Tesis Rodriges de Carvalho

i
UNIVERSIDADE FEDERAL DE GOIÁS

ESCOLA DE VETERINÁRIA
PROGRAMA DE PÓS-GRADUAÇÃO EM CIÊNCIA ANIMAL
VALOR NUTRITIVO DO GLICEROL E COMPORTAMENTO

INGESTIVO DE VACAS LEITEIRAS PERIPARTURIENTES
Eduardo Rodrigues de Carvalho

Orientador: Prof. Dr. Juliano José de Resende Fernandes
GOIÂNIA
2010
ii
iii
UNIVERSIDADE FEDERAL DE GOIÁS

ESCOLA DE VETERINÁRIA
PROGRAMA DE PÓS-GRADUAÇÃO EM CIÊNCIA ANIMAL
VALOR NUTRITIVO DO GLICEROL E COMPORTAMENTO

INGESTIVO DE VACAS LEITEIRAS PERIPARTURIENTES
Tese apresentada ao Programa de Pós-

Graduação em Ciência Animal da Escola de
Veterinária/UFG para obtenção do grau de
Doutor.
Área de Concentração:
Produção Animal
Orientador:
Prof. Dr. Juliano José de Resende Fernandes
Comitê de orientação:
Prof. Dr. Milton Luiz Moreira Lima
Prof. Dr. Shawn Scott Donkin
GOIÂNIA
2010
iv
v
vi
AGRADECIMENTOS
À Deus, pelo dom da vida, por estar à frente no meu caminho e por
todas as bênçãos concedidas, principalmente nos momentos de incerteza e
ansiedade.
À minha filha Mikaela, pela convivência harmoniosa e pela ajuda no
experimento coletando amostras.
À minha esposa Andrea, a quem eu dedico essa conquista, além de
ter sugerido a ideia da Pós-Graduação e ter me apoiado incondicionalmente
desde o Mestrado.
Ao meu pai Jayme (in memorian), a quem eu espero poder
reencontrar algum dia.
À minha avó Leonor (in memorian), pelos momentos maravilhosos
da minha infância e adolescência.
Ao meu avô Assuero (in memorian), pelos primeiros ensinamentos
na agricultura e por despertar o grande amor por vacas leiteiras.
Ao meu Orientador, Professor Juliano José de Resende Fernandes,
pelo aceite desde o Mestrado.
Ao meu Co-orientador, Professor Milton Luiz Moreira Lima, pela
amizade e oportunidade de publicarmos juntos.
Ao meu Co-orientador, Professor Shawn Scott Donkin, pelos
grandes ensinamentos em Nutrição de Ruminantes e pelo constante apoio na
Universidade de Purdue.
Aos colegas de Pós-Graduação da Purdue, Heather White e Nicole
Schmelz pela amizade e grande ajuda nas análises laboratoriais.
Ao estatístico Mark Einstein pela ajuda com o SAS e à Sophia
Wilcox pela ajuda com as análises dos vídeos.
Ao pessoal da Fazenda-Escola da Purdue, Mike Grott, Dee Hoffman,
Jeff Synesael, Jeremy, Steven Hendress e Gary Wernert pela ajuda na condução
do experimento.
Às estagiárias Amanda Slabaugh, Lynn Pezzanite e Kelli Kuehnert
pela ajuda nas coletas de amostras e no laboratório.
vii
Ao meu grande amigo João Carlos Cardoso da Silva (in memorian)

pela ótima convivência no período em que estivemos juntos.
Ao meu amigo John Lee Ferguson pela amizade e torcida nos
momentos de incerteza.
Aos meus grandes amigos Jon Cantey, James Johnson e Elvis
McNeely pela ótima convivência nos Estados Unidos.
À CAPES pela Bolsa de estudos no Brasil e nos Estados Unidos.
A todos que, de alguma forma, me ajudaram e torceram por mim na
realização desse trabalho.
Muito obrigado!
viii
SUMÁRIO
RESUMO GERAL xiii
CAPÍTULO 1 – CONSIDERAÇÕES GERAIS 1
1 INTRODUÇÃO 1
2 REVISÃO DE LITERATURA 2
2.1 Propriedades químicas do glicerol e seu uso potencial como ingrediente 2
nas rações de ruminantes
2.2 Glicerol no tratamento da cetose ou acetonemia 5
2.3 Metabolismo do glicerol no rúmen 6
2.4 A vaca leiteira no período de transição 8
3 REFERÊNCIAS 13
CAPÍTULO 2 – REPLACING CORN WITH GLYCEROL IN DIETS FOR 17
TRANSITION DAIRY COWS
RESUMO 17
ABSTRACT 18
1 INTRODUCTION 19
2 MATERIALS AND METHODS 20
2.1 Cows and dietary treatments 20
2.2 Blood collection and analysis 23
2.3 Rumen fluid sampling and analysis 23
2.4 Data Analysis 24
3 RESULTS 24
4 DISCUSSION 31
5 CONCLUSIONS 36
6 REFERENCES 36
CAPÍTULO 3 – FEEDING BEHAVIORS OF TRANSITION DAIRY COWS 41
FED GLYCEROL AS A REPLACEMENT FOR CORN.
RESUMO 41
ABSTRACT 42
1 INTRODUCTION 43
2.1 Housing and management 45
2.2 Diets and treatments 45
2.3 Video Taping and Feed Sorting 48
2.4 Data Analysis 49
ix
3 RESULTS 49
4 DISCUSSION 65
5 CONCLUSIONS 68
6 ACKNOWLEDGEMENTS 68
7 REFERENCES 69
CAPÍTULO 4 – AN ALTERNATIVE METHODOLOGY OF DETERMINING 72
FEED SORTING IN TRANSITION DAIRY COWS FED GLYCEROL.
RESUMO 73
ABSTRACT 74
1 INTRODUCTION 75
3 RESULTS 80
4 DISCUSSION 88
5 CONCLUSIONS 89
6 REFERENCES 90
CAPÍTULO 5 – CONSIDERAÇÕES FINAIS 92
x
LISTA DE FIGURAS
CAPÍTULO 1
FIGURA 1- Reação de transesterificação para produção de biodiesel (VAN 4
GERPEN et al., 2004)
CAPÍTULO 2
FIGURE 1- Dry matter intake (kg/day) of transition dairy cows fed control
(solid squares) or glycerol (open squares) during the pre- and 27
post-partum intervals.
FIGURE 2- Milk yield (kg/day) of transition dairy cows fed control (solid 27
squares) or glycerol (open squares).
FIGURE 3- Effect of glycerol on glucose, NEFA and BHBA in plasma for 30
cows fed control (solid squares) or glycerol (open squares).
CAPÍTULO 3
FIGURE 1- Eating rate (g DM/h) associated with particle size, diet, and time
after feed delivery. Cows were fed glycerol or control diets as a 53
TMR between 06:30 to 07:30 h on -16, -9, +9, +16 and +51
DRTC.
FIGURE 2- Effect of time after feed delivery and glycerol on eating rate.
Cows were fed glycerol (open squares) or control (solid 55
squares) diets as a TMR between 06:30 to 07:30 h on -16 and -
9 DRTC.
FIGURE 3- Distribution of particle sizes consumed during the prepartum
period for cows fed the glycerol or control diets. Cows were fed 56
glycerol or control diets as a TMR between 06:30 to 07:30 h on
-16 and -9 DRTC.
FIGURE 4- Time spent ruminating relative to treatment × time post feeding
effect of transition dairy cows fed control (solid squares) or 61
glycerol (open squares) during the prepartum interval. Pooled
SEM = 2.28
CAPÍTULO 4
FIGURE 1- Feed sorting of transition dairy cows fed control (bars 1 through
4) or glycerol (bars 5 through 8) during the prepartum interval 83
according to LEONARDI & ARMENTANO (2003).
4) or glycerol (bars 5 through 8) during the postpartum interval 84
according to LEONARDI & ARMENTANO (2003).
4) or glycerol (bars 5 through 8) during the prepartum interval 86
according to the methodology proposed in this study.
4) or glycerol (bars 5 through 8) during the postpartum interval 87
according to the methodology proposed in this study.
xi
LISTA DE TABELAS
CAPÍTULO 2
TABLE 1- Ingredient and nutrient composition of the pre- and post-partum 22
experimental diets.
TABLE 2- Effect of glycerol on feed intake, milk production, feed efficiency, 26
milk composition, BW change and BCS change.
29
TABLE 3- Effect of glycerol on plasma metabolites.
31
TABLE 4- Effect of glycerol on rumen parameters.
CAPÍTULO 3
experimental diets.
51
TABLE 2- Particle size distribution of the pre- and post-partum diets.
TABLE 3- Glycerol associated with particles in TMR containing glycerol and 51
fed during the pre- and post-partum periods.
TABLE 4- Probabilities associated with the effect of diet (control vs. glycerol),
DRTC, time after feed delivery, and interactions within particle 54
size for TMR for control or glycerol fed cows during the prepartum
period.
TABLE 5- Probabilities associated with the effect of diet (control vs. glycerol),
DRTC, time after feed delivery, and interactions within particle 57
size for TMR for control or glycerol fed cows during the
postpartum period.
TABLE 6- Effect of glycerol on prepartum eating, resting, and ruminating 59
activities.
TABLE 7- Probability values associated with an effect of treatment, DRTC,
time after feed delivery and interactions for eating, resting and 60
rumination for transition dairy cows fed glycerol or control diets
during the prepartum period.
TABLE 8- Effect of glycerol on postpartum eating, resting and ruminating 63
activities.
TABLE 9- Probability values associated with an effect of treatment, DRTC,
time after feed delivery and interactions for eating, resting and 64
rumination for transition dairy cows fed glycerol and control diets
during the postpartum period.
CAPÍTULO 4
experimental diets.
TABLE 2- Particle size distribution of the pre- and post-partum diets. 81
xii
LISTA DE ABREVIATURAS
ABHB ácido β-hidroxibutírico

ADF “acid detergent fiber”
AGNE ácidos graxos não esterificados
AGCC ácidos graxos de cadeia curta
AQP “aquaporins”
BHBA “β-hydroxybutyrate acid”
BCS “body condition score”
BEN balanço energético negativo
BW “body weight”
Ca “calcium”
CMS consumo de matéria seca
CP “crude protein”
DM “dry matter”
DMI “dry matter intake”
DRDP dias relativos à data de parição
DRTC “days relative to calving”
ECC escore de condição corporal
ELL energia líquida de lactação
MS matéria seca
NEFA “nonesterified fatty acids”
NDF “neutral detergent fiber”
NEL “net energy lactation”
P “phosphorus”
PCV “packed cell volume”
PSPS “Penn State Particle Separator”
RNAm ácido ribonucleico mensageiro
SCFA “short chain fatty acids”
TMR “total mixed ration”
VLDL “very low density lipoproteins”
xiii
RESUMO GERAL
A expansão da indústria de biocombustíveis tem gerado aumento na
disponibilidade do glicerol, o qual pode ser utilizado como ingrediente na ração de
vacas leiteiras. Objetivou-se nesse estudo determinar os efeitos do glicerol sobre
o CMS, produção de leite, produção de AGCC no rúmen, parâmetros sanguíneos,
consumo seletivo e comportamento ingestivo de vacas leiteiras periparturientes.
Vinte e seis vacas multíparas da raça Holandesa foram pareadas de acordo com
o desempenho na lactação anterior e data prevista de parição, e alimentadas com
dietas contendo glicerol ou milho grão úmido desde -28 até +56 DRDP. O glicerol
foi incluído em 11,5 e 10,8% do total da MS nas dietas pré e pós-parto,
respectivamente. O CMS não foi alterado (P>0,05) pela alimentação com glicerol
tanto no pré-parto (14,9 vs. 14,6 kg/dia, controle vs. glicerol) quanto no pós-parto
(19,8 vs. 20,7 kg/dia, controle vs. glicerol), assim como a produção (35,8 vs. 37
kg/dia, controle vs. glicerol) e composição de leite não diferiram (P>0,05) entre os
tratamentos. A concentração de glucose no sangue foi reduzida (P<0,05; 59,1 vs.
53,4 mg/dL, controle vs. glicerol) e de ABHB no sangue foi elevada (P<0,05; 0,58
vs. 0,82 mmol/L, controle vs. glicerol) nas vacas alimentadas com glicerol durante
o pré-parto. A concentração de AGNE no sangue não diferiu (P>0,05) entre os
grupos experimentais no pré-parto, e não houve efeito (P>0,05) do glicerol sobre
os parâmetros sanguíneos durante o pós-parto. A concentração total de AGCC no
rúmen (mmol/L) não diferiu (P>0,05) entre os tratamentos, mas houve aumento
(P<0,05) na proporção molar de propionato (22,7 vs. 28,6%, controle vs. glicerol)
e butirato (11,5 vs. 15,3%, controle vs. glicerol) e redução (P<0,05) na proporção
molar de acetato (61,5 vs. 51,5%, controle vs. glicerol) nas vacas alimentadas
com glicerol. Em relação ao consumo seletivo e comportamento ingestivo, houve
aumento (P<0,05) na taxa do CMS (94,2 vs. 144,4 g MS/h; controle vs. glicerol) e
no consumo preferencial (9,2 vs. 17,8%; controle vs. glicerol) de partículas longas
na dieta com glicerol, porém houve redução (P<0,05) na taxa do CMS de
partículas curtas (383,8 vs. 332,5 g MS/h; controle vs. glicerol) e muito curtas
(173,9 vs. 129,8 g MS/h; controle vs. glicerol) e aumento (P<0,05) na rejeição de
partículas curtas (42 vs. 37,3%; controle vs. glicerol) e muito curtas (17,9 vs.
13,6%; controle vs. glicerol) durante o pré-parto. Não houve efeito (P>0,05) dos
tratamentos sobre o consumo seletivo de partículas da dieta no pós-parto e
xiv
também sobre o comportamento ingestivo durante todo o período experimental.

Os dados obtidos na presente pesquisa indicaram que a substituição do milho
grão úmido pelo glicerol em dietas para vacas leiteiras periparturientes foi
satisfatória.
Palvras-chave: biocombustíveis, consumo seletivo, energia, preferência,

subproduto
1
CAPÍTULO 1 – CONSIDERAÇÕES GERAIS
1 INTRODUÇÃO
Apesar da taxa do crescimento da população mundial ter se

reduzido nos últimos anos, as estimativas ainda indicam um total de 8 bilhões de
pessoas no ano de 2025 (DEPARTMENT OF ECONOMIC AND SOCIAL
AFFAIRS OF THE UNITED NATIONS, 2009). Portanto, a fim de suprir alimentos
em quantidade e qualidade satisfatórias a essa população ainda crescente, são
necessárias pesquisas para se conhecer o valor nutritivo de subprodutos
agrícolas ou industriais, tais como o glicerol, que tem o potencial de substituir os
alimentos energéticos clássicos utilizados nas rações de ruminantes, diminuindo
nesse caso a utilização do milho nessas rações, trazendo como consequências o
aumento na eficiência do sistema de produção e redução dos recursos não
renováveis para a produção de grãos, além da possibilidade do milho ser utilizado
diretamente na alimentação humana nas regiões mais pobres do planeta.
Atualmente, o aumento na produção de biodiesel em todo o mundo tem gerado o
aumento concomitante na oferta de glicerol que não pode ser imediatamente
absorvido pelas indústrias química e farmacêutica, criando assim uma boa
oportunidade no uso desse subproduto como ingrediente primário nas rações de
ruminantes.
Até o momento, países como os Estados Unidos estão dependentes
basicamente da cultura da soja para produção de biodiesel, ao passo que o Brasil
tem um número muito maior de culturas oleaginosas para produção de biodiesel,
tais como a palma, babaçu e o pinhão manso na Região Norte, a soja, girassol e
o amendoim nas Regiões Sul, Sudeste e Centro-Oeste, e finalmente a mamona
como opção para a região Nordeste. A maior diversidade de culturas oleaginosas
para a produção de biodiesel permite a instalação de indústrias em vários locais
do país. Entretanto, a utilização em grande escala do glicerol como ingrediente de
rações para ruminantes irá depender de fatores como a distância entre as
indústrias de biodiesel e fazendas produtoras de leite e carne, logística de
2
transporte e armazenamento e ainda do preço do milho como balizador do preço

do glicerol.
2 REVISÃO DE LITERATURA
2.1 Propriedades químicas do glicerol e seu uso potencial como ingrediente

nas rações de ruminantes
O glicerol refinado é um líquido viscoso, levemente adocicado,

incolor, inodoro e higroscópico. Sua fórmula molecular é C3H5(OH)3, massa molar
de 92,09 g/mol e densidade de 1,26 kg/L (VAN GERPEN et al., 2004). Os termos
sinônimos para o glicerol incluem a glicerina, 1,2,3-propanotriol, 1,2,3-
trihidróxipropano e gliceritol. Por definição o glicerol é um álcool açucarado.
Devido às suas propriedades umectantes, conteúdo energético e alta solubilidade
em água, o glicerol tem sido amplamente utilizado na indústria alimentícia,
farmacêutica e de cosméticos. Com a expansão da indústria de biodiesel ao redor
do mundo, aumentou-se a disponibilidade de glicerol não refinado, um subproduto
valioso como fonte de energia que pode ser utilizado nas rações de ruminantes
(DONKIN & DOANE, 2007).
O biodiesel é atualmente produzido por uma reação de
transesterificação que utiliza uma base catalisadora que reage com a gordura
vegetal ou animal (FIGURA 1). No caso da produção de biodiesel à base de óleo
de soja, reage-se a mesma quantidade de óleo com um álcool de cadeia curta
(normalmente metanol, ou também etanol, dependendo das instalações
industriais) na presença de um catalisador (hidróxido de sódio ou hidróxido de
potássio) para a produção de biodiesel e glicerol não refinado. Essa reação
requer baixa temperatura e baixa pressão com alto índice de conversão de 98% e
reações secundárias mínimas, resultando na conversão direta de óleo vegetal em
biodiesel sem compostos intermediários. O biodiesel é então separado do glicerol
por gravidade ou centrifugação.
3
A maioria das plantas industriais utiliza uma razão molar álcool:óleo

de 6:1, a qual excede 100%, com o objetivo de completar a reação de
transesterificação, sendo que o excedente do álcool (até 80%) fica retido junto
com a camada de glicerol. O álcool é então removido das camadas de biodiesel e
glicerol por evaporação ou destilação instantânea, para depois ser recuperado e
reutilizado (VAN GERPEN et al., 2004).
Para cada 100 g de óleo de soja que são utilizados no início da
reação, são obtidos 12,25 g de glicerol não refinado (THOMPSON & HE, 2006). O
glicerol resultante contém catalisadores e sabões não utilizados que são
neutralizados por acidificação para produção de glicerol não refinado contendo 80
a 88% de glicerol. O posterior refino de glicerol não refinado para 99% ou mais de
pureza se faz necessário para o seu uso na indústria farmacêutica e de
cosméticos (VAN GERPEN et al., 2004). De acordo com a FDA (2006), o glicerol
é classificado como seguro para uso na alimentação animal, apesar do glicerol
definido nessa classificação ser refinado e historicamente produzido nas
indústrias de sabões. Por outro lado, o glicerol não refinado contém
contaminantes, dos quais o metanol é o mais preocupante, sendo que a U.S. FDA
emitiu um documento ressaltando que o glicerol proveniente da produção de
biodiesel deve conter no máximo 150 mg de metanol por kg de glicerol, e que
níveis acima desse limite são inapropriados para o uso na alimentação animal.
Entretanto, o governo alemão estabelece como limite máximo 5.000 mg de
metanol por kg de glicerol para alimentação animal, portanto bem acima do limite
imposto pelo governo estadunidense (SELLERS, 2008). Até a presente data não
existem limites estabelecidos pelas autoridades brasileiras em relação ao nível
máximo de metanol presente no glicerol para uso na alimentação animal.
Recente pesquisa da composição do glicerol não refinado indicou
76,2% de glicerol, 7,98% de lipídeos, 0,05% de proteína e 2,73% de cinzas (11
ppm de cálcio, 6,8 ppm de magnésio, 53 ppm de fósforo e 1,2% de sódio),
totalizando 86,96%, porém não houve relato da composição dos 13,04%
remanescentes (THOMPSON & HE, 2006).
4
Catalisador
NaOH ou KOH
Triglicerídeo Metanol Biodiesel Glicerol
FIGURA 1- Reação de transesterificação para produção de biodiesel (VAN

