Relationship between ruminal degradability and chemical composition of dehydrated lucerne
Relationship between ruminal degradability and chemical composition of dehydrated lucerne
José Luis REPETTO 0
r GONZÁLEZ 0
María Remedios ALVIR 0
Carlos Alberto RODRÍGUEZ 0
0 Departamento de Producción Animal, Escuela Técnica Superior de Ingenieros Agrónomos, Universidad Politécnica de Madrid , 28040 Madrid , Spain
- The rumen degradation characteristics and effective degradability (ED) of dry matter (DM) and crude protein (CP) of 21 samples of dehydrated lucerne from 5 different processing plants were studied in three rumen fistulated wethers using the nylon bag technique. The animals were fed at an intake level of 40 g DM·kg-1 BW0.75 with a mixed diet of lucerne (40% dehydrated and 60% hay) and concentrate (2:1 on DM). The mean values of the ED of DM and CP, calculated for rumen outflow rates determined in each animal (2.28%·h-1, as average), presented a low variation (mean = 61.2%; CV = 5.31% for DM and mean = 73.5%; CV = 4.95% for CP). Degradation of DM was directly related to lucerne quality, with negative and positive correlations with the contents of fibre and CP, respectively. The best prediction of the ED of DM was derived from the contents of acid detergent fibre, which explained 73.3% of the total variation. The best prediction of the ED of CP (R2 = 0.592) was related negatively to the proportion of insoluble nitrogen in neutral detergent fibre and positively to the CP concentration as the first and second predictive variables. However, the first variable allowed a good estimation of the ED of CP for the dehydrated lucerne samples presented in a long form (R2 = 0.818). Degradation studies of DM and CP also indicated that most of the N available to animals was derived from rumen microbial synthesis.
foin) et d’aliment concentré dans des proportions 2:1 (sur MS) à un niveau de 40 g MS·kg–1 P0,75. Les
valeurs de DT, calculées avec les taux de sortie des particules du rumen mesurés pour chaque animal
(moyenne = 2,28 %·h–1), ont été peu variables tant pour la MS (moyenne = 61,2 % ; CV = 5,31 %)
que pour les MAT (moyenne = 73,5 % ; CV = 4,95 %). La DT de la MS a été directement liée à la
qualité de la luzerne, avec des corrélations négatives et positives pour les teneurs en fibre et MAT,
respectivement. La meilleure prédiction de la DT de la MS a été obtenue avec la teneur en lignocellulose
(ADF), qui explique 73,3 % de la variation totale. Pour les MAT, la meilleure prédiction de la DT
(R2 = 0,592) a été négativement corrélée à la proportion d’azote insoluble dans la solution au
détergent neutre et positivement à la teneur en MAT, respectivement première et seconde variables
indépendantes. La première variable a permis une prédiction convenable de la DT des MAT pour la
luzerne déshydratée présentée sous forme de brins longs (R2 = 0.818). Les études de dégradation de la
MS et des MAT ont aussi montré que la plus grande partie de l’azote disponible pour l’animal est
fournie par la synthèse microbienne dans le rumen.
luzerne déshydratée / dégradabilité ruminale / protéine / composition chimique / ovins
Reliable estimates of rumen degradability
of feed protein are essential for the sensible
application of new systems for assessing
protein nutrition in ruminants. However,
these systems use fixed values of crude
protein (CP) degradability for each feed type
and do not integrate the possible intratype
variability existing for this trait.
Consequently, studies focused at predicting
degradability or determining the factors
affecting this variability are of interest,
specially for the most commonly used feeds.
Lucerne is characterised by a high content
of protein with an extensive rumen
degradation, which may reduce its protein value
as a consequence of important nitrogen
losses by ammonia absorption from the
]. Therefore, the knowledge of the
possible factors conditioning rumen
degradability are useful to improve
rationing and, consequently, the nitrogen
utilisation of this feed. Previous results on the
effective degradability (ED) of CP of
dehydrated lucerne show a great variability.
Mean values obtained by the NRC [
Verité et al. [
], Kowalski et al. [
Faria-Marmol et al. [
] and Repetto et al.
] were respectively: 41, 60, 64, 71 and
77.5%. Rumen degradability of dehydrated
lucerne should be affected by multiple
factors related to the crop, the post-harvest
forage management and the industrial
process of dehydration. The purposes of this
research were: (i) to obtain information on
the variation of rumen degradability of
dehydrated lucerne and (ii) to study the effect
of chemical composition and CP solubility
of dehydrated lucerne on rumen
2. MATERIALS AND METHODS
2.1. Experimental procedures
A total of 21 samples of dehydrated lu
cerne presented in a long form (17) or
pelleted (4) were studied. These samples
were obtained during the same harvest
season from 5 plants (A, B, C, D and E) located
in the Lleida province (north-east Spain).
The plants differed in technology and
management of the post-harvest and industrial
processes, except for the harvest stage,
which was always at the beginning of
flowering (about 5% of the plants). Plant A
processed the lucerne directly, whereas the
other plants made a field wilting period of
different duration (between 12 and 72 h).
