,

the only Archaea identified in the rumen are methanogens belonging to the phylum

Euryarchaeota. Within this phylum rumen methanogens have been identified in two

classes (Methanobacteria and Methanomicrobia). Methanobacteria and

Methanomicrobia consist of three orders, Methanobacteriales, Methanomicrobiales and

Methanosarcinales, which are described in more detail because they contain rumen

methanogens (Ferry et al., 1974; Bryant and Boone, 1987; Whitman et al., 1991; Garcia

et al., 2000).

The order Methanobacteriales, comprises non-motile methanogens, is divided

into two families: Family I, Methanobacteriaceae, contains four genera

(Methanobacterium, Methanobrevibacter, Methanosphaera and Methanothermus). For

this thesis the three most important genera in this family are Methanobacterium,

Methanobrevibacter and Methanosphaera , because they are the only genera that

contain methanogens observed in the rumen (Miller and Wolin, 1985).

Methanobrevibacter is in fact the major archaeal genus found in the rumen and contains

species like Mbr. ruminantium and Mbr. smithii (Smith and Hungate, 1958; Miller and

Wolin, 1986). Family II, Methanothermaceae, contains one genus (Methanothermus),

but it does not contain any methanogens that have been observed in the rumen.

The order Methanomicrobiales comprises three families (Methanomicrobiaceae,

Methanocorpusculaceae and Methanospirillaceae) and nine genera

(Methanocorpusculum, Methanoculleus, Methanofollis, Methanogenium,

Methanolacinia, Methanomicrobium, Methanoplanus, Methanospirillum and


8

Methanocalculus). Only methanogens belonging to the genera Methanomicrobium and

Methanospirillum have been identified from the rumen (Ferry et al., 1974).

The order Methanosarcinales is divided into two families (Methanosarcinaceae

and Methanosaetaceae) and nine genera (Methanosaeta, Methanimicrococcus,

Methanococcoides, Methanohalobium, Methanohalophilus, Methanolobus,

Methanomethylovorans, Metthanosalsum and Methanosarcina). The family

Methanosarcinaceae contains methanogens living in the rumen. These methanogens

belong to the genus Methanosarcina and are acetotrophic (Bryant and Boone, 1987).

Recently, the existence of a novel group of methanogens has been suggested as a

new order, but cultivars have to be characterised before a new order can be accepted

(Wright et al., 2004, 2006, 2007). It consists of uncultured archaeal sequences from

diverse anaerobic environments, which are distantly related (>20%) based on 16S rRNA

sequence similarity to Thermoplasma.

Pathways of methanogenesis

The production of methane gas is the major source of energy for growth of

methanogens. It might be expected that the reduction of primarily C-1 compounds

would be a simple reaction, but because of the structure and synthesis of many of the

coenzymes involved in methanogenesis, biochemically, it is a complex process

(Whitman et al., 1991). The pathway of methanogenesis is slightly different depending

on the substrate. The best-known synthesis of methane is the reduction of CO2 to CH4,

which has seven steps that involve a series of co-enzymes of which many are unique to

methanogens (Rouviere and Wolfe, 1988) (Figure 2.1). The source of electrons is either

H2 or formate (Whitman et al., 1991). The cycle starts with the activation of CO2, which

requires energy in the form of ATP (Figure 2.1). The energy for the activation is derived

from the final step of methane synthesis. After the initial step, several co-enzymes are


9

involved before CH4 is formed (Figure 2.1). The key intermediate in this process is

methyl-co-enzyme M, which is formed after the sixth step and is required in the methyl

reductase system that represents the completion of the cycle with the release of CH4 and

activation of CO2 for the next cycle (Figure 2.1). Methyl-co-enzyme M is important

because it is also required for methanogenesis from substrates other than CO2 (Whitman

et al., 1991).

Figure 2.1: The pathway of methane formation from acetate, methanol and CO2 goes through seven steps. The numbers refer to the seven steps of the cycle. MFR: Methanofuran, HS-HTP: 7-Mercaptoheptanoylthreonine phosphate, H4MPT: Tetrahydromethanopterin, F420: Coenzyme F420, HS-CoM: Coenzyme M, F430: Coenzyme F430 [adapted from Rouviere and Wolfe (1988)].

Substrate range

The substrate range of methanogens is limited despite the large phylogenetic diversity.

Methanogens are divided into three main nutritional categories, on the basis that some

of them can use more than one substrate for methanogenesis, which means that these

methanogens cannot be placed in a single category. The three categories are: i)

hydrogenotrophic methanogens (e.g. Methanobrevibacter), which oxidize H2 and reduce

CO2. This category also includes the utilization of formate, certain alcohols and has the


10

highest energy conversion during methanogenesis (Table 2.1); ii) methylotrops (e.g.

Methanosphaera), which utilise methyl compounds such as methylamines, methanol or

dimethylsulfide; and iii) acetotrophic methanogens (e.g. Methanosarcina), which can

produce methane from acetate (Whitman et al., 1991; Garcia et al., 2000).

Table 2.1: Overview of the most common methanogenic reactions and their energy yield (Whitman et al., 1991). Reactants Products ∆G˚`(kJ/mol CH4) 4 H2 + HCO3

- + H+ CH4 + 3 H2O - 136

4 HCO2- + H+ + H2O CH4 + 3 HCO3- - 130

2 CH3CH2OH + HCO3- CH4 + 2 CH3COO- + H+ + H2O - 116

CH3COO- + H2O CH4 + HCO3- - 31

4 CH3OH 3 CH4 + HCO3- + H+ + H2O - 105

CH3OH + H2 CH4 + H2O - 113

4 (CH3)3-NH+ + 9 H2O 9 CH4 + 3 HCO3- + 4 NH+ + 3 H+ - 74

2 (CH3)2-S + 3 H2O 3 CH4 + HCO3- + 2 H2S + H+ - 74

Other short chain alcohols, methylated amines and methyl mercaptan are utilized.

The most widespread catabolic reaction performed by methanogens is the

reduction of CO2 to CH4 using H2 as a reductant (Table 2.1). This is particularly

relevant to the rumen because many other ruminal microbes produce H2 as a major

fermentation end-product (Zinder, 1993). Formate is also used by hydrogenotrophic

methanogens but it can be hard to detect because it is rapidly converted to H2 and CO2

(Garcia et al., 2000). When short chain alcohols are used in methanogenesis, the

alcohols are often oxidized to volatile fatty acid (VFA) (Garcia et al., 2000) (Table 2.1).

In the rumen, all three categories of methanogens are represented, but the majority of

rumen methanogens belong to the hydrogenotrophic methanogens.


11

Ecology

Methanogens normally compete for hydrogen with three other major groups of bacteria.

The three competitors are sulphate reducing bacteria, acetogens and ferric iron (Fe3+)

reducers. In natural environments where substrates (electron donors) are limited, there is

a hierarchy of hydrogen utilisers. Ferric iron reducers are at the top of the hierarchy

followed by sulphate reducing bacteria, methanogens and then acetogens provided that

the respective electron acceptors are present (Zinder, 1993). This hierarchy is in

accordance with the energy yield from the reactions where Fe3+ reducing bacteria have

the highest ∆G˚ (Zinder, 1993). However, it has been shown that acetogens can compete

with methanogens in vitro (Joblin, 1999), but this observation needs to be validated

under in vivo conditions.

In the rumen, levels of hydrogen vary depending on factors like diet, retention

time and pH, and certain ruminal conditions can favour specific microorganisms, while

suppressing others. For example a high grain diet will enhance the abundance of

propionate producing bacteria and therefore propionate production, which requires

hydrogen, and thereby reduce methane production (Moss et al., 1995; Lana et al., 1998;

Tajima et al., 2001a). However, this competition between hydrogen utilisers may be

different in vivo as opposed to in vitro, due to the interaction between microorganisms.

Methanogens have been found to have a close association with ciliates in the rumen

(Vogels et al., 1980; Stumm et al., 1982; Stumm and Zwart, 1986) and this association

may change the hierarchy of competitors, as methanogens have been found associated

with hydrogen producing organelles (i.e. hydrogenosomes) in ciliates (Van Hoek et al.,

2000).


12

2.3 Taxonomy and characteristics of rumen ciliates

The main focus of this section relates to the digestion of other ruminal microorganisms,

carbohydrates and metabolism of dietary compounds by ciliates and the

interrelationships between species of rumen ciliates. This forms the foundation for the

coming section centred around the association between methanogens and ciliates. The

metabolism of rumen ciliates is an important factor when examining rumen methane

production, because there is evidence that both a change in the abundance of rumen

ciliates and a change in the generic composition can lead to a change in methane

production (Krumholz et al., 1983; Itabashi et al., 1994).

The largest of the rumen microorganisms are the ciliates. They are divided into

groups within the subclass Trichostomatia. These are the vestibuliferids (order

Vestibuliferida), and the entodiniomorphids (order Entodiniomorphida) (Williams and

Coleman, 1992). Analysis of the 18S rRNA genes confirmed that the entodiniomorphid

and vestibuliferids ciliates belong to two different orders of the class Litostomatea

(Wright et al., 1997; Wright and Lynn, 1997a; Wright and Lynn, 1997b). The two

groups are quite different and will be considered separately in this section.

Traditionally, ciliate taxonomy has relied on morphology and cellular

ultrastructures. Most rumen ciliates are difficult to identify because of their small size

and limited morphology, and most of the described species have never been cultured

(Williams and Coleman, 1997). Another problem with identifying rumen ciliates is that

the original classifications were based on both internal and external characteristics, but

unless samples are taken from an animal that has been starved for 24 hours, the ciliates

are full of starch and internal structures are difficult to observe (Williams and Coleman,

1997). Therefore, new molecular tools have been developed to examine rumen ciliate


13

diversity and abundance using the 18S rRNA gene (Regensbogenova et al., 2004b;

Sylvester et al., 2005; Skillman et al., 2006b).

2.3.1 The entodiniomorphid ciliates

Digestion and metabolism of dietary components

Bacteria - Bacteria are believed to be the most important single source of nitrogenous

compounds for the rumen ciliates. There is no consistent pattern across the ciliates

regarding their preference for certain bacteria, even though the bacteria Selenomonas

ruminantium and Butyrivibrio fibrisolvens are almost always taken up faster or at the

same rate as other bacteria. In contrast, bacterial species like Escherichia coli and

Prevotella ruminicola are not engulfed regularly or are taken up very slowly (Williams

and Coleman, 1997). The rate at which ciliates take up bacteria has been compared in

two ways: (a) the rate of uptake from an infinitely dense suspension, which measures

the rate at which ciliates can pass bacteria down the oesophagus and form food vesicles,

and (b) clearance-rate of bacteria from an infinitely dilute suspension, which measures

the ability of ciliates to find and capture prey. For the conditions in the rumen (a) would

be the most relevant (Williams and Coleman, 1997). It has been found that the uptake of

bacteria is relatively unaffected by changes in salt concentrations, whereas pH plays a

vital role, with pH 6 being the optimum, 75% uptake at pH 7 and no uptake at pH 5

(Coleman and Sandford, 1979). It is noteworthy that the uptake of the yeast

Saccharomyces fragilis is almost independent of population density as they are digested

at a steady rate (Williams and Coleman, 1992). After engulfment different bacteria are

digested at different rates. The rate of digestion depends on their cell wall. Bacteria with

a cell wall that is comparatively resistant to lysozyme have their cell contents digested


14

before there is extensive digestion of their cell wall. In contrast, Gram-negative bacteria,

like Escherichia coli, are digested very rapidly (Coleman and Hall, 1972).

Rumen fungi - There is evidence that fungal rhizoids, zoospores and sporangia

are all engulfed by ciliates, as there appears to be an inverse relationship between the

population densities in the rumen of ciliates and fungi (Williams and Coleman, 1997).

However, Newbold and Hillman (1990) believed that ciliates play a greater role in the

turnover of bacterial protein than in the turnover of fungal protein.

Carbohydrates - The uptake rate of starch grains varies between rumen ciliates.

Entodinium spp. engulf starch grains very rapidly initially and then much more slowly.

Epidinium spp. behave similarly, but at slower rates. In contrast, the larger

entodiniomorphid ciliates engulf starch grains more slowly, but at a constant rate for

several hours (Coleman, 1992). The fermentation of starch grains in ciliates are similar

and the principal products are hydrogen, carbon dioxide, acetic acid, butyric acid and

glycerol, depending on the oxygen and carbon dioxide in the gas phase (Ellis et al.,

1991a, b). The starch is digested to maltose and then glucose, which is phosphorylated

to glucose-6-phosphate, which can be stored or metabolised to produce energy

(Williams and Coleman, 1997).

2.3.2 The vestibuliferid ciliates

Digestion and metabolism of dietary components

Metabolic and biochemical studies of the vestibuliferid ciliates have been undertaken

only with Dasytricha ruminantium and the two species of Isotricha (Isotricha prostoma

and I. intestinalis) (Williams and Coleman, 1997).

Carbohydrates - The vestibuliferids are believed to be involved in the utilisation

of soluble sugars and non-structural polysaccharides. They are able to utilise fructose,


15

glucose and galactose and certain soluble oligomers and polysaccharides containing one

or more of these sugars; fructose containing carbohydrates are utilized most rapidly.

Furthermore, the range of carbohydrates metabolised is genus dependent (Williams and

Coleman, 1997). The rate at which sugars are taken up is affected by the nature and

concentration of the sugar and by the rumen pH and temperature (Williams and Harfoot,

1976). The vestibuliferid ciliates have also been found to have a chemotaxic attraction

to sources of sugar (Orpin and Letcher, 1978). These ciliates are able to maintain close

contact to carbohydrate sources by colonising plant tissue (Orpin and Letcher, 1978).

Despite differences in the range of disaccharides fermented by Dasytricha and Isotricha

their enzyme profiles are identical, but the activities of the different enzymes are

different (Williams and Coleman, 1997). The ciliates belonging to the order

Vestibuliferida are also able to store polysaccharides as a branched homoglucan, similar

to amylopectin, and it is believed that initially 75-80% of the sugar taken up is

converted into this storage polymer (Williams and Coleman, 1997).

Metabolites formed during carbohydrate fermentation by Dasytricha

ruminantium and Isotricha spp. are lactic acid, butyric acid, acetic acid, hydrogen,

carbon dioxide, storage polysaccharides and small amounts of propionic acid and

alanine (Ellis et al., 1991b). The metabolite formation is affected by diet, nutrient status

of the ciliate, rumen pH, temperature, presence of oxygen, headspace gas composition,

and metabolic interactions with other microbial groups (e.g. methanogens) (Williams

and Coleman, 1997).

2.3.3 Interrelationships between species of rumen ciliates

The different species of rumen ciliates are not all present in the rumen at the same time

because certain species of ciliates feed on other ciliate species. This means that different

population types have been identified and they have been designated Types A, B, O and


16

K (Table 2.2) (Eadie, 1962). The A-type population is characterised by the presence of

Polyplastron multivesiculatum (Table 2.2). The B-type population is characterised by

Epidinium spp. and/or Eudiplodinium maggii (Table 2.2). The K-type (found in cattle

only) is characterised by the presence of Elytroplastron bubali and O-type contains only

the vestibuliferids, Dasytricha and Isotricha (i.e. no entodiniomorpids) (Table 2.2).

However, within these groups individual ciliate species can appear and disappear for no

apparent reason (Williams and Coleman, 1992).

The different population types are not equally common. A- and B-type

populations are dominant in sheep and in a flock of sheep it would be likely to find

approximately equal numbers of A- and B-type populations. However, in New Zealand,

no B population has been found in their animals (Williams and Coleman, 1992).

Table 2.2: A classification scheme of different ciliate population types found in cattle, sheep and goats. Table was taken from Williams and Coleman (1992). Ciliates Type A Type B Type O Type K*

Entodinium X X X

Isotricha X X X X

Dasytricha X X X X

Diplodinium X X X

Eremoplastron X X X

Diploplastron affine X X

Eodinium X X

Enoploplastron X X

Ostracodinium X X

Ophryoscolex X

Polyplastron multivesiculatum X

Metadinium X

Epidinium X

Eudioplodinium maggii X

Elytroplastron bubali X

*Cattle only


17

2.4 Interaction between methanogens and rumen ciliates

A close symbiosis between methanogens and ciliates has been observed in the rumen

(Vogels et al., 1980; Stumm and Zwart, 1986; Embley and Finlay, 1993), and various

species of methanogens have been found to associate with rumen ciliates (Vogels et al.,

1980; Stumm et al., 1982; Finlay et al., 1994; Newbold et al., 1995; Tokura et al., 1997;

Sharp et al., 1998; Chagan et al., 1999; Tokura et al., 1999; Schonhusen et al., 2003;

Irbis and Ushida, 2004; Regensbogenova et al., 2004a). In this part of the literature

review I outline the reasons why the interaction is thought to exist and identify the

species of methanogens and ciliates that have been found to associate together.

Niches occupied by methanogens in the rumen

Methanogens occupy three different niches within the rumen. Some live freely in the

rumen digesta (planktonic), some are attached to the outer surface of the rumen ciliates

(ectosymbiotic), and some reside within the ciliates (endosymbiotic) (Vogels et al.,

1980; Stumm et al., 1982; Stumm and Zwart, 1986; Finlay et al., 1994; Newbold et al.,

1995; Tokura et al., 1997; Sharp et al., 1998; Chagan et al., 1999; Tokura et al., 1999;

Schonhusen et al., 2003; Irbis and Ushida, 2004; Regensbogenova et al., 2004a).

However, some methanogens can attach and detach themselves to ciliates depending on

the conditions in the rumen (Stumm et al., 1982; Tokura et al., 1997), which means that

they can be part of the free-living and ectosymbiotic pools.

Why is there an interaction between ciliates and methanogens?

Both ciliates and methanogens are thought to benefit from their association. The

advantages for the ciliates are thought to be that the methanogens keep the H2

concentration low, which enhances the energy yield per mol glucose converted by the


18

ciliates (Hino, 1982; Stumm and Zwart, 1986). At low hydrogen concentrations,

reoxidation of reduced coenzymes via ferrodoxin-linked hydrogenase is favoured over

reoxidation by fermentation reactions because of the higher energy yields (Stumm and

Zwart, 1986).

The endosymbiotic methanogens have access to substrates for methanogenesis

from the ciliate’s metabolism. Stumm et al. (1982) proposed that the attachment of

methanogens to the ciliates was dependent on the H2 pressure in the surroundings.

However, according to results obtained by Ushida et al. (1997), it is more likely that

methanogens use ciliates as an easy way to access substrates for methanogenesis, for

example H2 or formate. Methanogens most likely use formate produced by the ciliates,

as well as hydrogen, based on evidence provided by Hutten et al. (1980), Finlay and

Fenchel (1992) and Ushida et al. (1997). Hydrogen is considered to be the main

substrate and of greatest value to the endosymbionts because they have been observed

in close assemblages or association with hydrogenosomes. Hydrogenosomes are

membrane-bound organelles, like mitochondria, that produce ATP. They have only

been found in anaerobic organisms that cannot use oxygen as an electron acceptor, but

reduce protons to molecular hydrogen instead and are involved in terminal steps of

anaerobic energy metabolism (Van Hoek et al., 2000). Endosymbiotic methanogens

also have the advantage of living in a protected, oxygen-free, environment inside the

ciliates. Methanogens can actually tolerate small amounts of oxygen (Scott et al., 1983),

but their oxygen tolerance increases in the presence of rumen ciliates even if the

methanogens are not endosymbiotic, due to the use of oxygen by the hydrogenosomes

(Hillman et al., 1988).

The ectosymbiotic methanogens are thought to attach themselves to the surface

of the ciliates to get easy access to substrates for methanogenesis via interspecies


19

hydrogen transfer (Stumm and Zwart, 1986; Ushida et al., 1997). Hydrogenosomes are

generally located near the cell surface and hydrogen from the hydrogenosomes diffuses

to the cell surface where it is absorbed by methanogens (Van Hoek et al., 2000).

The ability of some methanogens to associate with rumen ciliates and not others

is not understood. Even though a broad range of methanogens have been identified to

associate with ciliates, there is a requirement for a better understanding of how diet

affects the ecto- and endosymbiotic association between methanogens and ciliates in the

rumen.

Methanogens and ciliate species that associate

Recently, studies have been conducted to examine the species of methanogens that

associate with rumen ciliates using molecular analyses. In most of these studies the

most abundant methanogens associating with rumen ciliates were Methanobrevibacter

smithii and Methanobrevibacter gottschalkii-like (>99% similarity), within the order

Methanobacteriales (Sharp et al., 1998; Chagan et al., 1999; Tokura et al., 1999; Irbis

and Ushida, 2004). Regensbogenova et al. (2004a) also found that the main

methanogens associated with the ciliates belonged to the order Methanomicrobiales, but

also the order Methanosarcinales. In all of these experiments the samples examined

came from a single animal fed different diets and limited data were available in relation

to the overall diversity of methanogens in the samples. With the exception of Sharp et

al. (1998), who examined the crude rumen fluid as well as the ciliate fraction isolated

from the rumen fluid, both the crude rumen fluid and the ciliate fraction was examined

using hybridisation probes. They found methanogens from the orders

Methanobacteriales, Methanomicrobiales and Methanosarcinales in both the ciliate

fraction and the rumen fluid, but there was only a very low representation of

Methanomicrobiales in the ciliate fraction. This indicates that it may not be a specific


20

group of methanogens that associate with the rumen ciliates, but a general

representation of the methanogens found in the rumen fluid. Furthermore,

Regensbogenova et al. (2004a) found no preference from any ciliates to associate with

specific groups of methanogens. Nevertheless, these studies were performed on rumen

fluid from very few animals, which could also explain the differences observed between

the studies. This also means that there is a paucity of knowledge about what effect a

change in diet and/or ruminal retention time has on the different methanogen

populations and their association with the rumen ciliates.

Using traditional microscopy, the rumen ciliates that associate with

methanogens have been reported to belong to the order of Entodiniomorphida (Vogels

et al., 1980), but in later studies using molecular techniques the vestibuliferid ciliates

have been found to interact also with methanogens (Irbis and Ushida, 2004). This

demonstrates that traditional methods may be limited in terms of giving a complete

description of the methanogen and ciliate association, when compared to molecular

tools.

Time after feeding and the interaction between ciliates and methanogens

It has been proposed by Stumm et al. (1982) that at least the ectosymbiotic association

between methanogens and rumen ciliates is controlled by the surrounding physico-

chemical conditions such as the hydrogen partial pressure, with more methanogens

associated with ciliates at low hydrogen pressure. Using microscopy, Smolenski and

Robinson (1988) determined that the association between methanogens and ciliates

decreases from 65% before feeding to 30% an hour after feeding and that hydrogen

levels increase in the rumen at that time. In contrast, Tokura et al. (1997) found, using

molecular techniques, that the number of methanogens per ciliate increased shortly (1h)

after feeding and then decreased. The difference between these two studies may be due


21

to the different methodologies that were used, as the microscopic counting would not

include endosymbiotic methanogens. The difference may also have been due to the

difference in diet between the two studies (Hegarty, 1999). In both studies, methane

production was found to be highest an hour after feeding (Stumm et al., 1982; Tokura et

al., 1997).