GERPEN et al., 2004)
Dentro do escopo dessa revisão de literatura, encontrou-se apenas

um experimento em que se utilizou glicerol não refinado na alimentação de vacas
leiteiras periparturientes, o qual continha 80,2% de glicerol e 1,3% de metanol,
portanto muito acima do limite estabelecido até mesmo pelo governo alemão
(5.000 mg/kg ou 0,5%). Verificou-se nesse estudo que as vacas suplementadas
com glicerol tanto na maior (0,86 kg/dia) quanto na menor (0,43 kg/dia + 0,43
kg/dia de amido) dose tiveram menor CMS do que as vacas suplementadas com
0,86 kg/dia de amido durante o pré-parto. Todavia, o CMS no pós-parto e a
produção de leite não foram afetados pelos tratamentos (DEFRAIN et al., 2004).
Em outra pesquisa, a utilização de três níveis (4, 8 e 12% do total da
MS) de glicerol não refinado (85,7% de glicerol e 0,09% de metanol) na
alimentação de bovinos de corte demonstrou não haver efeito negativo sobre o
CMS (8,27 ± 0,32 kg/dia) e o ganho de peso médio diário (1,36 ± 0,087 kg/dia)
desses animais comparado à dieta controle (MACH et al., 2009). Assim, há
necessidade da realização de novas pesquisas a fim de se determinar o limite
máximo de metanol presente no glicerol que não seja prejudicial ao desempenho
e saúde de ruminantes, uma vez que quanto maior o grau de refino executado
pela indústria, maior será o preço do glicerol pago pelo produtor rural, além do
fato de existir uma grande discrepância entre os limites máximos de metanol
estabelecidos pelos governos estadunidense e alemão. Por outro lado, a fim de
5
se evitar qualquer tipo de efeito deletério de contaminantes sobre o desempenho

das vacas, utilizou-se nesse estudo glicerol com 99,5% de pureza.
2.2 Glicerol no tratamento da cetose ou acetonemia
Nas pesquisas iniciais com glicerol, esse subproduto foi utilizado

como precursor glucogênico no tratamento da cetose ou acetonemia em
experimentos conduzidos na década de 50. À época, sabia-se que o glicerol era
um metabólito natural que é sintetizado na formação de tecido adiposo e oxidado
quando esse mesmo tecido é catabolizado (JOHNSON, 1955). JOHNSON (1955)
também reportou que quando vacas receberam o glicerol na forma oral, misturado
ao concentrado ou diretamente no rúmen através de cânula ruminal, não houve
nenhuma reação adversa nos animais, a não ser redução no CMS devido ao alto
conteúdo de energia do glicerol. Concluiu-se nessa pesquisa que o glicerol era
eficiente no tratamento de cetose ou acetonemia como precursor glucogênico,
porém seu uso não foi viabilizado à época devido ao alto custo de produção
(JOHNSON, 1955).
A avaliação do glicerol como tratamento para cetose ou acetonemia
foi posteriormente explorada na década de 70, quando FISHER et al. (1971)
adicionaram 3,3% de glicerol ou propileno glicol ao concentrado oferecido a vacas
leiteiras em lactação. O consumo de glicerol e propileno glicol foi respectivamente
de 472 e 376 g/dia, em razão do consumo de concentrado das vacas
suplementadas com glicerol ter sido maior (14,3 kg/dia) comparado às vacas
suplementadas com propileno glicol (11,4 kg/dia). Ainda nesse estudo, não houve
efeito de tratamento sobre o consumo de feno, produção e composição de leite,
balanço de energia, concentração de glucose no sangue, pH ruminal e
concentrações de acetato e propionato no rúmen, porém as vacas suplementadas
com glicerol aumentaram a concentração de butirato no rúmen e de ABHB no
sangue (FISHER et al., 1971).
Mais tarde, FISHER et al. (1973) investigaram a inclusão de 3% de
propileno glicol ou 3 e 6% de glicerol adicionados ao concentrado e verificaram
que o consumo de concentrado foi de 5 kg/dia, resultando portanto em um
6
consumo diário de 150 (3% de inclusão) e 300 g/dia (6% de inclusão) de glicerol.
Não houve diferença sobre a produção e composição do leite e balanço de
energia, sendo que os autores atribuíram a baixa incidência de cetose ou
acetonemia na ausência de resposta entre os tratamentos (FISHER et al., 1973).
Mais recentemente, o glicerol foi reexaminado quanto à sua máxima
dose tolerável e também no tratamento clínico de cetose ou acetonemia. A
administração de um, dois ou três litros de glicerol via tubo esofágico em vacas
não lactantes aumentou a concentração de glucose no sangue em 16, 20 e 25%
após 30 minutos, comparado aos valores pré-tratamentos, e permaneceram
elevados por oito horas, voltando aos níveis antes dos tratamentos após 24 horas.
Duas das três vacas que receberam três litros se mostraram cambaleantes, tendo
retornado à sua condição normal dentro de quatro horas. Em relação ao
tratamento da cetose ou acetonemia clínica, duas vacas lactantes previamente
tratadas por dois dias com glucose intravenosa, apresentando pouca melhora,
receberam um litro de glicerol, sendo que nos dois animais o nível de corpos
cetônicos na urina foi reduzido a traços após 24 horas, e a produção de leite foi
aumentada entre 1,8 a 2,7 kg/dia. Ademais, a concentração de glucose no sangue
em uma das vacas aumentou de 48 para 75 mg/dL 30 minutos após a ingestão de
glicerol e para 109 mg/dL cinco horas pós-tratamento, enquanto que na outra
vaca a concentração de glucose no sangue permaneceu estável durante quatro
horas pós-tratamento, quando então aumentou de 48 para 74 mg/dL e reduziu
para 64 mg/dL oito horas pós-tratamento (GOFF & HORST, 2001).
2.3 Metabolismo do glicerol no rúmen
No rúmen, o glicerol é convertido a AGCC ou existe a hipótese de

ser absorvido diretamente pelo epitélio ruminal. Relatos antigos indicaram que o
glicerol foi totalmente convertido a propionato (JOHNS et al., 1953; GARTON et
al., 1961), o que foi posteriormente confirmado por KIJORA et al. (1998). Outros
autores reportaram aumento em acetato e propionato (WRIGHT, 1969) ou
propionato e butirato (CZERKAWSKI & BRECKENRIDGE, 1972; KHALILI et al.,
1997; WANG et al., 2009).
7
A fermentação in vitro de glicerol utilizando-se fluido ruminal de

vacas adaptadas à alimentação com glicerol indicou aumento na produção de
propionato e butirato e redução de acetato. O mesmo padrão de resultado foi
confirmado in vivo, quando amostras de fluido ruminal foram coletadas de vacas
alimentadas com glicerol na ração (1,2 kg/dia) ou administrado diretamente no
rúmen (0,24 kg/dia) via cânula (RÉMOND et al., 1993). Corroborando os
resultados in vitro obtidos por RÉMOND et al. (1993), BERGNER et al. (1995)
incubaram fluido ruminal de ovinos não adaptados à alimentação com glicerol por
seis horas e utilizaram glicerol marcado com C14, verificando que a maior parte
do glicerol foi encontrada no propionato, assim como a relação acetato:propionato
diminuiu de 3,5 a 4 (sem adição de glicerol) para 2,5 com 25% de adição de
glicerol.
Como descrito anteriormente, existe a hipótese do glicerol ser
absorvido diretamente pelo epitélio ruminal, baseado nos resultados obtidos por
RÉMOND et al. (1993), em que dos 0,24 kg/dia administrados diretamente no
rúmen via cânula, 0,032 kg foram encontrados no abomaso, 0,105 kg foram
fermentados no rúmen e 0,103 kg foram absorvidos diretamente através da
parede do rúmen.
Em concordância com esses autores, KIJORA et al. (1998)
reportaram que somente 0,75% dos 0,2 kg de glicerol colocados diretamente no
rúmen chegaram ao duodeno, e que o aumento da concentração de glicerol no
sangue em até três vezes, comparado às amostras dos animais alimentados com
a dieta controle, sugerem que o epitélio ruminal apresenta capacidade de
absorção direta do glicerol.
Relatos recentes têm evidenciado a existência de AQP, as quais são
proteínas presentes nas membranas celulares que regulam o fluxo de água
(AGRE, 2006), e que potencialmente também podem absorver o glicerol. As AQP
existem em bactérias, plantas e em todo o reino animal, e até o momento 13
delas foram identificadas em tecidos mamíferos, sendo que pelo menos quatro
(AQP-3, AQP-7, AQP-9 e AQP-10) também têm a função de canais para a
passagem de glicerol e/ou ureia.
O tecido adiposo apresenta quantitativamente as maiores
quantidades da expressão de AQP-7 no RNAm, enquanto que em seres humanos
8
a maior expressão de AQP-7 no RNAm ocorre no tecido omental (WINTOUR &

HENRY, 2006). A importância da AQP-7 no transporte de glicerol foi destacada
em cobaias que não possuíam o gene funcional da AQP-7 e, como resultado,
desenvolveram obesidade e resistência à insulina, em parte devido ao aumento
da atividade da glicerolkinase no tecido adiposo e aumento da síntese de
triglicerídeos (HIBUSE et al., 2005). Em suma, a presença da AQP-7 no epitélio
ruminal ainda tem que ser comprovada por meio de estudos de biologia
molecular, o que trará maior conhecimento do metabolismo do glicerol no rúmen.
2.4 A vaca leiteira no período de transição
A vaca leiteira periparturiente ou em período de transição tem sido

definida quando o animal se encontra entre três semanas antes e três semanas
depois da parição (GRUMMER, 1995). Durante esse período de transição da
gestação para a lactação, a vaca leiteira periparturiente é submetida a enormes
mudanças fisiológicas, além de estar sob o mais alto risco de incidência de
doenças metabólicas e infecciosas, comparado a qualquer outro período do seu
ciclo de vida (GOFF & HORST, 1997).
Durante os estágios finais da gestação o feto em crescimento mais o
útero são responsáveis pela utilização de respectivamente 46, 72 e 12% de
glucose, aminoácidos e acetato (BELL, 1995). A importância da demanda por
nutrientes no fim da gestação se torna ainda maior por haver redução acentuada
do CMS à medida que a vaca se aproxima do dia da parição (GREENFIELD et al.,
2000; INGVARTSEN & ANDERSEN, 2000; HAYIRLI et al., 2002; DOUGLAS et
al., 2006).
As adaptações metabólicas da vaca leiteira periparturiente para o
ajuste dessa menor oferta de nutrientes devido a redução do CMS estão no
aumento da taxa de gluconeogênese no fígado (principalmente a partir de lactato
e aminoácidos), menor utilização de glucose pelos tecidos periféricos, redução do
catabolismo de aminoácidos e aumento da mobilização e utilização de AGNE com
consequente aumento dos níveis de glicerol e corpos cetônicos na circulação
9
sanguínea (BELL, 1995), sendo que existe correlação negativa entre

concentração de AGNE no sangue e CMS (INGVARTSEN & ANDERSEN, 2000).
Por outro lado, o início da lactação em vacas leiteiras com potencial
de produção em torno de 10.000 kg/lactação impõe dramáticos aumentos nas
exigências de glucose, aminoácidos e ácidos graxos que não podem ser
atendidos somente pelo consumo da dieta, onde as estimativas das exigências de
glucose, aminoácidos, ácidos graxos e energia pela glândula mamária a partir dos
primeiros dias da lactação são respectivamente de 2,7; duas; 4,5 e três vezes
maiores comparado ao útero gestante no final da gestação (BELL, 1995). O
grande aumento na demanda por esses nutrientes pela glândula mamária desde
a primeira semana de lactação associado ao atraso entre o máximo CMS
comparado ao pico de produção de leite, que pode durar entre seis a oito
semanas (INGVARTSEN & ANDERSEN, 2000), fazem com que a vaca leiteira
periparturiente entre em BEN (BELL, 1995).
Além das mudanças fisiológicas ocorridas durante o período de
transição, existem também alterações na população de microrganismos do rúmen
de vacas leiteiras periparturientes. A mudança de uma dieta rica em fibra durante
o período seco para uma dieta rica em concentrado ao início da lactação causa
alterações na população de microrganismos e no epitélio ruminal. Dietas ricas em
concentrado favorecem o crescimento de bactérias que utilizam amido como
substrato, que por sua vez aumentam a produção de propionato e lactato,
enquanto que dietas ricas em fibra favorecem o crescimento de bactérias
celulolíticas e aumento na produção de metano (NRC, 2001). Por outro lado, os
tipos de produtos finais da fermentação ruminal influenciam o crescimento das
papilas ruminais, as quais são responsáveis pela absorção de AGCC (DIRKSEN
et al., 1985).
O aumento de concentrado na dieta e consequente aumento na
concentração de propionato no rúmen favorecem a elongação das papilas
ruminais, enquanto que dietas ricas em fibra exercem efeito oposto (NRC, 2001).
Estima-se que em torno de 50% da área de absorção do rúmen possa ser perdida
durante as sete primeiras semanas do período seco, sendo que a elongação das
papilas ruminais pode tardar várias semanas após a reintrodução de concentrado
na dieta a partir do início da lactação (DIRKSEN et al., 1985). Dessa forma, a
10
introdução de grãos na dieta imediatamente após o parto tem efeitos deletérios

sobre a saúde do rúmen, pois a produção de lactato aumenta antes do
restabelecimento das bactérias que utilizam lactato como substrato, sendo o
lactato considerado o mais potente rebaixador de pH ruminal (GOFF & HORST,
1997). Além disso, o atraso na elongação das papilas ruminais após a
reintrodução de grãos na dieta logo após o parto faz com que a capacidade de
absorção dos AGCC pelo epitélio ruminal se torne limitada (GOFF & HORST,
1997).
Conforme citado anteriormente, uma das ocorrências mais
observadas em vacas leiteiras periparturientes é a queda gradativa do CMS à
medida que a vaca se aproxima da parição (GREENFIELD et al., 2000;
INGVARTSEN & ANDERSEN, 2000; HAYIRLI et al., 2002; DOUGLAS et al.,
2006). Uma das razões para a redução do CMS de vacas leiteiras periparturientes
semanas antes da parição é a compressão física exercida pelo útero gestante no
rúmen. Entretanto, esse não é o único fator envolvido, pois no momento da
parição a cavidade abdominal é aliviada do fluido amniótico, feto e envoltórios
fetais em cerca de 70 kg numa vaca da raça Holandesa. O desaparecimento de
tal massa deveria permitir, por si só, um rápido aumento do CMS nos primeiros
dias pós-parto. Entretanto, observa-se que o aumento do CMS é lento, ao
contrário do rápido aumento que ocorre na produção de leite (GREENFIELD et
al., 2000; PATTON et al., 2004; DOUGLAS et al., 2006). Dessa forma, fica claro
que as alterações metabólicas e endócrinas têm maior participação na redução do
CMS do que a compressão física do útero gestante exercida no rúmen de vacas
leiteiras periparturientes, sendo que esse fenômeno não é restrito somente aos
ruminantes, mas também relatado em ratos (INGVARTSEN & ANDERSEN,
2000).
Como mencionado acima, durante o período de transição a vaca
leiteira é mais vulnerável a doenças metabólicas do que em qualquer outra fase
do seu ciclo de vida, sendo que as mais importantes são a cetose ou acetonemia
e lipidose hepática ou fígado gorduroso. Devido à alta ocorrência dessas doenças
em rebanhos leiteiros de alta produção, as quais acarretam prejuízos econômicos
consideráveis, ambas têm sido amplamente estudadas a fim de proporcionar
11
melhor entendimento das taxas de incidência, patogênese, diagnóstico,

tratamento e estratégias de prevenção.
A cetose é uma condição fisiológica caracterizada pela elevação
anormal dos corpos cetônicos ácido acetoacético, acetona e ABHB, tanto nos
tecidos corporais quanto na corrente sanguínea (BERGMAN, 1971). Por
definição, corpos cetônicos são oriundos da oxidação incompleta de ácidos
graxos, sendo que no caso da vaca leiteira são produzidos no epitélio ruminal,
glândula mamária e fígado, e podem servir como fonte de energia para o tecido
muscular como efeito da liberação de AGNE durante a lipólise (BERGMAN,
1971). Portanto, desconsiderando-se os efeitos negativos da cetose, os corpos
cetônicos são um importante constituinte do metabolismo energético da vaca
leiteira. Clinicamente, a doença se manifesta por meio da diminuição do apetite,
perda da massa corporal, redução da produção de leite e motilidade ruminal
(DUFFIELD, 2000).
O diagnóstico da cetose também pode ser realizado por meio da
medição de parâmetros sanguíneos, em que as vacas acometidas pela doença
podem apresentar níveis de glucose no sangue abaixo de 20 a 40 mg/dL e
concentrações sanguíneas de ABHB acima de 1,4 mmol/L (DUFFIELD, 2000).
Outra forma indireta de se realizar o diagnóstico da doença, ainda que apenas a
nível experimental, está na relação triglicerídeos:glicogênio no fígado. Pesquisas
têm demonstrado que quando essa relação está abaixo de 2:1 no momento da
parição, as vacas parecem ser resistentes ao aparecimento da cetose, assim
como a indução voluntária da doença ocorre quando a relação está acima de 2:1.
Dessa forma, conclui-se que o desenvolvimento da lipidose hepática precede a
cetose, e que a minimização da mobilização de tecido adiposo em vacas leiteiras
periparturientes é essencial na prevenção de cetose (DRACKLEY et al., 1992).
O fígado desempenha papel essencial no metabolismo lipídico em
vacas leiteiras periparturientes ao metabolizar ácidos graxos, os quais são
especialmente importantes durante o período de BEN que ocorre no estágio inicial
da lactação, quando há falta de sincronismo entre a demanda por energia pela
glândula mamária e o CMS, fazendo com que grandes quantidades de AGNE
sejam liberadas pelo tecido adiposo e removidas da circulação sanguínea pelo
fígado (HERDT, 2000). Uma vez no fígado, os AGNE podem ser oxidados,
12
convertidos a corpos cetônicos dentro da mitocôndria ou ainda serem

reesterificados e removidos do fígado na forma de VLDL (HERDT, 2000). No caso
da lipidose hepática, essa condição patológica resulta quando a taxa de formação
de triglicerídeos no fígado através da reesterificação de ácidos graxos excede a
taxa de oxidação desses ácidos graxos associada à taxa de exportação de
triglicerídeos na forma de VLDL para a circulação periférica (BELL, 1995). Apesar
da lipidose hepática serem comumente descritos como uma doença metabólica
que ocorre no pós-parto, os níveis de triglicerídeos no fígado podem aumentar
ainda no pré-parto (BERTICS et al., 1992; GREENFIELD et al., 2000), e a
concentração de AGNE no sangue pode aumentar em até duas vezes nas três
semanas que antecedem ao parto (BERTICS et al., 1992).
Apesar de não diretamente envolvidos na etiologia da lipidose
hepática, existem alguns fatores de risco, entre eles a nutrição, manejo e a
genética, que vêm sendo identificados como responsáveis no surgimento da
lipidose hepática. Um dos maiores fatores de risco é quando a vaca leiteira
adquire ECC ≥ 4 à época da parição devido ao excesso de energia suprida
durante o pré-parto, resultando em elevado acúmulo de triglicerídeos no fígado
(FRONK et al., 1980). Embora a obesidade no pré-parto seja um fator de risco
para o desenvolvimento da lipidose hepática, há vacas que não desenvolvem
essa doença metabólica no pós-parto mesmo com ECC ≥ 4 (SMITH et al., 1997),
indicando a necessidade de mais estudos nessa área em relação ao papel
potencial exercido pela genética na susceptibilidade da doença, ou mesmo se há
variação de indivíduos dentro da mesma raça.
Além da cetose, outras doenças na vaca leiteira periparturiente têm
sua frequência de ocorrência aumentada devido à lipidose hepática, tais como a
retenção dos envoltórios fetais, endometrite e mastite (MORROW et al., 1979),
assim como maior intervalo entre partos, atraso no primeiro cio fértil e redução na
taxa de prenhez (JORRITSMA et al., 2000).
Atualmente, recomendam-se várias estratégias nutricionais com o
objetivo de minimizar a incidência de doenças metabólicas na vaca leiteira
periparturiente, as quais incluem estudos de novos ingredientes e suas
implicações quanto aos níveis nutricionais, digestibilidade e potencial de consumo
dessas rações. De modo geral, busca-se diminuir a magnitude do BEN de forma
13
direta ou indireta para minimizar a redução do CMS que ocorre próximo a parição
e se estende nas primeiras semanas de lactação. Assim, justificam-se pesquisas
com novos alimentos energéticos, como no caso do glicerol, na tentativa de
reduzir o BEN na vaca leiteira periparturiente e, consequentemente, diminuir a
incidência de cetose e lipidose hepática em rebanhos leiteiros de alta produção.
Além disso, o maior entendimento da fisiologia da vaca leiteira periparturiente se
torna fundamental para avanços na nutrição e manejo desses animais durante o
período de transição, o que, em contrapartida, poderá gerar benefícios no bem-
estar animal e aumento da rentabilidade das fazendas produtoras de leite.
3 REFERÊNCIAS
AGRE, P. The aquaporin water channels. Proceedings of the American

Thoracic Society, v.3, p.5-13, 2006.
BELL, A. W. Regulation of organic nutrient metabolism during transition from late

pregnancy to early lactation. Journal of Animal Science, v.73, p.2804-2819,
1995.
BERGMAN, E. N. Hyperketonemia-ketogenesis and ketone body metabolism.

Journal of Dairy Science, v.54, p.936-948, 1971.
BERGNER, H.; KIJORA, C.; CERESNAKOVA, Z.; SZAKÁCS, J. In vitro studies on

glycerol transformation by rumen microorganisms. Arch Tierernahr, v.48, p.245-
256, 1995.
BERTICS, S. J.; GRUMMER, R. R.; CADORNIGA-VALINO, C.; STODDARD, E. E.

Effect of prepartum dry matter intake on liver triglyceride concentration and early
lactation. Journal of Dairy Science, v.75, p.1914-1922, 1992.
CZERKAWSKI, J. W.; BRECKENRIDGE, G. Fermentation of various glycolytic

intermediates and other compounds by rumen microorganisms, with particular
reference to methane production. British Journal of Nutrition, v.27, p.131-146,
1972.
DEFRAIN, J. M.; HIPPEN, A. R.; KALSCHEUR, K. F.; JARDON, P. W. Feeding

glycerol to transition dairy cows: effects on blood metabolites and lactation
performance. Journal of Dairy Science, v.87, p.4195-4206, 2004.
DEPARTMENT OF ECONOMIC AND SOCIAL AFFAIRS OF THE UNITED

NATIONS. 2009. Population newsletter. On-line. Disponível em:
14
http://www.un.org/esa/population/publications/popnews/Newsltr_87.pdf. Acesso
em 16 de junho de 2010.
DIRKSEN, G. U.; LIEBICH, H. G.; MAYER, E. Adaptive changes of the ruminal

mucosa and their functional and clinical significance. Bovine Practice, v.20,
p.116-120, 1985.
DONKIN, S. S; DOANE, P. Glycerol as a feed ingredient in dairy rations.

Proceedings… Tri-State Dairy Nutrition Conference, Fort Wayne, Indiana, United
States, 2007, p. 97-103. On-line. Disponível em:
http://tristatedairy.osu.edu/Proceedings%202007/Donkin%20paper.pdf. Acesso
em 24 de agosto de 2010.
DOUGLAS, G. N.; OVERTON, T. R.; BATEMAN II, H. G.; DANN, H. M.;

DRACKLEY, J. K. Prepartal plane of nutrition, regardless of dietary energy source,
affects periparturient metabolism and dry matter intake in Holstein cows. Journal
of Dairy Science, v.89, p.2141-2157, 2006.
DRACKLEY, J. K.; RICHARD, M. J.; BEITZ, D. C.; YOUNG, J. W. Metabolic

changes in dairy cows with ketonemia in response to feed restriction and dietary
1,3-butanediol. Journal of Dairy Science, v.75, 1622-1634, 1992.
DUFFIELD, T. Subclinical ketosis in lactating dairy cattle. Veterinary Clinics of

North America: Food Animal Practice, v.16, p.231-253, 2000.
FDA. Code of Federal Regulations, 21 CFR 582.1320, Title 21, Vol. 6,

21CFR582.1320, 2006. On-line. Disponível em:
http://edocket.access.gpo.gov/cfr_2002/aprqtr/21cfr582.1320.htm. Acesso em 17
de junho de 2009.
FISHER, L. J.; ERFLE, J. D.; LODGE, G. A.; SAUER, F. D. Effects of propylene

glycol or glycerol supplementation of the diet of dairy cows on feed intake, milk
yield and composition, and incidence of ketosis. Canadian Journal of Animal
Science, v.53, p.289-296, 1973.
FISHER, L. J.; ERFLE, J. D.; SAUER, F. D. Preliminary evaluation of the addition

of glucogenic materials to the rations of lactating cows. Canadian Journal of
Animal Science, v.51, p.721-727, 1971.
FRONK, T. J.; SCHULTZ, L. H.; HARDIE, A. R. Effect of dry period

overconditioning on subsequent metabolic disorders and performance of dairy
cows. Journal of Dairy Science, v.63, p.1080-1090, 1980.
GARTON, G. A.; LOUGH, A. K.; VIOQUE, E. Glyceride hydrolysis and glycerol

fermentation by sheep rumen contents. Journal of General Microbiology, v.25,
p.215-225, 1961.
15
GOFF, J. P.; HORST, R. L. Physiological changes at parturition and their

relationship to metabolic disorders. Journal of Dairy Science, v.80, p.1260-1268,
1997.
GOFF, J. P.; HORST, R. L. Oral glycerol as an aid in the treatment of ketosis/fatty

liver complex. Journal of Dairy Science, v. 84 (Suplemento 1), p.153, 2001.
GREENFIELD, R. B.; CECAVA, M. J.; JOHNSON, T. R.; DONKIN, S. S. Impact of

dietary protein amount and rumen undegradability on intake, peripartum liver
triglyceride, plasma metabolites, and milk production in transition dairy cattle.
GRUMMER, R. R. Impact of changes in organic nutrient metabolism on feeding

the transition dairy cow. Journal of Animal Science, v.73, p.2820-2833, 1995.
HAYIRLI, A.; GRUMMER, R. R.; NORDHEIM, E. V.; CRUMP, P. M. Animal and

dietary factors affecting feed intake during the prefresh transition period in
Holsteins. Journal of Dairy Science, v.85, p.3430-3443, 2002.
HIBUSE, T.; MAEDA, N.; FUNAHASHI, T.; YAMAMOTO, K.; NAGASAWA, A.;
MIZUNOYA, W.; KISHIDA, K.; INOUE, K.; KURIYAMA, H.; NAKAMURA, T.;
FUSHIKI, T.; KIHARA, S.; SHIMOMURA, I. Aquaporin 7 deficiency is associated
with development of obesity through activation of adipose glycerol kinase.
Proceedings of the National Academy of Sciences of the United States of
America, v.102, p.10993-10998, 2005.
HERDT, T. H. Ruminant adaptation to negative energy balance. Veterinary

Clinics of North America: Food Animal Practice, v.16, p.215-230, 2000.
INGVARTSEN, K. L.; ANDERSEN, J. B. Integration of metabolism and intake

regulation: a review focusing on periparturient animals. Journal of Dairy Science,
v.83, p.1573-1597, 2000.
JOHNS, A. T. Fermentation of glycerol in the rumen of sheep. New Zealand

Journal of Science and Technology, v.35, p.262-269, 1953.
JOHNSON, R. B. The treatment of ketosis with glycerol and propylene glycol.

