Rumen bacterial, archaeal and anaerobic fungal communities of Holstein dairy heifers and
cows, in a tropical system of production, were characterized through sequencing the 16s
rRNA and the ITS genes. In addition, we investigated the relationship between these
communities and enteric methane (CH4) emissions and productive traits, such as digestible
dry matter intake (dDMI), digestible organic matter intake (dOMI), average body weight
(BW), rumen pH, volatile fatty acids (VFA) and its main components, acetate, propionate
and butyrate. Prepubertal heifers (PP), pubertal heifers (PB), and pregnant heifers (PG) were
used in Chapter 1. Pregnant heifers emitted more CH4 than others, followed by PB and PP.
Regarding CH4 emissions, the animals were split in high and low CH4 emitters. Heifers were
fed a diet composed by corn silage and concentrate (corn, soybean meal and minerals).
Prevotella, Ruminococcus, Coprococcus, Butyrivibrio, Clostridium, Shuttleworthia, SHD-
231, CF231, p-75-a5, Methanobrevibacter, Methanosphaera and Caecomyces communis
were detected to be the core microbiome of the evaluated heifers. Families Bifidobacteriaceae
and RF16 and genera Acetobacter and Coprococcus were strongly correlated with CH4
emissions. Genera Eubacterium, p-75-a5 and SHD-231 showed inverse correlations with
CH4 emissions, dDMI, dOMI, BW and rumen pH. Methanobrevibacter, in archaeal
community, and Orpinomyces, in anaerobic fungal, showed positive and weak correlations
with CH4 emissions. On the other hand, strong and negative correlations were observed
among Methanosphaera and this variable. Prepubertal and PG heifers were the most
divergent groups in relation to CH4 emissions. Surprisingly, they did not differ in relative
abundances of Firmicutes and Bacteroidetes, but PG had greater abundance of
Methanobrevibacter and Vadin CA11 and lower abundance of Methanosphaera. None of the
bacterial, archaea and anaerobic fungi which correlate with CH4 emissions showed
significant correlations (P>0.10) with VFA and the individual concentrations of acetate,
propionate and butyrate. Lastly, this work showed that bacterial, archaeal and anaerobic
fungal communities did not covaried and the microbial communities did not covaried with
volatile fatty acids concentration either. In Chapter 2, high-producing (HP), mediumproducing
(MP), low-producing (LP) and dry (DC) were evaluated. The forage:concentrate
ratios they were fed were 50:50 for HP, 70:30 for MP, 80:20 for LP, and 90:10 for DC.
Considering the intake of digestible fraction of feed, DC emitted more CH4, followed by MP,
HP and LP, but the HP and LP emissions were similar. The core microbiome of the evaluated
Holstein cows in tropical environment was composed by Prevotella, Ruminococcus,
Butyrivibrio, Clostridium, Coprococcus, Shuttleworthia, CF231, SHD-231,
Methanobrevibacter, and Methanosphaera. None of the anaerobic fungal operational
taxonomic units (OTU) were found in all samples. Firmicutes and Bacteroidetes were the
most abundant phyla found in the rumen of Holstein cows. For the archaeal community,
Methanobrevibacter genera was the most abundant and in anaerobic fungi, most of the
sequences were unclassified. The strongest negative correlations with CH4 emissions
detected were with the relative abundance of family Coriobacteriaceae and S24-7 and of
genera Butyrivibrio, Clostridium and Schwartzia. Positive correlations were found between
CH4 emissions and families RF16 and Succinivibrionaceae. In the archaeal community,
genera Methanosphaera relative abundance showed a strong negative correlation with CH4.
Surprisingly, no significant correlation between CH4 emissions and Methanobrevibacter
relative abundance was found. Relative abundance of genera Vadin CA11 (in archaea) and
Caecomyces (in anaerobic fungi) were detected to be positively correlated with CH4 in g/day.
Many families and genera from Firmicutes phylum showed positive correlations with dDMI
and dOMI. None of the bacterial, archaea and anaerobic fungi which correlate with CH4
emissions showed significant correlations (P>0.1) with VFA and the individual
concentrations of acetate, propionate and butyrate. The most opposite results observed in the
present study were among DC and HP. Dry cows showed greater CH4 emissions in g/kg
dDMI and g/kg of dOMI and, besides no differences were observed in relative abundances
of Firmicutes, Bacteroidetes and Firmicutes:Bacteroidetes ratio, DC had lower relative
abundance of Coriobacteriaceae, which was negatively correlated with CH4, and greater
relative abundance of Succinivibrionaceae, that was positively correlated with CH4. In
addition, DC had greater relative abundance of Methanobrevibacter and lower of
Methanosphaera. Lastly, bacterial, archaeal and anaerobic fungal communities did no covary
and VFA and microbial communities did not vary in a similar way either. Chapter 3 was
composed by two trials. In trial 1, CH4 emissions were estimated from the seven previously
described Holstein dairy cattle categories based on the SF6 tracer gas technique and on IPCC
(2006) equations. Enteric CH4 emission was higher for the PP heifers when estimated by the
equations proposed by the IPCC Tier 2. However, higher CH4 emissions were estimated by
the SF6 technique for MP, HP and DC. Pubertal heifers, PG, and LP had equal CH4 emissions
as estimated by both methods. In trial 2, two dairy farms were monitored for one year to
identify all activities that contributed in any way to GHG emissions. The total emission from
Farm 1 was 3.21 t CO2e/animal/yr, of which 1.63 t corresponded to enteric CH4. Farm 2
emitted 3.18 t CO2e/animal/yr, with 1.70 t of enteric CH4. For the carbon balance
calculations, when the carbon stock in pasture and other crops was considered, the carbon
balance suggested that both farms are sustainable for GHG, by both estimation methods. On
the other hand, carbon balance without carbon stock, by both estimation methods, suggests
that farms emit more carbon than the system is capable of stock. It was concluded that IPCC
estimations can underestimate CH4 emissions from some categories while overestimate
others. However, considering the whole property, these discrepancies were offset and we
would submit that the equations suggested by the IPCC properly estimate the total CH4
emission and carbon balance of the properties. Thus, the IPCC equations should be utilized
with caution, and the herd composition should be analyzed at the property level.