22

2.5 Grain diets and ruminal retention times influence hydrogen

availability, methanogens, rumen ciliates and their association

It is well established that the diversity of methanogens and rumen ciliates can be

influenced by diet (Williams and Coleman, 1992; Wright et al., 2004; 2006; 2007) and

rumen parameters also change depending on factors like ruminal retention time

(Faichney et al., 1999). I focus now on the effect of diet and retention time on hydrogen

availability, methanogens, ciliates and their interaction in the rumen.

Hydrogen availability

The overall microbial ecology in the rumen is influenced by increased grain content in

the diet. One reason for this is thought to be the changes in the availability of hydrogen

in the rumen. As established earlier, the main substrate for methanogenesis in the rumen

is hydrogen. This is supported by Van Nevel et al. (1969), who found that inhibiting

methanogens in sheep using chloral hydrate resulted in an accumulation of hydrogen

and an increase in propionate production. The changes in propionate concentrations on a

high grain diet have also been found in a number of other studies as its production

requires hydrogen and the ratio of acetate to propionate in the rumen has an inverse

relationship with methanogenesis (Van Kessel and Russell, 1996; Lana et al., 1998;

Russell, 1998). This increase in propionate is the main indicator of a shift in rumen

fermentation. It has been suggested that the decrease in methanogenesis when

propionate production is increased in the rumen could be because, for diets high in

grain, propionate is used as a ‘sink’ for metabolic hydrogen by organisms that grow

faster than methanogens under these conditions (Baker, 1997; Russell, 1998). One of

these organisms could be Selenomonas ruminantium, which has been found previously

to produce propionate (Scheifinger and Wolin, 1973), and have more than a two-fold


23

increase in abundance after an extended period of feeding a high grain diet (Tajima et

al., 2001a). Feeding a high grain diet also increases lactate production in the rumen

(Counotte et al., 1983) as well as concentrations of lactate producing and utilising

bacteria (Goad et al., 1998). Therefore, bacteria like Megasphaera elsdenii would

contribute to the increase in propionate, as it has the ability to ferment lactate to

butyrate and propionate. This process has been examined in Eubacterium hallii, which

was found to utilise hydrogen (Duncan et al., 2004) and limit the amount of hydrogen

available for methanogenesis.

Methanogens, rumen ciliates and their interaction

Diets and ruminal retention time are very closely linked; a change in diet, for example

from pasture to grain, changes the retention time in the rumen. The addition of grain to

diets changes rumen fermentation by changing a wide range of parameters such as

decreasing pH and acetate/propionate ratios, as well as methane production (Moss et al.,

1995; Baker, 1997; Lana et al., 1998). Evans (1981a,b) observed that, as the percentage

of grain in the diet increased, the retention time also increased. This enables the

methanogens and ciliates to stay in the rumen for longer. Ciliates are frequently

attached to plant material in the rumen because it allows them to remain in the rumen

when the liquid retention time is less than their growth rate (Bauchop and Clarke, 1976;

Williams and Coleman, 1992). Methanogens, on the other hand, are either free-living or

closely associated with the rumen ciliates, which mean that if retention time is

decreased, then some free-living methanogens are likely to be washed out of the rumen

because the turnover in the rumen is likely to be higher than the generation time of the

methanogens. The generation time for methanogens grown in vitro varies between 2.25

hours under optimal conditions for Methanobacterium spp. and 73 hours or more for an

acetate-fermenting strain of Methanosarcina on a basal medium (Mah et al., 1978;


24

Lovley et al., 1984; Morgan et al., 1997). Methanogens associated with the rumen

ciliates can remain in the rumen because the ciliates are attached to plant material.

Matsuyama et al. (2000) reported that by decreasing the retention time artificially with

“Rumen fibre” (pot scrubbers) the amount of methane produced decreased and the

levels of propionate increased. The decrease in methane from animals with pot

scrubbers in their rumen could be caused by the free-living methanogens being washed

out.

Apart from grain and retention time, pH is another important factor that can

influence methanogenesis. It has been reported that rumen methanogens are sensitive to

even modest decreases in pH and that below pH 6, methane production could not be

detected in vitro (Van Kessel and Russell, 1996). Changes in pH are thought to explain

up to 25% of the changes in the acetate/propionate ratio, but it is more subtle than the

effect of diet (Russell, 1998). Lana et al. (1998) concluded that the lower pH observed

in grain-fed animals is because of higher VFA production, lowered motility and higher

retention time in the rumen. When grain is introduced into the diet, pH and in vitro

methane production decreases (Moss et al., 1995), because of the pH sensitivity of

methanogens (Van Kessel and Russell, 1996). However, the decrease in methane

production due to decreased pH in the rumen might not be as significant as Van Kessel

and Russell (1996) reported since their studies were performed with ciliate-free rumen

fluid. This is supported by the findings that ciliates can tolerate pH levels as low as 5.3

(Dehority, 2005) and it has been estimated that methanogens associated with ciliates

produce between 9 - 37% of the methane (Finlay et al., 1994; Newbold et al., 1995).

This means that Van Kessel and Russell (1996) may have underestimated methane

production because methanogens associated with ciliates are less likely than free-living


25

methanogens to be affected by a drop in ruminal pH, especially the endosymbiotic

methanogens because they live in a protected environment inside the ciliates.

Despite a decrease in methanogenesis from animals fed a higher percentage of

grain in their diet, the results from in vitro studies have indicated that the methanogenic

population actually can increase or be maintained under these conditions (Baker, 1997).

Newbold et al. (1995) also suggested that the abundance of methanogenic archaea and

methane production may not be linked. This means that it is critical that not only the

abundance of methanogens is examined, but also their diversity as some methanogens

may produce more methane than others.

When a high grain diet is fed and conditions in the rumen change, it is likely that

selection for certain methanogens will occur. Zinder (1993) compared the affinity of

methanogens for hydrogen and found differences between species. These differences

show that different species are likely to respond differently to changes in the availability

of hydrogen in the rumen. As mentioned earlier, the production of propionate in the

rumen requires hydrogen, leaving less hydrogen for methane production. Therefore, it

would be expected that methanogens would associate with ciliates either ecto- or

endosymbiotically and take advantage of interspecies hydrogen transfer to survive in the

rumen when hydrogen becomes limited as a consequence of a high grain diet. The

methanogens should also have a high affinity for hydrogen to survive under these

conditions in the rumen. This means that there would be a selection for methanogens

with a high affinity for hydrogen and/or methanogens that can associate with ciliates.

Under these constraints, decreased methanogen diversity would be expected in the

rumen when animals are fed a higher percentage of grain.

The abundance of rumen ciliates change with increasing amounts of grain in the

diet. By increasing the percentage of grain in the diet, the abundance of ciliates,


26

especially Entodinium spp. (Williams and Coleman, 1992) is likely to increase due to

enhanced substrate availability (Mackie et al., 1978; Franzolin and Dehority, 1996;

Hristov et al., 2001). However, when grain is present at more than 70% of the diet, the

abundance of ciliates is likely to decline in sheep (Grubb and Dehority, 1975; Mackie et

al., 1978), probably due to a decrease in ruminal pH. Krumholz et al. (1983) found that

the methanogenic activity in rumen fluid was highest in fractions containing high

numbers of rumen ciliates so changes in the abundance of rumen ciliates would be

expected to affect methanogenesis.

The diversity of ciliates is also expected to change in the rumen of animals

consuming grain. As described earlier Entodinium spp. probably are more dominant in

the rumen on a high grain diet and other ciliates may subside because of this (Hristov et

al., 2001). Itabashi et al. (1994) observed that a change in the generic composition of

the ciliates can lead to a change in methane production and this could also explain the

reduction in methane production on grain-based diets. This is supported further by the

observations by Newbold et al. (1995), who found that the number of methanogens and

the amount of methane produced was decreased by decreasing the diversity of rumen

ciliates. This underlines the importance of examining both diversity and abundance of

ciliates since both have been shown to affect methane production.


27

2.6 Value and limitations of two molecular methods for examining

methanogens and ciliates in the rumen

The use of molecular techniques to study the rumen microbiome has great potential and

widespread application. Denaturing gradient gel electrophoresis (DGGE) and

quantitative real-time PCR are two techniques that have changed the level of detail and

the types of results that can be obtained from microbial communities in the rumen.

However, there are various limitations that can influence the accuracy and precision of

the results when using these techniques. It is important to consider these limitations so

that valid interpretations can be made from the results that are obtained from molecular

studies. In the final part of this literature review the value and limitations of both DGGE

and quantitative real-time PCR are discussed. These two methods are used in the

experimental sections of my thesis.

2.6.1 Denaturing gradient gel electrophoresis (DGGE)

The basis of DGGE is the detection of differences in a double-stranded nucleotide

sequence/amplicon being analysed. The actual denaturing gradient in a DGGE gel is

created using formamide and urea, otherwise the gel is a normal polyacrylamide gel.

The denaturing gradient can in theory range from 0 to 100%, but in practise it varies

somewhere between 20 and 80%. The different amplicons are separated in the DGGE

gel depending on their nucleotide sequence, and under the right conditions a difference

of one nucleotide can be detected. Therefore, dissimilar amplicons will separate at

different denaturing conditions, meaning in different positions on the DGGE gel.

Complete dissociation of the amplicon must be prevented, which is usually achieved by

adding a 40 bp GC-clamp on one of the primers during the PCR prior to running a

DGGE (Sheffield et al., 1989). Because DGGE depends on the melting behaviour of the


28

amplicon a prerequisite to successful DGGE is to find an amplicon with only one

optimal melting profile (i.e. one flat melting domain) (Nedergaard et al., 1997). Studies

have been conducted to optimise these conditions when the DNA strand of interest does

not fulfil the conditions, and a number of strategies can be followed including, adding

A/T, G/C nucleotides or altering the length of the GC-clamp (usually 40 bp) (Wu et al.,

1998). Furthermore, studies have been carried out specifically examining problems

associated with running a GC-rich DNA strand (Wu et al., 1999) and improvements can

be made to the DGGE conditions, such as changing the polyacrylamide percentages,

denaturing gradient, running time and voltage (Hayes et al., 1999).

DGGE was first used to examine a complex microbial population by Muyzer et

al. (1993). Since then DGGE has been used widely to examine microbial communities

in a wide range of environments, including the rumen (Kocherginskaya et al., 2001;

Regensbogenova et al., 2004b; Karnati et al., 2007). DGGE produces a fingerprint

banding pattern of the microbial population in the environment being examined, which

are valuable data if different conditions change the environment being studied.

Theoretically, each band on the gel represents one sequence/species of the

microorganisms represented in the sample and one lane represents one sample. The

differences found on a DGGE gel can then be analysed by sequencing the bands of

interest. The DGGE gels are analysed using specific software that uses defined

standards on the gel to normalise the gel and can assist in determining differences in

banding patterns between samples. Multiple gels can be compared if standards are run

on every gel and thereby normalised consistently. DGGE is a very useful tool to

examine microbial populations providing its limitations are known, because measures

can then be taken to eliminate and minimise these limitations.


29

Detecting changes in microbial ecology using DGGE is rapid and easy when the

DGGE assay is in place and validated. In contrast, developing a DGGE assay can be a

long and difficult process, as many factors affect the outcome. There are some technical

details that can limit the value of the results from DGGE if not solved, but they are

easily detected as a smeared gel or by the presence of fuzzy bands and are not discussed

further in this literature review.

The DGGE banding patterns observed in an experiment are always dependant on

the efficiency of previous steps in the procedure for example, sample handling, DNA

extraction and PCR including chimeric molecules, formation of heteroduplex molecules

and point mutations, which all affect the final DGGE banding pattern. All these

limitations are well covered in the review of Wintzingerode et al. (1997) and Muyzer

and Smalla (1998).

Limitations more specific to DGGE that affect the interpretation of the results

are discussed in detail here. There is a limitation on the length of the fragment that can

be used for DGGE, due to the one melting domain criteria. The fragment length is often

< 500 base pairs (Myers, 1985), which limits the sequence information for phylogenetic

analysis. This limitation is difficult to circumvent as the one melting domain criterion is

critical to obtain consistent DGGE results. However, with continual increases in the

availability of sequence data in various databases, the identification of DGGE bands

will become easier.

An obvious limitation of DGGE is the maximum number of DNA fragments that

can be separated from a sample containing thousands of different fragments. In general

no more then 30 different fragments of the thousands of fragments will be visualised.

This means that the DGGE will detect only the predominant species from highly

variable communities. The detection limit for DGGE-PCR for bacterial populations


30

using general primers has been estimated to be 1% or more of the total community

(Muyzer et al., 1993; Murray et al., 1996). To avoid this limitation in detection, a

smaller proportion of the microbial community can be examined. Primers can be

designed for more specific groups of microorganisms of interest and thereby a larger

proportion of the low abundance species will be detected. This means that general

bacterial DGGE primers should not be used if a certain group of microorganisms is of

particular interest.

Theoretically, one band should represent one organism, but co-migration of

DNA fragments has been known to occur, which causes a problem in retrieving clean

sequence data from individual bands (Muyzer and Smalla, 1998; Sekiguchi et al., 2001)

and underlines the bias associated with quantitative analyses based on band intensities

(Sekiguchi et al., 2001). This can be solved by cutting out bands and then re-running

them on a DGGE gel; it is believed that the bands separate on the secondary DGGE gel

because the sample contains fewer DNA fragments and there is less interference in the

migration of the DNA fragments.

Another problem that may occur when performing DGGE using the 16S rRNA

gene is the presence of multiple ribosomal rRNA operons in some bacteria (Nubel et al.,

1996). Multiple copies of the 16S rRNA in the genome means that the same gene (e.g.

16S rRNA) is repeated in the genome. When there are multiple copies of the 16S rRNA

gene in the microbial genome, some of the copies are silenced, which means that if

point mutations occur within the silenced gene they may not be corrected during normal

proofreading processes. The silenced copies are then different to the transcribed copy,

which creates more bands on the DGGE gel than there should be, leading to false

conclusions (i.e. increased diversity) (Muyzer and Smalla, 1998). Multiple copies of

genes can be detected using pure cultures and real-time PCR. A known number of cells


31

from a pure culture can be related to a standard curve made from a plasmid containing

one copy of the gene. The level of amplification from the pure culture detected by real-

time PCR is then correlated to the number of cells being analysed, resulting in an

estimate of the number of gene copies.

Degenerate primers in DGGE-PCR can cause a similar problem to multiple gene

copies (Kowalchuk et al., 1997); i.e., an overestimate of diversity. However, the

increase in community diversity indicated by degenerate primers can be detected by

sequencing because differences in the nucleotide sequence will only appear where the

primer degeneracy occurred. This problem can be eliminated by using non-degenerate

primers.

Assumptions and compromises are made when using DGGE to analyse

microbial communities. One of the most common assumptions is that a band in a

similar position on two gels is identical. This assumption is generally accepted, but it is

also easily verified by sequencing the two bands thought to be identical. Another issue

not so commonly used anymore is quantification of targeted species based on band

intensity, because there is evidence that this is associated with significant error

(Sekiguchi et al., 2001). This is mainly due to co-migration of amplicons and

differences in PCR efficiency, which means that band intensity does not correlate with

abundance. Probably the most critical general assumption made is that a similar amount

of DNA in the PCR prior to DGGE will generate the same amount of PCR product.

This assumption is not true for samples where PCR inhibitors can be co-extracted with

the DNA. Rumen samples fall under this category (Krause et al., 2001). To circumvent

this error the amount of PCR product loaded on a DGGE gel should be estimated when

DNA templates have been extracted from samples where PCR inhibitors can be a

problem. If not, loading the same volume of each sample may result in an uneven


32

loading of PCR product. Producing a DGGE containing lanes that are either bright or

faint will result in bands from less abundant microorganisms not being visible in the

faint lanes but visible in the brighter lanes. This makes photographing the gel difficult

and adds a potential error when the DGGE banding patterns are analysed and compared.

This is an example of the type of error that might occur when the amount of DGGE-

PCR products are loaded unevenly onto the DGGE gel. In this case, a detected

difference in DGGE banding patterns between two diets may not be due to actual

differences in the microbial population. What might have happened is that one of the

diets may have resulted in more PCR inhibitors being co-extracted with the DNA. This

would lead to differences in the DGGE-PCR efficiency and the amount of PCR product

amplified. If the same volume is loaded on a DGGE gel, then more bands would be

observed in bright lanes from the diet where less PCR inhibitors were co-extracted with

the DNA. Differences observed then may simply be because more bands are visible in

the brighter lanes and not because of actual differences between the diets.

2.6.2 Quantitative real-time PCR (real-time PCR)

The principal of quantitative real-time PCR is similar to conventional PCR. The

difference is that in real-time PCR the amplification can be observed in real-time by

using a fluorescent dye. The intensity of the fluorescent dye, measured as relative

fluorescent units, increases with each amplification and it is measured after each cycle.

Being capable of measuring the amplification in real-time means that it is possible to

get a measure for when the amplification becomes exponential by monitoring when the

fluorescence gets stronger than the background fluorescent. The cycle where this

happens is assigned the threshold cycle. The threshold cycle is then compared against a

standard curve made from a dilution series of samples containing a known number of

the target microorganisms or an estimated copy number of the amplicon of interest. This


33

can then be used to determine the abundance of the microorganism of interest in

environmental samples. Quantifying microorganisms in an environment using real-time

PCR has become a standard procedure because of its sensitivity and reproducibility. The

technique is also valuable for detecting ruminal microbes (Tajima et al., 2001a;

Sylvester et al., 2005; Ozutsumi et al., 2006; Skillman et al., 2006b), but care has to be

taken to eliminate or minimize some of the pitfalls and limitations of the method.

It is important to realise that the results from real-time PCR results also depend

on the efficiency of previous steps in the procedure (e.g. sample handling and the DNA

extraction method). To quantify the presence of certain microbes in a sample accurately,

several limitations have to be considered. The DNA extracted from rumen fluid samples

is often co-extracted with PCR inhibitors (Krause et al., 2001) resulting in poor quality

DNA (Krause et al., 1999). Therefore, it is important to estimate the quality or PCR

efficiency for each rumen sample for accurate quantitation (Pfaffl, 2001; Liu and Saint,

2002; Tichopad et al., 2003; 2004). PCR efficiency is traditionally calculated based on a

dilution series of the template and then the same efficiency is used for all dilutions.

Traditionally, the logarithmic function of the dilutions are then plotted against the

threshold cycle for the different dilutions and the PCR efficiency can then be calculated

from the slope of the line with the equation slope/110−=ε (Rasmussen, 2001). An

alternative way of calculating the PCR efficiency is to calculate it for each dilution by

plotting the threshold cycle against the logarithmic function of the relative fluorescent

units at the time the amplification becomes exponential. The efficiency is then

calculated from the slope with the equation slope10=ε (Liu and Saint, 2002). By

calculating it this way an individual PCR efficiency is obtained for all dilutions. PCR

efficiency has a large influence on the calculation of the final abundance of populations

(N) according to the equation nNN ε×= 0 (N0 is the number of methanogens/amount


34

of DNA present initially (before the PCR), ε is the efficiency of the PCR and n is the

number of cycles) (Rasmussen, 2001). To circumvent the problem with co-extraction of

PCR inhibitors Dionisi et al. (2003) examined the effect of DNA extraction methods

and real-time PCR procedures on assay variability. They concluded that increasing the

number of real-time PCR assays performed with a single DNA extract may have as

large an effect on statistical power as increasing the number of DNA extractions and

real-time PCR assays. This suggests that co-extraction of PCR inhibitors can be difficult

to avoid. Therefore, estimating PCR efficiency for each individual dilution of a sample

will be more accurate than estimating PCR efficiency based on a serial dilution of a

sample, as equal efficiency is assumed in the latter estimate (Liu and Saint, 2002). The

former will also reduce the number of dilutions necessary for accurate quantification.

The fundamental step in real-time PCR or any PCR is the design and selection

of primers. In this process the specificity of the primers is tested to avoid overestimation

of abundance by amplifying non-specific sequences. If general primers are designed for

a large group of microorganisms (e.g. methanogens) the preference of the primers for

certain species within this group of microbes should also be tested, preferably against

more than one genus, especially if primers cover more taxa. The preference of primers

should be tested for two main reasons. First, to ensure all species within this group are

detected by the primers and second, to ensure that the efficiency of amplification is

similar for all species within this group of microorganisms. If these tests are not

undertaken, then errors may occur when comparing samples containing different species

of the same group of microorganisms, because some species may be detected and not

others. The importance of primer design has been highlighted by Skillman et al.

(2006a), who found that primers that were published as general archaeal primers were


35

biased towards certain species of methanogens due to mismatches in the primer

sequence of some methanogen groups.

The design and the specificity of primers become pivotal in the actual detection

of amplification. Different strategies can be employed to detect amplification such as

double-stranded (ds)-DNA-binding dyes, Taqman probes and molecular beacons. The

advantages of dsDNA-binding dyes compared to using Taqman probe or molecular

beacon (Zhang and Fang, 2006) is that they can be applied with any PCR primers, and

using them is not complicated by the need to design an additional probe. The limitations

to using dsDNA binding dyes are the inability of the dye to discriminate between primer

specific and non-specific amplification, and the formation of primer-dimers that may

affect the detection sensitivity (Sharkey et al., 2004). Thus, accurate primer design and

testing is crucial when dsDNA-binding dyes are used (Zhang and Fang, 2006).

A difficult limitation to overcome when using quantitative real-time PCR to

study environmental samples is the fact that some microbes have multiple copies of the

16S rRNA gene. Similar to DGGE, this becomes an issue when molecular methods are

employed, because one microorganism is counted as three if it possesses three copies of

the 16S rRNA gene. The main problem arises when the general abundance of a group of

bacteria is to be enumerated, if the species within this group have different copy

numbers of the 16S rRNA gene, and a shift in diet changes the dominant species within

this group. The difference in abundance estimated may simply be due to a difference in

number of copies of the 16S rRNA gene.


36

2.7 Summary

In summary, the dominant methanogens in the rumen (Skillman et al., 2004; Wright et

al., 2004; 2006; 2007) have been identified, including the ones associated with the

rumen ciliates (Vogels et al., 1980; Stumm et al., 1982; Finlay et al., 1994; Newbold et

al., 1995; Tokura et al., 1997; Sharp et al., 1998; Chagan et al., 1999; Tokura et al.,

1999; Schonhusen et al., 2003; Irbis and Ushida, 2004; Regensbogenova et al., 2004a).