Cornell Veterinarian, v. 44, p.6-21, 1955.
JORRITSMA, R.; JORRITSMA, H.; SCHUKKEN, Y. H; WENTINK, G H.

Relationships between fatty liver and fertility and some periparturient diseases in
commercial Dutch dairy herds. Theriogenology, v.54, p.1065-1074, 2000.
KHALILI, H.; VARVIKKO, T.; TOIVONEN, V.; HISSA, K.; SUVITIE, M. The effects
of added glycerol or unprotected free fatty acids or a combination of the two on
silage intake, milk production, rumen fermentation and diet digestibility in cows
given grass silage based diets. Agricultural and Food Science in Finland, v.6,
p.349-362, 1997.
16
KIJORA, C.; BERGNER, H.; GÖTZ, K. P.; BARTELT, J.; SZAKÁCS, J.; SOMMER,
A. Research note: investigation on the metabolism of glycerol in the rumen of
bulls. Archives of Animal Nutrition, v.51, p.341-348, 1998.
MACH, N.; BACH, A.; DEVANT, M. Effects of crude glycerin supplementation on

performance and meat quality of Holstein bulls fed high-concentrate diets. Journal
of Animal Science, v.87, p.632-638, 2009.
MORROW, D. A.; HILLMAN, D.; DADE, A. W.; KITCHEN, H. Clinical investigation

of a dairy herd with the fat cow syndrome. Journal of the Animal Veterinary
Medical Association, v.174, p.161-167, 1979.
NATIONAL RESEARCH COUNCIL. Nutrient Requirements of Dairy Cattle. 7ed.

Washington, D.C.:National Academy Press, 2001. 381 p.
PATTON, R. S.; SORENSON, C. E.; HIPPEN, A. R. Effects of dietary glucogenic

precursors and fat on feed intake and carbohydrate status of transition dairy cows.
RÉMOND, B.; SOUDAY, E.; JOUANY, J. P. In vitro and in vivo fermentation of

glycerol by rumen microbes. Animal Feed Science and Technology, v.41, p.121-
132, 1993.
SELLERS, R. S. Glycerin as a feed ingredient, official definition(s) and approvals.

Journal of Dairy Science, v.91 (Suplemento 1), p.392, 2008.
SMITH, T. R.; HIPPEN, A. R.; BEITZ, D. C.; YOUNG, J. W. Metabolic

characteristics of induced ketosis in normal and obese dairy cows. Journal of
Dairy Science, v.80, p.1569-1581, 1997.
THOMPSON, J. C.; HE, B. B. Characterization of crude glycerol from biodiesel

production from multiple feedstocks. Applied Engineering in Agriculture, v.22,
p.261-265, 2006.
VAN GERPEN, J.; SHANKS, B.; PRUSZKO, R.; CLEMENTS, D.; KNOTHE, G.
Biodiesel production technology. National Renewable Energy Laboratory, 2004.
On-line. Disponível em: http://www.nrel.gov/docs/fy04osti/36244.pdf. Acesso em
24 de julho de 2009.
WANG, C.; LIU, Q.; HUO, W. J.; YANG, W. Z.; DONG, K. H.; HUANG, Y. X.;
GUO, G. Effects of glycerol on rumen fermentation, urinary excretion of purine
derivatives and feed digestibility in steers. Livestock Science, v. 121, p.15-20,
2009.
WINTOUR, E. M.; HENRY, B. A. Glycerol transport: an additional target for obesity

therapy? Trends in Endocrinology and Metabolism, v.17, p.77-78, 2006.
WRIGHT, D. E. Fermentation of glycerol by rumen microorganisms. New Zealand

Journal of Agricultural Research, v.12, p.281-286, 1969.
17
CAPÍTULO 2 – REPLACING CORN WITH GLYCEROL IN DIETS FOR

TRANSITION DAIRY COWS
RESUMO
A expansão da indústria de biocombustíveis tem gerado aumento na
disponibilidade do glicerol, o qual pode ser utilizado como ingrediente na ração de
vacas leiteiras. Objetivou-se nesse estudo determinar os efeitos do glicerol sobre
o CMS, produção de leite, produção de AGCC no rúmen e parâmetros
sanguíneos em vacas periparturientes. Vinte e seis vacas multíparas da raça
Holandesa foram pareadas de acordo com o desempenho na lactação anterior e
data prevista de parição, e alimentadas com dietas contendo glicerol ou milho
grão úmido desde -28 até +56 DRDP. O glicerol foi incluído em 11,5 e 10,8% do
total da MS nas dietas pré e pós-parto, respectivamente. O CMS não foi alterado
(P>0,05) pela alimentação com glicerol tanto no pré-parto (14,9 vs. 14,6 kg/dia,
controle vs. glicerol) quanto no pós-parto (19,8 vs. 20,7 kg/dia, controle vs.
glicerol). A produção de leite não diferiu (P>0,05) entre os tratamentos (35,8 vs.
37 kg/dia, controle vs. glicerol) e também não houve resposta (P>0,05) do glicerol
sobre a composição do leite, nitrogênio ureico no leite, contagem de células
somáticas e balanço de energia. Houve tendência (P<0,15) do efeito de
tratamento × dias de lactação que favoreceu maior produção de leite nas vacas
alimentadas com glicerol. A concentração de glucose no sangue foi reduzida
(P<0,05; 59,1 vs. 53,4 mg/dL, controle vs. glicerol) e de ABHB no sangue foi
elevada (P<0,05; 0,58 vs. 0,82 mmol/L, controle vs. glicerol) nas vacas
alimentadas com glicerol durante o pré-parto. A concentração de AGNE no
sangue não diferiu (P>0,05) entre os grupos experimentais no pré-parto, e não
houve efeito (P>0,05) do glicerol sobre os parâmetros sanguíneos durante o pós-
parto. A concentração total de AGCC no rúmen (mmol/L) não diferiu (P>0,05)
entre os tratamentos, mas houve aumento (P<0,05) na proporção molar de
propionato (22,7 vs. 28,6%, controle vs. glicerol) e butirato (11,5 vs. 15,3%,
controle vs. glicerol) e redução (P<0,05) na proporção molar de acetato (61,5 vs.
51,5%, controle vs. glicerol) nas vacas alimentadas com glicerol. Os dados
obtidos na presente pesquisa indicaram que a substituição do milho grão úmido
pelo glicerol em dietas para vacas leiteiras periparturientes foi satisfatória.
18
Palvras-chave: biocombustíveis, energia, vacas periparturientes
ABSTRACT
Expansion of the biofuels industry has increased the availability of glycerol as an
alternative feed for dairy cows. The objective of this study was to determine the
effects of glycerol on feed intake, milk production, rumen SCFA, and blood
parameters in transition dairy cows. Twenty-six multiparous Holstein cows were
paired by previous lactation performance and expected calving date, and fed diets
containing either high moisture corn or glycerol from -28 to +56 DRTC. Glycerol
was included at 11.5 and 10.8% of the ration DM for the pre- and post-partum
diets respectively. Prepartum feed intake was not affected (P>0.05) by glycerol
feeding (14.9 vs. 14.6 kg/day, control vs. glycerol) nor did postpartum feed intake
differ (P>0.05; 19.8 vs. 20.7 kg/day, control vs. glycerol). Overall milk yield did not
differ (35.8 vs. 37 kg/d, control vs. glycerol) and there were no effects of glycerol
on milk composition, milk urea nitrogen (MUN), somatic cell count (SSC), and
energy balance (P>0.05). There was a tendency (P<0.15) for a treatment x days of
lactation effect that supported greater milk production for cows fed glycerol. Blood
glucose was reduced (P<0.05; 59.1 vs. 53.4 mg/dL, control vs. glycerol) and blood
BHBA was increased (P<0.05; 0.58 vs. 0.82 mmol/L, control vs. glycerol) in cows
fed glycerol during the prepartum period. Concentrations of blood NEFA did not
differ between the treatment groups (P>0.05) during the prepartum interval, and
there was no response (P>0.05) to glycerol for blood metabolites during the
postpartum period. Total rumen SCFA concentrations (mmol/L) did not differ
(P>0.05) between treatments, but molar proportion of rumen propionate and
butyrate were greater (P<0.05) for cows fed glycerol (22.7 vs. 28.6% propionate,
control vs. glycerol and 11.5 vs. 15.3% butyrate, control vs. glycerol) at the
expense of molar proportion of acetate (61.4 vs. 51.5%, control vs. glycerol). The
data indicate that glycerol is a suitable replacement for corn grain in diets for
transition dairy cows.
Keywords: biofuels, energy, periparturient cows

19
1 INTRODUCTION
Expansion of the biofuels industry around the world has provided

opportunities for alternative energy sources for livestock. In many cases the
amount of glycerol co-product has exceeded the capacities of the pharmaceutical
and chemical industries for refinement and further processing (THOMPSON & HE,
2006). Consequently, opportunities may exist to utilize glycerol as an energy
source in diets for dairy cattle, however little is known about the feeding value of
glycerol or tolerable levels of glycerol that can be fed across the lactation cycle.
According to the FDA (2006), glycerol is recognized as a safe
ingredient for use in animal feeds. The value of glycerol as a treatment for ketosis
was realized over 50 years ago (JOHNSON, 1955) and evaluation of glycerol as a
ketosis treatment was explored 40 years ago (FISHER et al., 1973) and then again
more recently (GOFF & HORST, 2001). Glycerol as a potential preventative for
ketosis has been fed at levels ranging from 0.25 kg/day (CHUNG et al., 2007) to
0.86 kg/day (DEFRAIN et al., 2004) in diets for transition dairy cows, and studies
conducted to date indicate a lack of difference in DMI, health incidences, and milk
production (DEFRAIN et al., 2004; CHUNG et al., 2007). However, other authors
have reported an increase in DMI between the fifth and ninth week of lactation,
greater milk yield, greater blood concentration of BHBA and lower blood
concentration of glucose and NEFA when transition dairy cows were
supplemented either with 300 or 500 mL/day of glycerol (BODARSKI et al., 2005).
Recent studies indicate that glycerol can replace corn grain in diets
for mid-lactation dairy cows to as much as 15% of the ration DM (DONKIN et al.,
2009). The effects of feeding greater levels of glycerol in place of corn in diets fed
to transition cows have not been explored. Combined results of previous studies
indicating a lack of negative effect of glycerol on intake or production suggest that
glycerol may be a suitable energy source in diets of transition dairy cows. We
hypothesized that glycerol could replace corn grain in diets for transition cows.
The objective of the present experiment was to determine the effects
of replacing high moisture corn with glycerol in diets for transition dairy cows on
feed intake, metabolic and rumen parameters and postpartum milk production.
20
2 MATERIALS AND METHODS
2.1 Cows and Dietary Treatments
Twenty-six multiparous Holstein cows were paired by previous

lactation performance and expected calving date, and randomly assigned to
receive either a diet containing corn silage, alfalfa haylage, hay, high moisture
corn, cotton seed hulls, soybean hulls, vitamins, and minerals (control) or a diet in
which high moisture ear corn was replaced with a mixture of glycerol and soybean
meal (glycerol). Refined glycerol (99.5% USP-grade glycerine; Pt Sumi Ashi
Oleochemicals Industry, Jakarta, Indonesia) was included at 11.5 and 10.8% of
the ration DM for the pre- and post-partum diets respectively (Table 1). Soybean
meal was added to the prepartum glycerol diet and a greater amount of the same
ingredient was added to the postpartum glycerol diet in order to adjust for the
protein removed from the diet with high moisture corn. The diets were adjusted
weekly to account for forage DM fluctuation.
Cows were housed in individual tie stalls at the Purdue Dairy
Research and Education Center and fed diets formulated to meet or exceed the
NRC (2001) guidelines for 600 kg dairy cattle from -28 through +56 DRTC. Diets
were fed as a TMR once daily between 06:30 to 07:30 h in amounts that ensured
ad libitum consumption and approximately 10 to 15% feed refusals. Cows were
milked twice daily at approximately 07:00 and 18:30 h.
Samples of TMR were collected weekly, dried in a forced-air oven for
72 h at 55ºC and grounded using a Wiley mill to pass a 1-mm screen. Composite
samples were formed monthly and analyzed by a commercial laboratory (Dairy
One, Ithaca, NY) for DM, CP, ADF, starch and minerals by wet chemistry following
AOAC (2000) procedures, and for NDF following the method of GOERING & VAN
SOEST (1970). Target composition of the pre- and post-partum diets were
respectively 14.9 and 18.3% CP, 39.4 and 30% NDF , 25.5 and 19.5% ADF, 1.60
and 1.62 Mcal of NEL/kg DM, 0.9 and 1% Ca, and 0.3 and 0.4% P, with an
anticipated intake of 11.6 and 22.6 kg/day. The study was conducted from
21
May/28/2009 to Oct/23/2009 and animal use and handling protocols were

approved by the Purdue Animal Care and Use Committee.
22
TABLE 1- Ingredient and nutrient composition of the pre- and post-partum

experimental diets.
Prepartum Postpartum
Item Control Glycerol Control Glycerol
Ingredient, % of DM
Corn silage 35.4 35.4 39.0 39.0
Alfalfa haylage 8.0 8.0 15.5 15.5
Grass hay 13.0 13.0 3.5 3.5
Wheat straw - - 1.5 1.5
Cotton seed hulls 6.0 6.0 - -
Soybean hulls 7.8 7.8 2.0 2.0
High moisture corn 14.0 - 12.5 -
Glycerol - 11.5 - 10.8
Soybean meal - 2.5 10.0 11.0
Megalac R1 - - 0.7 0.7
2
Protein blend - - 5.3 6.0
Supplement3,4 15.8 15.8 10.0 10.0
Chemical composition
DM, % 50.9 49.4 46.8 46.0
CP, % of DM 16.6 ± 1.00 16.6 ± 1.35 18.2 ± 0.83 18.7 ± 1.00
ADF, % of DM 22.9 ± 1.75 25.5 ± 1.79 19.5 ± 1.77 20.8 ± 2.32
NDF, % of DM 38.0 ± 1.18 42.2 ± 1.35 31.4 ± 2.71 34.2 ± 1.67
Starch, % of DM 22.6 ± 2.64 15.0 ± 1.22 26.7 ± 1.73 19.2 ± 1.12
NEL, Mcal/kg of DM 1.58 ± 0.02 1.61 ± 0.05 1.65 ± 0.02 1.61 ± 0.02
Ca, % of DM 1.09 ± 0.17 1.02 ± 0.12 1.11 ± 0.07 1.11 ± 0.20
P, % of DM 0.36 ± 0.02 0.34 ± 0.02 0.43 ± 0.02 0.40 ± 0.04
Mg, % of DM 0.39 ± 0.04 0.36 ± 0.02 0.36 ± 0.04 0.35 ± 0.02
K, % of DM 1.22 ± 0.05 1.29 ± 0.09 1.47 ± 0.11 1.44 ± 0.03
Na, % of DM 0.15 ± 0.01 0.15 ± 0.01 0.32 ± 0.01 0.32 ± 0.02
1
Church & Dwight Co., Princeton, NJ.
2
Contained 44% Aminoplus®, 3% Menhaden fish meal, 53% ProvAAL STD 5000.
3
Prepartum: contained 38.29% soybean meal, 25.65% Bio-Chlor® (Church & Dwight
Co., Princeton, NJ), 5.4% CaCO3, 2.16% dicalcium phosphate, 1.08% MgO, 1.08%
NaCl, 1.65% mineral/vitamin premix (16.11% Ca, 2.11% S, 31,505 mg/kg Zn, 8,036
mg/kg Cu, 26,020 mg/kg Mn, 140 mg/kg Se, 473 mg/kg Co, 284 mg/kg I, 1,440 kg IU/kg
vitamin A, 416 kg IU/kg vitamin D, 6,647 kg IU/vitamin E), 2.16% MgSO4, 5.08%
Megalac R (Church & Dwight Co., Princeton, NJ), 0.49% Niacinamide® (99.5% niacin),
2.62% yeast culture (Diamond V Mills, Cedar Rapids, IA), 1.8% vitamin E 20,000,
0.08% Rumensin 80®, 2.62% Omingen-AF (Prince-Agri Products, Quincy, IL), 1.08%
urea, 4.38% blood meal, 3.81% Aminoplus®, 0.57% Menhaden fish meal.
4
Postpartum: contained 25% dried molasses, 42.75% finely ground corn, 7.5% CaCO3,
5% dicalcium phosphate, 6.2% NaHCO3, 2% MgO, 2% DCAD plus, 0.5% potassium
magnesium sulfate, 2.5% NaCl, 2.025% mineral/vitamin premix (16.11% Ca, 2.11% S,
31,505 mg/kg Zn, 8,036 mg/kg Cu, 26,020 mg/kg Mn, 140 mg/kg Se, 473 mg/kg Co,
284 mg/kg I, 1,440 kg IU/kg vitamin A, 416 kg IU/kg vitamin D, 6,647 kg IU/vitamin E),
0.25% Niacinamide® (99.5% niacin), 2% yeast culture (Diamond V Mills, Cedar Rapids,
IA), 0.213% vitamin E 20,000, 0.062% Rumensin 80®, 2% Omingen-AF (Prince-Agri
Products, Quincy, IL).
23
2.2 Blood collection and analysis
Feed refusals were measured daily and feed intake was determined
by difference. Milk yield was recorded daily, milk samples from individual cows
were obtained weekly at 2 consecutive milkings, preserved with 2-bromo-2-
nitropropane-1,3-diol, and analyzed for protein, fat, lactose, somatic cell count
(SSC), and total solids by a commercial laboratory (Dairy One, Ithaca, NY). Body
weights (BW) and body conditions scores (BCS) were obtained at -28, +1, +28
and +56 DRTC. Body condition was scored by 2 trained individuals using a 5-point
scale (WILDMAN et al., 1982) and scores were averaged for each cow within day
of observation. Blood samples were collected at -28, -14, -7, -5, -3, -1, +1, +3, +14,
+28 and +56 DRTC via venipuncture of the coccygeal vein or artery into
evacuated tubes (Becton Dickinson, Franklin Lakes, NJ) and centrifuged at 3000 x
g for 15 minutes to separate plasma. Tubes (6 mL) containing potassium oxalate
and 4% sodium fluoride were used for collection of plasma for analyses of glucose
(glucose test, code no. 439-90901, Wako Chemicals USA, Inc., Richmond, VA),
NEFA (HR Series NEFA-HR 2, code no. 999-34691, 991-34891, 993-35191,
Wako Chemicals USA, Inc., Richmond, VA) and BHBA (procedure no. 2440,
Stanbio Laboratory Inc., Boerne, TX). Tubes (10 mL) containing sodium heparin
were used to collect plasma for analyses of glycerol (Free Glycerol Determination
Kit, Sigma-Aldrich, St. Louis, MO). Packed cell volume (PCV) was determined at
the time of blood sampling using heparinized capillary tubes and a
microhematocrit centrifuge. Blood samples were transported on ice to the
laboratory and plasma separated within 1.5 h after sampling. Plasma samples
were stored frozen at -20ºC until further analyses.
2.3 Rumen fluid sampling and analysis
Ruminal fluid from individual cows was collected by tube extending

into the rumen at 4 hours post feeding on day +56 relative to calving. The pH of
rumen fluid was immediately measured and 2.5 mL of 25% metaphosphoric acid
was added to 7.5 mL of rumen fluid and stored for SCFA analyses. A separate
24
sample was acidified with 1.5 mL of 6 N sulfuric acid per 23.5 mL of rumen fluid
and stored pending rumen glycerol analyses.
Samples were analyzed for SCFA by gas chromatography (Model
7890A, Agilent Technologies, Santa Clara, CA) and flame ionization detection, a
Nukol capillary column (30 m in length, 0.25 mm ID, 25 µ phase, Supelco, Inc.,
Bellefonte, PA) and 2-ethylbutyric acid as an internal standard. Oven temperature
program was 90°C to 150°C at 10°C/min, using helium carrier gas maintained at
80 psi. The injector port was set at 270°C and split injections were made at a 30:1
split ratio. The FID detector conditions were: temperature 300°C, air at 400
mL/min, and H2 at 40 mL/min with makeup helium flow at 26 mL/min.
Concentrations of SCFA were determined from comparison with a SCFA standard
solution (Sigma-Aldrich, St. Louis, MO). Rumen fluid samples were analyzed for
glycerol using the Free Glycerol Determination Kit (Sigma-Aldrich, St. Louis, MO).
2.4 Data Analysis
The data were analyzed using the MIXED procedure of SAS (1999).
The model accounted for the effects of treatment, time (as either DRTC or week of
experiment) and the interaction of treatment by time (as DRTC or week).
Compound symmetry covariance structure was used to evaluate variables
measured. Feed intake, milk production and blood parameters were analyzed with
repeated measures by DRTC. Average milk composition was analyzed using
repeated measures by week. Means were different if P<0.05 and tended to differ if
0.05 ≤ P ≤ 0.15. Values reported are least squares means and associated
standard errors.
3 RESULTS
Twenty-three cows completed the study. One cow from the control
group and one cow from the glycerol group were removed due to a displaced
abomasum and one cow from the control group was removed due to uterine
25
torsion that occurred 7 days prior to parturition. Data reported are for 12 cows in
the glycerol group and 11 cows in the control group.
Cows fed glycerol had similar (P>0.05) daily feed intake during the
pre- and post-partum periods as cows fed the control (Table 2). There was a
treatment × time response for intake during the prepartum period that appears to
be due to reduced intake for the glycerol fed cows during the first 10 days of the
trial (Figure 1) and there was a tendency (P=0.09) for a treatment × time effect on
feed intake over the entire experiment that favored cows fed glycerol. Average
milk yield, 4% fat-corrected milk yield, and feed efficiency were not altered
(P>0.05) by glycerol feeding (Table 2), but there was a tendency (P<0.15) for a
treatment × day effect on milk yield due to greater milk production for the glycerol
fed cows during the last 14 days of the trial (Figure 2). There were no main effects
(P>0.05) of diet on milk composition (Table 2) but there was a tendency (P<0.10)
for a treatment × week effect for milk protein yield due to greater production for the
glycerol fed cows during the last 14 d of the trial.
Initial BW was 688.5 ± 21.8 and 718.4 ± 20.9 kg for cows fed the
control and glycerol diets, respectively, and BW change did not differ (P>0.05)
between treatments at any time prior to or after calving. The BW mean for cows
fed control and glycerol at calving, +28 and +56 DRTC was 653.9, 593.1, 597.5 ±
21.8 kg and 689.8, 626.5, 630.7 ± 20.9 kg, respectively. Initial BCS was 2.64 ±
0.17 and 2.69 ± 0.19 for the control and glycerol groups, respectively, and there
were no differences (P>0.05) due to treatment during the pre- or post-partum
intervals (Table 2).
26
TABLE 2- Effect of glycerol on feed intake, milk production, feed efficiency, milk
composition, BW change and BCS change.
Treatment P-values
Item Control Glycerol SEM Trt1 Time2 Trt × time
DMI, kg/d
Prepartum 14.9 14.6 0.44 0.62 <0.05 <0.05
Postpartum 19.8 20.7 0.51 0.24 <0.05 0.47
Pre- and post-partum 18.2 18.6 0.43 0.47 <0.05 0.09
Milk yield, kg/d 35.8 37.0 1.45 0.56 <0.05 0.15
4% FCM3, kg/d 35.3 36.7 1.67 0.56 <0.05 0.88
Milk:DMI, kg/kg 1.84 1.83 0.05 0.85 <0.05 0.43
Milk Composition
Milk fat, % 4.01 4.00 0.12 0.95 <0.05 0.41
Milk fat, kg/d 1.41 1.46 0.08 0.65 0.20 0.80
Milk protein, % 2.93 2.90 0.05 0.65 <0.05 0.15
Milk protein, kg/d 1.02 1.06 0.04 0.52 0.09 0.09
Milk lactose, % 4.72 4.69 0.02 0.28 0.14 0.39
Milk lactose, kg/d 1.67 1.73 0.07 0.59 <0.05 0.75
Milk solids, % 12.55 12.44 0.15 0.60 <0.05 0.31
Milk solids, kg/d 4.42 4.57 0.20 0.60 <0.05 0.86
SCC, × 1,000 429 624 287.7 0.64 0.32 0.07
cells/mL
MUN, mg/dL 16.01 14.94 0.63 0.24 0.39 0.84
BW change, kg -73.7 -69.4 6.63 0.64 <0.05 0.95
BCS change -0.18 -0.19 0.03 0.80 <0.05 0.94
1
Treatment.
2
Days relative to calving for DMI, milk yield and feed efficiency and weeks relative
to calving for 4% FCM and milk composition.
3
4% Fat-corrected milk yield.
27
30
25
20
DMI (kg/d)
15
10
0
-28 -21 -14 -7 0 7 14 21 28 35 42 49 56
Days relative to calving
FIGURE 1- Dry matter intake (kg/day) of transition dairy cows fed control (solid
squares) or glycerol (open squares) during the pre- and post-partum
intervals.
45
40
35
Milk yield (kg/d)
30
25
20
15
10
5
0
0 7 14 21 28 35 42 49 56
FIGURE 2- Milk yield (kg/day) of transition dairy cows fed control (solid squares)
or glycerol (open squares).
28
Cows fed glycerol had lower (P<0.05) mean blood glucose

concentrations for the entire experimental period. When the analysis was
separated to pre- and post-calving intervals blood glucose was reduced (P<0.05)
during the prepartum interval, and there was a tendency (P=0.10) of reduction
during the postpartum interval for cows fed glycerol (Table 3, Figure 3). There was
also a tendency (P<0.10) of reduction of blood NEFA during the combined pre-
and post-partum intervals for cows fed glycerol (Table 3, Figure 3). Overall blood
BHBA was increased (P<0.05) for cows fed glycerol and there were no time ×
treatment effects (Table 3, Figure 3).
Ruminal pH and glycerol concentration in the rumen were not
influenced (P>0.05) by treatments (Table 4). Likewise, total rumen SCFA did not
differ (P>0.05) between treatments, but rumen propionate, butyrate and valerate
(expressed as percentages of total SCFA) were greater (P<0.05) for cows fed
glycerol, at the expense of acetate and isobutyrate. Cows fed glycerol decreased
(P<0.05) the acetate to propionate ratio (Table 4).
29
TABLE 3- Effect of glycerol on plasma metabolites.

Treatment P-values
Item Control Glycerol SEM Trt1 Time2 Trt × time
PCV%3
Prepartum4 33.0 32.3 0.58 0.37 <0.05 <0.05
Postpartum5 31.6 31.7 0.45 0.90 <0.05 0.84
Pre- and post-partum6 32.5 32.1 0.47 0.60 <0.05 0.21
Glucose (mg/dL)
Prepartum 59.1 53.4 1.61 <0.05 <0.05 0.11
Postpartum 53.6 50.8 1.15 0.10 <0.05 0.63
Pre- and post-partum 56.5 52.5 1.00 <0.05 <0.05 0.21
NEFA (mmol/L)
Prepartum 0.22 0.18 0.02 0.22 <0.05 0.13
Postpartum 0.50 0.41 0.04 0.14 <0.05 0.62
BHBA (mmol/L)
Prepartum 0.58 0.82 0.04 <0.05 0.07 0.82
Postpartum 0.77 0.94 0.07 0.09 <0.05 0.49
Pre- and post-partum 0.66 0.87 0.04 <0.05 <0.05 0.75
Glycerol (mg/dL)
Prepartum 0.64 0.82 0.14 0.32 0.05 0.75
Postpartum 0.91 0.61 0.17 0.22 0.05 0.55
1
Treatment.
2
Days relative to calving.
3
Packed cell volume.
4
Data for -28 through calving.
5
Data for +1 through +56 DRTC.
6
Combined data for -28 through +56 DRTC.
30
80
70
Blood glucose (mg/dL)
60
50
40
30
20
10
0
-28 -21 -14 -7 0 7 14 21 28 35 42 49 56
0.8
0.7
Blood NEFA (mmol/L)
0.6
0.5
0.4
0.3
0.2
0.1
0
-28 -21 -14 -7 0 7 14 21 28 35 42 49 56
1.4
1.2
Blood BHBA (mmol/L)
0.8
0.6
0.4
0.2
0
-28 -21 -14 -7 0 7 14 21 28 35 42 49 56
FIGURE 3- Effect of glycerol on glucose, NEFA and BHBA in plasma for cows fed
control (solid squares) or glycerol (open squares).
31
TABLE 4- Effect of glycerol on rumen parameters.

Item Control Glycerol SEM P-values
pH 7.08 6.94 0.13 0.44
Glycerol (mg/mL) 0.08 0.07 0.01 0.81
Total SCFA, mmol/L 82.3 85.9 6.52 0.70
Individual SCFA, % of total
SCFA
Acetate (A) 61.4 51.5 1.08 <0.05
Propionate (P) 22.7 28.6 1.44 <0.05
Isobutyrate 1.11 0.92 0.05 <0.05
Butyrate 11.5 15.3 0.71 <0.05
Isovalerate 1.84 1.64 0.07 0.07
Valerate 1.48 2.06 0.06 <0.05
A:P ratio 2.76 1.88 0.14 <0.05
Rumen fluid was sampled 4 hours post feeding from individual cows at +56 DRTC.
4 DISCUSSION
Ongoing increases in the production of biofuels has also led to a

supply of glycerol that may not be absorbed by the pharmaceutical and chemical
industries, therefore creating an opportunity to use glycerol as an energy source in
rations fed to dairy cattle. Previous research examining the potential for glycerol as
a feed for dairy cattle indicates that as much as 15% of the ration DM can be fed
as glycerol to mid-lactation dairy cows (DONKIN et al., 2009). The onset of
lactation poses considerable challenges to the dairy cow as feed intake often lags
energy needs, and feeding management strategies must consider the impact of
feeds on intake and health status. Previous research indicates a mixed response
to glycerol feeding in transition cows (DEFRAIN et al., 2004; BODARSKI et al.,
2005; CHUNG et al., 2007). Top dressing with glycerol up to 7.6% of ration DM
resulted in a reduction in feed intake during the prepartum interval, but this effect
disappeared after parturition (DEFRAIN et al., 2004). In contrast, feeding 300 or
500 mL of glycerol from -14 to +70 DRTC had no effect on feed intake during the
prepartum interval but increased feed intake between +35 and +63 DRTC
32
(BODARSKI et al., 2005). A recent study (CHUNG et al., 2007) to evaluate the
effects of feeding a dry glycerin product from calving to +21 days postpartum
indicated no effect of glycerin on feed intake. Liquid glycerol from biodiesel
processing may contain a variety of contaminants including water, salt and
methanol (THOMPSON & HE, 2006) and the crude biodiesel glycerol used by
DEFRAIN et al (2004) contained 80.2% glycerol, 11.5% salt, 6.6% water, and
1.3% methanol. Feeding crude glycerol containing 85.7% glycerol, 8.6% water,
5.5% salt and 0.09% methanol to 12% of ration DM had no detrimental effect on
intake and performance of beef cattle (MACH et al., 2009). The presence of
contaminants in any feeds, including glycerol, clouds the interpretation feeding
experiments therefore we utilized food-grade glycerol in the present experiment to
eliminate any possible confounding effects of impurities on DMI, production or
health.
The present data indicate that glycerol can replace corn in the diet of
transition cows to at least 11.5% of ration DM without compromising intake during
the pre- or post-partum intervals. Previous studies indicate that feeding 15%
glycerol of ration DM to mid-lactation dairy cows may be accompanied by a
transient reduction in feed intake but a similar temporary reduction in intake was
not observed when 10% glycerol was fed (DONKIN et al., 2009). Although there is
debate regarding optimal levels of fermentable energy for transition dairy cows
(OVERTON & WALDRON, 2004) the data reported here indicates that glycerol
provides an alternative energy source for use in transition cow rations that does
not negatively impact intake. Additional studies are needed to determine the
effects of glycerol at levels in excess of 11.5% of ration DM in diets for transition
cows.
Milk production was similar for cows fed the glycerol and the control
diet. There was a tendency for a treatment × day response on milk production that
favored glycerol feeding. Glycerol resulted in a numerical increase in milk
production during the fourth, fifth and seventh week of lactation. The lack of effect
of glycerol on average milk production for the 56-day experiment agrees with
previous studies in which transition cows were fed glycerol (DEFRAIN et al., 2004;
CHUNG et al., 2007) but differs from reports where 300 or 500 mL of glycerol were
supplemented as a topdress fed to dairy cows once daily from 2 weeks before
33
calving through 10 weeks of lactation. In the latter experiment glycerol increased

milk production approximately 4.7 kg/d compared with cows fed a control diet
(BODARSKI et al., 2005). We also examined milk production for 56 days after
cows were released from the present study but failed to observe any similar
carryover effects despite a treatment × day effect on milk production that suggests
a tendency for increased milk production during the last 14 days of the trial for
cows fed glycerol.
There has been interest on the potential of glycerol in altering water
intake and hydration state in different species, although the results found in the
literature have been contradictory. In humans, the ingestion of a large volume of
fluid (39.2 or 51.1 mL/kg BW/day) with glycerol (2.9 to 3.1 g/kg BW/day) resulted
in a smaller urine volume voided, although no significant changes were observed
in PCV% (KOENIGSBERG et al., 1995). Furthermore, previous studies (LYONS et
al., 1990; MONTNER et al., 1996), as well as the reviews from ROBERGS &
GRIFFIN (1998) and WAGNER (1999) described improvements in hemodynamic
(lower heart rate) and thermoregulatory responses through greater sweating rate
and lower core temperature during endurance exercise with glycerol
hyperhydration in humans. However, LATZKA et al. (1997; 1998) subjected
humans to both compensable and uncompensable exercise heat stress and failed
to demonstrate any benefits of glycerol hyperhydration in hemodynamic or
thermoregulatory responses, as long as euhydration was maintained by replacing
fluid (sweat) losses by regular intake of water during the exercise test. In
euhydrated horses, the nasogastric administration of 1 g/kg BW glycerol in a total
volume of 8 L of 0.9% NaCl solution resulted in an additional intake of 5.2 L of
water (SCHOTT et al., 2001), but earlier studies also failed to reveal any benefits
of glycerol supplementation in stimulating voluntary water intake and maintaining
euhydration during the exercise test in horses (DÜSTERDIECK et al., 1999;
SCHOTT et al., 1999).
In midlactating cows, BW change increased with increasing level of
glycerol feeding, but this effect was not translated in a corresponding change in
BCS, which was possibly due to differences in water intake and hydration state of
cows fed glycerol (DONKIN et al., 2009). Measurements of water intake were not
performed in the present study; however; the lack of response of glycerol on
34
PCV% is a clear indication that hydration state of cows was not influenced by
glycerol. Besides, glycerol influenced neither BW nor BCS change, which is in
agreement with other studies with transition dairy cows (DEFRAIN et al., 2004;
BODARSKI et al., 2005; CHUNG et al., 2007). Nevertheless, future studies should
still investigate the possibility of changes in the hydration state of transition dairy
cows by glycerol feeding, especially in experiments performed during heat stress
or in the tropical zone.
Cows fed glycerol had a tendency for reduced blood NEFA
concentrations across the entire study but had increased blood BHBA. The former
suggests either an increased capacity to metabolize NEFA or reduced release of
NEFA from adipose tissue. The observed increases in BHBA suggest increased
appearance of BHBA from either partial oxidation of NEFA, increased production
by rumen epithelium in the metabolism of butyrate, or both that is in excess of the
capacity of extrahepatic tissue to utilize ketones. An increase fatty acid clearance
by liver in response to glycerol in diets for transition cows would support a
reduction in liver lipid at calving. The reduction in NEFA observed in the present
study is consistent with the early postpartum response to glycerol observed by
BODARSKI et al. (2005), but is in conflict with the lack of response in NEFA
observed in other experiments in which using glycerol was fed as a topdress
(CHUNG et al., 2007) or included in drinking water (OSBORNE et al., 2009).
Increased BHBA has also been observed for transition cows supplemented with
500 mL of glycerol during the third week of lactation (BORDARSKI et al., 2005)
and previous experiments indicated that both feeding and drenching with glycerol
increases ruminal butyrate (LINKE et al., 2004). The present data indicate a shift in
rumen metabolism resulting in greater butyrate concentrations.
The root cause of increased BHBA in plasma with glycerol feeding
can not be determined without additional physiological measures given that
ruminal butyrate concentrations do not accurately reflect production rates
(FIRKINS et al., 2006). However due to the role of butyrate in ketogenesis
(BERGMAN, 1990) caution should be exercised when using glycerol in transition
cow diets as an increase in BHBA in blood can be accompanied by reduced feed
intake and undesirable metabolic consequences. However it should be noted that
none of the cows fed glycerol in the current study had BHBA concentrations in
35
excess of 1.4 mmol/L, which has been described as a threshold value for
subclinical ketosis in early-lactation cows (DUFFIELD, 2000). Furthermore there
were no negative effects of glycerol on feed intake or milk production detected in
this study.
Cows fed glycerol experienced a reduction in blood glucose
concentration during the prepartum interval which persisted through +14 DRTC.
These data differ from previous work indicating that blood glucose is increased
when glycerol is drenched (GOFF & HORST, 2001; LINKE et al., 2004) or fed to
mid-lactation cows (DONKIN et al., 2009), but a similar tendency for reduced
blood glucose in response to glycerol feeding has been observed previously for
transition cows (DEFRAIN et al., 2004). Because blood glucose is a function of the
rates of glucose absorption, glucose production, and glucose utilization it is difficult
to determine the mechanism of glycerol action without additional physiological
measures. However the lack of overall differences in milk production, BW change
and BCS change do not support differences in glucose utilization.
Gluconeogenesis from propionate is impaired by the presence of butyrate
(AIELLO & ARMENTANO, 1987). The capacity of the ruminal epithelium to
metabolize butyrate is limited. If butyrate absorption exceeds the metabolic
capacity, it affects rumen epithelial and hepatic nutrient metabolism, and therefore
nutrient supply to peripheral tissues (KRISTENSEN & HARMON, 2004). An
increase in rumen butyrate concentrations in the present study, if linked to butyrate
absorption in excess of rumen epithelial metabolism could increase the supply of
butyrate to liver and consequently reduce hepatic gluconeogenesis from
propionate. However the metabolic capacity of the rumen epithelium to metabolize
butyrate to BHBA can also be adapted by increasing butyrate loads (SEHESTED
et al., 1999) therefore feeding glycerol might induce an adaptation of the metabolic
capacity of the ruminal epithelium.
The present data indicates that the greatest reduction in plasma
glucose occurred within 14 days after the initiation of glycerol feeding. A period of
four to six weeks has been suggested to adapt the absorptive capacity of the
rumen to increased grain in the diet (DIRKSEN et al., 1985). Adaptation of rumen
epithelium to enhance butyrate oxidation and reduce flux of butyrate to liver is
consistent with a recovery of glucose to control values at calving. Additional
36
experiments are necessary to determine the impact of glycerol feeding on rumen

butyrate production rates, and postabsorptive butyrate metabolism. Moreover, the
data from the present experiment indicate that the main fate of glycerol in the
rumen is the conversion to propionate and butyrate at the expense of acetate,
which is in agreement with previous authors (CZERKAWSKI & BRECKENRIDGE,
1972; KHALILI et al., 1997; WANG et al., 2009) and refute the possibility that
glycerol may be directly absorbed across the rumen epithelium (RÉMOND et al.,
1993; KIJORA et al., 1998), since there was no response to glycerol feeding on
blood glycerol.
4 CONCLUSIONS
Glycerol can be recommended to replace high moisture corn in

rations fed to transition dairy cows at 11.5 and 10.8% of the pre- and post-partum
rations DM, respectively, based on the lack of detrimental effects on feed intake,
milk production or milk composition. Reduced glucose and elevated BHBA in
blood suggest that additional information is needed to fully assess the metabolic
status of transition cows fed glycerol.
5 REFERENCES
AIELLO, R. J.; ARMENTANO, L. E. Effects of volatile fatty acids on propionate

metabolism and gluconeogenesis in caprine hepatocytes. Journal of Dairy
Science, v.70, p.2504-2510, 1987.
AOAC. Association of Official Analytical Chemistry. Official methods of analysis.

17th ed. AOAC Int., Gaithersburg, MD, 2000, 1298 p.
BERGMAN, E. N. Energy contributions of volatile fatty acids from the

gastrointestinal tract in various species. Physiological Reviews, v.70, p.567-590,
1990.
BODARSKI, R.; WERTELECKI, T.; BOMMER, F.; GOSIEWSKI, S. The changes

of metabolic status and lactation performance in dairy cows under feeding TMR
with glycerin (glycerol) supplement at periparturient period. Electronic Journal of
Polish Agricultural Universities, v.8, p.1-9, 2005.
37
CHUNG, Y. H.; RICO, D. E.; MARTINEZ, C. M.; CASSIDY, T. W.; NOIROT, V.;
AMES, A.; VARGA, G A. Effects of feeding dry glycerin to early postpartum
Holstein dairy cows on lactational performance and metabolic profiles. Journal of
Dairy Science, v.90, p.5682-5691, 2007.
CZERKAWSKI, J. W.; BRECKENRIDGE, G. Fermentation of various glycolytic

intermediates and other compounds by rumen microorganisms, with particular
reference to methane production. British Journal of Nutrition, v.27, p.131-146,
1972.
DEFRAIN, J. M.; HIPPEN, A. R.; KALSCHEUR, K. F.; JARDON, P. W. Feeding

glycerol to transition dairy cows: effects on blood metabolites and lactation
performance. Journal of Dairy Science, v.87, p.4195-4206, 2004.
DIRKSEN, G. U.; LIEBICH, H. G.; MAYER, E. Adaptive changes of the ruminal

mucosa and their functional and clinical significance. Bovine Practice, v.20,
p.116-120, 1985.
DONKIN, S. S.; KOSER, S. L.; WHITE, H. M.; DOANE, P. H.; CECAVA, M. J.

Feeding value of glycerol as a replacement for corn grain in rations fed to lactating
dairy cows. Journal of Dairy Science, v.92, p.5111-5119, 2009.
DUFFIELD, T. Subclinical ketosis in lactating dairy cattle. Veterinary Clinics of

North America: Food Animal Practice, v.16, p.231-253, 2000.
DÜSTERDIECK, K. F.; SCHOTT, H. C.; EBERHART, S. W.; WOODY, K. A.;

COENEN, M. Electrolyte and glycerol supplementation improve water intake by
horses performing a simulated 60 km endurance ride. Equine Veterinary
Journal, v.30, p.418-424, 1999.
FDA. Code of Federal Regulations, 21 CFR 582.1320, Title 21, Vol. 6,

21CFR582.1320, 2006. On-line. Disponível em:
http://edocket.access.gpo.gov/cfr_2002/aprqtr/21cfr582.1320.htm. Acesso em 17
de junho de 2009.
FIRKINS, J. L.; HRISTOV, A. N.; HALL, M. B.; VARGA, G. A.; ST-PIERRE, N. R.

Integration of ruminal metabolism in dairy cattle. Journal of Dairy Science, v.89,
p.31-51, 2006.
FISHER, L. J.; ERFLE, J. D.; LODGE, G. A.; SAUER, F. D. Effects of propylene

glycol or glycerol supplementation of the diet of dairy cows on feed intake, milk
yield and composition, and incidence of ketosis. Canadian Journal of Animal
Science, v.53, p.289-296, 1973.
GOERING, H. K.; VAN SOEST, P. J. Forage fiber analysis (Apparatus,

Reagents, Procedures and Some Applications). Agricultural Handbook nº 379,
ARS-USDA, Washington, D. C., 1970, 19 p.
38
GOFF, J. P.; HORST, R. L. Oral glycerol as an aid in the treatment of ketosis/fatty

liver complex. Journal of Dairy Science, v. 84 (Suplemento 1), p.153, 2001.
JOHNSON, R. B. The treatment of ketosis with glycerol and propylene glycol.

Cornell Veterinarian, v. 44, p.6-21, 1955.
KHALILI, H.; VARVIKKO, T.; TOIVONEN, V.; HISSA, K.; SUVITIE, M. The effects
of added glycerol or unprotected free fatty acids or a combination of the two on
silage intake, milk production, rumen fermentation and diet digestibility in cows
given grass silage based diets. Agricultural and Food Science in Finland, v.6,
p.349-362, 1997.
KOENIGSBERG, P. S.; MARTIN, K. K.; HLAVA, H. R.; RIEDESEL, M. L.

Sustained hyperhydration with glycerol ingestion. Life Sciences, v.57, p.645-653,
1995.
KRISTENSEN, N. B.; HARMON, D. L. Effect of increasing ruminal butyrate

absorption on splanchnic metabolism of volatile fatty acids absorbed from the
washed reticulorumen of steers. Journal of Animal Science, v.82, p.3549-3559,
2004.
LATZKA, W. A.; SAWKA, M. N.; MONTAIN, S. J.; SKRINAR, G. S.; FIELDING, R.

A.; MATOTT, R. P.; PANDOLF, K. B. Hyperhydration: thermoregulatory effects
during compensable exercise-heat stress. Journal of Applied Physiology, v.83,
p.860-866, 1997.
LATZKA, W. A.; SAWKA, M. N.; MONTAIN, S. J.; SKRINAR, G. S.; FIELDING, R.

A.; MATOTT, R. P.; PANDOLF, K. B. Hyperhydration: tolerance and
cardiovascular effects during uncompensable exercise-heat stress. Journal of
Applied Physiology, v.84, p.1858-1864, 1998.
LINKE, P. L.; DEFRAIN, J. M.; HIPPEN, A. R.; JARDON, P. W. Ruminal and

plasma responses in dairy cows to drenching or feeding glycerol. Journal of Dairy
Science, v.87 (Suplemento 1), p.343, 2004.
LYONS, T. P.; RIEDESEL, M. L.; MEULI, L. E.; CHICK, T. W. Effects of glycerol-

induced hyperhydration prior to exercise in the heat on sweating and core
temperature. Medicine & Science in Sports & Exercise, v.22, p.477-483, 1990.
MACH, N.; BACH, A.; DEVANT, M. Effects of crude glycerin supplementation on

performance and meat quality of Holstein bulls fed high-concentrate diets. Journal
of Animal Science, v.87, p.632-638, 2009.
39
MONTNER, P.; STARK, D. M.; RIEDESEL, M. L.; MURATA, G.; ROBERGS, R.;
TIMMS, M.; CHICK, T. W. Pre-exercise glycerol hydration improves cycling
endurance time. International Journal of Sports Medicine, v.17, p.27-33, 1996.
NATIONAL RESEARCH COUNCIL. Nutrient Requirements of Dairy Cattle. 7ed.