However, the effect that changes in rumen parameters, diet and retention times have on

the abundance, diversity and association between methanogens and ciliates remains

unexplored. It has been established that a lower retention time in the rumen and

increased grain in the diet results in decreased methane emissions (Moss et al., 1995;

Lana et al., 1998; Matsuyama et al., 2000), but the reason for this reduction at the

microbial ecology level, particularly the ecology of the methanogens and ciliates, is

unknown. The reduction could result from changes in the diversity and/or abundance of

methanogens in the different niches they occupy in the rumen, or be due to a change in

the diversity or abundance of ciliates, or a combination of both. Decreasing the retention

time in the rumen would be expected to “washout” mainly free-living methanogens, as

they are not attached to plant material or associated with ciliates attached to the plant

material in the rumen. However, with increasing grain in the diet, ciliates would be

expected to increase in abundance, whereas the grain would mainly target free-living

methanogens as the availability of hydrogen in the rumen is believed to be reduced due

to competition from other hydrogen utilising bacteria (e.g. propionate-producers). The

differences in hydrogen thresholds between methanogen species would add to the

selection pressure on the methanogen populations (Zinder, 1993). Despite changes in

the diversity of methanogens, the total abundance of methanogens in vitro has been

found to remain stable after grain has been introduced, even though lower methane


37

production was reported (Baker, 1997). Although limitations apply to the use of

molecular tool, the data obtained using these techniques are superior to traditional

culture based data for the type of study being proposed. The use of molecular

techniques, like DGGE and real-time PCR, makes it possible to examine the association

between methanogens and rumen ciliates in greater detail than achieved before. To do

this a new method for examining numbers and diversity of methanogens has to be

established, integrating steps to minimise the limitations of molecular techniques. These

steps would include loading similar amounts of the DGGE-PCR product on the DGGE

gel and assessing the PCR efficiency for every real-time PCR sample being analysed.

The general hypothesis tested in this thesis was that increasing the amount of

grain in the diet and reducing the retention time would affect the abundance diversity of

methanogens in their different niches, as well as their association with ruminal ciliates.

More specific hypotheses were developed and tested in the three experimental Chapters

(4-6):

• Chapter 4: The combined effect of increasing the grain content in the diet and

reducing retention time in the rumen would decrease methane production and

acetate/propionate ratios more than the individual effect of each treatment.

• Chapter 5: (i) that the diversity of methanogens would decrease and the abundance

of methanogens associated with ciliates would increase in the rumen of sheep

consuming grain; and (ii) that these effects would be more pronounced in sheep with

pot scrubbers.

• Chapter 6: Increasing the amount of grain in the diet would increase the abundance

and reduce the diversity of rumen ciliates, whereas the abundance and diversity of

the rumen ciliates would not be affected by the addition of pot scrubbers.

Chapter 3: General materials and methods

38

CHAPTER 3

Materials and methods


39

3.1 Introduction

This Chapter covers the experimental design, which is common to the three

experimental Chapters in this thesis.

3.2 Experimental design

During the experiment, 24 five year old merino rams, surgically fitted with rumen

cannulae, were individually penned indoors at the ‘Large Animal Facility’ at The

University of Western Australia. The animals were fed once daily in the morning and

had free access to water at all times. Animal ethics approval (Protocol: RA/3/100/239)

was obtained from the University of Western Australia prior to the commencement of

experiments.

Before and during the experiment all animals were weighed and fed at

maintenance according to their individual weights. All animals consumed 95% or more

of their rations during the whole experiment. The 24 animals were divided into four

equal groups based on live weight (average 67.4 ± 0.267 kg (SEM)), making up a

complete factorial design. All animals were then fed an oaten-chaff diet (ME: 8.7

MJ*kg-1DM and DM: 86%), supplemented with Siromin, (Narrogin Vitamin and

Mineral Stock Mix, Narrogin, WA, Australia) vitamins and minerals mix (White et al.,

1992), during the 2-wk acclimatization period. After the acclimatization period two of

the four animal groups had six pot scrubbers (5 cm diameter) inserted into their rumen

as a way of decreasing the retention time of digesta without changing the diet

composition (Matsuyama et al., 2000). All animals were started on the first of three diet

phases each lasting three weeks. During the first diet phase all animal groups were

maintained on the oaten-chaff diet. After the first diet phase one group with pot

scrubbers and one group without pot scrubbers were assigned to the oaten-chaff diet


40

throughout the experiment. The animals in the other two groups, one with and one

without pot scrubbers, were assigned to an oaten-grain diet (ME: 10.7 MJ*kg-1 and DM:

86%). In the second diet phase (3-wk duration), the animals assigned to the grain diet

were offered an oaten-chaff diet supplemented with oaten-grain accounting for 35% of

energy intake (low grain). In the third diet phase, (3-wk duration) the oaten-chaff diet

was supplemented with oaten-grain accounting for 70% of energy intake (high grain).

Rumen fluid samples were collected through the rumen cannulae at the end of

each diet phase with a 50 cm long plexiglass tube; the exact time is specified in the

different Chapters. Samples were used to analyse: the rumen retention time in the liquid

and particulate matter, methane production in vitro, pH, acetate, propionate and total

VFA concentration. DNA was also extracted to examine the abundance and diversity of

methanogens and rumen ciliates and their association with each other.

Chapter 4: Rumen parameters

* This Chapter has been published in Journal of Animal Science, in press. 41

CHAPTER 4

In vitro methane emission and acetate/propionate

ratio are decreased when artificial stimulation of

the rumen wall is combined with increasing grain

diets in sheep*


42

4.1 Introduction

Methane has a global warming potential 23 times more potent than carbon dioxide, which

makes methane one of the most important greenhouse gases (Wuebbles and Hayhoe, 2002).

The microbes producing methane, methanogenic Archaea (i.e. methanogens), compete with

bacteria in the rumen for substrates such as hydrogen (Zinder, 1993). Furthermore, during

fermentation, hydrogen is produced and the removal of hydrogen is important for the

efficiency of rumen fermentation (Stewart et al., 1997). Propionate is an end-product of

fermentation that requires hydrogen for production and the acetate/propionate ratio in the

rumen has an inverse relationship with methanogenesis (Lana et al., 1998; Russell, 1998).

The addition of grain to the diet increases the amount of starch in the rumen and

changes rumen fermentation. Methane production is generally reduced per unit feed intake

when grain content in the diet is increased. This reduction is indicated by a lower

acetate/propionate ratio and pH (Moss et al., 1995; Lana et al., 1998). Rumen fermentation

is also affected by decreasing retention time. The type of diet can reduce retention time

(Evans, 1981b, a) with high grain diets having shorter retention times compared to

roughage diets. Retention time has also been decreased in the past using nylon mesh balls,

commonly known as pot scrubbers, where the physical contact and ‘scratching’ against the

ruminal wall is thought to increase ruminal turnover (Matsuyama et al., 2000). Matsuyama

et al. (2000) found that decreasing ruminal retention time decreased methanogenesis.

My experiment was conducted to examine the interaction of pot scrubbers and

feeding a concentrate-based diet on methane production in vitro, acetate/propionate ratio,

and pH. It was hypothesized that the interaction of increasing the grain content in the diet

and reducing retention time in the rumen would decrease methane production and

acetate/propionate ratios more than the individual effect of each treatment.


43

4.2 Materials and methods

Experimental design

As described in Chapter 3, with the exception that only strained crude rumen fluid samples

from diet phase 2 and 3 were analysed in this Chapter.

Estimation of mean retention time of liquid and particulate matter

Chromium (Cr) and ytterbium (Yb) were used to measure the retention time of digesta in

the rumen for each sheep using the double marker dose technique (Faichney, 1992a).

Ytterbium, as YbCl3 was used to associate with the particulate matter in the rumen

(Faichney, 1992b) and Cr-EDTA with the liquid fraction (Faichney, 1992a). A single dose

of 50 mL of Cr/Yb solution was added to the rumen 1 h after feeding. The disappearance of

these markers per unit time was used to calculate the mean retention time (MRT) of the

particulate and liquid fractions from the rumen according to Faichney (1975).

Preparation of Cr and Yb solution. Cr-EDTA and YbCl3 stock solutions were

prepared in the following manner: A total of 36.5 g of Na2EDTA was boiled in 1 L of

deionized H2O for 1 h and 33.3 g of CrCl3-6H2O was dissolved in 200 mL of deionized

H2O and slowly added to the boiling Na2EDTA solution. The mixture was boiled for 4 h

and allowed to cool to room temperature before 125 mL of concentrated NH4OH was added

and mixed. The solution was left overnight. The solution was then filtered through

Whatman No. 1 filter paper and pH was adjusted to pH 7 with 10 N HCl. Finally, 36.25 g

of YbCl3 was added to the solution and made up to 2 L with deionized H2O.

Rumen sample collection and processing. Samples for measuring retention time of

fluid and particulate matter in the rumen were collected on d 17 to 19 in diet phase 1 and 2.

The sheep were fed 1 h before 50 mL of the Cr/Yb solution was administered. The solution


44

was delivered to the rumen through the cannulae using a syringe and flexible tubing

situated approximately in the middle of the rumen. To detect the disappearance of Cr and

Yb from the rumen, samples were collected at 2, 4, 6, 8, 14, 26, 36, 48, 58 and 72 h after

administration. At every sampling, 120 mL were collected at different locations of the

rumen and mixed. After mixing, approximately 20 mL were stored at -20 ºC. The

remaining 100 mL were put back into the rumen through the cannulae.

Rumen samples were prepared for Cr and Yb analysis according to the following

protocol: Rumen samples were thawed and weights were recorded. The samples were then

lyophilized for 96 h using a freeze dryer (Heto, FD4.0, Thermo Fisher Scientific, Waltham,

MA). When dry, the weights of the containers and the dried samples were recorded. The

dried samples were then transferred to a conical flask and concentrated nitric acid was used

to remove all carbon. The slurry of dried salts was then dissolved in 20 mL of deionized

H2O.

Chromium and Ytterbium assay. The Cr standards were prepared from K2Cr2O7

and the Yb standards were prepared from Yb2O3. The concentrations used were equivalent

to 0 to 12 µg*mL-1 for both trace elements. The diluted digested samples were aspirated

into an atomic absorption spectrophotometer (Varian, AA300, Palo Alto, CA) and

concentrations of Cr were measured using the following parameters: lamp current 7 mA,

slit width 0.2 nm, wavelength 357.9 nm, and an air/acetylene oxidizing flame;

concentrations of 1.0 ppm yielded absorbance of 0.100. The concentrations of Yb were

measured via atomic absorption spectrophotometery. The parameters used were: lamp

current 5 mA, slit width 0.5 nm, wavelength 398.8 nm and a nitrous oxide/acetylene

reducing flame. All calculations were done according to the protocols published by

Faichney (1975) and Faichney et al. (1999).


45

Volatile fatty acids, pH and in vitro methane production

Rumen sampling. On the last day of each of the 3-wk diet phases (d 21 and 42), samples

were collected 1 h after feeding for analyzing VFA and methane production in vitro. The

sample for VFA measurement was strained through 2 layers of muslin cloth and a 9 mL

sample was added to 1 mL of 1 N NaOH and stored at -20 ºC until used. Approximately

100 mL of crude rumen fluid was collected for measuring pH before the sample was

prepared for measuring methanogenesis in vitro.

VFA. Frozen samples were allowed to reach room temperature before being

vortexed for 30 s, and the particulate matter was allowed to settle before 0.5 mL of the

supernatant was centrifuged (Eppendorf, 5415C, Hamburg, Germany) at maximum speed

for 20 min. The concentrations of acetic and propionic acids and the total amount of VFA

were determined by GLC (Agilent 6890 series GC with HP 6890 injector and using HP

chemstation software) using the standard procedure for separation of VFA (Supelco

Bulletin no. 749D, Supelco, Inc., Bellefonte, PA). The GLC was fitted with a HP-FFAP

capillary column 30 m x 0.53 mm x 1 µm (Agilent Technologies, Palo Alto, CA, US). All

measurement and calculations were performed by the Department of Agriculture Western

Australia, Perth, Australia.

In vitro methane production. Methane production from sampled digesta was

performed in 100 mL serum bottles, which were purged with N2 before 30 mL of whole

rumen fluid was added to each bottle. Afterwards, N2 was bubbled through each bottle for 1

min to ensure anaerobic conditions. The experiment was performed in triplicate for each

sheep. The bottles were then incubated for 24 h at 39 ºC in an Innova 4080 shaking

incubator (New Brunswick Scientific, Edison, NJ) at 100 rpm. The incubation was

terminated with injection of 3 mL of 17.5% formalin. Gas pressures in the serum bottles


46

were measured with a pressure transducer and the composition of the headspace in the

bottles and the method to calculate the amount of methane produced was done according to

Klein and Wright (2006) with the following modifications: column temperature was 190 °C

and injector and detector temperatures were 250 °C.

Statistical analyses

All statistical analyses were done using the multivariate statistical package PRIMER v6

(Clarke and Gorley, 2006). The data used had been transformed according to draftsman

plots where necessary. Multivariate data analysis were normalized and reassembled in a

Euclidean distance matrix before using ANOSIM; a multivariate analogue to ANOVA.

Non-metric multi-dimensional scaling (MDS ordination) was used to investigate the effect

of grain and pot scrubbers on methane production, pH, retention time of digesta, and VFA;

on MDS plots, most similar communities are grouped closer together (Clarke, 1993; Clarke

and Warwick, 2001).

4.3 Results

Effect of diet

High grain-fed sheep without pot scrubbers showed a significantly lower acetate/propionate

ratio than oaten-chaff-fed sheep (P < 0.01). Additionally, there was a significant decrease

in acetate/propionate ratio for grain-fed sheep without pot scrubbers (P < 0.05) when grain

content was increased from low to high (Figure 4.1). High grain-fed sheep without pot

scrubbers had significantly greater propionate concentrations than oaten-chaff-fed sheep (P

< 0.01). Furthermore, going from a low to a high grain diet significantly increased ruminal

propionate concentration (P < 0.05) (Table 4.1).


47

Grain was found to affect rumen fermentation, but it did not change rumen retention

time for either the liquid or the particulate phase (Table 4.1). High grain-fed sheep without

pot scrubbers produced significantly (P < 0.01) less methane in vitro than both groups fed

oaten-chaff diets. Similarly, the increase from low to high grain resulted in a 16% decrease

in in vitro methane production (P < 0.01) in the sheep without pot scrubbers (Table 4.1).

When a multivariate statistical method was used to do an analysis including all data,

high grain-fed sheep without pot scrubbers were found to differ significantly from the

oaten-chaff-fed sheep (P < 0.01) (Table 4.1). The analysis also showed an overall change

when increasing the grain content in the diet from low to high as grain-fed sheep without

pot scrubbers changed significantly (P < 0.05) (Table 4.1).

Oa

Gc*OPb

GPd**

0.00.51.01.52.02.53.03.54.04.5

Diet phase 1 (35% grain) Diet phase 2 (70% grain)

Ace

tate

:pro

pion

ate

ratio

Figure 4.1: Acetate/propionate ratios (mol*mol-1) for sheep fed oaten-chaff, low grain (phase 1) and high grain (phase 2) diets, with and without pot scrubbers. (O: oaten-chaff diet; G: grain diet; OP: oaten-chaff diet + pot scrubber; GP: grain diet + pot scrubber). Groups with different superscript letters within a diet phase were found to be significantly different (P < 0.05). Groups with a superscript * in diet phase 2 were found to change significantly when going from low to high grain (*: P < 0.05; **: P < 0.01). Error bars are SEM.

Table 4.1: In vitro methane production, pH, VFA concentration, and ruminal retention time (mean ± SEM) in sheep fed oaten-chaff, low grain and high grain diets, with and without pot scrubbers. The overall significance between treatment groups when combining data in a multivariate statistical analysis is also presented.

VFA, mmol*L-1 Mean retention time, h Methane,

mmol*mL-1 digesta Acetate Propionate Total pH

Liquid Particulate Overall

significance

Diet phase 1

Oaten-chaff 4.25a ± 0.16 59 ± 3.48 19 ± 2.60 93 ± 7.26 6.2 ± 0.07

13.5 ± 0.68 19.5 ± 1.82 a

Oaten-chaff + pot scrubbers 4.02a,b ± 0.10 58 ± 2.05 19 ± 1.68 92 ± 4.13 6.3 ± 0.09

13.3 ± 0.99 15.1 ± 0.79 a

Low grain 4.11a,b ± 0.11 62 ± 1.77 19± 0.54 98 ± 2.11 6.2 ± 0.05

13.6 ± 1.08 16.6 ± 0.90 a

Low grain + pot scrubbers 3.71b ± 0.14 58 ± 1.80 21 ± 1.65 93 ± 2.64 6.1 ± 0.07

14.3 ± 0.69 16.0 ± 0.55 a

Diet phase 2

Oaten-chaff 4.53a ± 0.20 56 ± 2.43 16a ± 1.52 87a ± 3.19 6.3a ± 0.06

15.6 ± 1.52 19.2 ± 0.92 a

Oaten-chaff + pot scrubbers 4.16a,b ± 0.10 55 ± 1.36 20a,b ± 1.02 92a,b ± 4.59 6.2a ± 0.09

13.6 ± 1.47 19.0 ± 1.25 a

High grain 3.47c ± 0.08** 57 ± 3.01 24b ± 1.52* 95a,b ± 4.44 6.1a,b ± 0.11

16.6 ± 0.92 17.9 ± 1.15 b*

High grain + pot scrubbers 3.71b,c ± 0.19 54 ± 1.14 35c ± 3.23** 103b ± 3.54 5.9b ± 0.09

14.1 ± 0.76 18.2 ± 0.68 c**

a,b,c: Values in the same column within the same period not showing the same superscript letters differ, P < 0.05. * P < 0.05; ** P < 0.01: Indicates a significant change from a low to a high grain diet.


49

Effect of pot scrubbers

There was no effect of pot scrubbers on liquid or particulate retention time in the rumen

(Table 4.1). However, oaten-chaff-fed sheep with pot scrubbers had a significantly lower

acetate/propionate ratio than sheep fed the same diet but without pot scrubbers (P < 0.05)

(Figure 4.1). The same was observed for high grain-fed sheep with and without pot

scrubbers (Figure 4.1).

Combined effect of diet and pot scrubbers

The combination of diet and pot scrubbers was found to alter rumen fermentation, but no

effect was recorded on liquid or particulate retention time. There was a significantly lower

level of methane produced in vitro when a low grain diet was combined with pot scrubbers

than when the oaten-chaff was fed to sheep without pot scrubbers (P < 0.05) (Table 4.1).

The significantly lower in vitro methane production between these 2 groups also remained

after a high grain diet was fed (P < 0.05) (Table 4.1).

When a high grain diet was fed to sheep with pot scrubbers they showed a

significantly (P < 0.05) lower acetate/propionate ratio than grain-fed sheep without pot

scrubbers (Figure 4.1). The acetate/propionate ratio was also significantly (P < 0.01) lower

than in the 2 groups offered an oaten-chaff diet (Figure 4.1). Furthermore, there was a

significant decrease in acetate/propionate ratio (P < 0.05) when grain content was increased

from low to high (Figure 4.1), with a larger and more significant (P < 0.01) decrease for

sheep with pot scrubbers. Propionate concentrations showed similar responses as did the

acetate/propionate ratio when a high grain diet was fed to sheep with pot scrubbers. The

change from low to high grain also resulted in a significant increase in propionate

concentrations, with a larger increase for sheep with pot scrubbers (Table 4.1). The high


50

grain diet combined with pot scrubbers also increased total VFA concentration significantly

compared to sheep given oaten-chaff without pot scrubbers (P < 0.05) (Table 4.1).

Furthermore, when a high grain diet was offered, rumen pH was significantly less in sheep

fed grain with pot scrubbers compared to oaten-chaff-fed sheep with or without pot

scrubbers (Table 4.1).

When differences were tested using a multivariate statistical analysis including all

data, sheep fed high grain with pot scrubbers were significantly different from grain-fed

sheep without pot scrubbers (P < 0.05), as well as from both groups given oaten-chaff (P <

0.01). Furthermore, when the diet was shifted from a low to a high grain diet, the sheep

with pot scrubbers fed the high grain diet were found to change more than the grain-fed

sheep without pot scrubbers (Table 4.1).

4.4 Discussion


It was hypothesized that the combination of increasing the grain content in the diet and

decreasing retention time of digesta from the rumen would result in decreased methane

production and acetate/propionate ratios more than either one of these strategies

individually. My data partly supports this hypothesis. When a high grain diet was combined

with pot scrubbers the acetate/propionate ratio was found to be significantly lower than

other sheep groups (Figure 4.1). However, high grain-fed sheep without pot scrubbers

exhibited slightly, but not significantly, lower in vitro methane production (Table 4.1). In

vitro methane production was less in low grain-fed sheep with pot scrubbers compared to

oaten-chaff-fed sheep without pot scrubbers. This reduction was not accompanied by a

significant alteration in actetate/proprionate ratio or pH, which means that other factors


51

may be responsible for the reduction in methane production during this phase. However, I

expected a further reduction in in vitro methane production in high grain-fed sheep with pot

scrubbers compared to when they were fed a low grain diet. A possible explanation for why

this did not occur could be because the balance between methanogens and bacteria

competing for hydrogen had already been manipulated and established by the combination

of pot scrubbers and low grain. To fully support this argument, further investigations to

enumerate methanogens and other hydrogen utilizing bacteria under different dietary

conditions would be necessary.

Reasons for the lower acetate/propionate ratio observed in high grain-fed sheep with

pot scrubbers are unknown, but the lower methane production in vitro in low grain-fed

sheep with pot scrubbers could explain why more propionate was produced when a high

level of grain was offered. In this situation, the microorganisms competing with

methanogens for hydrogen to produce propionate in the rumen of sheep with pot scrubbers

would have had more time to adjust to the increasing amount of hydrogen available.

The lower pH measured in the rumen of high grain-fed sheep with pot scrubbers

may also have had an effect on both methane production and the acetate/propionate ratio.

Rumen methanogens are sensitive to even modest decreases in pH (Van Kessel and Russell,

1996), and changes in pH can explain up to 25% of the changes in the acetate/propionate

ratio, although the effect of pH is more subtle than the effect of diet (Russell, 1998). Lana

et al. (1998) concluded that lower pH in grain-fed sheep is due to higher VFA production,

less motility, and slower dilution rate of the rumen. My results indicate that the total

concentration of VFA influences ruminal pH more than dilution rate.

The multivariate statistical analysis, which can be illustrated by the non-metric

multi-dimensional scaling (MDS) plot, supports the above findings and shows that the 4


52

groups were different (Figure 4.2). The MDS plot (Figure 4.2) illustrates the differences

found between the groups when a high grain diet was offered, showing the oaten-chaff-fed

sheep further from grain-fed sheep with pot scrubbers than from grain-fed sheep, which I

also found in my analysis (Table 4.1). Further analyses carried out by overlaying the MDS

plot with the individual sheep’s methane production, acetate/propionate ratio, and pH

confirmed that these were the main factors separating the sheep groups (data not shown).

Overall, the MDS plot in Figure 2 illustrates that pot scrubbers and grain influenced the

rumen fermentation and that there was an combined effect of these 2 variables.

Figure 4.2: Non-metric multi-dimensional scaling (MDS) plot of effects of in vitro methane production, pH, VFA concentration, and ruminal retention time on sheep fed an oaten-chaff and a high grain diet with and without pot scrubbers. O: oaten-chaff diet; G: grain diet; OP: oaten-chaff diet + pot scrubber; GP: grain diet + pot scrubber.

Effect of diet

The decline in in vitro methanogenesis when sheep were fed a high grain diet and the

significantly lower acetate/propionate ratio is consistent with previous observations

(Hodgson and Thomas, 1975; Van Kessel and Russell, 1996; Baker, 1997; Lana et al.,

1998; Russell, 1998; Hristov et al., 2001).