Washington, D.C.:National Academy Press, 2001. 381 p.
OSBORNE, V. R.; ODONGO, N. E.; CANT, J. P.; SWANSON, K. C.; MCBRIDE,

B. W. Effects of supplementing glycerol and soybean oil in drinking water on feed
and water intake, energy balance, and production performance of periparturient
OVERTON, T. R.; WALRON, M. R. Nutritional management of transition dairy

cows: strategies to optimize metabolic health. Journal of Dairy Science, v.87,
p.105-119, 2004.
RÉMOND, B.; SOUDAY, E.; JOUANY, J. P. In vitro and in vivo fermentation of

glycerol by rumen microbes. Animal Feed Science and Technology, v.41, p.121-
132, 1993.
ROBERGS, R. A.; GRIFFIN, S. E. Glycerol: biochemistry, pharmacokinetics and

clinical and practical applications. Sports Medicine, v.26, p.145-167, 1998.
SAS INSTITUTE. SAS/STAT User’s Guide: Statistics, Version 9.2 Edition. SAS
Inst. Inc., Cary, NC, 1999.
SCHOTT, H. C.; DÜSTERDIECK, K. F.; EBERHART, S. W.; WOODY, K. A.;

REFSAL, K. R.; COENEN, M. Effect of electrolyte and glycerol supplementation
on recovery from endurance exercise. Equine Veterinary Journal, v.30, p.384-
393, 1999.
SCHOTT, H. C.; PATTERSON, K. S.; EBERHART, S. W. Glycerol hyperhydration

in resting horses. Veterinary Journal, v.161, p.194-204, 2001.
SEHESTED, J.; DIERNAES, L.; MØLLER, P. D.; SKADHAUGE, E. Ruminal

transport and metabolism of short-chain fatty acids (scfa) in vitro: effect of scfa
chain length and pH. Comparative Biochemistry and Physiology Part A, v.123,
p.359-368, 1999.
THOMPSON, J. C.; HE, B. B. Characterization of crude glycerol from biodiesel

p.261-265, 2006.
WAGNER, D. Hyperhydrating with glycerol: implications for athletic performance.

Journal of the American Dietetic Association, v.99, p.207-212, 1999.
WANG, C.; LIU, Q.; HUO, W. J.; YANG, W. Z.; DONG, K. H.; HUANG, Y. X.;
GUO, G. Effects of glycerol on rumen fermentation, urinary excretion of purine
40
derivatives and feed digestibility in steers. Livestock Science, v. 121, p.15-20,

2009.
WILDMAN, E. E.; JONES, G. M.; WAGNER, P. E.; BOMAN, R. L.; TROUTT, H. F.;
LESCH, T. N. A dairy cow body condition scoring system and its relationship to
selected production characteristics. Journal of Dairy Science, v.65, p.495-501,
1982.
41
CAPÍTULO 3 – FEEDING BEHAVIORS OF TRANSITION DAIRY COWS FED

GLYCEROL AS A REPLACEMENT FOR CORN.
RESUMO
O consumo seletivo de partículas da ração é um comportamento natural de vacas
leiteiras que pode resultar em inconsistências no valor nutritivo da dieta ao longo
do dia. Objetivou-se nesse estudo determinar os efeitos da substituição do milho
grão úmido pelo glicerol sobre o consumo seletivo de partículas da dieta e
comportamento ingestivo de vacas leiteiras periparturientes. Vinte e seis vacas
multíparas da raça Holandesa foram pareadas de acordo com o desempenho na
lactação anterior e data prevista de parição, e alimentadas com dietas contendo
glicerol ou milho grão úmido desde -28 até +56 DRDP. O glicerol foi incluído em
11,5 e 10,8% do total da MS nas dietas pré e pós-parto, respectivamente.
Filmaram-se as vacas ininterruptamente por 24 horas aos -17, -10, +8, +15 e +50
DRDP. Avaliou-se o comportamento ingestivo das vacas durante intervalos de
uma hora entre 0-1, 1-2, 5,5-6,5 e 11-12 horas pós-alimentação. Vinte e quatro
horas após cada sessão de vídeo, determinou-se a taxa do CMS (g MS/h) entre
0-4, 4-8, 8-12 e 12-24 horas pós- alimentação, por meio da divisão do CMS de
cada intervalo pela sua correspondente duração em horas. Determinou-se o perfil
de distribuição de partículas da dieta no momento da alimentação e em cada
intervalo pós-alimentação utilizando-se o PSPS para originar partículas longas
(>19 mm), médias (<19, >8 mm), curtas (<8, >1,18 mm) e muito curtas (<1,18
mm). Determinou-se então o consumo seletivo mensurando-se a distribuição do
tamanho de partículas da dieta consumida entre 0-4, 4-8, 8-12 e 12-24 horas pós-
alimentação. Na ração oferecida durante o pré-parto, a adição de glicerol
aumentou (P<0,05) a porcentagem de MS retida das partículas longas (>19 mm)
e reduziu (P<0,05) a porcentagem de MS retida das partículas curtas (<8, >1,18
mm) e muito curtas (<1,18 mm), mas não alterou (P>0,05) a porcentagem de MS
retida das partículas médias (<19, >8 mm). O CMS não diferiu (P>0,05) entre as
dietas e foi de 14,7 ± 0,4 e 20,2 ± 0,5 kg/dia no pré e pós-parto, respectivamente.
Durante o pré-parto, houve aumento (P<0,05) na taxa do CMS (94,2 vs. 144,4 g
MS/h; controle vs. glicerol) e no consumo preferencial (9,2 vs. 17,8%; controle vs.
glicerol) de partículas longas na dieta com glicerol, porém houve redução (P<0,05)
42
na taxa do CMS de partículas curtas (383,8 vs. 332,5 g MS/h; controle vs. glicerol)
e muito curtas (173,9 vs. 129,8 g MS/h; controle vs. glicerol) e aumento (P<0,05)
na rejeição de partículas curtas (42 vs. 37,3%; controle vs. glicerol) e muito curtas
(17,9 vs. 13,6%; controle vs. glicerol). Não houve efeito (P>0,05) dos tratamentos
sobre o consumo seletivo de partículas da dieta no pós-parto e também sobre o
comportamento ingestivo durante todo o período experimental. Concluiu-se que
embora não houve diferença no CMS entre os tratamentos, o glicerol
proporcionou aumento pela preferência por partículas longas (>19 mm) durante a
fase pré-parto, o que pode trazer benefícios para a saúde ruminal.
Palavras-chave: consumo seletivo, preferência, vacas periparturientes
ABSTRACT
Feed sorting is a natural behavior of dairy cows that can result in inconsistencies
in nutritive value of a TMR. The objective of this study was to determine the effects
of replacing high moisture corn with glycerol on feed sorting and feeding behavior
of transition dairy cows. Twenty-six Holstein multiparous cows were paired by
previous lactation performance and expected calving date, and fed diets
containing either glycerol or high moisture corn from -28 through +56 DRTC.
Glycerol was included at 11.5 and 10.8% of the ration DM for the pre- and post-
partum diets, respectively. Cow activity was continuously videotaped for 24 hours
on -17, -10, +8, +15 and +50 DRTC. Feeding behavior was evaluated during 1-h
intervals between 0-1, 1-2, 5.5-6.5 and 11-12 hours relative to feed delivery.
Twenty-four hours later each videotaping session, eating rate (g DM/h) was
determined between 0-4, 4-8, 8-12 and 12-24 hours relative to feed delivery by
dividing the DMI in each interval by its correspondent length of time. The TMR
profile at feeding and at each time point post-feed delivery was determined using
the Penn State Particle Separator (PSPS) to yield long (>19 mm), medium (<19,
>8 mm), short (<8, >1.18 mm) and fine (<1.18 mm) particles. Feed sorting was
then determined by measuring the particle size distribution of feed consumed
between 0-4, 4-8, 8-12 and 12-24 hours post feeding. Adding glycerol to the
offered prepartum diet increased (P<0.05) the DM% retained as long particles
43
(>19 mm) and reduced (P<0.05) the DM% retained as short (<8, >1.18 mm) and
fine (<1.18 mm) particles, but did not change (P>0.05) the DM% retained as
medium particles (<19, >8 mm). Feed intake did not differ (P>0.05) between diets
and was 14.7 ± 0.4 and 20.2 ± 0.5 kg/day for the pre- and post-partum intervals,
respectively. During the prepartum period, glycerol increased (P<0.05) the eating
rate (94.2 vs. 144.4 g DM/h, control vs. glycerol) and preferential consumption (9.2
vs. 17.8%, control vs. glycerol) for long particles, but decreased (P<0.05) the
eating rate of short (383.8 vs. 332.5 g DM/h, control vs. glycerol) and fine particles
(173.9 vs. 129.8 g DM/h, control vs. glycerol), and increased (P<0.05) sorting
against short (42 vs. 37.3%, control vs. glycerol) and fine particles (17.9 vs.
13.6%, control vs. glycerol). There was no effect (P>0.05) of diet on feed sorting
after parturition or time spent eating, resting or ruminating across the whole
experimental length. The data indicate that although there was no effect on overall
DMI, glycerol enhanced the preference for long particles (>19 mm) during the
prepartum interval, which may be beneficial for rumen health.
Keywords: feed sorting, preference, periparturient cows
1 INTRODUCTION
Dairy cattle have the intrinsic ability to select specific and needed
nutrients when feeds are offered separately (STRICKLIN & KAUTZ-SCANAVY,
1983). Feeding rations as a TMR is a common practice on high yielding dairy
farms; however; one concern about TMR feeding is the ability of cows to
selectively consume or sort various feed components from their ration. Dairy cows
have been shown to have preference for the grain component of the TMR and sort
against the longer forage components (LEONARDI & ARMENTANO, 2003) which
can lead to consumption of rations that fail to meet the cows daily nutrient needs
or result in suboptimal rumen fermentation patterns (DEVRIES et al., 2007).
Transition dairy cows are at a greater risk for developing metabolic and infectious
diseases than at any other time during their life (GRUMMER, 1995; DRACKLEY,
1999). Therefore, special attention must be paid to diet formulation and feeding
44
management practices for transition cows to minimize feed sorting behaviors and
avoid consumption of a ration with compromised nutritive value (STONE, 2004).
When dominant and subordinate cows are grouped together in free-
stalls, feed sorting by dominant cows is also likely to impact the nutritional value of
feeds available for other cows in the group and therefore reduce the feeding value
of the ration (KRAUSE & OETZEL, 2006). Likewise, with a greater interval of time
during the post feed delivery phase there is a greater likelihood of feed sorting,
therefore increasing the deviation from the target nutrient and particle profile for a
particular TMR (DEVRIES et al., 2005). The addition of water to a TMR has been
investigated to reduce feed sorting by decreasing discrimination against long
particles and reducing the preferential consumption of short particles (LEONARDI
et al. 2005a). However, a more recent study (MILLER-CUSHON & DEVRIES,
2009) demonstrated that the addition of water decreased the DMI and increased
sorting against long particles and increased preference for short and fine particles.
Molasses has been recognized for its property to conglomerate small feed
particles to larger particles; the addition of molasses to corn silage based diets
decreased feed sorting, suggesting that molasses might be beneficial to enhance
uniformity of TMR consumption for group-fed cows (OELKER et al., 2009).
Glycerol is a byproduct of the biodiesel industry and has been
recently demonstrated to be a suitable primary feed ingredient to replace corn
grain in rations fed to mid-lactation dairy cows (DONKIN et al., 2009). Glycerol is a
sweet tasting (LEE, 1987) viscous liquid that has been used in the food industry
due to its ability to enhance water holding capacity (FARAHNAKY et al., 2009) and
sweetness (ROPER, 2007) of foods. Furthermore, glycerol has been shown to
stimulate gustatory chemoreceptors that sense ‘sweet’ taste in ruminants (BELL &
KITCHELL, 1966) and early lactation dairy cattle have been shown to prefer
‘sweet’ tasting feeds (NOMBEKELA et al., 1984). Therefore we hypothesized that
glycerol in rations for transition cows would alter preference for particle size due to
its ability to coat feed particles in a TMR and alter the feed sorting and feeding
behavior of transition cows.
45
2.1 Housing and management
Twenty-six Holstein multiparous cows were paired by previous

lactation performance and expected calving date, and housed in individual tie
stalls at the Purdue Dairy Research and Education Center. Cows were randomly
assigned to diets containing either high moisture corn (control) or glycerol, and fed
diets formulated to meet or exceed the NRC (2001) guidelines for 600 kg dairy
cattle.
2.2 Diets and treatments
The ingredients and nutrient composition of the pre- and post-partum

diets are presented in Table 1. Description of the diets, production response and
metabolic profiles of cows has been detailed elsewhere (CARVALHO et al., 2010),
but some of the data is repeated here to assist in evaluation of the feeding
behavior. Diets were fed once daily between 06:30 to 07:30 h in amounts that
ensured ad libitum intake and 10 to 15% weighbacks from -28 through +56 DRTC.
The diets were adjusted weekly to account for forage DM fluctuation. Feed
refusals were measured daily and feed intake was determined by difference.
Samples of TMR were collected weekly, dried in a forced-air oven for
72 h at 55ºC and grounded using a Wiley mill to pass a 1-mm screen. Composite
samples were formed monthly and analyzed by a commercial laboratory (Dairy
One, Ithaca, NY) for DM, CP, ADF, starch and minerals by wet chemistry following
AOAC (2000) procedures, and for NDF following the method of GOERING & VAN
SOEST (1970). Target composition of the pre- and post-partum diets were
respectively 14.9 and 18.3% CP, 39.4 and 30% NDF , 25.5 and 19.5% ADF, 1.60
and 1.62 Mcal of NEL/kg DM, 0.9 and 1% Ca, and 0.3 and 0.4% P, with an
anticipated intake of 11.6 and 22.6 kg/day. The study was conducted from
46
May/28/2009 to Oct/232009 and animal use and handling protocols were

approved by the Purdue Animal Care and Use Committee.
47

experimental diets.
Ingredient, % of DM
Corn silage 35.4 35.4 39.0 39.0
Grass hay 13.0 13.0 3.5 3.5
Glycerol - 11.5 - 10.8
Soybean meal - 2.5 10.0 11.0
Megalac R1 - - 0.7 0.7
2
Supplement3,4 15.8 15.8 10.0 10.0
DM, % 50.9 49.4 46.8 46.0
CP, % of DM 16.6 ± 1.00 16.6 ± 1.35 18.2 ± 0.83 18.7 ± 1.00
ADF, % of DM 22.9 ± 1.75 25.5 ± 1.79 19.5 ± 1.77 20.8 ± 2.32
NDF, % of DM 38.0 ± 1.18 42.2 ± 1.35 31.4 ± 2.71 34.2 ± 1.67
Starch, % of DM 22.6 ± 2.64 15.0 ± 1.22 26.7 ± 1.73 19.2 ± 1.12
NEL, Mcal/kg of DM 1.58 ± 0.02 1.61 ± 0.05 1.65 ± 0.02 1.61 ± 0.02
Ca, % of DM 1.09 ± 0.17 1.02 ± 0.12 1.11 ± 0.07 1.11 ± 0.20
P, % of DM 0.36 ± 0.02 0.34 ± 0.02 0.43 ± 0.02 0.40 ± 0.04
Mg, % of DM 0.39 ± 0.04 0.36 ± 0.02 0.36 ± 0.04 0.35 ± 0.02
K, % of DM 1.22 ± 0.05 1.29 ± 0.09 1.47 ± 0.11 1.44 ± 0.03
Na, % of DM 0.15 ± 0.01 0.15 ± 0.01 0.32 ± 0.01 0.32 ± 0.02
1
Church & Dwight Co., Princeton, NJ.
2
3
4
Products, Quincy, IL).
48
2.3 Video Taping and Feed Sorting
Ten cows in each treatment group were continuously videotaped for

24 h using 10 ceiling-mounted black and white video cameras (Panasonic WV –
BP – 330) that allowed complete visualization of two cows per camera. Cameras
were linked to a central monitor and a four-channel stand-alone digital video
recorder for collection of data. Targeted days of video recording of behaviors were
-17, -10, +8, +15 and +50 DRTC. Feeding behaviors from individual cows were
later evaluated using Observer 5.0 (Noldus Information Technology 2009)
software for 1-h intervals between 0-1, 1-2, 5.5-6.5 and 11-12 hours relative to
feed delivery. An ethogram was developed to determine a time budget for eating,
ruminating (both standing and lying) and resting. Behaviors were scored from
continuous real time video. For each scan the behavior of eating was defined as
obtaining or manipulating feed, chewing feed with head in the feed bunk, or
chewing feed with head away from the feed bunk. The end of eating was defined
as the cessation, for one minute, of the feeding behaviors described above. An
eating bout was defined as a continuous period of feeding separated from a
different period of feeding by at least one interval of nonfeeding. Rumination was
defined as manipulating a cud or repetitive mouth movements that were not
categorized as feeding based on the description above. The end of a rumination
bout was defined as the cessation of rumination activities for one minute. A
rumination bout was defined as a continuous period of rumination separated from
a different period of rumination by at least one interval of nonrumination. Resting
(both standing and lying) was defined as inactivity, drinking or grooming following
a bout of either eating or rumination and was terminated with the initiation of either
rumination or eating. These scans were used to calculate the total time, number of
bouts and bout duration for eating, ruminating, and resting.
Feed sorting was determined during the 24-h period immediately
following each videotaping on days -16, -9, +9, +16 and +51 relative to calving.
Samples of diets were collected at 0, 4, 8, 12 and 24 hours after feed delivery and
frozen at -20ºC pending analysis. Upon thawing, samples were size separated
using the Penn State Particle Separator (PSPS) through 19, 8, and 1.8 mm
screens to yield long (>19 mm), medium (<19, >8 mm), short (<8, >1.18 mm) and
49
fine (<1.18 mm) particles as described previously (LAMMERS et al., 1996).

Separated materials were then dried in a forced-air oven for 72 h at 55ºC. Feed
intake of the individual particle sizes during each interval was determined by
subtracting the particle size distribution available at the beginning of the interval
from the particle size distribution remaining at the end of the interval. Eating rate (g
DM/h) was determined between 0-4, 4-8, 8-12 and 12-24 hours relative to feed
delivery by dividing the DMI of each interval by its corresponding length of time,
and feed sorting was determined by measuring the particle size distribution of feed
consumed between 0-4, 4-8, 8-12 and 12-24 hours relative to feed delivery.
Recovery of glycerol for each fraction of the PSPS separated samples was
determined by mixing 2.5 g of feed in water to a volume of 25 mL. After 24-h
incubation at room temperature the resulting suspension was centrifuged at 2,000
× g for 15 minutes and glycerol was determined in the supernatant using Free
Glycerol Determination Kit FG0100-1KT (Sigma – Aldrich, St Louis, MO).
2.4 Data Analysis
The data were analyzed using the MIXED procedure of SAS (1999).
The model accounted for the effects of treatment, DRTC and times post feeding
(4, 8, 12 and 24 hours for eating rate and feed sorting or 0, 1, 5.5 and 11 hours for
feeding behavior), as well as the interaction effect of treatment by DRTC,
treatment by time, DRTC by time and treatment by DRTC by time. Means were
different if P<0.05 and tended to differ when 0.05 ≤ P ≤ 0.15. Values reported are
least square means and associated standard errors.
3 RESULTS
Twenty-three cows completed the study. One cow from the control
group and one cow from the glycerol group were removed due to a displaced
abomasum and one cow from the control group was removed due to uterine
50
torsion that occurred 7 days prior to parturition. Data reported are for 12 cows in
the glycerol group and 11 cows in the control group.
Due to the imprecision associated with target calving dates the actual
mean and associated standard deviations for video recordings were -17 ± 3.6, -10
± 3.7, +10 ± 1.6, +17 ± 2.4 and +52 ± 1.8 DRTC. Adding glycerol to the offered
prepartum diet increased (P<0.05) the DM% retained as long particles (>19 mm)
and reduced (P<0.05) the DM% retained as short (<8, >1.18 mm) and fine (<1.18
mm) particles, but did not change (P>0.05) the DM% retained as medium particles
(<19, >8 mm). There was no effect (P>0.05) of glycerol on particle size distribution
of the offered postpartum diets (Table 2). Average glycerol content of the pre- and
post-partum glycerol diets was 9.20 ± 0.48% and 10.26 ± 0.48%, respectively, and
relatively close to the dietary composition as shown in Table 1. There was no
difference (P>0.05) in regard to the glycerol concentration among the four
individual screens of the PSPS, although there was a tendency (P=0.14) for the
postpartum diet to contain more glycerol, especially in particles greater than 19
mm (Table 3). Glycerol was not detected in the control diets. Data for feeding
behaviors, eating rate and feed sorting were analyzed separately for the pre- and
post-partum periods. Feed intake did not differ (P>0.05) between treatments and
was 14.7 ± 0.4 and 20.2 ± 0.5 kg/day for the pre- and post-partum intervals,
respectively.
51

Item Control Glycerol SEM P
Prepartum
% of DM retained on screen
>19 mm 9.6 16.9 0.4 <0.05
<19, >8 mm 31.0 31.5 0.7 0.58
<8, >1.18 mm 41.7 37.5 0.6 <0.05
<1.18 mm 17.7 14.1 0.4 <0.05
Postpartum
% of DM retained on screen
>19 mm 5.7 7.1 0.5 0.10
<19, >8 mm 36.8 37.7 0.9 0.47
<8, >1.18 mm 42.9 41.5 0.6 0.16
<1.18 mm 14.6 13.7 0.7 0.36
TABLE 3- Glycerol associated with particles in TMR containing glycerol and fed
during the pre- and post-partum periods1.
Prepartum Postpartum SEM P2
Glycerol3, % of DM
>19 mm 8.5 11.2 1.0 0.06
<19, >8 mm 10.1 11.6 1.0 0.30
<8, >1.18 mm 9.5 9.6 1.0 0.57
<1.18 mm 8.8 8.6 1.0 0.94
1
Main effects: prepartum vs. postpartum diet P=0.14; screen P=0.22; diet × screen
P=0.47.
2
P value for prepartum vs. postpartum diet within particle size.
3
Glycerol associated with TMR particle size separated using the PSPS.
Before parturition cows fed glycerol increased (P<0.05) the eating

rate of long particles (>19 mm) but reduced (P<0.05) the eating rate of short (<8,
>1.18 mm) and fine particles (<1.18 mm), and there was no effect (P>0.05) of
prepartum glycerol on the eating rate of medium (<19, >8 mm) particles (Figure 1,
Table 4). Overall, cows had the greatest eating rate of each individual particle size
and the sum of all particles within the first 4 h post feeding (Table 4). There was a
52
treatment × time after feed delivery response (P<0.05) on the eating rates for
medium (<19, >8 mm), short (<8, >1.18 mm), fine (<1.18 mm) and sum of all
particles (Figure 2, Table 4).
There was no effect (P>0.05) of treatments on eating rate after
parturition (Figure 1, Table 5). Regardless of the diet consumed, cows increased
(P<0.05) the eating rate of medium (<19, >8 mm), short (<8, >1.18 mm), fine
(<1.18 mm) and the sum of all particles when they reached +51 DRTC, but there
was no response (P>0.05) of DRTC on the eating rate of long particles (>19 mm)
(Table 5). There was a treatment × DRTC × time effect (P<0.05) on the eating rate
of short (<8, >1.18 mm) particles and sum of all particles (Table 5).
Cows fed glycerol increased (P<0.05) the preference for long
particles (>19 mm) and reduced (P<0.05) for short (<8, >1.18 mm) and fine (<1.18
mm) particles during the prepartum interval (Figure 1). When the prepartum intake
data are expressed as a percentage of total particles consumed and collapsed
across DRTC and time post feeding, there is an effect of diet that indicates a
preference (P<0.05) for long particles (>19 mm), no response (P>0.05) for medium
particles (<19, >8 mm), and a rejection against (P<0.05) short (<8, >1.18 mm) and
fine (<1.18 mm) particles for cows fed glycerol (Figure 3). Glycerol increased
(P<0.05) consumption of long particles (>19 mm) from 9.2 to 17.8% (control vs.
glycerol) of total particles consumed but increased (P<0.05) sorting against short
particles (<8, >1.18 mm) from 42 to 37.3% (control vs. glycerol) and against fine
particles (<1.18 mm) from 17.9 to 13.6% (control vs. glycerol). Feed sorting during
the postpartum interval did not differ (P>0.05) between treatments.
53
Eating rate (g DM/h)
Time 4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
Treatment Control Glycerol Control Glycerol Control Glycerol Control Glycerol Control Glycerol
DRTC -16 -9 9 16 51
FIGURE 1- Eating rate (g DM/h) associated with particle size, diet, and time after feed delivery. Cows were fed
glycerol or control diets as a TMR between 06:30 to 07:30 h on -16, -9, +9, +16 and +51 DRTC.
Eating rate was determined between 0-4, 4-8, 8-12 and 12-24 hours relative to feed delivery by
dividing the DMI of each interval by its corresponding length of time. Areas within each stack
represent mean long (lowest cross hatched boxes), medium (shaded boxes), short (upper cross
hatched boxes), and fine (open boxes) particles and associated standard errors.
54
TABLE 4- Probabilities associated with the effect of treatment (control vs. glycerol), DRTC, time after feed delivery, and
interactions on the eating rate (g DM/h) of particle sizes for transition dairy cows fed control or glycerol during
the prepartum period.
P values
Trt1 DRTC2 Time3 Trt×DRTC Trt×time DRTC×time Trt×DRTC×time
Eating rate, g DM/h
Particle size
>19 mm <0.05 0.44 <0.05 0.26 0.34 0.62 0.45
<19, >8 mm 0.56 0.13 <0.05 0.55 <0.05 0.16 0.54
<8, >1.18 mm <0.05 0.81 <0.05 0.71 <0.05 0.88 0.66
<1.18 mm <0.05 0.17 <0.05 0.41 <0.05 0.77 0.94
Sum of all particles 0.27 0.24 <0.05 0.63 <0.05 0.61 0.42
1
Treatment.
2
Days relative to calving (-16 and -9).
3
Time after feed delivery (4, 8, 12 and 24 h).
55
600 900
Eating rate (g DM/h) of medium
800
Eating rate (g DM/h) of short

500
particles (<19, >8 mm)
particles (<19, >8 mm)

700
400 600
500
300
400
200
*
300
200
100
100
0 0
4 8 12 24 4 8 12 24
Time after feed delivery (h) Time after feed delivery (h)
450
Eating rate (g DM/h) of fine particles
2000
Eating rate (g DM/h) of all particle

400 1800
350 1600
300 1400
(<1.18 mm)
250 1200
sizes
1000
200
800 *
150
* 600
100
400
50 200
0 0
4 8 12 24 4 8 12 24
Time after feed delivery (h) Time after feed delivery (h)
FIGURE 2- Effect of time after feed delivery and glycerol on eating rate. Cows were fed glycerol (open squares) or control (solid squares) diets as a TMR
between 06:30 to 07:30 h on -16 and -9 DRTC. Eating rate was determined between 0-4, 4-8, 8-12 and 12-24 hours relative to feed delivery by
dividing the DMI of each interval by its corresponding length of time. Data are least squared means and standard errors for eating rate of medium,
short, fine and sum of all particles from top to bottom respectively. Data indicate a treatment × time effect (P<0.05). Stars below the means indicate
differences (P<0.05) between treatments within the same time post feeding.
56
Control Glycerol
FIGURE 3- Distribution of particle sizes consumed during the prepartum period

for cows fed the glycerol or control diets. Cows were fed glycerol
or control diets as a TMR between 06:30 to 07:30 h on -16 and -9
DRTC. Feed sorting was determined by measuring the particle
size distribution of feed consumed between 0-4, 4-8, 8-12 and 12-
24 hours relative to feed delivery using the Penn State Particle
Separator through 19, 8, and 1.8 mm screens to yield long (>19
mm), medium (<19, >8 mm), short (<8, >1.18 mm) and fine (<1.18
mm) particles. Areas within each stack represent mean long
(lowest cross hatched boxes), medium (shaded boxes), short
(upper cross hatched boxes), and fine (open boxes) particles and
associated standard errors.
57
interactions on the eating rate (g DM/h) of particle sizes for transition dairy cows fed control or glycerol during
the postpartum period.
P values
Trt1 DRTC2 Time3 Trt×DRTC Trt×time DRTC×time Trt×DRTC×time
Eating rate, g DM/h
Particle size
>19 mm 0.32 0.11 <0.05 0.31 0.42 0.13 0.97
<19, >8 mm 0.89 <0.05 <0.05 0.96 0.25 <0.05 0.16
<8, >1.18 mm 0.73 <0.05 <0.05 0.83 0.19 <0.05 <0.05
<1.18 mm 0.78 <0.05 <0.05 0.96 0.91 <0.05 0.84
Sum of all particles 0.91 <0.05 <0.05 0.96 0.46 <0.05 <0.05
1
Treatment.
2
Days relative to calving (+9, +16 and +51).
3
Time after feed delivery (4, 8, 12 and 24 h).
58
The activities of eating, resting, and rumination were not influenced

(P>0.05) by treatments for either the pre- or post-partum periods (Tables 6, 7, 8, and 9).
During the prepartum interval (Tables 6 and 7), cows decreased (P<0.05) the time spent
eating as calving approached (15.7 vs. 12.9 ± 1.0 min/h for -17 vs. -10 DRTC,
respectively). The number of eating bouts decreased as calving approached (11.2 vs.
7.6 ± 0.75 bouts/h for -17 vs. -10 DRTC, respectively) but eating bout duration was
similar (P>0.05) between these 2 days. There was no effect (P>0.05) of DRTC on time
spent resting, but the number of resting bouts decreased (P<0.05) towards parturition
(11.8 vs. 8.3 ± 0.72 bouts/h for -17 vs. -10 DRTC, respectively), and resting bout
duration tended to be greater (P<0.10) at -10 DRTC (9.9 ± 1.4 minutes). Time spent
ruminating tended to be longer (P<0.10) at -10 DRTC (11 ± 1.2 min/h) and there was no
response (P>0.05) for DRTC on rumination bouts, but rumination bout duration was
greater (P<0.05) at -10 DRTC (9.4 ± 1 minute). Regardless of the treatment, cows spent
more time (P<0.05) eating between 0-1 hour past feed delivery (33.7 ± 1.3 min/h).
Eating bouts (19 ± 0.9 bouts/h) and eating bout duration (1.9 ± 0.2 minutes) were
greater (P<0.05) during this interval as well. Time spent resting was increased (P<0.05)
between 1-2 hours post feeding (46.4 ± 1.7 min/h), but resting bouts were greater
(P<0.05) between 0-1 hour after cows were fed (18.8 ± 0.9 bouts/h). Resting bout
duration was longer (P<0.05) between 1-2, 5.5-6.5 and 11-12 (9.6, 10.2 and 11.5 ± 1.8
minutes, respectively) hours relative to feed delivery. Cows spent the longest (P<0.05)
time ruminating between 5.5-6.5 and 11-12 (17.8 and 14 ± 1.6 min/h, respectively)
hours after feed was delivered, and rumination bouts (1.1 and 1.2 ± 0.12 bouts/h for 5.5-
6.5 and 11-12 hours post feeding, respectively) and rumination bout duration (15.3 and
10.4 ± 1.4 minutes for 5.5-6.5 and 11-12 hours post feeding, respectively) followed the
same pattern as time spent ruminating (P<0.05). There was a tendency (P<0.15) for a
treatment × DRTC effect that reduced the eating bout duration in cows fed prepartum
glycerol at -10 DRTC (1.7 vs. 1.3 ± 0.2 minutes for control vs. glycerol, respectively),
and a tendency (P<0.15) for a treatment × time response for time spent resting, time
spent ruminating (Figure 4) and rumination bout duration.
59
TABLE 6- Effect of glycerol on prepartum eating, resting, and ruminating activities.

Control Glycerol
Time relative to feeding, h
0-1 1-2 5.5-6.5 11-12 0-1 1-2 5.5-6.5 11-12 SEM1
Item
-17 DRTC2
Eating
Duration, min/h 35.8 8.8 6.2 10.2 34.0 6.9 8.3 16.0 2.35
Bouts/h 21.2 9.6 6.3 8.3 21.0 6.9 4.3 11.9 1.67
Bout duration, min 1.9 0.8 1.1 1.4 1.8 1.0 1.3 1.3 0.37
Resting
Duration, min/h 24.2 47.4 37.1 35.9 26.1 52.3 30.6 36.6 3.37
Bouts/h 20.9 10.6 6.9 9.3 20.7 7.8 5.6 12.5 1.62
Ruminating
Duration, min/h 0.0 3.9 16.8 13.9 0.0 0.8 21.2 7.5 3.25
Bouts/h 0.0 0.5 1.4 1.4 0.0 0.2 1.2 1.0 0.25
-10 DRTC
Eating
Duration, min/h 34.6 9.1 6.5 6.0 30.6 5.8 4.8 6.1 2.35
Bouts/h 17.5 6.3 4.0 4.3 16.5 5.2 3.7 3.4 1.67
Resting
Duration, min/h 24.3 39.8 40.2 35.6 29.4 46.1 35.4 37.9 3.37
Bouts/h 17.4 7.0 5.1 5.3 16.3 6.3 4.7 4.7 1.62
Ruminating
Duration, min/h 1.1 11.2 13.3 18.6 0.2 7.9 19.9 16.0 3.25
Bouts/h 0.1 0.9 0.9 1.1 0.1 0.6 0.9 1.1 0.25
1
Standard error associated with treatment × day relative to calving × time relative to feed
delivery.
2
60
interactions on eating, resting and rumination activities for transition dairy cows fed glycerol or control diets
during the prepartum period.
Item Trt1 DRTC2 Time3 Trt×DRTC Trt×time DRTC×time Trt×DRTC×time
Eating
Duration, min/h 0.73 <0.05 <0.05 0.17 0.27 0.09 0.89
Bouts /h 0.66 <0.05 <0.05 0.78 0.49 0.06 0.32
Bout duration, min 0.52 0.17 <0.05 0.13 0.89 0.97 0.89
Resting
Duration, min/h 0.54 0.93 <0.05 0.55 0.08 0.11 0.99
Bouts /h 0.69 <0.05 <0.05 0.79 0.53 0.07 0.42
Ruminating
Duration, min/h 0.69 0.09 <0.05 0.72 0.11 0.08 0.94
Bouts /h 0.28 0.98 <0.05 0.52 0.87 0.10 0.89
Bout duration, min 0.85 <0.05 <0.05 0.98 0.11 0.06 0.92
1
Treatment.
2
Days relative to calving (-17 and -10).
3
Time after feed delivery (0, 1, 5.5 and 11 h).
61
25
Time spent ruminating (min)

20
15
10
0
0 1 5.5 11
Time after feed delivery (h)
FIGURE 4- Time spent ruminating relative to treatment × time post feeding effect
of transition dairy cows fed control (solid squares) or glycerol (open
squares) during the prepartum interval. Pooled SEM = 2.28
After parturition, cows spent a similar (P>0.05) amount of time eating

on +8, +15 and + 50 (15.2, 14.1 and 16.5 ± 1.2 min/h, respectively) DRTC, but eating
bout duration was longest (P<0.05) at +50 DRTC (6.4 ± 0.8 minutes) as the number
of eating bouts decreased (P<0.05) from 9.3 and 8 ± 0.7 bouts/h for 8 and 15 DRTC,
respectively, to 3 ± 0.7 bouts/h at +50 DRTC (Tables 8 and 9). Time spent resting
(28.5 ± 1.6 min/h) and number of resting bouts (3.7 ± 0.7 bouts/h) decreased
(P<0.05) as lactation progressed to +50 DRTC, and there was a tendency for
reduction (P<0.15) of resting bout duration at +8 DRTC (8.7 ± 1.8 minutes) as a
consequence of greater (P<0.05) number of resting bouts at +8 DRTC (10.2 ± 0.7
bouts/h) compared with +50 DRTC (3.7 ± 0.7 bouts/h). Cows increased (P<0.05)
time spent ruminating (15 ± 1.5 min/h) and rumination bout duration (14.4 ± 1.4
minutes) at +50 DRTC. Likewise in the prepartum stage, time spent eating (28.8 ±
1.6 min/h), eating bouts (11.2 ± 0.8 bouts/h) and eating bout duration (5.3 ± 0.8
minutes) were increased (P<0.05) between 0-1 hour past feed delivery. Time spent
resting increased (P<0.05) between 1-2, 5.5-6.5 and 11-12 (37.2, 30.8 and 34.1 ±
2.1 min/h, respectively) hours post feeding, the number of resting bouts increased
(P<0.05) between 0-1 and 11-12 (11.3 and 9.3 ± 0.8 bouts/h, respectively) hours
after cows were fed, and resting bout duration increased (P<0.05) between 1-2 and
5.5-6.5 (18.7 and 12.9 ± 2.2 minutes) hours after feed was delivered. Cows
62
increased (P<0.05) time spent ruminating between 1-2 and 5.5-6.5 (16.4 and 20.4 ±
1.8 min/h, respectively) hours post feeding; furthermore rumination bouts (0.8 and 1
± 0.1 bouts/h for 1-2 and 5.5-6.5 hours post feeding, respectively) and rumination
bout duration (15 and 18.8 ± 1.7 minutes for 1-2 and 5.5-6.5 hours post feeding,
respectively) were also greater (P<0.05) at these times. There was a tendency
(P<0.15) for a treatment by DRTC effect for time spent resting and tendencies
(P<0.15) of treatment by time response for the number of eating and resting bouts.
63
TABLE 8- Effect of glycerol on postpartum eating, resting and ruminating activities.

Control Glycerol
Time relative to feeding, h
0-1 1-2 5.5-6.5 11-12 0-1 1-2 5.5-6.5 11-12 SEM1
Item
+8 DRTC2
Eating
Duration, min/h 31.6 10.3 6.6 17.3 29.0 4.6 5.5 16.8 3.77
Bouts /h 12.4 6.9 5.1 13.2 16.3 4.4 4.9 11.6 1.87
Resting
Duration, min/h 21.6 33.5 30.6 37.7 30.9 44.9 31.0 35.3 4.83
Bouts /h 12.2 8.0 6.5 14.1 16.1 5.9 6.0 12.7 1.89
Ruminating
Duration, min/h 6.8 16.2 22.8 5.0 0.1 10.2 23.0 8.0 4.66
Bouts /h 0.3 1.0 1.3 0.6 0.0 0.9 1.0 1.0 0.26
+15 DRTC
Eating
Duration, min/h 30.6 8.9 8.1 12.5 30.5 1.6 3.1 17.8 3.77
Bouts /h 12.8 6.2 4.7 9.8 16.8 1.1 2.7 10.1 1.87
Resting
Duration, min/h 29.0 39.0 33.2 36.9 29.6 47.3 38.2 38.2 4.83
Bouts /h 12.5 7.2 5.4 10.4 16.7 2.0 3.5 10.6 1.89
Ruminating
Duration, min/h 0.5 11.8 18.9 10.8 0.0 10.9 18.4 3.7 4.66
Bouts /h 0.1 0.7 0.8 0.7 0.0 0.6 1.0 0.3 0.26
+50 DRTC
Eating
Duration, min/h 26.1 5.7 14.7 20.5 25.0 6.3 14.0 19.8 3.77
Bouts /h 4.6 0.6 3.5 3.0 4.6 2.2 2.0 3.7 1.87
Resting
Duration, min/h 30.5 32.0 28.5 30.6 30.5 26.7 23.4 25.9 4.83
Bouts /h 5.4 1.6 4.4 3.5 4.9 2.8 2.7 4.3 1.89
Ruminating
Duration, min/h 3.6 22.2 16.9 8.9 4.2 27.4 22.6 14.4 4.66
Bouts /h 0.5 0.9 0.9 0.5 0.6 0.9 1.0 0.7 0.26
1
Standard error associated with treatment × day relative to calving × time relative to feed
delivery.
2
64
interactions on eating, resting and rumination activities for transition dairy cows fed glycerol or control diets
during the postpartum period.
Item Trt1 DRTC2 Time3 Trt×DRTC Trt×time DRTC×time Trt×DRTC×time
Eating
Duration, min/h 0.37 0.24 <0.05 0.76 0.67 <0.05 0.81
Bouts /h 0.83 <0.05 <0.05 0.88 0.10 <0.05 0.37
Resting
Duration, min/h 0.47 <0.05 <0.05 0.10 0.67 0.11 0.78
Bouts /h 0.80 <0.05 <0.05 0.93 0.13 <0.05 0.35
Ruminating
Duration, min/h 0.94 <0.05 <0.05 0.19 0.88 0.23 0.85
Bouts /h 0.82 0.12 <0.05 0.54 0.94 0.42 0.73
1
Treatment.
2
Days relative to calving (+8, +15 and +50).
3
Time after feed delivery (0, 1, 5.5 and 11 h).
65
4 DISCUSSION
In the present study, our primary objective was to determine the

effect of glycerol in the diet of transition cows on preference for particles of
different sizes in the TMR. In addition we were interested in determining the effect
of glycerol on feeding, resting and ruminating behaviors. The analysis of glycerol
recovered in particles that were size separated using the PSPS indicate a uniform
distribution of glycerol per unit of feed dry matter among these fractions. These
data would suggest that the adherent and conglomerate properties of glycerol are
not specific to a class of particle sizes within the diets fed in this study as a TMR.
The lack of differences in intake between the diets provides an opportunity to
determine the impact of physical coating of particles with a sweet tasting material
on the particle size distribution of intake.
Glycerol increased the eating rate and preference for long particles
(>19 mm) of the prepartum diet but had no impact on preference in the postpartum
diet. The differential effect of glycerol in this regard may be due to a greater
proportion of long-stem forage in the glycerol prepartum diet compared with the
postpartum diet therefore enhancing the opportunity for particle selection within
the prepartum diet. Nevertheless, cows fed the prepartum diet containing glycerol
favored the consumption of long stemmed particles. Furthermore, selection for
long stemmed particles was greatest during the first 4 hours post feed delivery.
Therefore, the data from the prepartum interval support the hypothesis that
glycerol alters preference for particle size in a TMR fed to transition cows.
Social structure influences feeding behaviors especially when access
to feed resources is reduced, consequently subordinate cows are limited by feed
bunk access. Due to sorting for feed that favors consumption of short particles by
more competitive cows in the group, subordinate cows are at risk for consuming a
diet with reduced energy density relative to demands (HOSSEINKHANI et al.,
2008). Within-day sorting in transition cow diets may negatively influence
nutritional status for cows that face competition for feed bunk access by reducing
the energy density of the residual TMR because dominant cows tend to select
against long particles (HOSSEINKHANI et al., 2008). Thus, selection for long
particles when glycerol is included in the diet may reduce the effects of social
66
structure and competitiveness. However because feed sorting for individually

housed cows (LEONARDI & ARMENTANO, 2007) and competitively housed cows
(DEVRIES et al., 2004) differs, the effects of glycerol on feeding sorting and
preference for long particles must be further evaluated to determine the impact of
glycerol on variation in nutrient consumption in cows housed in groups and given
access to a common feed bunk.
The treatment by time relative to feed delivery effect on the eating
rates of medium (<19, >8 mm), short (<8, >1.18 mm), fine (<1.18 mm) and sum of
all particles during the prepartum interval reveals that cows fed glycerol had a
lower eating rate from 4 to 12 hours post feeding, but this pattern was shifted in
the remaining hours of the day. Determining the biology underlying this shift was
not a primary objective in the present study, but we speculate this change is a
consequence of a difference in digestion rates between corn starch and glycerol.
Available estimates of glycerol fermentation rate differ considerably (GARTON et
al., 1961; KIJORA et al., 1998) depending on animal type and adaptation to
glycerol feeding; more accurate estimate is needed to assess the potential effects
of glycerol on rumen fermentation patterns. Nonetheless, the data presented in the
treatment by time post feeding effect on the eating rates of medium (<19, >8 mm),
short (<8, >1.18 mm), fine (<1.18 mm) and sum of all particles indicate a more
balanced eating rate for all particles in cows fed glycerol during the prepartum
period, which may prove beneficial for rumen health, especially during the
transition period when sudden changes of diet and shifts in microbial growth
population are thought to be problematic.
Feed sorting was determined in the present study relative to the
amount of feed that was available in the beginning of each interval and profile of
feed consumed at the end of the interval (CARVALHO et al., 2010). This approach
differs from the methods introduced by LEONARDI and ARMENTANO (2003), and
followed subsequently (LEONARDI et al., 2005a; LEONARDI et al., 2005b;
DEVRIES et al., 2007; BHANDARI et al., 2008; DEVRIES et al., 2008;
HOSSEINKHANI et al., 2008) where feed sorting is calculated as the actual intake
of each screen of the PSPS expressed as a percentage of the predicted intake for
that screen based on the initial profile of the diet. The actual intake for each
individual particle size equals the product of the DMI in each interval multiplied by
67
the particle size distribution of each interval, whereas the predicted intake for each
individual particle size equals the product of the DMI in each interval multiplied by
the particle size distribution of the offered TMR. The final outcome is that values
equal to 100% indicate no sorting, <100% indicate selective refusals (sorting
against), and >100% indicate preferential consumption (sorting for). We have used
an alternative method in order to better represent feed sorting based on the feed
available to the cow at the beginning of each interval of measurement because it
reflects the sequential choice of particles sorted for or against in the ration
(CARVALHO et al., 2010).
Although there was no effect of treatments on the feeding behavior in
the present study, the data presented herein contribute to the base information on
feeding behavior of dairy cows in the periparturient period. The fact that time spent
eating was not affected by treatments can be corroborated by the absence of
response of treatments on the DMI in the days when cows were videotaped.
Cows spent significantly less time eating and had fewer eating bouts as parturition
approached regardless of the type of diet ingested, which can be explained by two
reasons: first, there is the effect of less space allowable for the rumen to be filled
with digesta due to the maximum growth of the calf in the final period of gestation
(GRANT & ALBRIGHT, 1995), and secondly, physiological factors, such as lipid
mobilization and hormonal changes, can contribute for feed intake reduction as
cows approach parturition (DRACKLEY, 1999). Nevertheless, our results are in
agreement with HUZZEY et al. (2005), who also reported decreased eating times
in the days close to parturition.
The tendency for the treatment by time post feeding response on
time spent ruminating indicates that cows fed prepartum glycerol had a different
pattern of rumination in comparison with cows fed prepartum control. The fact that
cows fed prepartum glycerol spent more time ruminating between 5.5-6.5 hours
post feeding can be considered a desirable behavior in terms of buffering the
SCFA production in the rumen after cows had a large meal just after fresh feed
was delivered.
The lack of response of postpartum DRTC on time spent eating was
surprising based on the result that cows increased the eating rate of the sum of all
particles at +50 DRTC, indicating an overall increase in DMI. However, increases
68
in eating bout duration and time spent ruminating, as well as a decrease in time
spent resting points to an increase in meal size as lactation progressed to +50
DRTC as previously described for early lactation cows (GRANT & ALBRIGHT,
1995). Overall, time spent eating, both before and after parturition, was greatest
within the first hour after fresh feed was delivered, in agreement with the data from
this study that the eating rate of all particle sizes was greatest within the first 4
hours post feeding, and also with BHANDARI et al. (2008), who reported that cows
consumed the highest amount of feed within the first 3 hours post feeding, and
decreased feed intake in the subsequent times throughout the day.
5 CONCLUSIONS
Glycerol can be recommended as a primary feed ingredient to

replace high moisture corn in rations fed to transition dairy cows. The addition of
glycerol had no deleterious consequences on intake or feeding behaviors. The
data supports an increased preference for long particles (>19 mm) when diets
containing glycerol are fed during the prepartum period which may prove beneficial
for health of transition cows.
6 ACKNOWLEDGEMENTS
The authors acknowledge the support of Dr. Perry Doane and Archer
Daniels Midland for the generous gift of glycerol for this project, the help afforded
by Mike Grott and the Purdue Dairy Research and Education Center staff, and the
laboratory expertise of S. Koser. This project was supported by the Agricultural
Food Research Initiative of the National Institute of Food and Agriculture, USDA,
Grant # 2007-55618-18237. Scholarship support for E. R. Carvalho was from
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES),
Ministério da Educação, Brasil.
69
7 REFERENCES
AOAC. Association of Official Analytical Chemistry. Official methods of analysis.