53

In the present study, methane levels were reduced in vitro for both groups of high

grain-fed sheep with or without pot scrubbers, but it is possible that a similar decrease was

a result of a different balance of microorganisms competing for hydrogen, as I found in

vitro methane production from low grain-fed sheep with pot scrubbers was already

significantly less than that of oaten-chaff-fed sheep (Table 4.1). The differences found

using multivariate statistical analysis confirm the above mentioned effects of grain.

Effect of pot scrubbers

The possible explanations for the observed effect of pot scrubbers on acetate/propionate

ratio when a high grain diet was offered would be similar to the reasons described above.

Although, pot scrubbers did not affect rumen retention time as hypothesized and observed

by Matsuyama et al. (2000), this is supported by the findings of Loerch (1991). Who

observed that there was no effect of pot scrubbers on ruminal pH, VFA concentrations,

dilution rate, and rumen volume for cattle on a high grain diet (100%). Instead, he

concluded that the beneficial effects observed were due to stimulation of the rumen wall.

Based upon my findings, it is reasonable to conclude that the main benefit of pot scrubbers

may be stimulation of the rumen wall, but there is no doubt that adding pot scrubbers also

affected microbial fermentation in the rumen.

In conclusion, in this experiment I changed the fermentation characteristics in the

rumen of sheep by using combinations of pot scrubbers and grain and examined the effect

this has on in vitro methanogenesis. I found that pot scrubbers, in combination with low

grain in the diet significantly decreased the amount of methane produced (P < 0.05), and I

observed a similar effect in sheep consuming a high grain diet without pot scrubbers. The

acetate/propionate ratio was lower in high grain-fed sheep with pot scrubbers compared to

sheep fed the same diet but without pot scrubbers in their rumen, indicating that I altered


54

the balance of hydrogen utilizing bacteria. These findings were supported by my multi-

variant statistical analysis including all data, which showed that sheep given the combined

treatment of high grain and pot scrubbers were different from all other sheep groups in this

experiment (Table 2). Overall, this indicates that the combined effect was important and

that further effort should be made to clarify fully the relationship between increasing grain

supplement in the diet and pot scrubbers and the microbes inhabiting the rumen, with a

focus especially on methanogens and their role. In conclusion, introducing pot scrubbers

into the rumens of livestock consuming low levels of grain may be a way to lower methane

emissions.

Chapter 5: Methanogen abundance and diversity

55

CHAPTER 5

Grain and artificial stimulation of the rumen wall

changes the association between methanogens and

rumen ciliates


56

5.1 Introduction

Methanogens that associate with rumen ciliates have been studied previously (Sharp et al.,

1998; Chagan et al., 1999; Tokura et al., 1999; Irbis and Ushida, 2004) and their diversity

spans three orders: the Methanobacteriales, the Methanomicrobiales, and the

Methanosarcinales, with the most abundant methanogens being Methanobrevibacter smithii

and Methanobrevibacter gottschalkii-like (>99% similarity). Methanogens occupy three

different niches (i.e. free-living, ecto- or endosymbiotic) within the rumen (e.g. Vogels et

al., 1980; Stumm et al., 1982). However, some methanogens attach and detach from ciliates

depending on the conditions in the rumen (Tokura et al., 1997). The reason for this close

association with the ciliates is thought to be because the methanogens get direct access to

hydrogen for methanogenesis from the ciliates’ metabolism (Stumm and Zwart, 1986;

Zinder, 1993), coupled with an added advantage of living in a protected environment for

the endosymbiotic methanogens.

The removal of the hydrogen produced during fermentation is important for the

efficiency of rumen fermentation (Stewart et al., 1997). One of the hydrogen “sinks” for

ruminants, especially on a grain diet, is propionate production, because propionate

formation requires hydrogen. Methanogens are thought to compete for hydrogen with

propionate producers and this is supported by the findings that inhibition of methanogens in

sheep results in an accumulation of hydrogen and an increase in propionate production

(Van Nevel et al., 1969). Furthermore, the acetate to propionate ratio in the rumen has been

found to have an inverse relationship with methanogenesis (Van Kessel and Russell, 1996;

Lana et al., 1998; Russell, 1998). Therefore, as the grain content of the diet increases one

would expect the competition for hydrogen between the methanogens, particularly the


57

free-living population, and the propionate producers to increase and the diversity and

abundance of methanogens in the three niches to change. The main competition would be

between the free-living methanogens and the propionate producers and it is likely that the

diversity in this niche would change because it would favour those methanogen species that

have a higher affinity for hydrogen. In addition, more methanogens would be expected to

associate with the ciliates, where they gain exclusive access to hydrogen.

In Chapter 4, I found that a combination of low levels of grain (i.e. 35% of energy

intake) in the diet and pot scrubbers (Loerch, 1991) reduced methane production compared

with sheep fed oaten chaff, without changing the acetate to propionate ratio. There was no

additional reduction in methane production when the sheep were fed a high grain diet (i.e.

70% of energy intake), but the acetate/propionate ratio in sheep fed the high grain diet with

pot scrubbers was lower than in the sheep without pot scrubbers. The pot scrubbers did not

reduce ruminal retention rates as they were expected to do, but their physical presence still

decreased methane production and acetate to propionate ratios when in combination with

low or high grain, respectively. This indicates that, in both cases, the pot scrubbers add to

the competition for hydrogen in the rumen of sheep fed grain.

In this study, I examined the effect of grain and pot scrubbers on the diversity and

abundance of methanogens in the different rumen niches. I tested two hypotheses: (i) that

the diversity of methanogens would decrease and the abundance of methanogens associated

with ciliates would increase in the rumen of sheep consuming grain; and (ii) that these

effects would be more pronounced in sheep with pot scrubbers.


58

5.2 Materials and Methods

Experimental design

As described in Chapter 3.

Rumen sampling

On the last day of each diet phase all sheep were sampled. Rumen samples were collected 1

h after feeding, when the number of methanogens associated with ciliates should be at their

highest (Tokura et al., 1997). Samples were collected from at least four different locations

in the rumen, which were then pooled. Approximately, 130 mL of rumen digesta was taken

from each sheep and each sample was strained through two layers of muslin cloth. The

particulate matter was washed with 120 mL of phosphate-buffered saline (PBS) (137 mM

NaCl; 2.7 mM KCl; 10 mM Na2HPO4; 2 mM KH2PO4; pH 7.4) to release the

microorganisms attached to plant material; the strained rumen fluid and PBS buffer wash

were combined.

The rumen samples were divided into three different fractions to examine and

quantify methanogens. Separation of the ciliate fraction in the rumen fluid was performed

according to Williams and Coleman (1997). A brief summary of the procedure follows: A

20 mL sample of strained rumen fluid was stored from all sheep for DNA extraction. This

fraction contained all groups of methanogens (free-living, ecto- and endosymbiotic

methanogens). To eliminate the free-living methanogens the remaining rumen sample

(~200 mL) was divided equally into four 50 mL falcon tubes with lids fixed loosely and the

tubes incubated for 60 min at 39 °C to allow the ciliate fraction to settle. The ciliate

sediment was then collected from all four tubes with a hand-drawn Pasteur pipette and


59

transferred into one new 50 mL falcon tube containing 45 mL PBS buffer. The remaining

supernatant in each of the four tubes was centrifuged (Beckman GS-6R centrifuge) at 1,000

x g for 10 min and the resulting pellet was collected and added to the 50 mL falcon tube

containing the suspended sediment and a 15 mL sample was taken for DNA extraction.

This fraction included ecto- and endosymbiotic methanogens. To eliminate the

ectosymbiotic methanogens, the sample was centrifuged and washed in PBS buffer

(Beckman, GS-6R) at 200 x g for 30 s. This procedure was repeated four times and, after

the last spin, the ciliate pellet was resuspended in 30 mL of PBS buffer and stored at -20 °C

for DNA extraction. This fraction contained endosymbiotic methanogens. All samples were

stored in 6.25% isobutanol and snap-frozen in liquid nitrogen.

DNA extraction and quantification

Total DNA was extracted from the collected samples using an ultra clean™ fecal DNA kit

(Mo-Bio Laboratories Inc., CA). To maximise the DNA yield, different amounts of starting

material were used for the different sample fractions based on preliminary experiments

(data not shown). For strained crude rumen fluid, a 0.5 mL sample was used for DNA

extraction and a 1.0 mL sample was used for the two remaining fractions in PBS buffers

(i.e. one containing ecto-and endosymbiotic methanogens and the other endosymbiotic

methanogens). All samples were taken with a cut off Eppendorf pipette tip. Before DNA

extraction, all samples were washed with water-saturated diethyl ether to remove the

isobutanol. Samples were centrifuged at 16,100 x g for 5 min in an Eppendorf bench top

centrifuge (Model: 5415C, Hamburg, Germany), the supernatant was removed and the

pellet resuspended in 200 µL of water-saturated diethyl ether. This step was repeated and

the pellet resuspended in 200 µL millipore water, which was centrifuged again for 2 min


60

at 16,100 x g. The supernatant was removed and the pellet used for DNA extractions. DNA

was extracted using the manufacturer’s directions with the following modifications: a flat

top multi-tube vortex (Baxter Scientific Products, model 2600, USA) was used and the

amount of S3 buffer was doubled. This was followed by an extra wash of the column with

S4 buffer.

DNA concentration was determined using SYBR green gel stain (Sigma cat. No.

S9430) and a Polarstar Galaxy Fluorimeter (Labtechnologies GmbH, Offenburg, Germany)

according to the method and recommendations of Rengarajan et al. (2002). Estimates were

the mean of at least two replicates.

Denaturing Gradient Gel Electrophoresis (DGGE)

The extracted DNA was used in PCR amplifications to target the methanogen 16S rRNA

gene using newly designed methanogen-specific primers designed for DGGE. For all PCR-

DGGE amplifications, a proof reading, high fidelity DNA polymerase was used to lower

the risk of point mutations. The forward primer Met630F*GC, with a 40 bp GC-clamp (5`-

CGC CCG CCG CGC CCC GCG CCC GGC CCG CCG CCC CCG CCC G GGA TTA

GAT ACC CSG GTA GT-`3) and the reverse primer Met803R (5`-GTT GAR TCC AAT

TAA ACC GCA-`3) [numbered according to Methanomicrobium mobile (AY196679)]

were included in a PCR mixture (25 µL) containing 1.0 µL template DNA (10-200 ng), 3.0

µL dNTP mixture (200 µM of each dNTP), 2.5 µL 10X Qiagen PCR buffer (with 15 mM

MgSO4), 2.5 µL of forward primer (200 nM), 2.5 µL of reverse primer (200 nM), 1.0 µL of

25 mM MgSO4, 1.0 µL of proofstart™ DNA polymerase (2.5 Units) (Qiagen GmbH,

Germany), and 11.5 µL sterile Millipore H2O. After a 5 min hot start at 95ºC, one PCR

cycle consisted of the following parameters: 95 ºC for 30 s, 58 ºC for 30 s and 72 ºC for


61

90 s. This cycle was repeated 40 times, and on the 40th and last cycle, the primer extension

step was extended for 10 min.

The PCR product (25 µL) was run on a 1.5% agarose gel for 60 min at 100 volts

and stained with ethidium bromide. Bands were visualised using a Gel-Doc system

(BioRad, Hercules, CA) using the Quantitative One software (BioRad, Hercules, CA). A

PCR band circa 200 bp was excised from the gel using Extract gel cutters (Geneworks,

Adelaide, Australia) and the DNA was extracted from the agarose gel using QIAquick gel

extraction kit (Qiagen, Doncaster, Australia) according to the manufacturer’s instructions,

and quantified (Rengarajan et al., 2002). The extracted product was then stored at -20 ºC.

The extracted PCR products were quantified to ensure similar amounts of PCR product

were loaded onto DGGE gels and to eliminate single stranded bands on the DGGE gel.

DGGE was performed using the Dcode Universal Mutation Detection system (16

cm, BioRad, Hercules, CA). The amplicons were run using an 8% wt/vol gel [Acrylamide:

bisacrylamide 37.5:1 (BioRad, Hercules, CA)] with a denaturing gradient ranging from 40 -

55% of urea and formamide (100% corresponds to 7 M urea and 40% wt/vol formamide).

DGGE was performed at a constant 150 volts for 7 h and a constant temperature of 59 ºC.

A final concentration ranging from 3 – 6 ng amplicon/µL of gel-extracted PCR product in

15 µL of deionised H2O and 10 µL 2X loading buffer (0.05% (w/w) bromophenol blue,

0.05% (w/w) xylene cyanol, 73.61% glycerol, 26.29% Millipore H2O) was loaded on the

gel. This means that all products within the range were loaded undiluted, whereas products

with a concentration higher than 6 ng/µL were diluted to a concentration of 5 ng/µL with

Millipore water. At the completion of the run, the gels were stained for 20 min with SYBR

green (Sigma-Aldrich) and DGGE bands were visualised using the Bio-Rad’s gel-doc


62

system using the Bio-Rad Quantitative one software.

To standardise and reduce differences between gels, four markers were run on every

gel. Three of the markers were the 50 bp gel marker (Fermentas Life Sciences, SM0371),

which were loaded on each side of the gel and in the middle. On the left side of the centre

marker, the same sample was loaded on every gel to correct for between-gel variations. The

dominant bands were excised from the gel and sequenced. The amplicon was extracted

from the acrylamide gel using a combination of QIAEX II (step 1–4) and Qiaquick gel

extraction kit. The amplicons were then re-amplified using Met630F and Met803R without

the GC-clamp on the forward primer and a reduced annealing temperature of 45 ºC for only

30 cycles. The re-amplified amplicons were then sequenced in both directions using the

same primers with an ABI Prism 373 automated DNA sequencer (Applied Biosystems Inc.,

Foster City, Ca.) using Big Dye terminator. Methanogen sequences were confirmed by

using the Basic Local Alignment Search Tool (BLAST) (Altschul et al., 1997) in GenBank.

Bands in similar positions on different gels were excised to verify the assumption that these

bands were the same. DGGE gels were analysed using the GelCompar II software (Applied

Maths, Inc., Texas).

Phylogenetic analysis

Sixty-nine 16S rRNA gene sequences representing the Methanobacteriales,

Methanomicrobiales, and Methanosarcinales were included in the phylogenetic analysis.

All sequences were aligned using Clustal W (Thompson et al., 1994). PHYLIP (ver. 3.62C)

(Felsenstein, 2004) was used to calculate the evolutionary distances between pairs of

nucleotide sequences using the Kimura two-parameter correction model (Kimura, 1980). A

distance matrix tree was then constructed by using the neighbor-joining method (Saito


63

and Nei, 1987) and was bootstrap re-sampled 100 times (Felsenstein, 1985).

Methanococcus vannielii was used as the outgroup, having been shown previously to

branch outside the clade consisting of the rumen methanogens (Wright et al., 2004; 2006;

2007). The phylogenetic relationships between the sequenced DGGE bands, the

methanogens known to be closely associated with the rumen ciliates, and the other rumen

methanogens were determined. The only methanogens found previously to be associated

with the rumen ciliates that were used in the phylogenetic tree were clones identified by

Regensbogenova et al. (2004a). Other available sequences did not span the full region of

the 16S rRNA gene that was used for these analyses and were therefore omitted.

Quantitative real-time PCR

The three different sample fractions representing free-living, ecto- and endosymbiotic and

endosymbiotic methanogens were analysed using real-time PCR to determine the numbers

of methanogens in the extracted DNA. Real-time PCR amplification was carried out with

the Bio-Rad Icycler using the same DGGE primers, but without the 40 bp GC-clamp on

Met630F. Reactions were done in a 25 µL volume containing the following reagents: 12.5

µL SYBR green mix (QuantiTect™ SYBR® Green PCR, Qiagen), 9.5 µL sterile Millipore

H2O, 1.0 µL forward primer (10 µM concentration), 1.0 µL reverse primer (10 µM

concentration) and 1.0 µL template DNA (10-200 ng). Real-time PCR amplification was

initiated by a hot start at 95°C for 15 min, followed by 40 cycles of 95 °C for 30 s, 60 °C

for 30 s, and 72 °C for 60 s. Threshold cycles were calculated automatically by the Icycler

software (version 3.5).

PCR efficiency for each reaction was calculated from the logarithmic portion of the


64

sigmoid shaped curve in real-time PCR reactions according to the methods described by

Lui and Saint (2002). Three dilutions of DNA were amplified and the threshold cycle (Ct)

of the most efficient PCR (PCR efficiency close to 2 is doubling of DNA each cycle and

the slope of the log relative fluorescent units (RFU) vs. cycle number closest to 0.301) was

recorded. Real-time PCR can be described by the formula: nNN ε×= 0

Where N0 is the number of methanogens/amount of DNA present initially (before the PCR),

ε is the efficiency of the PCR and n is the number of cycles (Rasmussen, 2001). The

amount of DNA required to reach the threshold of fluorescence detection is the same for all

PCR reactions. Therefore, the amount of DNA present in any PCR reaction at its Ct is the

same, a constant, N.

To compare samples from different fractions and different dilutions, the theoretical

Ct for undiluted DNA extracted was calculated according to the equation:

( )ε

εlog

log dilutionActual Ct

÷

After the numbers of methanogens were calculated for individual samples they were

divided by the total amount of DNA to compare among samples within and between

fractions.

The constant N was determined for sample fractions in rumen fluid and PBS buffer

using a mixture of pure cultures of characterised methanogenic species. The following

species were used: Methanobacterium (Mb) formicicum, Methanobrevibacter (Mbr)

ruminantium, Mbr. smithii, Mbr. woeseii, Methanococcus (Mcc) vannielii and

Methanospirillum (Msl) hungateii. Each of the pure isolates was cultured and the

Theoretical Ct =


65

numbers of cells were counted using a coulter counter (MultisizerTM 3, Beckman, CA).

Two, 1.0 mL aliquots of culture were taken out, mixed and centrifuged. One millilitre was

resuspended in clarified and autoclaved rumen fluid and the remaining 1.0 mL was

resuspended in PBS buffer, followed by DNA extraction as previously described. A series

of dilutions ranging from 103 – 108 were made up and real-time PCR was performed as

previously described. From these dilutions, N was calculated using the equation above.

Evaluation of the real-time PCR assay was done by spiking samples containing

crude strained rumen fluid and PBS buffer with cultured methanogens. A mixture of four

different samples from the strained crude rumen fluid fraction was spiked with a known

number of methanogen cells. Another mixture of four samples from the two fractions in

PBS buffer containing ecto- and endosymbiotic and endosymbiotic methanogens,

respectively, were also spiked with methanogen cells ranging from 106 – 108. The DNA

from the spiked samples was then extracted and real-time PCR was performed as described

previously. The predicted numbers of 16S rRNA gene copies present in the spiked sample

were then regressed against the same copies recovered from the corresponding mixed

sample to evaluate the real-time PCR procedure. Furthermore, threshold cycle values of

DNA extracted from eight pure cultures of methanogens were recorded and compared to

non-target DNA from bacterial cultures, a ciliate clone and plant material to detect the

difference in threshold cycles between samples containing target and non-target DNA.

Melting curves were also used to examine the differences between target and non-target

DNA.


66

Statistical analysis and diversity index

Quantitative real-time PCR data were reassembled in a Euclidean distance matrix before

using ANOSIM in the multi-variable statistical packed PRIMER v6 (Clarke and Gorley,

2006). The DGGE banding patterns were also analysed using the PRIMER v6 statistical

package using the Bray-Curtis similarity matrix instead of a Euclidean distance matrix

(Clarke, 1993).

The Shannon index of general diversity (Shannon and Weaver, 1949) was

calculated using the DGGE banding patterns and significant differences in diversity

between treatments were found by using a student’s t-test.

Nucleotide sequence accession number

The sequenced DGGE bands reported in this paper have been deposited in the GenBank

database under accession numbers EF513256 to EF513270.

5.3 Results

Effect of diet

The diversity of methanogens in the different rumen niches did change because of diet.

However, independent of diet, the abundance of methanogens did not change in any of the

niches during the study, although there was an initial difference in diet phase 1 between the

2 groups of sheep fed oaten-chaff without pot scrubbers (Group ‘O’ and ‘G’) (Figure 5.1 A,

B and C).


67

All methanogens

0

2

4

6

8

Diet phase 1(Oaten-chaff)

Diet phase 2(35% Grain)


Log

(10)

of m

etha

noge

ns p

er

ng D

NA

ext

ract

ed

OOPGGP

Ecto- and endo-symbiotic methanogens

0

2

4

6

8




Log

(10)

of m

etha

noge

ns p

er

ng D

NA

ext

ract

ed

OOPGGP

Endo-symbiotic methanogens

0

2

4

6

8




Log

(10)

of m

etha

noge

ns p

er

ng D

NA

ext

ract

ed

OOPGGP

Figure 5.1: A, B and C: Logarithm-transformed (base 10) values of mean numbers of methanogen cells found in each treatment group over the 3 diet phases with SEM. Groups with different superscript within a diet were found to be significant different (P < 0.05). Groups with a different number of superscript * were found to change significantly between diet phases. O: oaten-chaff diet; OP: oaten-chaff diet + pot scrubber; G: grain diet; GP: grain diet + pot scrubber.

a

b**

b a, b a, ba, b a*

b A

B

a, b a

b bC


68

The main effect of diet was seen on the diversity of methanogens. In samples containing

ecto- and endosymbiotic methanogens; the DGGE banding patterns of methanogens

changed in the ‘grain-fed’ group of sheep when they were moved from an oaten chaff diet

to a low grain diet (diet phase 1 to 2; P < 0.05) and also between the these sheep when they

were fed oaten chaff compared to a high grain diet (diet phase 1 and 3; P < 0.05). The

DGGE banding pattern of endosymbiotic methanogens also changed between diet phase 1

and 3 for grain-fed sheep (P < 0.01) and were different from sheep fed oaten chaff (P <

0.05). The corresponding Shannon indices of endosymbiotic methanogens in samples from

sheep fed grain decreased from diet phase 1 to 2 (P < 0.05) and increased when the diet

shifted from a low to a high grain diet (P < 0.05) (Table 5.1).


The combination of pot scrubbers with either oaten-chaff or grain in the diet affected both

the types and numbers of methanogens living freely or associated to rumen ciliates in the

rumen. The abundance of methanogens was found to change in different niches. For low

grain-fed sheep without pot scrubbers the total abundance of methanogens was higher than

in sheep fed oaten-chaff with pot scrubbers (P < 0.05) (Figure 5.1A). While the abundance

of endosymbiotic methanogens in high grain-fed sheep with and without pot scrubbers was

higher when compared to oaten-chaff-fed sheep with pot scrubbers (Figure 5.1C). Changes

were also observed when pot scrubbers were combined with oaten-chaff. The total

abundance of methanogens in sheep fed oaten-chaff with pot scrubbers decreased between

diet phase 1 and 2 (P < 0.05), but numbers returned close to original levels by the end of

diet phase 3 (Figure 5.1A). Otherwise, the abundance did not change between diet phases

for any of the sheep groups.