17th ed. AOAC Int., Gaithersburg, MD, 2000, 1298 p.
BELL, F. R.; KITCHELL, R. L. Taste reception in the goat, sheep and calf. Journal
of Physiology, v.183, p.145-151, 1966.
BHANDARI, S. K.; Li, S.; OMINSKI, K. H.; WITTENBERG, K. M.; PLAIZIER, J. C.

Effects of the chop lengths of alfalfa silage and oat silage on feed intake, milk
production, feeding behavior, and rumen fermentation of dairy cows. Journal of
Dairy Science, v.91, p.1942-1958, 2008.
CARVALHO, E. R.; LIMA, M. L. M.; FRANÇA, A. F. S.; FERNANDES, J. J. R.;

WHITE, H. M.; DONKIN, S. S. An alternative methodology of determining feed
sorting in transition dairy cows fed glycerol. Ciência Animal Brasileira, v.11,
p.825-834, 2010.
CARVALHO, E. R.; SCHMELZ, N. S.; WHITE, H. M.; DONKIN, S. S. Replacing

corn with glycerol in diets for transition dairy cows. Journal of Dairy Science,
2010 (submitted for review).
DEVRIES, T. J.; BEAUCHEMIN, K. A.; KEYSERLINGK, M. A. G. Dietary forage

concentration affects the feed sorting behavior of lactating dairy cows. Journal of
Dairy Science, v.90, p.5572-5579, 2007.
DEVRIES, T. J.; DOHME, F.; BEAUCHEMIN, K. A. Repeated ruminal acidosis

challenges in lactating dairy cows at high and low risk for developing acidosis:
feed sorting. Journal of Dairy Science, v.91, p.3958-3967, 2008.
DEVRIES, T. J.; KEYSERLINGK, M. A. G.; BEAUCHEMIN, K. A. Frequency of

feed delivery affects the behavior of lactating dairy cows. Journal of Dairy
Science, v.88, p.3553-3562, 2005.
DEVRIES, T. J.; KEYSERLINGK, M. A. G.; WEARY, D. M. Effect of feeding space

on inter-cow distance, aggression, and feeding behavior of free-stall housed
lactating dairy cows. Journal of Dairy Science, v.87, p.1432 -1438, 2004.
DONKIN, S. S.; KOSER, S. L.; WHITE, H. M.; DOANE, P. H.; CECAVA, M. J.

DRACKLEY, J. K. Biology of dairy cows during the transition period: the final
frontier? Journal of Dairy Science, v.82, p.2259-2273, 1999.
FARAHNAKY, A.; ANSARI, S.; MAJZOOBI, M. Effect of glycerol on the moisture

sorption isotherms of figs. Journal of Food Engineering, v.93, p.468-473, 2009.
70
GARTON, G. A.; LOUGH, A. K.; VIOQUE, E. Glyceride hydrolysis and glycerol

fermentation by sheep rumen contents. Journal of General Microbiology, v.25,
p.215-225, 1961.
GOERING, H. K.; VAN SOEST, P. J. Forage fiber analysis (Apparatus,

Reagents, Procedures and Some Applications). Agricultural Handbook nº 379,
ARS-USDA, Washington, D. C., 1970, 19 p.
GRANT, R. J.; ALBRIGHT, J. L. Feeding behavior and management factors during

the transition period in dairy cattle. Journal of Animal Science, v.73, p.2791-
2803, 1995.
GRUMMER, R. R. Impact of changes in organic nutrient metabolism on feeding

the transition dairy cow. Journal of Animal Science, v.73, p.2820-2833, 1995.
HOSSEINKHANI, A.; DEVRIES, T. J.; PROUDFOOT, K. L.; VALIZADEH, R.;

VEIRA, D. M.; KEYSERLINGK, M. A. G. The effects of feed bunk competition on
the feed sorting behavior of close-up dry cows. Journal of Dairy Science, v.91,
p.1115-1121, 2008.
HUZZEY, J. M.; KEYSERLINGK, M. A. G.; WEARY, D. M. Changes in feeding,

drinking, and standing behavior of dairy cows during the transition period. Journal
of Dairy Science, v.88, p.2454-2461, 2005.
KRAUSE, K. M.; OETZEL, G. Understanding and preventing subacute ruminal

acidosis in dairy herds: A review. Animal Feed Science and Technology, v.126,
p.215-236, 2006.
LAMMERS, B. P.; BUCKMASTER, D. R.; HEINRICHS, A. J. A simple method for

the analysis of particle sizes of forage and total mixed rations. Journal of Dairy
Science, v.79, p.922-928, 1996.
LEE, C. K. The Chemistry and Biochemistry of the Sweetness of Sugars.

Advances in Carbohydrate Chemistry and Biochemistry, v.45, p.199-351,
1987.
LEONARDI, C.; ARMENTANO, L. E. Effect of quantity, quality, and length of

alfalfa hay on selective consumption by dairy cows. Journal of Dairy Science,
v.86, p.557–564, 2003.
LEONARDI, C.; ARMENTANO, L. E. Feed selection by dairy cows fed individually

in a tie-stall or as a group in a free-stall barn. Journal of Dairy Science, v.90,
p.2386-2389, 2007.
71
LEONARDI, C.; GIANNICO, F.; ARMENTANO, L. E. Effect of water addition on

selective consumption (sorting) of dry diets by dairy cattle. Journal of Dairy
Science, v.88, p.1043-1049, 2005a.
LEONARDI, C.; SHINNERS, K. J.; ARMENTANO, L. E. Effect of different dietary

geometric mean particle length and particle size distribution of oat silage on
feeding behavior and productive performance of dairy cattle. Journal of Dairy
Science, v.88, p.698-710, 2005b.
MILLER-CUSHON, E. K.; DEVRIES, T. J. Effect of dietary dry matter

concentration on the sorting behavior of lactating dairy cows fed a total mixed
ration. Journal of Dairy Science, v.92, p.3292-3298, 2009.
NOMBEKELA, S. W.; MURPHY, M. R.; GONYOU, H. W.; MARDEN, J. I. Dietary

preferences in early lactation cows as affected by primary tastes and some
common feed flavors. Journal of Dairy Science, v.77, p.2393-2399, 1994.
NATIONAL RESEARCH COUNCIL. Nutrient Requirements of Dairy Cattle. 7th

revised edition. Washington, D.C.:National Academy Press, 2001. 381 p.
OELKER, E. R.; REVENEAU, C.; FIRKINS, J. L. Interaction of molasses and

monensin in alfalfa hay- or corn silage-based diets on rumen fermentation, total
tract digestibility, and milk production by Holstein cows. Journal of Dairy
Science, v.92, p.270-285, 2009.
ROPER, S. D. Signal transduction and information processing in mammalian taste

buds. Pflügers Archiv European Journal of Physiology, v.454, p.759-776,
2007.
SAS INSTITUTE. SAS/STAT User’s Guide: Statistics, Version 9.2 Edition. SAS
Inst. Inc., Cary, NC, 1999.
STONE, W. C. Nutritional approaches to minimize subacute ruminal acidosis and

laminitis in dairy cattle. Journal of Dairy Science, v.87, p.13-26, 2004.
STRICKLIN, W. R.; KAUTZ-SCANAVY, C. C. The role of behavior in cattle

production: a review of research. Applied Animal Ethology, v.11, p.359-390,
1983.
72
CAPÍTULO 4 – AN ALTERNATIVE METHODOLOGY OF DETERMINING FEED

SORTING IN TRANSITION DAIRY COWS FED GLYCEROL.
73
RESUMO
Objetivou-se nessa pesquisa comparar a metodologia padrão com um método
alternativo na determinação do consumo seletivo em vacas leiteiras durante o
período de transição. Vinte e seis vacas multíparas da raça Holandesa foram
pareadas de acordo com o desempenho na lactação anterior e data prevista de
parição, e alimentadas com dietas contendo glicerol ou milho grão úmido desde
os -28 até +56 dias relativos à data de parição (DRDP). Determinou-se o consumo
seletivo aos -16, -9, +9, +15 e +51 DRDP por duas maneiras. Na primeira, como o
consumo real de cada peneira do PSPS entre 0-4, 4-8, 8-12 e 12-24 horas pós-
alimentação, e expresso como a porcentagem do consumo predito da peneira
correspondente. Na segunda maneira, por meio da mensuração da distribuição do
tamanho de partículas do alimento consumido entre 0-4, 4-8, 8-12 e 12-24 horas
pós-alimentação. Tanto no momento da alimentação quanto em cada tempo pós-
alimentação, a ração foi separada por tamanho utilizando-se um conjunto com
três peneiras (19, 8 e 1,18 mm) e um fundo liso denominado PSPS, gerando
partículas longas (>19 mm), médias (<19, >8 mm), curtas (<8, >1,18 mm) e muito
curtas (<1,18 mm), respectivamente. As vacas alimentadas com glicerol no pré-
parto aumentaram (P<0,05) a preferência pelas partículas longas (>19 mm) na
metodologia padrão (77,2 vs. 101,5%, controle vs. glicerol) e também na
metodologia alternativa (9,2 vs. 17,8%, controle vs. glicerol). As vacas
alimentadas com glicerol no pré-parto rejeitaram (P<0,05) as partículas curtas (<8,
>1,18 mm) na metodologia padrão (102,6 vs. 94,2%, controle vs. glicerol) assim
como na metodologia alternativa (42 vs. 37,3%, controle vs. glicerol). Não houve
efeito (P>0,05) de tratamento sobre o consumo seletivo de partículas muito curtas
(<1,18 mm) de acordo com a metodologia padrão na fase pré-parto, mas as vacas
alimentadas com glicerol nessa fase diminuíram (P<0,05) a preferência pelas
partículas muito curtas (<1,18 mm) na metodologia alternativa (17,9 vs. 13,6%,
controle vs. glicerol). As vacas alimentadas com glicerol na fase pós-parto
aumentaram (P<0,05) a preferência por partículas médias (<19, >8 mm) de
acordo com a metodologia padrão (108,6 vs. 116,5%, controle vs. glicerol), porém
não houve efeito (P>0,05) de tratamento sobre o consumo seletivo de partículas
médias (<19, >8 mm) na metodologia alternativa. As vacas alimentadas com
glicerol no pós-parto rejeitaram (P<0,05) as partículas curtas (<8, >1,18 mm) de
74
acordo com a metodologia padrão (100,6 vs. 96,6%, controle vs. glicerol), mas
não houve resposta da dieta (P>0,05) sobre o consumo seletivo de partículas
curtas (<8, >1,18 mm) na metodologia alternativa. Concluiu-se que a metodologia
alternativa proposta nesse estudo é mais confiável do que a metodologia padrão
na determinação do consumo seletivo em vacas leiteiras periparturientes
alimentadas com glicerol.
Palavras-chave: biodiesel, preferência, subproduto, tamanho de partícula
ABSTRACT
The objective of this study was to compare the standard methodology with an
alternative method to determine feed sorting in dairy cows during the transition
period. Twenty-six Holstein multiparous cows were paired by previous lactation
performance and expected calving date, and fed diets containing either glycerol or
high moisture corn from -28 through +56 days relative to calving (DRTC). Feed
sorting was determined on -16, -9, +9, +15 and +51 DRTC in two different ways.
Firstly, as the actual intake of each screen of the PSPS consumed between 0-4, 4-
8, 8-12 and 12-24 hours post feeding, and expressed as a percentage of the
predicted intake of that correspondent screen. Secondly, by measuring the particle
size distribution of feed consumed between 0-4, 4-8, 8-12 and 12-24 hours post
feeding. The TMR at feeding and at each time post feeding was separated by size
using three screens (19, 8, and 1.18 mm) and a bottom pan (PSPS) to yield long
(>19 mm), medium (<19, >8 mm), short (<8, >1.18 mm), and fine particles (<1.18
mm). Cows fed prepartum glycerol increased (P<0.05) the preference for long
particles (>19 mm) according to the standard methodology (77.2 vs. 101.5%,
control vs. glycerol) and also in the alternative methodology (9.2 vs. 17.8%, control
vs. glycerol). Cows fed prepartum glycerol rejected (P<0.05) short particles (<8,
>1.18 mm) in the standard methodology (102.6 vs. 94.2%, control vs. glycerol) as
well as in the alternative methodology (42 vs. 37.3%, control vs. glycerol). There
was no response (P>0.05) of diet on feed sorting of fine particles (<1.18 mm)
according to standard methodology during the prepartum interval, but cows fed
prepartum glycerol decreased (P<0.05) the preference for fine particles (<1.18
75
mm) in the alternative methodology (17.9 vs. 13.6%, control vs. glycerol). Cows
fed postpartum glycerol increased (P<0.05) the preference for medium particles
(<19, >8 mm) according to standard methodology (108.6 vs. 116.5%, control vs.
glycerol), but did not (P>0.05) according to the alternative methodology. Cows fed
postpartum glycerol rejected (P<0.05) short particles (<8, >1.18 mm) according to
the standard methodology (100.6 vs. 96.6%, control vs. glycerol), but did not
(P>0.05) according to the alternative methodology. The alternative methodology
proposed in this study is more reliable than the standard methodology to
determine feed sorting in transition dairy cows fed glycerol.
Keywords: biodiesel, byproduct, particle size, preference
1 INTRODUCTION
Cattle have the intrinsic ability to select specific and needed nutrients
when feeds are offered separately (STRICKLIN & KAUTZ-SCANAVY, 1983).
However, feeding dairy cows with rations in the form of a TMR is a preferred
practice compared with component feeding systems on most dairy farms, but
despite the aim of providing feed as a homogenous mixture in TMR, dairy cattle
have selectively consumed or sorted the grain component of a TMR and
discriminated against the longer forage components (LEONARDI & ARMENTANO,
2003).
The periparturient or transition dairy cow has been defined when the
cow is between three weeks prepartum and three weeks postpartum (GRUMMER,
1995). During this period of transition from gestation to lactation the cow is at
greater risk for developing metabolic and infectious diseases than at any other
time during her life (DRACKLEY, 1999). Therefore special attention must be paid
when formulating diets, as well as feeding and nutritional strategies to minimize
the feed sorting behavior of dairy cows, which can result in the consumption of a
ration with inconsistent nutritive value (STONE, 2004) and increasing risks of
developing subacute ruminal acidosis (COOK et al., 2004; STONE, 2004),
particularly when early lactating cows are fed low forage diets. Likewise, the feed
76
sorting behavior leads to variations in nutritive values of the TMR in the feed bunk
with a greater interval of time post feed delivery (DEVRIES et al., 2005), especially
when dominant and subordinate cows are grouped together, where feed sorting by
some groups of cows is also likely to impact the nutritional value of feeds available
for other cows in the group, and may reduce the feeding value of the ration
(KRAUSE & OETZEL, 2006).
Several additions to rations fed as a TMR have been investigated in
an attempt to reduce feed sorting. The addition of water to a TMR has been
investigated to reduce feed sorting by decreasing the refection for long particles
and reducing the preferential consumption of short particles (LEONARDI et al.,
2005a). However, a more recent study (MILLER-CUSHON & DEVRIES, 2009)
demonstrated that the addition of water decreased DM intake and increased
sorting against long particles and increased preference for short and fine particles.
Molasses has been recognized for its property to conglomerate small feed
particles to larger particles. The addition of molasses to corn silage based diets
decreased feed sorting, suggesting that molasses might be beneficial to enhance
uniformity of TMR consumption for group-fed cows (OELKER et al., 2009) and
addition of molasses reduces the percentage of fine particles found in calf starter
(LESMEISTER & HEINRICHS, 2005).
Glycerol is a byproduct of the biodiesel industry that has been
currently produced by a reaction that utilizes a base-catalyzed transesterification
of oil in the formation of methyl and ethyl fatty acid esters in the production of
biodiesel (THOMPSON & HE, 2006), and is a main byproduct of ethanol
fermentation processing (MICHNICK et al., 1997). For each 100 g of soybean oil
input, there is a yield of 12.25 g of glycerol (THOMPSON & HE, 2006).
Furthermore, glycerol has been recently demonstrated to be a suitable primary
feed ingredient to replace corn grain in rations fed to mid-lactating dairy cows
(DONKIN et al., 2009).
Feed sorting has been previously measured according to LEONARDI
& ARMENTANO (2003) and followed subsequently by several authors
(LEONARDI et al., 2005a; LEONARDI et al., 2005b; DEVRIES et al., 2007;
BHANDARI et al., 2008; DEVRIES et al., 2008; HOSSEINKHANI et al., 2008). The
objective of this study was to compare the methodology according to LEONARDI
77
& ARMENTANO (2003) with an alternative method to determine feed sorting in

periparturient dairy cows fed glycerol.
Twenty-six Holstein multiparous were paired by previous lactation

performance and expected calving date, and housed in individual tie stalls at the
Purdue Dairy Research and Education Center. Cows were randomly assigned to
diets containing either high moisture corn (control) or glycerol, and fed diets
formulated to meet or exceed the NRC (2001) guidelines for 600 kg dairy cattle
from -28 through +56 DRTC.
The ingredients and nutrient composition of the pre- and post-partum
diets are shown in Table 1. The diets were adjusted weekly to account for forage
DM fluctuation. The addition of soybean meal in the prepartum diet and the higher
participation of this ingredient in the postpartum diet were intended to adjust for
the protein that was removed with high moisture corn. Diets were offered once
daily from 06:30 to 07:30 for ad libitum intake (10 to 15% weight backs). Feed
refusals were measured daily and feed intake was determined by difference. The
study was conducted from May/28/2009 to Oct/232009 and animal use and
handling protocols were approved by the Purdue Animal Care and Use
Committee.
78

experimental diets.
Ingredient, % of DM
Corn silage 35.4 35.4 39.0 39.0
Grass hay 13.0 13.0 3.5 3.5
Glycerol - 11.5 - 10.8
Soybean meal - 2.5 10.0 11.0
Megalac R1 - - 0.7 0.7
2
Supplement3,4 15.8 15.8 10.0 10.0
DM, % 50.9 49.4 46.8 46.0
CP, % of DM 16.6 ± 1.00 16.6 ± 1.35 18.2 ± 0.83 18.7 ± 1.00
ADF, % of DM 22.9 ± 1.75 25.5 ± 1.79 19.5 ± 1.77 20.8 ± 2.32
NDF, % of DM 38.0 ± 1.18 42.2 ± 1.35 31.4 ± 2.71 34.2 ± 1.67
Starch, % of DM 22.6 ± 2.64 15.0 ± 1.22 26.7 ± 1.73 19.2 ± 1.12
NEL, Mcal/kg of DM 1.58 ± 0.02 1.61 ± 0.05 1.65 ± 0.02 1.61 ± 0.02
Ca, % of DM 1.09 ± 0.17 1.02 ± 0.12 1.11 ± 0.07 1.11 ± 0.20
P, % of DM 0.36 ± 0.02 0.34 ± 0.02 0.43 ± 0.02 0.40 ± 0.04
Mg, % of DM 0.39 ± 0.04 0.36 ± 0.02 0.36 ± 0.04 0.35 ± 0.02
K, % of DM 1.22 ± 0.05 1.29 ± 0.09 1.47 ± 0.11 1.44 ± 0.03
Na, % of DM 0.15 ± 0.01 0.15 ± 0.01 0.32 ± 0.01 0.32 ± 0.02
1
Church & Dwight Co., Princeton, NJ
2
3
4
Products, Quincy, IL)
79
Samples of corn silage, alfalfa haylage and TMR were collected

weekly, dried in a forced-air oven for 72 h at 55ºC and ground using a Willey mill to
pass a 1-mm screen. Composite samples were formed monthly and analyzed by a
commercial laboratory (Dairy One, Ithaca, NY) for DM, crude protein (CP), acid
detergent fiber (ADF), starch and minerals by wet chemistry following AOAC
(2000) procedures, and for NDF following the method of GOERING & VAN
SOEST (1970).
Feed sorting was determined at -16, -9, +9, +16 and +51 DRTC
(targeted days) by two methodologies. In both, samples of diets were taken at 0, 4,
8, 12 and 24 hours after feed delivery and frozen at -20ºC. Upon thawing, samples
were separated using three screens (19, 8, and 1.18 mm) and a bottom pan
(PSPS) to yield long (>19 mm), medium (<19, >8 mm), short (<8, >1.18 mm) and
fine (<1.18 mm) particles (LAMMERS et al., 1996). Post separated materials were
then dried in a forced-air oven for 72 h at 55ºC. Feed sorting was determined
according to LEONARDI & ARMENTANO (2003) as the actual intake of each
screen of the PSPS consumed between 0-4, 4-8, 8-12 and 12-24 hours post
feeding, and expressed as a percentage of the predicted intake of that
correspondent screen. The actual intake for each individual particle size equals
the product of the DMI between 0-4, 4-8, 8-12 and 12-24 hours post feeding
multiplied by the particle size distribution within the same intervals, whereas the
predicted intake for each individual particle size equals the product of the DMI
between 0-4, 4-8, 8-12 and 12-24 hours post feeding multiplied by the particle size
distribution of the offered TMR. The final outcome is that values equal to 100%
indicate no sorting, <100% indicate selective refusals (sorting against), and >100%
indicate preferential consumption (sorting for). In the alternative methodology
proposed in this study, feed intake of the individual particle sizes during each
interval was determined by subtracting the particle size distribution available at the
beginning of each interval from the particle size distribution remaining at the end of
the interval, and feed sorting was determined by measuring the particle size profile
of feed consumed between 0-4, 4-8, 8-12 and 12-24 hours relative to feed
delivery.
The data were analyzed using the MIXED procedure (SAS, 1999).
The model accounted for the effects of treatments, DRTC and times post feeding
80
(0-4, 4-8, 8-12 and 12-24 hours), as well as the interaction effect of treatment by
DRTC, treatment by time, DRTC by time and treatment by DRTC by time. Means
were different if P<0.05 and values are reported as least square means and
associated standard errors.
3 RESULTS
Twenty-three cows completed the study (12 in the glycerol group and
11 in the control group). Two cows were removed due to displaced abomasum just
after parturition (3935; control group and 4090; glycerol group) and one cow
(3846; control group) experienced uterine torsion 7 days prior to parturition. Data
from these cows were used in the prepartum parameters. Due to imprecision
associated with target calving dates, the actual means and associated standard
deviations of feed sorting sampling days were -16 (±3.6), -9 (±3.7), +11 (±1.6), +18
(±2.4) and +53 (±1.8) DRTC. Feed intake did not differ (P>0.05) between
treatments and was 14.7 ± 0.4 and 20.2 ± 0.5 kg/d for the pre- and post-partum
intervals, respectively.
There was effect of treatments on the particle size distribution of the
prepartum experimental diets. Adding glycerol increased (P<0.05) the DM%
retained as long particles (>19 mm). Contrarily, the DM% retained as short (<8,
>1.18 mm) and fine (<1.18 mm) particles was reduced (P<0.05) with glycerol
inclusion. There was no response (P>0.05) to prepartum glycerol on the DM%
retained as medium particles (<19, >8 mm) and no response (P>0.05) of
treatments on the particle size distribution of the postpartum diets (Table 2).
81

Item Control Glycerol SEM P-values
Prepartum
% of DM retained on sieve
>19 mm 9.6 16.9 0.4 <0.05
<19, >8 mm 31.0 31.5 0.7 0.58
<8, >1.18 mm 41.7 37.5 0.6 <0.05
<1.18 mm 17.7 14.1 0.4 <0.05
Postpartum
% of DM retained on sieve
>19 mm 5.7 7.1 0.5 0.10
<19, >8 mm 36.8 37.7 0.9 0.47
<8, >1.18 mm 42.9 41.5 0.6 0.16
<1.18 mm 14.6 13.7 0.7 0.36
Results of feed sorting according to the methodology from

LEONARDI & ARMENTANO (2003) are presented in Figures 1 and 2, whereas
according to the alternative methodology proposed in this study are shown in
Figures 3 and 4. Cows fed prepartum glycerol increased (P<0.05) the preference
or sorting for long particles (>19 mm) according to the methodology from
LEONARDI & ARMENTANO (2003) (77.2 vs. 101.5%, control vs. glycerol, Figure
1) and also in the alternative methodology (9.2 vs. 17.8%, control vs. glycerol,
Figure 3). There was no response (P>0.05) of treatments on feed sorting for
medium particles (<19, >8 mm) during the prepartum period in both methodologies
(116 vs. 117.4%, control vs. glycerol, LEONARDI & ARMENTANO, 2003, Figure
1; 30.9 vs. 31.3%, control vs. glycerol, alternative methodology, Figure 3). Cows
fed prepartum glycerol sorted against (P<0.05) short particles (<8, >1.18 mm) in
both methodologies (102.6 vs. 94.2%, control vs. glycerol, LEONARDI &
ARMENTANO, 2003, Figure 1; 42 vs. 37.3%, control vs. glycerol, alternative
methodology, Figure 3). Conversely, there was no response (P>0.05) of diet on
feed sorting for fine particles (<1.18 mm) according to the methodology from
LEONARDI & ARMENTANO (2003) (78.3 vs. 76.1%, control vs. glycerol, Figure
1), but cows fed prepartum glycerol decreased (P<0.05) the preference for fine
82
particles (<1.18 mm) in the alternative methodology (17.9 vs. 13.6%, control vs.
glycerol, Figure 3).
83
Long particles (>19 mm) Medium particles (<19, >8 mm)
120
Particle size distribution
160

100 140
80 120
100
60 80
40 60
20 40
20
0 0
4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
Time post feeding (h) Time post feeding (h)
Short particles (<8, >1.18 mm) Fine particles (<1.18 mm)
120
100

100
80
80
60
60
40 40
20 20
0 0
4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
FIGURE 1- Feed sorting of transition dairy cows fed control (solid bars) or glycerol (open bars) during the prepartum interval according to
LEONARDI & ARMENTANO (2003). Long particles – treatment (P<0.05), DRTC (P>0.05), time (P<0.05), treatment × DRTC
(P>0.05), treatment × time (P<0.05), treatment × DRTC × time (P>0.05). Medium particles – treatment (P>0.05), DRTC (P>0.05),
time (P<0.05), treatment × DRTC (P>0.05), treatment × time (P>0.05), treatment × DRTC × time (P>0.05). Short particles –
treatment (P<0.05), DRTC (P>0.05), time (P>0.05), treatment × DRTC (P>0.05), treatment × time (P<0.05), treatment × DRTC ×
time (P>0.05). Fine particles – treatment (P>0.05), DRTC (P>0.05), time (P<0.05), treatment × DRTC (P>0.05), treatment × time
(P>0.05), treatment × DRTC × time (P>0.05).
84
140
160

120 140
100 120
80 100
60 80
60
40
40
20
20
0 0
4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
120
120

100 100
80 80
60 60
40 40
20 20
0 0
4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
FIGURE 2- Feed sorting of transition dairy cows fed control (solid bars) or glycerol (open bars) during the postpartum interval according to
LEONARDI & ARMENTANO (2003). Long particles – treatment (P>0.05), DRTC (P<0.05), time (P>0.05), treatment × DRTC
(P>0.05), treatment × time (P>0.05), treatment × DRTC × time (P>0.05). Medium particles – treatment (P<0.05), DRTC (P<0.05),
time (P<0.05), treatment × DRTC (P>0.05), treatment × time (P<0.05), treatment × DRTC × time (P>0.05). Short particles –
treatment (P<0.05), DRTC (P<0.05), time (P>0.05), treatment × DRTC (P>0.05), treatment × time (P<0.05), treatment × DRTC ×
time (P>0.05). Fine particles – treatment (P>0.05), DRTC (P<0.05), time (P<0.05), treatment × DRTC (P>0.05), treatment × time
(P<0.05), treatment × DRTC × time (P>0.05).
85
Feed sorting during the postpartum interval did not differ (P>0.05)
between treatments in both methodologies for long particles (>19 mm) (98.6 vs.
108.9%, control vs. glycerol, LEONARDI & ARMENTANO, 2003, Figure 2; 5.8 vs.
7%, control vs. glycerol, alternative methodology, Figure 4). However, cows fed
postpartum glycerol increased (P<0.05) the preference for medium particles (<19,
>8 mm) according to the methodology from LEONARDI & ARMENTANO (2003)
(108.6 vs. 116.5%, control vs. glycerol, Figure 2), but did not (P>0.05) according to
the alternative methodology (36.1 vs. 37%, control vs. glycerol, Figure 4).
Conversely, cows fed postpartum glycerol sorted against (P<0.05) short particles
(<8, >1.18 mm) according to the methodology from LEONARDI & ARMENTANO
(2003) (100.6 vs. 96.6%, control vs. glycerol, Figure 2), and again did not (P>0.05)
according to the alternative methodology (42.2 vs. 41.6%, control vs. glycerol,
Figure 4). There was no effect (P>0.05) of diet on feed sorting for fine particles
(<1.18 mm) in both methodologies (85.4 vs. 72.4%, control vs. glycerol,
LEONARDI & ARMENTANO, 2003, Figure 2; 15.8 vs. 14.4%, control vs. glycerol,
alternative methodology, Figure 4).
86
25
40

20 35
30
15 25
20
10
15
5 10
5
0 0
4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
50
25

40 20
30 15
20 10
10 5
0 0
4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
FIGURE 3- Feed sorting of transition dairy cows fed control (solid bars) or glycerol (open bars) during the prepartum interval according to the
alternative methodology proposed in this study. Long particles – treatment (P<0.05), DRTC (P>0.05), time (P>0.05), treatment ×
DRTC (P>0.05), treatment × time (P>0.05), treatment × DRTC × time (P>0.05). Medium particles – treatment (P>0.05), DRTC
(P>0.05), time (P<0.05), treatment × DRTC (P>0.05), treatment × time (P>0.05), treatment × DRTC × time (P>0.05). Short
particles – treatment (P<0.05), DRTC (P>0.05), time (P>0.05), treatment × DRTC (P>0.05), treatment × time (P>0.05), treatment ×
DRTC × time (P>0.05). Fine particles – treatment (P<0.05), DRTC (P>0.05), time (P<0.05), treatment × DRTC (P>0.05), treatment
× time (P>0.05), treatment × DRTC × time (P>0.05).
87
12
50

10
40
8
30
6
4 20
2 10
0 0
4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
50
25

40 20
30 15
20 10
10 5
0 0
4 8 12 24 4 8 12 24 4 8 12 24 4 8 12 24
FIGURE 4- Feed sorting of transition dairy cows fed control (solid bars) or glycerol (open bars) during the postpartum interval according to the
alternative methodology proposed in this study. Long particles – treatment (P>0.05), DRTC (P<0.05), time (P>0.05), treatment ×
DRTC (P>0.05), treatment × time (P>0.05), treatment × DRTC × time (P>0.05). Medium particles – treatment (P>0.05), DRTC
(P>0.05), time (P<0.05), treatment × DRTC (P>0.05), treatment × time (P>0.05), treatment × DRTC × time (P>0.05). Short
particles – treatment (P>0.05), DRTC (P<0.05), time (P>0.05), treatment × DRTC (P>0.05), treatment × time (P>0.05), treatment ×
DRTC × time (P>0.05). Fine particles – treatment (P>0.05), DRTC (P>0.05), time (P<0.05), treatment × DRTC (P>0.05), treatment
× time (P>0.05), treatment × DRTC × time (P>0.05).
88
4 DISCUSSION
The response to prepartum glycerol on the preference for long

particles (>19 mm) in both methodologies can be explained by the fact that the
proportion of DM retained as long particles (>19 mm) in the prepartum glycerol diet
was increased compared with the prepartum control diet (Table 2). This happened
because the glycerol adhesive property increased the weight only on the long
particles (>19 mm) of the prepartum diet, but not on the medium (<19, >8 mm),
short (<8, >1.18 mm) and fine particles (<1.18 mm), and yet, no effect of glycerol
coating was observed on the particle size distribution of the postpartum diet. Thus,
it can be inferred that the greater response to glycerol coating on the long particles
(>19 mm) was related to a higher proportion of long-stem forage in the glycerol
prepartum diet (13% of hay of the total DM ration) compared with the glycerol
postpartum diet (3.5% plus 1.5% of hay and straw of the total DM ration,
respectively), which may have resulted in a greater density of the long particles
(>19 mm) due to more surface exposure for glycerol coating on the long-stemmed
forage.
Overall, discrepancies of determinations of feed sorting between
methodologies occurred three times. During the prepartum stage, there was no
effect of treatments on feed sorting of fine particles (<1.18 mm) according to the
methodology from LEONARDI & ARMENTANO (2003), but cows fed glycerol
decreased the preference for those particles according to the alternative
methodology. During the postpartum interval, cows fed glycerol increased the
preference for medium particles (<19, >8 mm) according to the methodology from
LEONARDI & ARMENTANO (2003), but did not according to the alternative
methodology, as well as cows fed glycerol sorted against short particles (<8, >1.18
mm) according to the methodology from LEONARDI & ARMENTANO (2003), and
again did not according to the alternative methodology.
Finally, it can be stated that the alternative methodology proposed in
this study is more reliable than the standard method based on the fact that feed
sorting was determined as the particle size distribution relative to feed intake of the
individual particle sizes during each interval, and considering the amount of feed
that was available to the cow exclusively in the beginning of each interval as a
89
baseline to determine feed sorting in the end of each interval. However,

LEONARDI & ARMENTANO (2003) considered the predicted intake as a baseline
to determine feed sorting, and the problem is that cows will always sort for or
against the different components of the diet over the course of the day, and
consequently alter the particle size profile of the TMR, thereby impacting the
results of feed sorting in the end of the first interval and also in the following
intervals. Another strong reason to corroborate that the alternative methodology is
more reliable is that cows fed prepartum glycerol had significant differences on
feed sorting only when there were significant variations on the particle size
distribution of the experimental diets (Table 2). As reported previously, the addition
of prepartum glycerol increased the DM proportion of long particles (>19 mm) and
decreased the DM proportion of short (<8, >1.18 mm) and fine (<1.18 mm)
particles. Likewise, cows fed prepartum glycerol increased the preference for long
particles (>19 mm) and decreased the preference for short (<8, >1.18 mm) and
fine (<1.18 mm) particles. The same pattern was observed during the postpartum
interval, when there was neither effect of diet on the particle size distribution of the
experimental diets, nor on feed sorting according to the alternative methodology.
Nonetheless, additional research is needed to investigate whether or not the
alternative methodology proposed in this study will bring the same pattern of
results through experiments using cows fed other ingredients other than glycerol
that do not have adhesive properties on long-stemmed forage particles.
5 CONCLUSIONS
The alternative methodology proposed in this study is more reliable

than the standard methodology in the determination of feed sorting.
The glycerol adhesive property enhanced the preference for long
particles (>19 mm) of the diet during the prepartum interval, which may prove
beneficial to rumen health.
90
6 REFERENCES
AOAC. Official methods of analysis. 17th ed. AOAC Int., Gaithersburg, MD.
2000.
BHANDARI, S.K.; LI, S.; OMINSKI, K.H.; WITTENBERG, K.M.; PLAIZIER, J.C.
Effects of the chop lengths of alfalfa silage and oat silage on feed intake, milk
production, feeding behavior, and rumen fermentation of dairy cows. Journal of
Dairy Science, v.91, p.1942-1958, 2008.
COOK, N.B.; NORDLUND, K.V.; OETZEL, G.R. Environmental influences on claw

horn lesions associated with laminitis and subacute ruminal acidosis in dairy cows.
DEVRIES, T.J.; BEAUCHEMIN, K.A.; KEYSERLINGK, M.A.G. von. Dietary forage

concentration affects the feed sorting behavior of lactating dairy cows. Journal of
Dairy Science, v.90, p.5572-5579, 2007.
DEVRIES, T.J.; DOHME, F.; BEAUCHEMIN, K.A. Repeated ruminal acidosis

challenges in lactating dairy cows at high and low risk for developing acidosis:
feed sorting. Journal of Dairy Science, v.91, p.3958-3967, 2008.
DEVRIES, T.J.; KEYSERLINGK, M.A.G. von; BEAUCHEMIN, K.A. Frequency of

feed delivery affects the behavior of lactating dairy cows. Journal of Dairy
Science, v.88, p.3553-3562, 2005.
DONKIN, S.S.; KOSER, S.L.; WHITE, H.M.; DOANE, P.H.; CECAVA, M.J.
DRACKLEY, J.K. Biology of dairy cows during the transition period: the final
frontier? Journal of Dairy Science, v.82, p.2259-2273, 1999.
GOERING, H.K.; VAN SOEST, P.J. Forage fiber analysis (Apparatus,

Reagents, Procedures and Some Applications). Agric. Handbook No. 379.
ARS-USDA, Washington, DC. 1970.
GRUMMER, R.R. Impact of changes in organic nutrient metabolism on feeding the

transition dairy cow. Journal of Animal Science, v.73, p.2820-2833, 1995.
HOOSSEINKHANI, A.; DEVRIES, T.J.; PROUDFOOT, K.L.; VALIZADEH, R.;

VEIRA, D.M.; KEYSERLINGK, M.A.G. von. The effects of feed bunk competition
on the feed sorting behavior of close-up dry cows. Journal of Dairy Science,
v.91, p.1115-1121, 2008.
KRAUSE, K.M.; OETZEL, G. Understanding and preventing subacute ruminal

acidosis in dairy herds: A review. Animal Feed Science and Technology, v.126,
p.215-236, 2006.
91
LAMMERS, B.P.; BUCKMASTER, D.R.; HEINRICHS, A.J. A simple method for

the analysis of particle sizes of forage and total mixed rations. Journal of Dairy
Science, v.79, p.922-928, 1996.
LEONARDI, C.; ARMENTANO, L.E. Effect of quantity, quality, and length of alfalfa
hay on selective consumption by dairy cows. Journal of Dairy Science, v.86,
p.557–564, 2003.
LEONARDI, C.; GIANNICO, F.; ARMENTANO, L.E. Effect of water addition on

selective consumption (sorting) of dry diets by dairy cattle. Journal of Dairy
Science, v.88, p.1043-1049, 2005a.
LEONARDI, C.; SHINNERS, K.J.; ARMENTANO, L.E. Effect of different dietary

geometric mean particle length and particle size distribution of oat silage on
feeding behavior and productive performance of dairy cattle. Journal of Dairy
Science, v.88, p.698-710, 2005b.
LESMEISTER, K.E.; HEINRICHS, A.J. Effects of adding extra molasses to a

texturized calf starter on rumen development, growth characteristics, and blood
parameters in neonatal dairy calves. Journal of Dairy Science, v..88, p.411-418,
2005.
MICHNICK, S.; ROUSTAN, J.L.; REMIZE, F.; BARRE, P.; DEQUIN, S. Modulation
of glycerol and ethanol yields during alcoholic fermentation in Saccharomyces
cerevisiae strains overexpressed or disrupted for GPD1 encoding glycerol 3-
phosphate dehydrogenase. Yeast, v.13, p.783-793, 1997.
MILLER-CUSHON, E.K.; DEVRIES, T.J. Effect of dietary dry matter concentration

on the sorting behavior of lactating dairy cows fed a total mixed ration. Journal of
Dairy Science, v.92, p.3292-3298, 2009.
NRC. Nutrient Requirements of Dairy Cattle. 7th rev. ed. National Academy of
Sciences, Washington, DC. 2001.
OELKER, E.R.; REVENEAU, C.; FIRKINS, J.L. Interaction of molasses and

monensin in alfalfa hay- or corn silage-based diets on rumen fermentation, total
tract digestibility, and milk production by Holstein cows. Journal of Dairy
Science, v.92, p.270-285, 2009.
SAS Institute. 1999. SAS/STAT User’s Guide: Statistics, Version 9.2 Edition. SAS
Inst. Inc., Cary, NC.
STONE, W.C. Nutritional approaches to minimize subacute ruminal acidosis and

laminitis in dairy cattle. Journal of Dairy Science, v.87, p.13-26, 2004.
STRICKLIN, W.R.; KAUTZ-SCANAVY, C.C. The role of behavior in cattle

production: a review of research. Applied Animal Ethology, v.11, p.359-390,
1983.
92
THOMPSON, J.C.; HE, B.B. Characterization of crude glycerol from biodiesel

p.261-265, 2006.
CAPÍTULO 5 – CONSIDERAÇÕES FINAIS
Os resultados do presente estudo demonstram que o glicerol pode

ser recomendado na substituição do milho grão úmido em rações de vacas
leiteiras periparturientes em 11,5 e 10,8% do total da MS durante o pré e pós-
parto, respectivamente, baseado na ausência de efeitos deletérios sobre o CMS,
produção e composição do leite, parâmetros sanguíneos e comportamento
ingestivo. Além disso, o glicerol demonstrou ter efeito adesivo sobre as partículas
longas (>19 mm) da dieta, o que trouxe aumentos na taxa do CMS e preferência
por essas partículas, podendo implicar em melhoria na saúde do rúmen, o que
ainda deve ser comprovado.
Todavia, são necessárias mais pesquisas sobre os motivos do
aumento do butirato como produto final da fermentação ruminal, o que pode trazer
consequências metabólicas indesejáveis às vacas leiteiras no período de
transição, tais como o aumento de corpos cetônicos e maior propensão à cetose
ou acetonemia. Além disso, é necessário também conhecer o limite máximo de
metanol tolerável aos ruminantes, tendo em vista que quanto maior o grau de
refino do glicerol realizado nas indústrias de biocombustíveis, maior será o preço
final pago pelo produtor rural, fazendo com que o glicerol seja menos competitivo
economicamente na sua substituição ao milho.
Em relação ao efeito do glicerol sobre o aumento no consumo e
preferência por partículas longas (>19 mm) da dieta verificado nesse trabalho,
resta saber se esse padrão de comportamento seria mantido no caso de
alojamento de vacas em grupos e não individualmente, em que a hierarquia entre
animais passa a ser determinante sobre o consumo seletivo de determinadas
partículas da dieta. Outra possibilidade de uso do glicerol está na ensilagem da
mistura entre o bagaço de cana-de-açúcar in natura e o glicerol na alimentação de
bovinos de corte em confinamento, o que poderia reduzir o custo de produção em
93
função da utilização simultânea de dois subprodutos como fonte de volumoso e

concentrado energético.

Tesis Rodriges de Carvalho

Enviado por

Dados do documento

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Tesis Rodriges de Carvalho

Enviado por

Direitos autorais:

Formatos disponíveis

i

UNIVERSIDADE FEDERAL DE GOIÁS

VALOR NUTRITIVO DO GLICEROL E COMPORTAMENTO

Eduardo Rodrigues de Carvalho

UNIVERSIDADE FEDERAL DE GOIÁS

VALOR NUTRITIVO DO GLICEROL E COMPORTAMENTO

Tese apresentada ao Programa de Pós-

Ao meu grande amigo João Carlos Cardoso da Silva (in memorian)

ABHB ácido β-hidroxibutírico

também sobre o comportamento ingestivo durante todo o período experimental.

Palvras-chave: biocombustíveis, consumo seletivo, energia, preferência,

CAPÍTULO 1 – CONSIDERAÇÕES GERAIS

Apesar da taxa do crescimento da população mundial ter se

transporte e armazenamento e ainda do preço do milho como balizador do preço

2.1 Propriedades químicas do glicerol e seu uso potencial como ingrediente

O glicerol refinado é um líquido viscoso, levemente adocicado,

A maioria das plantas industriais utiliza uma razão molar álcool:óleo

Triglicerídeo Metanol Biodiesel Glicerol

FIGURA 1- Reação de transesterificação para produção de biodiesel (VAN

Dentro do escopo dessa revisão de literatura, encontrou-se apenas

se evitar qualquer tipo de efeito deletério de contaminantes sobre o desempenho

2.2 Glicerol no tratamento da cetose ou acetonemia

Nas pesquisas iniciais com glicerol, esse subproduto foi utilizado

2.3 Metabolismo do glicerol no rúmen

No rúmen, o glicerol é convertido a AGCC ou existe a hipótese de

A fermentação in vitro de glicerol utilizando-se fluido ruminal de

a maior expressão de AQP-7 no RNAm ocorre no tecido omental (WINTOUR &

2.4 A vaca leiteira no período de transição

A vaca leiteira periparturiente ou em período de transição tem sido

sanguínea (BELL, 1995), sendo que existe correlação negativa entre

introdução de grãos na dieta imediatamente após o parto tem efeitos deletérios

melhor entendimento das taxas de incidência, patogênese, diagnóstico,

convertidos a corpos cetônicos dentro da mitocôndria ou ainda serem

AGRE, P. The aquaporin water channels. Proceedings of the American

BELL, A. W. Regulation of organic nutrient metabolism during transition from late

BERGMAN, E. N. Hyperketonemia-ketogenesis and ketone body metabolism.

BERGNER, H.; KIJORA, C.; CERESNAKOVA, Z.; SZAKÁCS, J. In vitro studies on

BERTICS, S. J.; GRUMMER, R. R.; CADORNIGA-VALINO, C.; STODDARD, E. E.

CZERKAWSKI, J. W.; BRECKENRIDGE, G. Fermentation of various glycolytic

DEFRAIN, J. M.; HIPPEN, A. R.; KALSCHEUR, K. F.; JARDON, P. W. Feeding

DEPARTMENT OF ECONOMIC AND SOCIAL AFFAIRS OF THE UNITED

DIRKSEN, G. U.; LIEBICH, H. G.; MAYER, E. Adaptive changes of the ruminal

DONKIN, S. S; DOANE, P. Glycerol as a feed ingredient in dairy rations.

DOUGLAS, G. N.; OVERTON, T. R.; BATEMAN II, H. G.; DANN, H. M.;

DRACKLEY, J. K.; RICHARD, M. J.; BEITZ, D. C.; YOUNG, J. W. Metabolic

DUFFIELD, T. Subclinical ketosis in lactating dairy cattle. Veterinary Clinics of

FDA. Code of Federal Regulations, 21 CFR 582.1320, Title 21, Vol. 6,

FISHER, L. J.; ERFLE, J. D.; LODGE, G. A.; SAUER, F. D. Effects of propylene

FISHER, L. J.; ERFLE, J. D.; SAUER, F. D. Preliminary evaluation of the addition

FRONK, T. J.; SCHULTZ, L. H.; HARDIE, A. R. Effect of dry period

GARTON, G. A.; LOUGH, A. K.; VIOQUE, E. Glyceride hydrolysis and glycerol

GOFF, J. P.; HORST, R. L. Physiological changes at parturition and their

GOFF, J. P.; HORST, R. L. Oral glycerol as an aid in the treatment of ketosis/fatty

GREENFIELD, R. B.; CECAVA, M. J.; JOHNSON, T. R.; DONKIN, S. S. Impact of

GRUMMER, R. R. Impact of changes in organic nutrient metabolism on feeding

HAYIRLI, A.; GRUMMER, R. R.; NORDHEIM, E. V.; CRUMP, P. M. Animal and

HERDT, T. H. Ruminant adaptation to negative energy balance. Veterinary

INGVARTSEN, K. L.; ANDERSEN, J. B. Integration of metabolism and intake

JOHNS, A. T. Fermentation of glycerol in the rumen of sheep. New Zealand

JOHNSON, R. B. The treatment of ketosis with glycerol and propylene glycol.

JORRITSMA, R.; JORRITSMA, H.; SCHUKKEN, Y. H; WENTINK, G H.

MACH, N.; BACH, A.; DEVANT, M. Effects of crude glycerin supplementation on

MORROW, D. A.; HILLMAN, D.; DADE, A. W.; KITCHEN, H. Clinical investigation

NATIONAL RESEARCH COUNCIL. Nutrient Requirements of Dairy Cattle. 7ed.