69

The combined effect of grain and pot scrubbers on methanogens in different niches

was also evident on the diversity of methanogens measured by both DGGE banding

patterns and the Shannon index. The DGGE banding patterns of all methanogens from

strained crude rumen fluid, representing all methanogens, in the ‘grain-fed’ group with pot

scrubbers were different when they were fed an oaten-chaff diet (diet phase 1) compared to

a high grain diet (diet phase 3; P < 0.05) (Figure 5.2). Furthermore, the DGGE banding

patterns for the endosymbiotic methanogens in oaten-chaff-fed sheep with pot scrubbers

changed over time (diet phase 1and 2; P < 0.05), which was also reflected as an increase in

the Shannon index (P < 0.05) (Table 5.1).

Diet phase 1 Diet phase 3 . M 1 5 6 8 10 12 M 1 5* 6 8 10 12 M

Figure 5.2: Compiled DGGE picture showing the banding patterns of rumen methanogens from sheep in the grain and pot scrubber treatment group, and the change in DGGE banding patterns from an oaten-chaff diet (diet phase 1) to a high grain diet (diet phase 3). M: marker. The numbers correspond to the sheep in the treatment group. * Lane not included in the analysis. Gel lanes are not normalized according to marker.


70

The Shannon index also changed between diet phases in samples containing ecto- and

endosymbiotic methanogens, where the Shannon index increased when sheep with pot

scrubbers moved from an oaten-chaff diet to a low grain diet (diet phase 1 to 2; P < 0.01)

(Table 5.1). Changes were also observed within diet phases. In diet phase 3 the DGGE

banding patterns from strained crude rumen fluid, representing all niches of methanogens,

were different between high grain- and oaten-chaff-fed sheep with pot scrubbers (P < 0.05).

Furthermore, the Shannon index for samples containing ecto- and endosymbiotic

methanogens from sheep on the low grain diet were higher than in sheep on the same diet

without pot scrubbers (P < 0.05) (Table 5.1). However, the Shannon index for

endosymbiotic methanogens in sheep on the low grain diet without pot scrubbers was lower

than in sheep fed the oaten-chaff diet with pot scrubbers (P < 0.05) (Table 5.1).

Table 5.1: Average Shannon index (mean ± SD) calculated for each treatment group over the three diet phases and sample fractions with std. deviation. Groups with different superscript within a diet (column) were found to be significant different (P < 0.05). Groups with a different number of superscript * were found to change significantly between diet phases. O: oaten-chaff diet; OP: oaten-chaff diet + pot scrubber; G: grain diet; GP: grain diet + pot scrubber.

All methanogens Ecto- and endosymbiotic Endosymbiotic

Diet phases Diet phases Diet phases 1 2 3 1 2 3 1 2 3

O 0.26 ±0.010

0.48 ±0.014

0.47 ±0.011 0.48b

±0.013 0.49a,b ±0.015

0.31 ±0.011 0.34

±0.011 0.49a,b ±0.017

0.41 ±0.017

OP 0.42 ±0.014

0.52 ±0.014

0.51 ±0.013 0.45b

±0.015 0.56a,b ±0.016

0.38 ±0.011 0.29*

±0.012 0.58b,** ±0.015

0.40 ±0.013

G 0.36 ±0.014

0.39 ±0.013

0.47 ±0.014 0.50b

±0.015 0.40a

±0.014 0.30

±0.010 0.39* ±0.014

0.31a,** ±0.008

0.38* ±0.015

GP 0.36 ±0.014

0.51 ±0.014

0.39 ±0.011 0.32a,*

±0.008 0.48b,** ±0.014

0.39 ±0.013 0.37

±0.013 0.59a,b ±0.015

0.41 ±0.014


71

Identification of DGGE bands and their phylogenetic relationship

In total, 35 different DGGE bands were identified across all gels and 15 of the most

dominant bands were sequenced and identified presumptively using sequence alignments

(Table 5.2). Ten of the 15 different bands had >98% identity to a Methanobrevibacter spp.,

whereas the other five bands were found to be between 82% and 96% similar to

Methanobrevibacter (Table 5.2). Of the remaining 20 DGGE bands, 11 were unique to one

or two sample fractions. Five of the 11 unique DGGE bands were only found in samples

from strained crude rumen fluid and four bands were identified in both strained crude

rumen fluid and the ecto- and endosymbiotic fractions. One band was only seen in the ecto-

and endosymbiotic fraction and the final band was seen in strained crude rumen fluid and

the endosymbiotic fraction. None of the unique bands was observed in more than a total of

three sheep. Despite finding some unique bands, the symbiotic methanogens were similar to

the dominant methanogen species found in the strained crude rumen fluid.


72

Table 5.2: Sequenced and identified DGGE bands, their nearest valid neighbour and their presence in the different sample fractions.

Observed in % of Fraction Band No.

Sequence (bp) Nearest valid taxon % Identity All Ecto/endo Endo

1 151 Mbr. thaueri CW 99.0 100.0 95.2 100.0 2 150 Mbr. olleyae 100.0 75.4 60.3 72.6

3 150 Mbr. acididurans 98.0 7.2 9.5 11.3

4 151 Mbr. gottschalkii PG 92.7 72.5 74.6 67.7

5 158 Mbr. ruminantium; Mbr. olleyae 82.3 94.2 93.7 87.1

6 151 Mbr. thaueri CW 99.0 53.6 55.6 56.5

7 150 Mbr. thaueri CW 100.0 66.7 65.1 61.3

8 125* Mbr. thaueri CW; Mbr. millerae 99.0 52.2 49.2 29.0

9 151 Mbr. thaueri CW 99.0 44.9 44.4 33.9

10 128* Mbr. Smithii PS 85.2 8.7 7.9 9.7

11 149 Mbr. thaueri CW 99.0 81.2 79.4 83.9

12 150 Mbr. thaueri CW; Mbr. gottschalkii PG 99.0 17.4 14.3 3.2

13 151 Mbr. thaueri CW 96.0 14.5 7.9 4.8

14 149 Mbr. thaueri CW 99.0 14.5 14.3 4.8

15 152 Mbr. thaueri CW 94.0 8.7 14.3 8.1 - Accession numbers for DGGE bands and valid species refer to phylogenetic tree. - Mbr: Methanobrevibacter. - * sequenced in one direction only.


73

Figure 5.3: Phylogenetic relationships of the methanogens derived from 16S rRNA gene evolutionary distances produced by the Kimura two-parameter model (Kimura, 1980) and constructed by using neighbor-joining method (Saito and Nei, 1987) with 100 bootstrap resamplings of the data. Bootstrap values are indicated at the nodes. Bootstrap values less then 50% are not indicated.

Methanobrevibacter wolinii SH (MSU55240) Methanobacterium formicicum (AY196659)

Methanosphaera stadtmanae (AY196684) Methanobrevibacter sp. OCP (AY615203)

Methanobrevibacter sp. Z4, Z6, Z8 (AY196670-2) ARC 50 (AF029193) ARC 66 (AF029194) GB17 (AY422970) DGGE band 2 (EF513257) Methanobrevibacter sp. NT7 (AJ009959) PE-CAN.04 (DQ123864) Methanobrevibacter olleyae (AY615201) ON-CAN.01 (DQ123873)

Methanobrevibacter ruminantium (AY196666) ARC 61 (AF029184) PE-CAN.03 (DQ123863) ARC 44 (AF029180) ARC 52 (AF029182) ARC 40 (AF029179) DGGE band 3 (EF513258) Methanobrevibacter sp. ATM (AF242652)

Methanobrevibacter sp. FMB1 - 3 (AJ243838 - 40) Methanobrevibacter smithii PS (AY196669) ON-CAN.10 (DQ123880)

Methanobrevibacter woesei (MSU55237) DGGE band 5 (EF513260)

DGGE band 6 (EF513261) DGGE band 13 (EF513268) ON-CAN.12 (DQ123882) CSIRO3.08 (AY351496) PE-CAN.07 (DQ123867) Methanobrevibacter sp. SM9 (AJ009958)CSIRO3.12 (AY351500) DGGE band 7 (EF513262)

Methanobrevibacter gottschalkii PG (MSU55239) DGGE band 4 (EF513259) DGGE band 12 (EF513267)

CSIRO2.05 (AY351471) CSIRO2.07 (AY351473)

DGGE band 15 (EF513270) DGGE band 8 (EF513263)

CSIRO-Qld02 (AY995276) GB13 (AY422966) Methanobrevibacter sp. 1Y (DQ135988) DGGE band 1 (EF513256) DGGE band 9 (EF513264) ON-CAN.11 (DQ123881) Methanobrevibacter thaueri CW (MSU55236)

DGGE band 14 (EF513269) DGGE band 11 (EF513266)

Oph-12 (AJ606400) Met-01 (AJ606401)

Ent-01 (AJ606413) Ent-11 (AJ606415)

Ent-14 (AJ606416) Met-05 (AJ606405)

Met-11 (AJ606409) Met-12 (AJ606406) Ent-18 (AJ606417) Methanomicrobium mobile (AY196679) Met-14 (AJ606410) Oph-05 (AJ606411) Met-19 (AJ606408) Met-04 (AJ606407)

Methanospirillum hungatei (AY196683) Methanosarcina barkeri (AJ012094)

Oph-22 (AJ606403) Oph-24 (AJ606404)

Methanococcus vanielii (AY196675)

100 100

86

84 55

71

57

99

91 99

100

93

53

Methanobrevibacter sp. FM1 (AJ550157)

100

0.05


74

The phylogenetic relationships between the sequenced DGGE bands, the

methanogens known to be closely associated with the rumen ciliates, and the other rumen

methanogens were determined (Figure 5.3). All of the DGGE bands that were sequenced

and aligned clustered among the Methanobrevibacter clade.

Validation of real-time PCR assay

Methanogens representing eight different genera and four orders all amplified within a few

threshold cycles of each other and they also started amplifying about 10 cycles before the

non-target DNA templates (Table 5.3).

Table 5.3: Specificity of primers Met630F and Met803R for quantification of rumen methanogens Template DNA Group PCR efficiency Threshold Cycle Methanosphaera stadtmanae methanogen 1.89 18.9 Methanobrevibacter woeseii methanogen 1.93 19.6 Methanobrevibacter smithii methanogen 1.99 20.0 Methanococcus vannielii methanogen 1.81 20.3 Methanobrevibacter sp. 1Y methanogen 1.89 20.7 Methanomicrobium mobile methanogen 1.93 20.8 Methanospirillum hungateii methanogen 1.84 21.4 Methanobrevibacter ruminantium methanogen 1.84 21.5 ciliate clone mixture ciliate 2.18 29.8 Streptococcus bovis bacteria 1.91 32.4 Prevotella ruminicola ruminicola bacteria 1.97 33.8 Plant material (mixed) plant 0.84 36.0 Negative control water 3.05 34.9

The mix of four samples from the strained rumen fluid fraction and a mix of four

samples from the two fractions in PBS buffer containing ecto- and endosymbiotic and

endosymbiotic methanogens were spiked with a known number of methanogenic cells from

pure cultures. The regression analysis of the number of methanogens added against the

abundance estimated using real-time PCR demonstrated that the increased concentration of

cells could be detected using real-time PCR. The regression equations of methanogen


75

cells added versus spiked sample showed that the recovery of spiked methanogens was

76% for the strained rumen fluid fraction (Figure 5.4A) and 98% for the fraction in PBS

buffer (Figure 5.4B). Therefore, all quantitative real-time PCR data obtained were corrected

for these losses.

y = 0.7626x + 2.3299R2 = 0.9584

0

2

46

8

10

log

spik

ed c

rude

ru

men

flui

d sa

mpl

e

y = 0.9803x + 0.9784R2 = 0.9203

0

2

4

6

8

10

6 7 8 9Log methanogens added

log

spik

ed E

nto/

endo

-an

d En

do-s

ampl

e

Figure 5.4: Validation of real-time PCR using regression analysis of the logarithm-transformed (base 10) number of methanogens added versus the logarithm-transformed (base 10) of the number of methanogens recovered in the DNA extracted from spiked samples measured using real-time PCR. A: Mixture of crude rumen fluid from four samples containing all methanogen groups spiked with methanogenic cells from pure cultures, and B: A mixture of four samples containing ecto/endo- and endosymbiotic methanogens spiked with methanogenic cells from pure cultures.

5.4 Discussion

To my knowledge this is the first study where the effect of changes in rumen conditions on

the diversity and abundance of methanogens in their different niches has been examined.

In this study, I examined the effect of increasing the amount of grain in the diet of

sheep and adding pot scrubbers to their rumen on the diversity and abundance of

A

B


76

methanogens in the different rumen niches. I tested two hypotheses: (i) that the diversity of

methanogens would decrease and the abundance of methanogens associated with ciliates

would increase in the rumen of sheep consuming grain; and (ii) that these effects would be

more pronounced in sheep with pot scrubbers.

The first hypothesis was partially supported as the addition of low grain was found

to change the DGGE banding pattern and the Shannon index in the endosymbiotic

methanogens decreased. At the same time the Shannon index in the ecto- and

endosymbiotic niche was significantly lower in sheep fed grain without pot scrubbers than

those with pot scrubbers. In addition, there was an increase in the number of methanogens

in the endosymbiotic niche when sheep were fed the high grain diet with and without pot

scrubbers when compared to sheep fed oaten-chaff with pot scrubbers. The second

hypothesis was also supported in part because the diversity of the ecto- and endo symbiotic

methanogens in sheep fed low grain with pot scrubbers decreased. However, the diversity

of the endosymbiotic methanogens increased in sheep that were fed the high grain diet

without pot scrubbers.

It has been suggested that methanogens increase their association with rumen

ciliates to gain better access to hydrogen for methanogenesis or when there is a change in

diversity or abundance of rumen ciliates (Krumholz et al., 1983; Stumm and Zwart, 1986;

Itabashi et al., 1994). One of the hydrogen “sinks” for ruminants, especially on a grain diet,

is propionate production, because propionate formation requires hydrogen. Methanogens

are thought to compete for hydrogen with propionate producers and this is supported by the

findings that inhibition of methanogens in sheep results in an accumulation of hydrogen

and an increase in propionate production (Van Nevel et al., 1969). Furthermore, these


77

findings are consistent with the general observation that propionate levels are increased

when ruminants are fed high grain diets (Hodgson and Thomas, 1975; Baker, 1997;

Russell, 1998; Hristov et al., 2001). Additionally, Tajima et al. (2001a), found more than a

two-fold increase of Selenomonas ruminantium, a propionate producer, after an extended

period of feeding a high grain diet. In this Chapter, the total abundance of methanogens in

sheep fed the high grain diet, with or without pot scrubbers, was not different to the other

groups of sheep, but the abundance of the endosymbionts was higher. Therefore, one

explanation for this could be that more of the hydrogen available in the rumen was used for

propionate production and not for methanogenesis. The data obtained in Chapter 4, where

significantly higher acetate/propionate ratios were observed for sheep eating the high grain

diet compared to sheep eating oaten chaff, would support this conclusion. This would

reduce the amount of hydrogen available in the rumen fluid, forcing the methanogens to

associate with ciliates to survive under these ruminal conditions.

The increased association of methanogens with the ciliates was also coupled with a

significant reduction in methane produced in vitro from both high grain-fed groups

(Chapter 4). This also suggests that there was competition for hydrogen between the

methanogens and propionate production and that the methanogens associated

endosymbiotically with ciliates may produce less methane.

An alternative and simpler explanation for the increased abundance of methanogens

associated endosymbiotically with rumen ciliates on a high grain diet could be that there

was an increase in the abundance of rumen ciliates. To confirm this it would be necessary

to enumerate the rumen ciliates, which is the subject of Chapter 6.

The majority of the changes that were observed in the DGGE banding patterns and


78

Shannon indices were in samples where methanogens were associated with protozoa. This

also supports the argument that if hydrogen is limited in the rumen then methanogens

increase their association with ciliates either ecto- or endosymbiotically to get better access

to hydrogen. Overall, the changes in DGGE banding patterns and the Shannon indices

support the findings of Zinder (1993), who reported differences between methanogen

species in their affinity for hydrogen. If there is competition for hydrogen then the diversity

of methanogens would be expected to change because the environment would favour those

methanogens that have higher affinities for hydrogen or that can associate with the ciliates

to access hydrogen.

The pot scrubbers did not change ruminal retention time as they were expected to

(Chapter 4), but their physical presence still influenced the abundance and diversity of

methanogens in the rumen. This effect was most noticeable on the diversity and abundance

of methanogen populations associated with the protozoa in the sheep fed oaten chaff or low

grain diets. In Chapter 4, sheep that did not have pot scrubbers produced the same amount

of methane in vitro when they were fed oaten chaff or a low grain diet. However, sheep

with a combination of pot scrubbers and a low grain diet produced less methane in vitro

than sheep fed oaten chaff diet sheep with or without pot scrubbers (Chapter 4). These

results suggest that pot scrubbers, like grain, increase the competition for hydrogen and

probably target the free living niche. Results supporting the added pressure on methanogens

when low grain was combined with pot scrubbers can be found in the diversity data. For

grain-fed sheep without pot scrubbers an increase in the Shannon index was observed for

endosymbiotic methanogens between the low and a high grain diet. A similar increase in

the Shannon index was observed for the ecto- and endosymbiotic methanogens between


79

diet phase 1 and 2 for grain-fed sheep with pot scrubbers. The reason for the increase in the

Shannon index in diet phase 2 when pot scrubbers and grain are present is because then

there is an increased pressure for the methanogens to associate with rumen ciliates, similar

to when high grain was fed. This increase in Shannon index on both occasions

corresponded with the main decrease in methane production in vitro observed for the two

groups (Chapter 4). However, grain appears to dominate the effect of pot scrubbers on the

high grain diet because, regardless of pot scrubbers, the abundance of endosymbionts was

higher, and methane production in vitro lower, in sheep fed a high grain diet compared to

sheep fed oaten chaff.

Newbold et al. (1995) and Baker (1997) have suggested that methane production

and abundance may not be correlated and may instead depend upon the ruminal conditions

and the type of methanogens present. The data from my study support this suggestion

because there was no decrease in the total abundance of methanogens in sheep fed a high

grain diet, but they produced less methane in vitro (Chapter 4).

In the present study, an increase in diversity within a treatment was frequently

associated with a reduction in methanogen abundance in the same treatment. This suggests

that methanogens are opportunistic organisms, such that when one or a group of dominant

species are reduced, others flourish. This also means that sheep are very likely to harbour a

wide range of methanogens in small numbers, which are only detected when conditions

change in the rumen and they have the opportunity to prosper. If this is true, then it explains

why a dominant group of methanogens has been identified with few outliers (Tajima et al.,

2001b; Skillman et al., 2004; Wright et al., 2004; 2006).

Twenty-nine of the 35 bands were isolated from various DGGE gels and sequenced


80

and 15 different DGGE bands were identified. In order to use DGGE an assumptions was

made that bands in similar positions on different gels would be identical. This was

confirmed by sequencing bands in similar positions on different gels, which then allows for

comparisons between gels to be made confidently.

The phylogenetic relationship between methanogens, some previously found to

associate with rumen ciliates, and the DGGE bands and their nearest valid neighbours was

also examined. Twenty-nine of the 35 bands were isolated from various DGGE gels and

sequenced and 15 different DGGE bands were identified. Methanogens identified in the

rumen have been found to belong to three orders: the Methanobacteriales, the

Methanomicrobiales and the Methanosarcinales, with the dominant genus being

Methanobrevibacter (Miller and Wolin, 1986; Garcia et al., 2000). I identified five bands

that were 82-96% identical to various species within the Methanobrevibacter (Table 5.2), if

they had been full length sequences, it is likely that they would have represented new

species of methanogens. However, all DGGE bands that were identified clustered with

Methanobrevibacter genus (Figure 5.3), which fits well with previous studies where the

dominant methanogens belong to this genus (Lin et al., 1997; Stewart et al., 1997; Wolin et

al., 1997; Whitford et al., 2001; Wright et al., 2004). Furthermore, the genus

Methanobrevibacter has been reported to be associated with rumen ciliates (Sharp et al.,

1998; Chagan et al., 1999; Tokura et al., 1999; Irbis and Ushida, 2004). In contrast,

Regensbogenova et al. (2004a) found methanogens associated with the rumen ciliates

grouped with the genera in two different orders, the Methanomicrobiales and the

Methanosarcinales. Sharp et al. (1998) found all three methanogens from all the three

orders to be associated with rumen ciliates. The variation in methanogen profiles found


81

associated with ciliates could simply be due to dietary or regional differences. Differences

in the rumen methanogen populations between regions has been observed before when

major differences were found between methanogens in sheep from Western Australia and

from Queensland, Australia (Wright et al., 2004; 2006).

Methanogens have been found previously to live in three niches in the rumen, free-

living or ecto- or endosymbiotically (Vogels et al., 1980; Stumm et al., 1982). Eleven of

the DGGE bands identified were not represented in all sample fractions, which could

indicate that certain methanogens prefer or can only exist in one or two of the niches in

which the methanogens in the rumen exist. This would suggest that the five bands found

only in the strained crude rumen fluid represent methanogens that may prefer living freely

in the rumen, whereas the five bands found in the strained crude rumen fluid or the ecto-

and endosymbiotic fraction represent methanogens that were ectosymbionts of the rumen

ciliates. However, the band that was only found in the strained crude rumen fluid or the

endosymbiotic fraction may represent an endosymbiotic methanogen.

The DGGE bands and their phylogenetic relationships are similar to the findings by

Sharp et al. (1998) who reported that the dominant group of methanogens in the crude

rumen sample (i.e. the Methanobacteriaceae) was also the dominant group in the ciliate

fraction. Despite finding some unique bands, my results also indicate that the symbiotic

methanogens were similar to the dominant methanogen species found in the strained crude

rumen fluid.

Validation of real-time PCR and DGGE

Validation of real-time PCR and DGGE primers was done concurrently for the two

techniques because the primers were identical, except for the GC-clamp added for


82

DGGE. The development of new methanogen primers for both quantitative real-time PCR

and DGGE suitable for use with rumen samples was important to make direct comparisons

between the quantitative data and the DGGE data. The new primers were not biased

towards any of the methanogen groups as all of the methanogens amplified had threshold

cycles within a few cycles of each other (Table 5.3). The specificity was also satisfactory

because the methanogen target DNA started amplifying about 10 cycles before any of the

non-target DNA templates (Table 5.3). This was also confirmed using the melting curves

for the target and non-target DNA. This is important as SYBR green detection is non-

specific. The specificity and the primer’s efficacy for quantification were also tested by

spiking mixed rumen samples with known numbers of methanogens and constantly being

able to recover detect the difference in number of methanogens in the spiked samples.

Furthermore, the R2-value was consistently near 1. This was obtained for both the mixed

strained crude rumen fluid sample and the samples in PBS buffer containing ecto- and

endosymbiotic, and endosymbiotic only methanogens (Figure 5.4A & B). The calculation

of methanogen numbers was based upon replicate runs using the same DNA extraction and

correcting each run for PCR efficiency following the argument of Skillman et al. (2006b).

In conclusion, both a DGGE and a quantitative real-time PCR assay were developed

and validated to examine methanogen diversity and to quantify methanogens in the rumen

accurately. By applying these two techniques I was able to detect that both groups of high

grain-fed sheep had more endosymbiotic methanogens, but there was no change in total

abundance of methanogens. These findings correlated with lower methane production in

vitro and lower acetate/propionate ratios in vivo (Chapter 4), indicating that the abundance

of methanogens was not the deciding factor when it came to methanogenesis in the rumen,


83

but the level of association between methanogens (endosymbiotically) with the rumen

ciliates and competition for hydrogen probably was. Furthermore, my data indicate strongly

that the availability of hydrogen in the rumen for methanogenesis dictates whether the

methanogens were associated with rumen ciliates or not. Finally, Methanobrevibacter was

the dominant methanogen genus in all sample fractions. However, some bands were unique

to certain sample fractions.

Chapter 6: Ciliate abundance and diversity

84

CHAPTER 6

Grain changes the diversity of rumen ciliates but

not their abundance


85

6.1 Introduction

Methanogens associate symbiotically with rumen ciliates by attaching to their surface, or

residing within the ciliate (Vogels et al., 1980; Stumm et al., 1982; Finlay et al., 1994;

Newbold et al., 1995; Tokura et al., 1997; Sharp et al., 1998; Chagan et al., 1999; Tokura

et al., 1999; Schonhusen et al., 2003; Irbis and Ushida, 2004; Regensbogenova et al.,

2004a). The advantage for the rumen ciliate is thought to be that the methanogens keep the

hydrogen concentration low, which enhances the energy yield per mole of glucose

converted by the ciliate (Hino, 1982; Stumm and Zwart, 1986).

The diversity and abundance of rumen ciliates influences methane production.

Krumholz et al. (1983) found the methanogenic activity in rumen fluid was highest in

fractions containing high numbers of ciliates. Therefore, it is likely that if ciliate numbers

increase, then methanogenesis will also increase. Itabashi et al. (1994) found that a change

in the generic composition of the rumen ciliates can also influence methane production.

Increasing the percentage of grain in the diet can increase the abundance of ciliates, due to

enhanced substrate availability (Mackie et al., 1978; Franzolin and Dehority, 1996; Hristov

et al., 2001), and reduce ciliate diversity because some species, especially Entodinium spp.,

are favoured (Williams and Coleman, 1992).

Increasing rumen retention time through artificial stimulation of the rumen wall

using pot scrubbers has been demonstrated to change the diversity of methanogen

populations in the rumen (Chapter 5). However, the abundance and diversity of rumen

ciliates are less likely to be affected because ciliates are mainly attached to plant material

undergoing digestion in the rumen (Bauchop and Clarke, 1976; Williams and Coleman,


86

1992). This helps the ciliates to remain in the rumen when retention times are less than their

growth rate.

As reported in Chapter 5, the abundance of endosymbiotic methanogens increased

when a 70% grain diet was offered to sheep. However, this increase was correlated to a

significant decrease in methane production in vitro and a significant drop in

acetate/propionate ratios (Chapter 4). Therefore, it is important to establish whether the

increase in endosymbiotic methanogens was due to an increase in ciliate number, or

whether grain influences the number of endosymbiotic methanogens directly, to help

establish why methane production in vitro was lowered.

In this experiment I examined the effect of step-wise increases in grain content in

the diet and the addition of pot scrubbers to the rumen on the abundance and diversity of

rumen ciliates. It was hypothesised that increasing the amount of grain in the diet would

increase the abundance and reduce the diversity of rumen ciliates, whereas the abundance

and diversity of the rumen ciliates would not be affected by the addition of pot scrubbers.

6.2 Materials and methods

Experimental design

As described in Chapter 3.

Rumen sampling, extraction and quantification of DNA

Rumen sampling and extraction and quantification of DNA were performed according to

the methods described in Chapter 5, with the exception that only strained crude rumen

samples were analysed in this experiment.


87

Denaturing gradient gel electrophoresis (DGGE)

The extracted DNA was used in PCR targeting the 18S rRNA gene of the rumen ciliates

using specific primers designed for DGGE (P-SSU-316f and P-SSU-539r-GC; (Sylvester et

al., 2005). Both primers were included in a PCR mixture (50 µL) containing 5 µL dNTP

mixture (200 µM of each dNTP), 5 µL 10X Qiagen PCR buffer (with 15 mM MgCl2), 5 µL

of forward primer (200 nM), 5 µL of reverse primer (200 nM), 2 µL of 25 mM MgCl2, and

0.25 µL of HotStarTaq DNA Polymerase (5U/µL) (Qiagen GmbH, Germany). PCR

amplifications were performed according to the protocol published by Sylvester et al.

(2005) except that a 15 min hot start was used. The PCR product was run on a 1.5%

agarose gel and bands were extracted and quantified following the methods described in

Chapter 5.

DGGE was performed using the Dcode Universal Mutation Detection system (16

cm, BioRad, Hercules, CA) and gel denaturing gradients and concentrations were optimised

for the amplicon. The amplicons were run using an 8% wt/vol gel

(Acrylamide:bisacrylamide 37.5:1(BioRad)) with a denaturing gradient ranging from 28 -

43% of urea and formamide (100% corresponds to 7 M urea and 40% wt/vol formamide)

increasing in the direction of the electrophoresis run. Electrophoresis was performed at a

constant 240 volts for 6 h and a constant temperature of 60 ºC. A final concentration

ranging from 3 – 6 ng amplicon/µL in 15 µL of deionised H2O and 10 µL 2x loading buffer

[0.05% (w/w) bromophenol blue, 0.05% (w/w) xylene cyanol, 73.61% glycerol and 26.29%

Millipore H2O] was loaded on the gel. This means that all amplicons that were within the

range of 3 - 6 ng/µL were loaded undiluted, whereas products with a concentration higher

than 6 ng/µL were diluted to a concentration of 5 ng/µL with Millipore water. Gels were


88

stained for 20 min with SYBR green (Sigma-Aldrich) and DGGE bands were visualised

using Bio-Rad’s gel-doc system and Quantitative one software.

To standardise and reduce between-gel differences, four markers were run on each

gel. Three of the markers were a 50 bp marker (Fermentas Life Sciences, SM0371) and

they were loaded on each side of the gel and in the middle. The same sample was loaded on

the left side of the centre marker on every gel to correct for between-gel variation. Random

bands of interest were excised from the gel for sequencing and the nucleic acids were

extracted from the acrylamide gel using a combination of QIAEX II (step 1–4) and

Qiaquick gel extraction kit (Chapter 4). The amplicons were then re-amplified using the

same primers without the GC-clamp on the forward primer and a reduced annealing

temperature of 45 ºC for only 30 cycles. The re-amplified amplicons were then sequenced

in both directions using the same primers and an ABI Prism 373 automated DNA sequencer

(Applied Biosystems Inc., Foster City, Calif.) using Big Dye terminator. Ciliate sequences

were confirmed by using the Basic Local Alignment Search Tool (BLAST) (Altschul et al.,

1997) in GenBank. Bands in similar positions on different gels were exercised to verify the

assumption that these bands were the same. The banding patterns of the DGGE gels were

analysed using the GelCompar II software (Applied Maths, Inc.Texas, U.S.A).

Real-time PCR

Samples were analysed to determine the numbers of rumen ciliates in the extracted DNA

using real-time PCR. Real-time PCR amplification was carried out with the Bio-Rad

Icycler using the same DGGE primers, but without the 40 bp GC-clamp on the reverse

primer (P-SSU-539r). Reactions were done in a 25 µL volume containing the following


89

reagents: 12.5 µL SYBR green mix (QuantiTect™ SYBR® Green PCR, Qiagen), 9.5 µL

sterile Millipore H2O, 1.0 µL forward primer (10 µM concentration), 1.0 µL reverse primer

(10 µM concentration) and 1.0 µL template DNA (10-200 ng).

Real-time PCR amplification parameters were similar to the conditions used by

Sylvester et al. (2004), with the exception that a 15 min hot start was used. A final melting

curve analysis was carried out by continuously monitoring fluorescence between 60 °C and

95 °C with 0.5 °C increments every 10 s. Threshold cycles were calculated automatically

by the BioRad Icycler software (version 3.5). PCR efficiencies were calculated according to

the methods described in Chapter 4. External standards were prepared for quantification. A

series of ciliate-enriched rumen samples, enumerated by microscopy with a final density

between 7.7 x 103 and 1.2 x 106 cells/mL were prepared. The DNA from these samples was

then extracted as outlined above.

Statistical analysis and diversity index

Real-time PCR data were reassembled in a Euclidean distance matrix before using

ANOSIM in the multi-variable statistical packed PRIMER v6 (Clarke and Gorley, 2006).

The DGGE banding patterns were also analysed using the PRIMER v6 statistical package

using the Bray-Curtis similarity matrix. Non-metric multidimensional scaling (MDS

ordination) was used to investigate the effect of grain and pot scrubbers on DGGE banding

patterns (Clarke, 1993; Clarke and Warwick, 2001).

The Shannon index (Shannon and Weaver, 1949) of general diversity was

calculated using DGGE banding patterns and significant differences in diversity between

treatments were found by using a Student’s t-test.


90

Nucleotide sequence accession number

The sequences of the DGGE bands reported in this Chapter have been deposited in the

GenBank database under accession numbers EF595961 to EF595971.

6.3 Results

Effect of treatments

There was no effect of step-wise addition of grain or the presence of pot scrubbers on the

number of rumen ciliates between or within diet phases (Figure 6.1). In contrast, the

addition of grain had a profound effect on DGGE banding patterns, whereas pot scrubbers

had a minor effect. The DGGE banding patterns within diet phases were different for low

and high grain-fed sheep with and without pot scrubbers compared to banding patterns

observed in oaten-chaff-fed sheep (P < 0.01).

0

1

2

3

4

5




Log

(10)

of c

iliat

es p

erng

DN

A e

xtra

cted

OOPGGP

Figure 6.1: Logarithm-transformed (base 10) values of mean number of ciliates found in each diet phase with SEM. O: oaten-chaff diet; OP: oaten-chaff + pot scrubber; G: grain diet; GP: grain diet + pot scrubber.


91

In contrast, pot scrubbers only affected DGGE banding patterns in diet phase 2 where the

sheep with pot scrubbers that were fed oaten-chaff had significantly different DGGE

banding patterns compared to the sheep without pot scrubbers (P < 0.05).

DGGE banding patterns observed over time between diet phases were found to

change for all sheep groups between diet phase 1 and 3 (min P < 0.05) (Figure 6.2). For

grain-fed sheep with and without pot scrubbers, changes were also seen in DGGE banding

patterns between diet phase 1 and 2 (P < 0.01), and for grain-fed sheep with pot scrubbers

differences were also observed between diet phase 2 and 3 (P < 0.05).

Diet phase 1 Diet phase 3 . M 1 5 6 8 10 12 M 1 5 6 8 10 12 M

Figure 6.2: Compiled DGGE picture showing the banding patterns of rumen ciliates from sheep in the grain and pot scrubber treatment group, and the change in DGGE banding patterns from an oaten-chaff diet (diet phase 1) to a high grain diet (diet phase 3). M: marker. The numbers correspond to the sheep in the treatment group. Gel lanes are not normalised according to marker.


92

The variation found in DGGE banding patterns between diet phases for grain-fed

sheep were also reflected in the Shannon indexes. The Shannon index for the grain-fed

groups was lower (P < 0.05) in diet phase 1 (when they were fed oaten chaff) compared to

diet phase 2 (when they were fed the low grain diet), but higher than in diet phase 3 (high

grain diet) (P < 0.05) (Table 6.1).

Table 6.1: Mean number of the Shannon index calculated for each diet phase ± SD. Groups with different superscript letters between diet phases were found to be significant different (P < 0.05). O: oaten-chaff diet; OP: oaten-chaff + pot scrubber; G: grain diet; GP: grain diet + pot scrubber.

Diet phase 1 Diet phase 2 Diet phase 3

O 0.41 ±0.014 0.44 ±0.016 0.40 ±0.015

OP 0.42 ±0.013 0.41 ±0.014 0.38 ±0.013

G 0.42a ±0.010 0.45b ±0.010 0.32b ±0.013

GP 0.41a ±0.014 0.46b ±0.015 0.33b ±0.015

The multivariate statistical analysis of the DGGE banding patterns, which is

illustrated by the MDS plot below (Figure 6.3), demonstrates that the changes in DGGE

banding patterns observed between diet phase 1 and 3 for all sheep groups are different for

the two diets. All the observations in diet phase 1 (when all groups were fed the oaten-chaff

diet) are closely grouped together, which is expected when all animals are on the same diet.

When the treatments were imposed on the different sheep groups the DGGE banding

patterns of the individual sheep, and thereby their position on the MDS plot, change. This is

best illustrated by the two grain fed sheep groups (G and GP), which are all, except one

observation, on the right-hand side of the inserted line in Figure 6.3. The line is simply

added to illustrate the observation. Apart from the major changes caused by the high grain


93

diet, some minor changes can be observed for the sheep fed oaten-chaff (O and OP) in diet

phase 3, but they are still grouped in close proximity to each other.

Figure 6.3: Non-metric multi-dimensional scaling (MDS) plot showing the changes in DGGE banding patterns for diet phase 1 and diet phase 3 to illustrate the change occurring for all sheep (one dot is one sheep). Numbers correspond to diet phase and O: oaten-chaff diet; OP: oaten-chaff + pot scrubber; G: grain diet; GP: grain diet + pot scrubber (e.g. G3 represents sheep in the grain-fed group in diet phase 3).

Identification of DGGE bands

The DGGE bands that were sequenced matched the sequences for rumen ciliates closely

and included all the ciliates identified by Sylvester et al. (2005), with the addition of five

other species including: Cycloposthium sp., Ophryoscolex purkynjei, Diplodinium

dentatum, Dasytricha ruminantium and Isotricha intestinalis (Table 6.2). Some of the

bands were only present in low and high grain diets and not in all treatment groups (Table

6.2). A band found to represent Dasytricha ruminantium was present in all sheep groups in

diet phases 2 and 3, except the high grain-fed group without pot scrubbers (Table 6.2).

Similarly, the band representing Isotricha intestinalis was found to be absent or below

detectable levels in high grain-fed sheep with pot scrubbers (Table 6.2).


94

Table 6.2: Accession numbers of the bands that were extracted from the DGGE gels and sequenced, length of the sequenced fragment, in which treatment groups they were found and in total how many times they were observed, and the species corresponding to the sequences in GenBank that have the highest identity to those bands.

Band # (Genbank

accession #) Nearest species

%

identity

Fragment

length (bp)

Observed in

Treatment* Total

1 (EF595961) Epidinium caudatum 98 153 All 53

2 (EF595962) Cycloposthium sp. 98 194 O2; G2+3; GP3 6

3, 9, 11 (EF595963) Ophryoscolex purkynjei 97 153 O2+3; OP2+3;

G2; GP2+3 48

Epidinium caudatum 97 153 O2+3; OP2+3; G2; GP2+3 48

4, 13 (EF595964) Ophryoscolex purkynjei 99 152 All 70

Eudiplodinium maggii 99 152 All 70

Diplodinium dentatum 99 152 All 70

Epidinium caudatum 99 152 All 70

5, 16 (EF595965) Entodinium caudatum 98 195 All 46

6, 8, 10 (EF595966) Diplodinium dentatum 99 194 O2+3; OP2+3;

G2; GP2+3 49

7, 12 (EF595967) Dasytricha ruminantium 99 193 O2+3; OP2+3;

G2; GP2+3 48

14 (EF595968) Polyplastron multivesiculatum 100 194 All 33

15 (EF595969) Polyplastron multivesiculatum 99 193 O2+3; OP3;

G2+3; GP3 17

17 (EF595970) Isotricha intestinalis 99 195 O2+3; OP2+3; G2+3;GP2 21

18 (EF595971) Isotricha intestinalis 99 195 O3; OP3; G2+3; GP2 6

- Numbers correspond to diet phase and O: oaten-chaff diet; G: grain diet; OP: oaten-chaff + pot scrubber; GP: grain diet + pot scrubber (e.g. G2 was sheep on grain diet in diet phase 2).

- Band 17 and 18 are different as sequence differences were found at different positions in the sequence.

6.4 Discussion

It was hypothesised that increasing the grain content in the diet would increase the

abundance and reduce the diversity of rumen ciliates, while the abundance and diversity of

the rumen ciliates would not be affected by the addition of pot scrubbers. The expectation

that the diversity of the rumen ciliates would be reduced by feeding grain was supported


95

when sheep were fed high levels of grain with or without pot scrubbers. However, the

expectations that the abundance of ciliates would increase as the proportion of grain in the

diet increased and that pot scrubbers would not affect ciliate diversity were not supported.

The diversity of rumen ciliates was anticipated to decrease with increasing grain in

the diet, with Entodinium spp. being favoured because of increased substrate availability

(Williams and Coleman, 1992). The expected reduction in diversity of rumen ciliates with

increased grain in the diet was observed for sheep given high grain with and without pot

scrubbers, as indicated by the reduction in the Shannon index between diet phases 1 to 3

(Table 6.1). This reduction in the Shannon index in diet phase 3 followed a significant

increase in diversity of rumen ciliates for both low grain-fed groups of sheep in diet phase 2

when compared to diet phase 1 (when they had been fed oaten chaff), which suggests there

is an interaction between the level of grain and diversity of ciliates. It is possible that the

substrate availability and diversity in the animals on a low grain – high oaten-chaff mix

would be greater and promote a more complex microbial diversity. In contrast, the

diversity of substrates on the high grain diet (much lower proportion of oaten-chaff) would

be less and dominated by the starch in the grain. This appears to favour or provide a

competitive advantage to some ciliates, which reduces their diversity. The results from the

multivariate statistical test indicated that DGGE banding patterns changed in all of the

sheep groups between diet phase 1 and 3, but the DGGE banding patterns from sheep fed

oaten-chaff were different to the changes observed in both the grain-fed groups; clearly

demonstrated on the MDS plot (Figure 6.3). The MDS plot demonstrates that the DGGE

patterns change in all groups, but in different ways, and this is supported by the

multivariate statistical test of the DGGE patterns for the individual groups in diet phase 3


96

(min. P < 0.05).

There are other reasons why the diversity of rumen ciliates may have decreased

when high levels of grain were fed. Hristov et al. (2001) observed, when changing the diet

from a medium to a high grain diet, that several species of rumen ciliates were no longer

present in the rumen. In the present study I made a similar finding. The Shannon indices

were reduced and the number of DGGE bands observed for the individual treatment groups

were nearly reduced by 20%, when a high grain diet was fed.

The reason for the change in DGGE banding patterns in sheep with pot scrubbers

fed oaten-chaff in diet phase 2 is unclear. There was a 15% decrease in the number of bands

between diet phases 1 and 2, but half of that reduction was recovered in diet phase 3. The

most obvious interpretation of this result is that the pot scrubbers, even though they did not

modify retention rates in the rumen, were stimulating conditions in the rumen that required,

or resulted in, modifications to their rumen ciliate population, as the DGGE banding

patterns were found to be different between diet phase 1 and 3. It is possible that some of

the oaten-chaff interacted with the pot scrubbers, for example became lodged in the pot

scrubber, which created micro-environments that influenced ciliate diversity.

High grain diets are often accompanied with an increase in the number of rumen

ciliates, due to increased substrate availability (Mackie et al., 1978; Franzolin and Dehority,

1996). In the present study no increase in the number of rumen ciliates, measured using

real-time PCR, was observed when a high grain diet was offered (Figure 6.1). The reason

could be that the 70% grain was not high enough to support a significant increase in

abundance. Franzolin and Dehority (1996) found that at least 75% concentrate was needed

to significantly increase rumen ciliate counts.


97

The DNA sequence analysis of DGGE bands (Table 6.2) confirmed that bands in

similar positions on different gels were identical, which provided justification for the

validity of making comparisons between gels. On a high grain diet the three most

significant changes in the species composition identified by DGGE band sequences were

the presence or absence of Dasytricha ruminantium, Isotricha intestinalis, and

Cycloposthium sp. Dasytricha ruminantium and Isotricha intestinalis were not detected in

sheep fed high grain diets with or without pot scrubbers, respectively. Whether these

species were eliminated from the rumen or were at numbers below the detection level of the

DGGE system (~102) is unknown. The observation of a DGGE band 98% similar to a

Cycloposthium sp. (Table 6.2) on a high grain diet was unexpected. This ciliate has not

been found in the rumen previously, but they have been identified in digesta samples from

horses (Ito et al., 2002). This is the first report of a Cycloposthium-like sp. inhabiting the

rumen.

The DGGE primers and conditions used in this study to detect rumen ciliates was

adapted from Sylvester et al. (2005), who found DGGE bands with sequences similar to

Entodinium spp., Eudiplodium spp., Epidinium spp, and Polyplastron spp. in their study.

The identification of additional DGGE bands in the this study compared to the results of

Sylvester et al. (2005) is probably due to different running conditions, as Sylvester et al.

(2005) ran their gels for 18 h at 80 volts, and I used 6 h at 240 volts to give the same

number of volts x hours. Altering the voltage and the hours has resulted in detection of

additional bands in other studies (Wu et al., 1998). However, it is more likely that the

differences were because of the difference in diet between the two experiments and because

Sylvester et al. (2005) used rumen fluid from cattle, whereas I used rumen fluid from


98

sheep.

In conclusion, the major finding in this Chapter was that grain changed the diversity

of the ruminal ciliates without affecting their abundance. In addition, by identifying the

main DGGE bands by DNA sequencing it was possible to show that Dasytricha

ruminantium and Isotricha intestinalis disappeared in high grain-fed sheep with and

without pot scrubbers, respectively, and that a Cycloposthium-like species inhabited the

rumen. Furthermore, changes in diversity of rumen ciliates due to step-wise increases in

grain content could be detected by DGGE. The diet-dependent changes in diversity and the

maintenance of the abundance of rumen ciliates have implications for the interpretation of

the results in Chapters 4 and 5. Krumholz et al. (1983) and Itabashi et al. (1994), found the

abundance and diversity of rumen ciliates influences the level of methane production. The

full implications of the findings in this chapter in relation to the findings in Chapters 4 and

5 will be discussed in the general discussion (Chapter 7).

Chapter 7: General discussion

99

CHAPTER 7

General discussion


100

The general hypothesis tested in this thesis was that reducing the retention time and

increasing the amount of grain in the diet would affect the abundance and diversity of

methanogens in their different niches, as well as their association with ruminal ciliates.

More specific hypotheses were tested in each chapter. In summary, it was hypothesised that

the combined effect of increasing the grain content in the diet and reducing retention time

in the rumen would decrease methane production and acetate/propionate ratios. In Chapter

5 I hypothesised that the diversity of methanogens would decrease and the abundance of

methanogens associated with ciliates would increase in the rumen of sheep consuming

grain, and that these effects would be more pronounced in sheep with pot scrubbers. In the

final experimental chapter I hypothesised that increasing the amount of grain in the diet

would increase the abundance and reduce the diversity of rumen ciliates, but that the

abundance and diversity of the rumen ciliates would not be affected by the addition of pot

scrubbers. Some of these expectations were supported even though pot scrubbers did not

reduce ruminal retention time significantly, but the artificial stimulation of the rumen wall

by the pot scrubbers still had an effect on rumen parameters (Chapter 4).

There are three key results in this thesis that extend our understanding about how

the diversity and abundance of methanogens in different niches and rumen ciliates respond

to changes in diet in particular, but also to artificial stimulation of the rumen wall: first, the

increase in abundance of endosymbiotic methanogens observed on a high grain diet

regardless of whether the sheep had pot scrubbers or not in their rumen (Chapter 5); second,

the effect of pot scrubbers despite the fact that they did not change retention rate in the

rumen, which appears to be more important in sheep fed an oaten-chaff or low grain diet

(Chapter 4); and finally, the lower acetate/propionate ratios and methane production in vitro


101

in sheep with pot scrubbers fed the high level of grain and the changes observed in the

abundance and diversity of rumen ciliates (Chapters 4, 5 and 6). These three key results

form the basis of this general discussion. In addition, I will comment on the steps that were

taken to minimise the impact of some of the limitations associated with using DGGE and

quantitative real-time PCR techniques to study microbial ecology, and conclude with a

comment on some future work that I think would extend the findings in this thesis.

The most significant result in this thesis was the increase in abundance of

endosymbiotic methanogens in both groups of high grain-fed sheep (i.e. with and without

pot scrubbers) (Chapter 5). The main evidence for this increase comes directly from the

real-time PCR results, but it is also supported indirectly by the results from the DGGE

banding patterns and Shannon Diversity Index, because the majority of these changes were

observed in methanogens associated with rumen ciliates (Table 5.1).

The increase in abundance of endosymbionts supports the hypothesis that when

hydrogen is limited in the rumen then methanogens try to associate with ciliates either ecto-

or endosymbiotically to get better access to hydrogen (Czerkawski et al., 1972; Stumm et

al., 1982; Stumm and Zwart, 1986; Ushida et al., 1997). In my experiment, the increase in

endosymbionts occurred without changes occurring in the abundance of methanogens in

sample fractions containing “all methanogens” or “ecto-and endosymbiotic methanogens”.

This suggests that the increase in endosymbiotic methanogens came from previously free-

living or ectosymbiotic methanogens, which may have been because of a reduction in

hydrogen availability in the rumen fluid. It is possible that the hydrogen availability was

limited in sheep fed the high grain diets because more of the hydrogen was used for

propionate production. This is supported but the results in Chapter 4, where I observed


102

lower in vitro methane production and acetate/propionate ratios when the grain content in

the diet increased from 35% to 70%. The lower methane production from ruminants on a

high grain diet is consistent with previous findings (Moss et al., 1995; Baker, 1997). These

results are also consistent with the general observation that propionate levels increase when

ruminants are fed high grain diets (Hodgson and Thomas, 1975; Baker, 1997; Russell,

1998; Hristov et al., 2001). In addition, Tajima et al. (2001a), found more than a two-fold

increase of Selenomonas ruminantium, a propionate producer, after an extended period of

feeding a high grain diet. The higher amount of propionate produced in the present

experiment would have reduced the amount of hydrogen available in the rumen fluid and

may have increased the need for some methanogens to establish themselves in close

association with rumen ciliates to access hydrogen.

Zinder (1993) demonstrated that different methanogens have different affinities for

hydrogen and my results would indicate that this becomes important on a high grain diet.

Although speculative, the results from Chapters 4 and 5 suggest that it is mainly

methanogens with a low affinity for hydrogen that associate with rumen ciliates. The total

abundance of methanogens did not change, based on the real-time PCR, yet methane

production in vitro decreased and the number of endosymbionts increased significantly. It

would appear from this that endosymbiotic methanogens produce less methane.

Newbold et al. (1995) and Baker (1997) have also reported that the numbers of

methanogens do not always correlate well with the level of methane produced. They

suggest that the availability of hydrogen in the rumen, the type of methanogens, and/or their

association with rumen ciliates appear to be the determinants of the amount of methane

produced. My results suggest that the type of methanogens present may not be that


103

important, because the DGGE bands that were extracted and identified by sequencing in

Chapter 5 were represented in almost all the sample fractions. These results are similar to

those of Sharp et al. (1998), who found that the dominant methanogens in the crude rumen

fluid were also the dominant methanogens associating with the rumen ciliates.

An alternative and much simpler explanation for the increase in endosymbiotic

methanogens would be that the total number of rumen ciliates available for the

methanogens to associate with endosymbiotically increased. However, I used real-time

PCR to quantify the number of rumen ciliates in Chapter 6 and their numbers did not

increase in animals fed high levels of grain. The diversity of ciliates in the rumen is likely

not to influence the association between methanogens rumen ciliates, since methanogens

can associate with ciliates from both orders of rumen ciliates, the Entodiniomorphida and

the Vestibuliferida. It is more likely that the rumen conditions (e.g. hydrogen availability)

during high grain diets dictate their association.

The second key result was the effect of pot scrubbers in animals fed oaten-chaff and

low grain diets. Pot scrubbers decreased methane production in vitro when combined with

low grain (Chapter 4), and reduced total numbers of methanogens between diet phase 1 and

2 in the sheep that were given an oaten-chaff diet (Chapter 5). Pot scrubbers also affected

diversity by changing the DGGE banding patterns and the Shannon index observed for

sheep with pot scrubbers fed oaten-chaff or low levels of grain (Chapter 5). The pot

scrubbers were expected to reduce the ruminal retention time and target mainly the free-

living methanogens by “washing them out” of the rumen. What was most interesting was

that the effect of pot scrubbers could not be detected as a change in retention time of liquid

or particulate matter (Chapter 4), but their physical presence was enough to have an effect.


104

The reason why pot scrubber has these effects is difficult to explain but it is possible that

there is an interaction between the oaten-chaff and the pot scrubbers that creates localised

changes in the ruminal environment. As the amount of oaten-chaff in the diet is reduced,

the effect of pot scrubbers also becomes less obvious to the point where there is virtually no

effect of pot scrubbers in the animals fed a high grain diet. This may also be because pot

scrubbers and grain target methanogens in the same niche, i.e. the free-living methanogens,

and the grain was expected to reduce free-living methanogens by increasing the

competition for hydrogen. However, pot scrubbers influence methanogen diversity on low

grain or oaten-chaff diets, through some other mechanism than their effects on ruminal

retention time and it would be interesting to investigate how inert objects could be used to

influence methane production on different diets.

The third key result was the lower acetate/propionate ratio observed in sheep with

pot scrubbers fed the high grain diet, which was a result of higher propionate production

(diet phase 3; Chapter 4). The lower in vitro methane production in sheep with pot

scrubbers fed the low grain diet in diet phase 2 can help to explain why more propionate

was produced in diet phase 3, when the level of grain was increased. In diet phase 2 when

only 35% grain was fed, the pot scrubbers in (e.g. propionate producers) in the rumen of

grain-fed sheep seemed to give the hydrogen utilisers a competitive advantage for hydrogen

over the methanogens, based on the lower methane production in vitro. There was also a

trend towards higher propionate production in the rumen of these sheep. Therefore, the

population of hydrogen utilisers in the rumen of sheep with pot scrubbers fed grain would

have been in a stronger position to take advantage of any increase in the amount of


105

hydrogen available associated with moving onto a high grain diet (diet phase 3), compared

to the sheep without pot scrubbers.

The changes in methanogen diversity observed in Chapter 5 also help explain the

lower acetate/propionate ratio for sheep with pot scrubbers fed the high grain diet, based on

the assumption that grain and pot scrubbers target the free-living methanogens. If this

assumption is accepted then the changes in methanogen diversity for sheep with pot

scrubbers fed grain should appear in different sample fractions compared to the sheep

without pot scrubbers, because the two groups would be at different stages in adapting to

the changes occurring in the rumen. The results in Chapter 5 indicate that this is indeed the

case; the diversity of methanogens found in animals with pot scrubbers were observed in

different sample fractions than the changes for sheep on a grain diet without pot scrubbers.

The changes in DGGE banding patterns for the combined treatment were observed in the

strained crude rumen fluid fraction, whereas the corresponding changes for the sheep

without pot scrubbers and fed grain were observed in the ecto- and endosymbiotic, and the

endosymbiotic fraction. Similar differences were observed when the Shannon index was

used; differences for the combined effect of diet and pot scrubbers were found in the ecto-

and endosymbiotic fraction, whereas they were only observed in the endosymbiotic fraction

for the grain-fed sheep without pot scrubbers. It is likely that different species are changing

in the different niches and, since different methanogens can have different affinities for

hydrogen (Zinder, 1993), these changes could affect the amount of hydrogen available in

the rumen for other hydrogen utilisers.

The molecular techniques used to study microbial ecology and diversity have some

limitations. In particular, the influence of differences in PCR efficiency is often overlooked,


106

ignored or assumed to be similar for all samples analysed. However, I used a number of

strategies to minimise the limitations of the DGGE and the real-time PCR techniques and

am confident in the conclusions that can be made from the results obtained in this study.

For example, the PCR efficiency was calculated for individual samples and used in the final

calculations of abundance for each sample as opposed to using the same PCR efficiency for

all samples. The advantage of using individual PCR efficiencies is to eliminate or reduce

the effect of PCR inhibitors on the final calculations of abundance. The main strategies that

were used to enhance the reliability and reproducibility of DGGE were to estimate the

amount of PCR product loaded onto the DGGE gel for each sample and the use of a proof-

reading polymerase. The proof-reading polymerase was used to reduce PCR-induced point

mutations during amplification. The amount of PCR product was estimated, using SYBR-

green, to get an even brightness of bands on the DGGE gel to reduce the risk of detecting

changes in populations because of differences in PCR efficiency, or the amount of PCR

inhibitors co-extracted with the DNA.

I also used degenerate primers in the DGGE analyses in this study. The use of

degenerate primers has lead to an overestimation of the number of bacteria within bacterial

communities (Muyzer and Smalla, 1998). However, I am confident this did not occur in

this study because a total of 29 DGGE bands were sequenced (Chapter 5) and none of the

bands, in different positions on the gel, was found to be identical. For the ciliate DGGE

gels, 18 bands were sequenced and none of them, in different positions on the gel, was

found to be identical. An obvious limitation of both the ciliate and methanogen DGGE

primers used in this study was the length of the fragment analysed. Both fragments were <

200 bp and the limited phylogenetic information within the fragments may have prevented


107

an accurate identification of the species linked to the fragment. Examples of this can be

seen in both Chapters 5 and 6, where one DGGE band has been found to have equally high

similarity to two or more sequences in the ribosomal database. Furthermore, in Chapter 6

one DGGE band was identified as a Cycloposthium sp., which is not a normal rumen

inhabitant and mostly found in horses and kangaroos. Although it is possible that this is the

first report of Cycloposthium sp. being identified in the rumen, this needs to be confirmed

by more extensive sequence analysis. Otherwise, none of the other limitations discussed in

the literature review (Chapter 2) seemed to influence this study.

Future studies

From the results of my study I can only speculate about the role hydrogen utilising bacteria

play in the reduction of methanogenesis in the rumen when a high grain diet is fed. The

ruminal conditions may favour hydrogen utilising bacteria and they may compete with the

methanogens. There is a need for studies where the key hydrogen utilisers are examined at

the same time the methanogen and ciliate populations are monitored to get a clearer picture

of the interactions between these groups of organisms. There is also a need to investigate

how the hydrogen-utilising bacteria can be enhanced. For example, they could be

introduced or their numbers increased by enriching the diet with probiotics (e.g. freeze

dried hydrogen utilising bacteria, propionate producers, and acetogens). This may reduce

the amount of hydrogen available for methane production. The perseverance of the

introduced species in this way would need to be monitored using molecular techniques (e.g.

quantitative real-time PCR) to establish whether a supplement would have to be supplied

on a regular basis. A micro-array approach would make it possible to monitor diversity


108

changes to thousands of different bacteria simultaneously, which could be made more

powerful if the this approach was coupled with the development of quantitative real-time

PCR assays mainly of hydrogen utilising bacteria for more accurate quantification. Most

recently, there is evidence that using RNA in DGGE analyses may give a better

understanding of the more active groups of bacteria as oppose to DNA. For example, if

DGGE gels using DNA samples give a picture of the most abundant bacteria, then DGGE

gels using RNA should indicate the most active bacteria. These differences have been

found with DGGE (Licht et al., 2006) and should therefore be considered for future

experiments.

Conclusion

It appears that in sheep fed a high grain diet, methanogens associate endosymbiotically with

rumen ciliates to gain better access to hydrogen. The number and diversity of ciliates did

not change in these sheep, which suggests that the association between methanogens and

rumen ciliates is dictated by the availability of hydrogen in the rumen and not the generic

composition of the ciliate population. The methanogens that do associate endosymbiotically

with the ciliates appear to produce less methane. Further, it would be interesting to examine

the hydrogen affinity and efficiency of methane production of these endosymbiotic

methanogen species in more detail. It remains unclear how changes in retention time

influences the ecology of methanogens and ruminal ciliates because the pot scrubbers did

not change retention time. However, their physical presence was enough to lower

acetate/propionate measurements observed in sheep on the high grain diet. This result was

unexpected and opens up many questions about why an indigestible, ball of plastic mesh


109

affects the end products of fermentation, and why it would be diet specific. The findings in

this thesis are novel and have contributed to our understanding of how methanogens and

ciliates, and their interaction, respond to dietary changes and influence methanogenesis

from the rumen. Understanding these changes at a functional level is critical to controlling

enteric greenhouse gas emissions.

References

110

References

Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J.

Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein

database search programs. Nucl. Acids Res. 25: 3389–3402.

Australian Greenhouse Office. 2007. National greenhouse gas inventory 2005, Australian

Greenhouse Office, Canberra, Australia.

Baker, S. K. 1997. Gut microbiology and its consequences for the ruminant. Proc. Nutr.

Soc. Aust. 21: 6-13.

Balch, W. E., G. E. Fox, L. J. Magrum, C. R. Woese, and R. S. Wolfe. 1979. Methanogens:

Reevaluation of a unique biological group. Microbiol. Rev. 43: 260-296.

Barns, S. M., C. F. Delwiche, J. D. Palmer, and N. R. Pace. 1996. Perspectives on archaeal

diversity, thermophily and monophyly from environmental rRNA sequences. Proc.

Natl. Acad. Sci. USA 93: 9188-9193.

Barns, S. M., R. E. Fundyga, M. W. Jeffries, and N. R. Pace. 1994. Remarkable archaeal

diversity detected in a Yellostone National Park hot spring environment. Proc. Natl.

Acad. Sci. USA 91: 1609-1613.

Bauchop, T., and R. T. J. Clarke. 1976. Attachment of the ciliate Epidinium Crawley to

plant fragments in the sheep rumen. Appl. Environ. Microbiol. 32: 417-422.

Boone, D. R., and W. B. Whitman. 1988. Proposal of minimal standards for describing new

taxa of methanogenic bacteria. Int. J. Syst. Bacteriol. 38: 212-219.

Bryant, M. P., and D. R. Boone. 1987. Emended description of strain MS (DMS 800), the

type strain of Methanosarcina barkeri. Int. J. Syst. Bacteriol. 37: 169-170.

References

111

Burggraf, S., P. Heyder, and N. Eis. 1997a. A pivotal Archaea group. Nature 385: 780-780.

Burggraf, S., H. Huber, and K. O. Stetter. 1997b. Reclassification of the Crenarchael orders

and families in accordance with 16S rRNA sequence data. Int. J. Syst. Bacteriol. 47:

657-660.

Chagan, I., M. Tokura, J.-P. Jouany, and K. Ushida. 1999. Detection of methanogenic

archaea associated with rumen ciliate protozoa. J. Gen. Appl. Microbiol. 45: 305-

308.

Clarke, K. R. 1993. Non-parametric multivariate analyses of change in community

structure. Aust. J. Ecol. 18: 117-143.

Clarke, K. R., and R. N. Gorley. 2006. PRIMER v6: user manual/tutorial. PRIMER-E Ltd,

Plymouth, UK.

Clarke, K. R., and R. M. Warwick. 2001. Change in marine communities: an approach to

statistical analysis and interpretation. 2nd edition ed. PRIMER-E, Plymouth, UK.

Coleman, G. S. 1992. The rate of uptake and metabolism of starch grains and cellulose

particles by Entodinium species, Eudiplodinium maggii, some other

entodiniomorphid protozoa and natural protozoal populations taken from the ovine

rumen. J. Appl. Bacteriol. 73: 507-513.

Coleman, G. S., and F. J. Hall. 1972. Fine structural studies on the digestion of bacterial

species in the rumen ciliate, Entodinium caudatum. Tissue Cell 4: 37-48.

Coleman, G. S., and D. C. Sandford. 1979. The engulfment and digestion of mixed rumen

bacteria and individual bacteria species by single and mixed species of rumen ciliate

protozoa grown in vivo. J. Agric. Sci. 92: 729-742.

References

112

Counotte, G. H., A. Lankhorst, and R. A. Prins. 1983. Role of DL-lactic acid as an

intermediate in the rumen metabolism of dairy cows. J. Anim. Sci. 56: 1222-1235.

Czerkawski, J. W., C. G. Harfoot, and G. Breckenridge. 1972. The relationship between

methane production and concentrations of hydrogen in the aqueous and gaseous

phases during rumen fermentation in vitro. J. Appl. Bacteriol. 35: 537-551.

Dehority, B. A. 2005. Effect of pH on viability of Entodinium caudatum, Entodinium

exiguum, Epidinium caudatum, and Ophryoscolex purkynjei in vitro. J. Eukaryot.

Microbiol. 52: 339-342.

Dionisi, H. M., G. Harms, A. C. Layton, I. R. Gregory, J. Parker, S. A. Hawkins, K. G.

Robinson, and G. S. Sayler. 2003. Power analysis for real-time PCR quantification

of genes in activated sludge and analysis of the variability introduced by DNA

extraction. Appl. Environ. Microbiol. 69: 6597-6604.

Duncan, S. H., P. Louis, and H. J. Flint. 2004. Lactate-utilizing bacteria, isolated from

human feces, that produce butyrate as a major fermentation product. Appl. Environ.

Microbiol. 70: 5810-5817.

Eadie, J. M. 1962. Inter-relationships between certain rumen ciliate protozoa. J. Gen.

Microbiol. 29: 579.

Ellis, J. E., P. S. McIntyre, and M. Saleh. 1991a. Influence of CO2 and low concentrations

of O2 on fermentative metabolism of the rumen entodiniomorphid ciliate

Polyplastron multivesiculatum. Appl. Environ. Microbiol. 57: 1400-1407.

Ellis, J. E., P. S. McIntyre, and M. Saleh. 1991b. The influence of ruminal concentrations

of O2 and CO2 on fermentative metabolism of the rumen entodiniomorphid ciliate

Eudiplodinium maggii. Curr. Microbiol. 23: 245-251.

References

113

Embley, T. M., and B. J. Finlay. 1993. Systematic and morphological diversity of

endosymbiotic methanogens in anaerobic ciliates. Anton. Leeuw. Int. J. G. 64: 261-

271.

Evans, E. 1981a. An evaluation of the relationships between dietary parameters and rumen

liquid turnover rate. Can. J. Anim. Sci. 61: 91-96.

Evans, E. 1981b. An evaluation of the relationships between dietary parameters and rumen

solid turnover rate. Can. J. Anim. Sci. 61: 97-103.

Faichney, G. J. 1975. The use of markers to partition digestion within the gastro-intestinal

tract of ruminants. p 277-291 in Proc. IV Int. Symp. Rum. Physiol., Sydney,

Australia.

Faichney, G. J. 1992a. Application of the double-marker method for measuring digesta

kinetics to rumen sampling in sheep following a dose of the markers or the end of

their continuous infusion. Aust. J. Agric. Res. 43: 277-284.

Faichney, G. J. 1992b. Effect of non-uniform distribution of particle-associated markers on

the measurement of duodenal digesta flow by the double-marker technique. J.

Agric. Sci. 118: 119-120.

Faichney, G. J., N. M. Graham, and D. M. Walker. 1999. Rumen characteristics, methane

emissions, and digestion in weaned lambs reared in isolation. Aust. J. Agric. Res.

50: 1083-1089.

Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap.

Evolution 39: 783–791.

Felsenstein, J. 2004. PHYLIP - Phylogeny Inference Package (Version 3.62c) No. 5. p 164-

166. Dept. Genetics, University of washington, Seattle, Washington.

References

114

Ferry, J. G., P. H. Smith, and R. S. Wolfe. 1974. Methanospirillum, a new genus of

methanogenic bacteria, and characterization of Methanosririllum hungatei sp. nov.

Int. J. Syst. Bacteriol. 24: 465-469.

Finlay, B. J., G. Esteban, K. J. Clarke, A. G. Williams, T. M. Embley, and R. P. Hirt. 1994.

Some rumen ciliates have endosymbiotic methanogens. FEMS Microbiol. Lett. 117:

157-162.

Finlay, B. J., and T. Fenchel. 1992. Methanogens and other bacteria as symbionts of free-

living anaerobic ciliates. Symbiosis 14: 375 - 390.

Food and Agriculture Organization of the United Nations. 2000. Global impact domain:

methane emissions, Livestock Environment and Development (LEAD) Initiative.

Franzolin, R., and B. A. Dehority. 1996. Effect of prolonged high-concentrate feeding on

ruminal protozoa concentrations. J. Anim. Sci. 74: 2803-2809.

Garcia, J. L., B. K. C. Patel, and O. B. 2000. Taxonomic, phylogenetic and ecological

diversity of methanogenic Archaea. Anaerobe 6: 205-226.

Goad, D. W., C. L. Goad, and T. G. Nagaraja. 1998. Ruminal microbial and fermentative

changes associated with experimentally induced subacute acidosis in steers. J.

Anim. Sci. 76: 234-241.

Grubb, J. A., and B. A. Dehority. 1975. Effects of an abrupt change in ration from all

roughage to high concentrate upon rumen microbial numbers in sheep. Appl.

Microbiol. 30: 404-412.

Hayes, V., Y. Wu, J. Osinga, I. Mulder, P. van der Vlies, P. Elfferich, C. Buys, and R.

Hofstra. 1999. Improvements in gel composition and electrophoretic conditions for

References

115

broad-range mutation analysis by denaturing gradient gel electrophoresis. Nucl.

Acids Res. 27: e29 i-e29v.

Hegarty, R. S. 1999. Reducing rumen methane emissions through elimination of rumen

protozoa. Aus. J. Agric. Res. 50: 1321-1327.

Hillman, K., D. Lloyd, and A. G. Williams. 1988. Interactions between the methanogen

Methanosarcina barkeri and the rumen holotrich ciliate protozoa. Lett. Appl.

Microbiol. 7: 49-53.

Hino, T. 1982. Influence of hydrogen on the fermentation in rumen protozoa, Entodinium

species. Jap. J. Zootech. Sci. 54: 320-328.

Hodgson, J. C., and P. C. Thomas. 1975. A relationship between the molar proportion of

propionic acid and the clearance rate of the liquid phase in the rumen of the sheep.

Brit. J. Nutr. 33: 447-456.

Hristov, A. N., M. Ivan, L. M. Rode, and T. A. McAllister. 2001. Fermentation

characteristics and ruminal ciliate protozoal populations in cattle fed medium- or

high-concentrate barley-based diets. J. Anim. Sci. 79: 515-524.

Huber, H., M. J. Hohn, R. Rachel, T. Fuchs, V. C. Wimmer, and K. O. Stetter. 2002. A new

phylum of Archaea represented by a nanosized hyperthermophilic symbiont. Nature

417: 63-67.

Hungate, R. 1966. The rumen and its microbes. Academic Press, New York, USA.

Hutten, T. J., H. C. Bongaerts, C. van der Drift, and G. D. Vogels. 1980. Acetate, methanol

and carbon dioxide as substrates for growth of Methanosarcina barkeri. Anton.

Leeuw. Int. J. G. 46: 601-610.

References

116

Irbis, C., and K. Ushida. 2004. Detection of methanogens and proteobacteria from single

cell of rumen ciliate protozoa. J. Gen. Appl. Microbiol. 50: 203-212.

Itabashi, H., Y. Wahio, A. Takenaka, O. S., and T. Ishibashi. 1994. Effects of a controlled

fauna on methanogenesis, fibre digestion and rate of growth in calves. Proc. Soc.

Nutr. Physiol. 3: 179.

Ito, A., Y. Miyazaki, and S. Imai. 2002. Descriptions of new Parentodinium ciliates in the

family Parentodiniidae n. fam. from Hippopotamus amphibius in comparison with

some entodiniomorphs from horses and cattle. Eur. J. Protistol. 37: 405-426.

Joblin, K. N. 1999. Ruminal acetogens and their potential to lower ruminant methane

emissions. Aust. J. Agric. Res. 50: 1307-1313.

Johnson, K. A., and D. E. Johnson. 1995. Methane emissions from cattle. J. Anim. Sci. 73:

2483-2492.

Karnati, S. K. R., J. T. Sylvester, S. M. Noftsger, Z. Yu, N. R. St-Pierre, and J. L. Firkins.

2007. Assessment of ruminal bacterial populations and protozoal generation time in

cows fed different methionine sources. J. Dairy Sci. 90: 798-809.

Kimura, M. 1980. A simple method of estimating evolutionary rates of base substitutions

through comparative studies of nucleotide sequences. J. Mol. Evol. 16: 111–120.

Klein, L., and A.-D. G. Wright. 2006. Construction and operation of open-circuit methane

chambers for small ruminants. Aust. J. Exp. Agric. 46: 1257-1262.

Kocherginskaya, S. A., R. I. Aminov, and B. A. White. 2001. Analysis of the rumen

bacterial diversity under two different diet conditions using denaturing gradient gel

electrophoresis, random sequencing, and statistical ecology approaches. Anaerobe

7: 119-134.

References

117

Kowalchuk, G. A., J. R. Stephen, W. DeBoer, J. I. Prosser, T. M. Embley, and J. W.

Woldendorp. 1997. Analysis of ammonia-oxidizing bacteria of the beta subdivision

of the class Proteobacteria in coastal sand dunes by denaturing gradient gel

electrophoresis and sequencing of PCR-amplified 16S ribosomal DNA fragments.

Appl. Environ. Microbiol. 63: 1489-1497.

Krause, D. O., W. J. Smith, and C. S. McSweeney. 2001. Extraction of microbial DNA

from rumen contents containing plant tannins. Biotech. 31: 294-298.

Krause, D. O., W. J. M. Smith, F. M. E. Ryan, R. I. Mackie, and C. S. McSweeney. 1999.

Use of 16S-rRNA based techniques to investigate the ecological succession of

microbial populations in the immature lamb rumen: tracking of a specific strain of

inoculated Ruminococcus and interactions with other microbial populations in vivo.

Microbiol. Ecol. 38: 365-376.

Krumholz, L. R., C. W. Forsberg, and D. M. Veira. 1983. Association of methanogenic

bacteria with rumen protozoa. Can. J. Microbial. 29: 676-680.

Lana, R. P., J. B. Russell, and M. E. van Amburgh. 1998. The role of pH in regulating

ruminal methane and ammonia production. J. Anim. Sci. 76: 2190-2196.

Licht, T., M. Hansen, M. Poulsen, and L. Dragsted. 2006. Dietary carbohydrate source

influences molecular fingerprints of the rat faecal microbiota. BMC Microbiol. 6:

98.

Lin, C., L. Raskin, and D. A. Stahl. 1997. Microbial community structure in gastrointestinal

tracts of domestic animals: comparative analyses using rRNA-targeted

oligonucleotide probes. FEMS Microbiol. Lett. 22: 281-294.

References

118

Liu, W., and D. A. Saint. 2002. Validation of a quantitative method for real time PCR

kinetics. Biochem. Biophys. Res. Comm. 294: 347-353.

Loerch, S. C. 1991. Efficacy of plastic pot scrubbers as a replacement for roughage in high-

concentrate cattle diets. J. Anim. Sci. 69: 2321-2328.

Lovley, D. R., R. C. Greening, and J. G. Ferry. 1984. Rapidly growing rumen

methanogenic organism that synthesizes coenzyme M and has high affinity for

formate. Appl. Environ. Microbiol. 48: 81-87.

Mackie, R. I., F. M. C. Gilchrist, A. M. Robberts, P. E. Hannah, and H. M. Schwartz. 1978.

Microbiological and chemical changes in the rumen during the stepwise adaption of

sheep to high concentrate diets. J. Agric. Sci. 90: 241-254.

Mah, R. A., M. R. Smith, and L. Baresi. 1978. Studies on an acetate-fermenting strain of

Methanosarcina. Appl. Environ. Microbiol. 35: 1174 - 1184.

Matsuyama, H., K. Horiguchi, T. Takahashi, M. Ishida, S. Ando, and T. Nishida. 2000.

Control of methane production from expiratory gas by ruminal dosing with

mechanical stimulating goods in holstein steer. p 215 in Proc. 9th Congr. Asian-

Aus. Assoc. Anim. Prod. Soc. and 23rd Biennial Conf. Aus. Soc. Anim. Prod.,

Sydney, Australia.

McAllister, T. A., E. K. Okine, G. W. Mathison, and K. J. Cheng. 1996. Dietary,

environmental, and microbiological aspects of methane production in ruminants.

Can. J. Anim. Sci. 76: 231-243.

Miller, T. L., and M. J. Wolin. 1985. Methanospaera stadtmaniae gen. nov., sp. nov.: a

species that forms methane by reducing methanol with hydrogen. Arch. Microbiol.

141: 116-122.

References

119

Miller, T. L., and M. J. Wolin. 1986. Methanogens in human and animal intestinal tracts.

Syst. Appl. Microbiol. 7: 223-229.

Morgan, R. M., T. D. Pihl, J. Nolling, and J. N. Reeve. 1997. Hydrogen regulation of

growth, growth yields, and methane gene transcription in Methanobacterium

thermoautotrophicum ∆H. J. Bact. 179: 889-898.

Moss, A. R., D. I. Givens, and P. C. Garnsworthy. 1995. The effect of supplementing grass

silage with barley on digestibility, in sacco degradability, rumen fermentation and

methane production in sheep at two levels of intake. Anim. Feed Sci. & Tech. 55: 9-

33.

Murray, A. E., J. T. Hollibaugh, and C. Orrego. 1996. Phylogenetic compositions of

bacterioplankton from two California estuaries compared by denaturing gradient gel

electrophoresis of 16S rDNA fragments. Appl. Environ. Microbiol. 62: 2676-2680.

Muyzer, G., E. de Waal, and A. Uitterlinden. 1993. Profiling of complex microbial

populations by denaturing gradient gel electrophoresis analysis of polymerase chain

reaction-amplified genes coding for 16S rRNA. Appl. Environ. Microbiol. 59: 695-

700.

Muyzer, G., and K. Smalla. 1998. Application of denaturing gradient gel electrophoresis

(DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology.

Anton. Leeuw. Int. J. G. 73: 127-141.

Myers, R. M. 1985. Nearly all single base substitutitions in DNA fragments joined to a GC-

clamp can be detected by denaturing gradient gel-electrophoresis. Nucl. Acids Res.

13: 3131-3145.

References

120

Nedergaard, T., P. Guldberg, E. Ralfkier, and J. Zeuthen. 1997. A one-step DGGE scanning

method for detection of mutations in the K-, N-, and H-ras oncogenes: Mutations at

codons 12, 13 and 61 are rare in B-cell non-Hodgkin's lymphoma. Int. J. Cancer 71:

364-369.

Newbold, C. J., and K. Hillman. 1990. The effect of ciliate protozoa on the turnover of

bacterial and fungal protein in the rumen of sheep. Lett. Appl. Microbiol. 11: 100-

102.

Newbold, C. J., B. Lassalas, and J.-P. Jouany. 1995. The importance of methanogens

associated with ciliate protozoa in ruminal methane production in vitro. Lett. Appl.

Microbiol. 21: 230-234.

Nubel, U., B. Engelen, A. Felske, J. Snaidr, A. Wieshuber, R. I. Amann, W. Ludwig, and

H. Backhaus. 1996. Sequence heterogeneities of genes encoding 16S rRNAs in

Paenibacillus polymyxa detected by temperature gradient gel electrophoresis. J.

Bact. 178: 5636-5643.

Orpin, C. G., and A. J. Letcher. 1978. Some factors controlling the attachment of the rumen

holotrich protozoa Isotricha intestinalis and I. prostoma to plant particles in vitro. J.

Gen. Microbiol. 106: 33-40.

Ozutsumi, Y., K. Tajima, A. Takenaka, and H. Itabashi. 2006. Real-time PCR detection of

the effects of protozoa on rumen bacteria in cattle. Curr. Microbiol. 52: 158-162.

Pfaffl, M. W. 2001. A new mathematical model for relative quantification in real-time RT-

PCR. Nucl. Acids Res. 29: e45.

References

121

Rasmussen, R. 2001. Quantification on the Light cycler instrument. In: S. Meuer, C.

Wittwer and K. Nakagawera (eds.) Rapid cycle real-time PCR: Methods and

applications. p 21-34. Springer, Heidelberg.

Regensbogenova, M., N. R. McEwan, P. Javorsky, S. Kisidayova, T. Michalowski, C. J.

Newbold, J. H. P. Hackstein, and P. Pristas. 2004a. A re-appraisal of the diversity of

the methanogens associated with the rumen ciliates. FEMS. Microbiol. Lett. 238:

307-313.

Regensbogenova, M., P. Pristas, P. Javorsky, S. Y. Moon-van der Staay, G. W. van der

Staay, J. H. Hackstein, C. J. Newbold, and N. R. McEwan. 2004b. Assessment of

ciliates in the sheep rumen by DGGE. Lett. Appl. Microbiol. 39: 144-147.

Rengarajan, K., S. M. Cristol, M. Mehta, and J. M. Nickerson. 2002. Quantifying DNA

concentrations using fluorometry: A comparison of fluorophores. Mol. Vis. 8: 416-

421.

Rouviere, P. E., and R. S. Wolfe. 1988. Novel Biochemistry of Methanogenesis. J. Biol.

Chem. 263: 7913-7916.

Russell, J. B. 1998. The importance of pH in the regulation of ruminal acetate to propionate

ratio and methane production in vitro. J. Dairy Sci. 81: 3222-3230.

Saito, N., and M. Nei. 1987. The neighbor-joining method: a new method for constructing

phylogenetic trees. Mol. Biol. Evol. 4: 406-425.

Scheifinger, C. C., and M. J. Wolin. 1973. Propionate formation from cellulose and soluble

sugars by combined cultures of Bacteroides succinogenes and Selenomonas

ruminantium. Appl. Microbiol. 26: 789-795.

References

122

Schonhusen, U., R. Zitnan, S. Kuhla, W. Jentsch, M. Derno, and J. Voigt. 2003. Effects of

protozoa on methane production in rumen and hindgut of calves around time of

weaning. Arch. Tierernahr. 57: 279-295.

Scott, R. I., N. Yarlett, K. Hillman, T. N. Williams, A. G. Williams, and D. Lloyd. 1983.

The presence of oxygen in rumen liquor and its effects on methanogenesis. J. Appl.

Bact. 55: 143-149.

Sekiguchi, H., N. Tomioka, T. Nakahara, and H. Uchiyama. 2001. A single band does not

always represent single bacterial strains in denaturing gradient gel electrophoresis

analysis. Biotech. Lett. 23: 1205-1208.

Shannon, C. E., and W. Weaver. 1949. The mathematical theory of communication.

University of Illinois Press, Urbana, Il.

Sharkey, F. H., I. M. Banat, and R. Marchant. 2004. Detection and quantification of gene

expression in environmental bacteriology. Appl. Environ. Microbiol. 70: 3795-

3806.

Sharp, R., C. J. Ziemer, M. D. Stern, and D. A. Stahl. 1998. Taxon-specific associations

between protozoal and methanogen populations in the rumen and a model rumen

system. FEMS Microbiol. Ecol. 26: 71-78.

Sheffield, V. C., D. R. Cox, L. S. Lerman, and R. M. Myers. 1989. Attachment of a 40-

base-pair G+C-rich sequence (GC-Clamp) to genomic DNA fragments by the

polymerase chain reaction results in improved detection of single-base changes.

Proc. Natl. Acad. Sci. USA 86: 232-236.

References

123

Skillman, L. C., P. N. Evans, G. E. Naylor, B. Morvan, G. N. Jarvis, and K. N. Joblin.

2004. 16S ribosomal DNA-directed PCR primers for ruminal methanogens and

identification of methanogens colonising young lambs. Anaerobe 10: 277-285.

Skillman, L. C., P. N. Evans, C. Strömpl, and K. N. Joblin. 2006a. 16S rDNA directed PCR

primers and detection of methanogens in the bovine rumen. Lett. Appl. Microbiol.

42: 222-228.

Skillman, L. C., A. F. Toovey, A. J. Williams, and A.-D. G. Wright. 2006b. Development

and validation of a real-time PCR method to quantify rumen protozoa and

examination of variability between Entodinium populations in sheep offered a hay-

based diet. Appl. Environ. Microbiol. 72: 200-206.

Smith, P. H., and R. E. Hungate. 1958. Isolation and characterization of Methanobacterium

ruminantium n sp. J. Bact. 75: 713-718.

Smolenski, W. J., and J. A. Robinson. 1988. In situ rumen hydrogen concentrations in

steers fed 8 times daily, measured using a mercury reduction detector. Fems

Microbiol. Ecol. 53: 95-100.

Stewart, C. S., H. J. Flint, and M. P. Bryant. 1997. The rumen bacteria. In: P. N. Hobson

and C. S. Stewart (eds.) The rumen microbial ecosystem. p 10-72. Blackie

Acacemic & Professional, London.

Stumm, C. K., H. J. Gijzen, and G. D. Vogels. 1982. Association of methanogenic bacteria

with ovine rumen ciliates. Brit. J. Nutr. 47: 95-99.

Stumm, C. K., and K. B. Zwart. 1986. Symbiosis of protozoa with hydrogen-utilizing

methanogens. Microbiol. Sci. 3: 100-105.

References

124

Sylvester, J. T., S. K. Karnati, Z. Yu, M. Morrison, and J. L. Firkins. 2004. Development of

an assay to quantify rumen ciliate protozoal biomass in cows using real-time PCR.

J. Nutr. 134: 3378-3384.

Sylvester, J. T., S. K. Karnati, Z. Yu, C. J. Newbold, and J. L. Firkins. 2005. Evaluation of

a real-time PCR assay quantifying the ruminal pool size and duodenal flow of

protozoal nitrogen. J. Dairy Sci. 88: 2083-2095.

Tajima, K., R. I. Aminov, T. Nagamine, H. Matsui, M. Nakamura, and Y. Benno. 2001a.

Diet-dependent shifts in the bacterial population of the rumen revealed with real-

time PCR. Appl. Environ. Microbiol. 67: 2766-2774.

Tajima, K., T. Nagamine, H. Matsui, M. Nakamura, and R. I. Aminov. 2001b. Phylogenetic

analysis of archaeal 16S rRNA libraries from the rumen suggests the existence of a

novel group of archaea not associated with known methanogens. FEMS Microbiol.

Lett. 200: 67-72.

Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the

sensitivity of progressive multiple sequence alignment through sequence weighting,

position-specific gap penalties and weight matrix choice. Nucl. Acids Res. 22:

4673-4680.

Tichopad, A., A. Didier, and M. W. Pfaffl. 2004. Inhibition of real-time RT-PCR

quantification due to tissue-specific contaminants. Mol. Cell Probe. 18: 45-50.

Tichopad, A., M. Dilger, G. Schwarz, and M. W. Pfaffl. 2003. Standardized determination

of real-time PCR efficiency from a single reaction set-up. Nucl. Acids Res. 31:

e122.

References

125

Tokura, M., I. Chagan, K. Ushida, and Y. Kojima. 1999. Phylogenetic study of

methanogens associated with rumen ciliates. Curr. Microbiol. 39: 123-128.

Tokura, M., K. Ushida, K. Miyazaki, and Y. Kojima. 1997. Methanogens associated with

rumen ciliates. FEMS Microbiol. Ecol. 22: 137-143.

Ushida, K., C. J. Newbold, and J. P. Jouany. 1997. Interspecies hydrogen transfer between

the rumen ciliate Polyplastron multivesiculatum and Methanosarcina barkeri. J.

Gen. Appl. Microbiol. 43: 129-131.

Van Hoek, A. H. A. M., T. A. Van Alen, V. S. I. Sprakel, J. A. M. Leunissen, T. Brigge, G.

D. Vogels, and J. H. P. Hackstein. 2000. Multiple acquisition of methanogenic

archaeal symbionts by anaerobic cilliates. Mol. Biol. Evol. 17: 251-258.

Van Kessel, J. A. S., and J. B. Russell. 1996. The effect of pH on ruminal methanogenesis.

FEMS Microbiol. Ecol. 20: 205-210.

Van Nevel, C. J., and D. I. Demeyer. 1996. Control of rumen methanogenesis. Environ.

Monit. Assess. 42: 73-97.

Van Nevel, C. J., D. I. Henderickx, D. I. Demeyer, and J. Martin. 1969. Effect of chloral

hydrate on methane and propionic acid in the rumen. Appl. Microbiol. 17: 695-700.

Vogels, G. D., W. F. Hoppe, and C. K. Stumm. 1980. Association of methanogenic bacteria

with rumen ciliates. Appl. Environ. Microbiol. 40: 608-612.

White, C. L., D. G. Masters, D. W. Peter, D. B. Purser, S. P. Roe, and M. J. Barnes. 1992.

A multi element supplement for grazing sheep. I. Intake, mineral status and

production responses. Aust. J. Agric. Res. 43: 795-808.

Whitford, M., R. Teather, and R. Forster. 2001. Phylogenetic analysis of methanogens from

the bovine rumen. BMC Microbiol. 1: 1-5.

References

126

Whitman, W. B., T. L. Bowen, and D. R. Boone. 1991. The methanogenic bacteria. In: A.

Balows, H. G. Truper, M. Dworkin, W. Harder and K. J. Schleifer (eds.) The

Prokaryotes. p 719-767. Springer Verlag, New York.

Williams, A. G., and G. S. Coleman. 1992. The rumen protozoa. Springer-Verlag, New

York.

Williams, A. G., and G. S. Coleman. 1997. The rumen protozoa. In: P. N. Hobson and C. S.

Stewart (eds.) The rumen microbial ecosystem. p 73-139. Blackie acacemic &

professional.

Williams, A. G., and C. G. Harfoot. 1976. Factors affecting the uptake and metabolism of

soluble carbohydrates by the rumen ciliate Dasytricha ruminantium isolated from

the ovine rumen contents by filtration. J. Gen. Microbiol. 96: 125-136.

Wintzingerode, F., U. B. Gobel, and E. Stackebrandt. 1997. Determination of microbial

diversity in environmental samples: pitfalls of PCR-based rRNA analysis. FEMS

Microbiol. Rev. 21: 213-229.

Woese, C. R., O. Kandler, and M. L. Wheelis. 1990. Towards a natural system of

organisms: Proposal for the domains Archaea, Bacteria, and Eucarya. Proc. Natl.

Acad. Sci. USA 87: 4576-4579.

Wolin, M. J., T. L. Miller, and C. S. Stewart. 1997. Microbe-microbe interactions. In: P. N.

Hobson and C. S. Stewart (eds.) The rumen microbial ecosystem. p 467-491.

Blackie acacemic & professional, London.

Wright, A.-D. G., C. H. Auckland, and D. H. Lynn. 2007. Molecular diversity of

methanogens in feedlot cattle from Ontario and Prince Edward Island, Canada.

Appl. Environ. Microbiol. 73: 4206-4210.

References

127

Wright, A.-D. G., B. A. Dehority, and D. H. Lynn. 1997. Phylogeny of the rumen ciliates

Entodinium, Epidinium and Polyplastron (Litostomatea: Entodiniomorphida)

inferred to small subunit ribosomal RNA sequences. J. Eukaryot. Microbiol. 44: 61-

67.

Wright, A.-D. G., and D. H. Lynn. 1997a. Phylogenetic analysis of the rumen ciliate family

Ophryoscolecidae based on 18S ribosomal RNA sequences, with new sequences

from Diplodinium, Eudiplodinium, and Ophryoscolex. Can. J. Zool. 75: 963-970.

Wright, A.-D. G., A. F. Toovey, and C. L. Pimm. 2006. Molecular identification of

methanogenic archaea from sheep in Queensland, Australia reveal more uncultured

novel archaea. Anaerobe 12: 134-139.

Wright, A.-D. G., A. J. Williams, B. Winder, C. T. Christophersen, S. L. Rodgers, and K.

D. Smith. 2004. Molecular diversity of rumen methanogens from sheep in Western

Australia. Appl. Environ. Microbiol. 70: 1263-1270.

Wright, A. D. G., and D. H. Lynn. 1997b. Monophyly of the trichostome ciliates (Phylum

Ciliophora: Class Litostomatea) tested using new 18S rRNA sequences from the

vestibuliferids, Isotricha intestinalis and Dasytricha ruminantium, and the haptorian,

Didinium nasutum. Eur. J. Protistol. 33: 305-315.

Wu, Y., V. Hayes, J. Osinga, I. Mulder, M. Looman, C. Buys, and R. Hofstra. 1998.

Improvement of fragment and primer selection for mutation detection by denaturing

gradient gel electrophoresis. Nucl. Acids Res. 26: 5432-5440.

Wu, Y., R. P. Stulp, P. Elfferich, J. Osinga, C. H. Buys, and R. M. Hofstra. 1999. Improved

mutation detection in GC-rich DNA fragments by combined DGGE and CDGE.

Nucl. Acids Res. 27: e9i-e9iii.

References

128

Wuebbles, D. J., and K. Hayhoe. 2002. Atmospheric methane and global change. Earth-Sci.

Rev. 57: 117-210.

Zhang, T., and H. H. P. Fang. 2006. Applications of real-time polymerase chain reaction for

quantification of microorganisms in environmental samples. Appl. Microbiol.

Biotechnol. 70: 281-289.

Zinder, S. H. 1993. Physiological ecology of methanogens. In: J. G. Ferry (ed.)

Methanogenesis. Ecology, Physiology, Biochemistry and Genetics. p 128-206.

Chapman and Hall, New York.

grain and artificial stimulation of the rumen change the...

Documents