All plants employed rotational driers, with
one pass (plants B, D and E) or three passes
(plants A and C). The mean technical
specifications of the dehydration process and
the initial and final DM content of the
vegetable material are exposed in Table I.
These latter values were determined on a
total of 48 samples employed to select the
17 non conditioned samples employed in
this study. All granulated samples were
obtained from plant E. All materials were
ground to pass a 2 mm screen for
degradability trials and 1 mm screen for
chemical analysis and buffer CP solubility
All forages were incubated by the in situ
method into the rumen of three rumen
cannulated wethers of the Manchega breed,
which were fed with a 2/3 chopped (6–8 cm)
forage (40% dehydrated lucerne and 60%
lucerne hay) and 1/3 concentrate diet,
distributed at an intake level of 40 g DM·kg–0.75
in two equal-weight meals at 9:00 and
17:00 h. The NDF and CP contents of this
ration were 435 and 161 g·kg–1 (on DM),
The samples of dehydrated lucerne were
incubated in nylon blutex bags (46 µm pore
size; reference: 120 T, Tissages Tissues
Techniques, France) of 11 × 7 cm (internal
dimensions), heat-sealed and filled with
approximately 3 g (air dry basis) of feed
samples. The bags were incubated in the rumen
for each wether at times of 3, 6, 12, 24, 48
and 72 h. Two series of incubation were
conducted for each feed in two successive
periods, in order to have two bags per
animal and incubation time. Within each
series, the rings of 12 bags were
sequentially incubated. Each ring encompassed a
randomly associated pair of dehydrated
lucerne samples. All bags of each ring were
inserted at the morning feeding time. After
incubation, the bags were removed, washed
with tap water, and stored frozen. After
being defrosted, the bags were washed 3 times
for 5 min in a turbine washing machine,
dried in an air forced oven for 48 h at 80 oC
and weighed for DM determination. The
residues were homogenised and analysed
for N. Three additional bags of each sample
were reserved for the zero incubation,
which involved the washing procedure
without prior rumen incubation. The
specifications of the bags, their incubation and
the post-incubation treatment of the bags
were in agreement with the method
proposed by Michalet-Doreau et al. [
The degradation characteristics of DM
and CP were described using the model of
Ørskov and McDonald [
]. To determine
the ED of all samples, between both
incubation series of the studied sample, the rate
constant describing the passage from the
rumen of dietary particles (kp) was
determined for the dehydrated lucerne included
in the diet, which had been chopped at
6–8 cm and marked with Ytterbium (Yb) by
], at 10 mg Yb·g–1 of feed. The
procedure of this determination and the
resultant kp values (2.28 ± 0.22%·h–1) has
been previously published [
The samples of dehydrated lucerne were
analysed for dry matter (DM), organic
matter (OM), crude protein (CP) [
detergent fibre (NDF), acid detergent fibre
(ADF) and acid detergent lignin (ADL)
]. Fibre fractions were calculated ash
free. Neutral detergent insoluble nitrogen
(NDIN) and acid detergent insoluble
nitrogen (ADIN) were performed by Kjeldahl
analysis on the NDF and ADF residues,
respectively, and the contents are expressed
as percentage of total nitrogen. The
solubility of CP was determined in Mc Dougall
buffer for 6 h according to the procedure
described by Alvir et al. [
2.3. Statistical analysis
A least-square non-linear iterative process was used to fit the kinetics of DM and CP disappearance by the NLIN procedure of the Statistical Analysis System for
Windows software, version 6.12 [
same programme was employed to perform
correlation analyses between ruminal
degradation characteristics and chemical
composition of the samples. Then,
multivariate regression equations for ED were
obtained by using the stepwise procedure.
The chemical composition of
dehydrated lucerne samples (Tab. II) showed a
moderate variability, except for the
proportion of NDIN (CV = 26.3%), which varied
from 12.3 to 32.9% of total N. The
solubility of CP in buffer also showed an important
variability (from 20.2 to 44.6%). This last
value was related to the proportions of
ADIN (r = –0.577; P < 0.01) and NDIN
(r = –0.550; P < 0.01).
Ruminal degradation characteristics of
DM and CP for dehydrated lucerne samples
are shown in Table III. The disappearance
data of DM and CP fitted well with the
model used, and did not show evidence of
lag time for any sample. Both in DM and
especially in CP results, the fractional
1 % on total N or CP.
C.V.: coefficient of variation; NDIN: neutral detergent insoluble nitrogen; ADIN: acid detergent insoluble
a, b, and u represent soluble, non-soluble degradable, and undegradable fractions, respectively; kd: fractional
degradation rate of fraction b; ED: effective degradability; C.V.: coefficient of variation.
degradation rate (kd) showed higher
variations than the different feed fractions. The
lowest relative dispersion was observed for
the ruminal ED of both DM and CP, which
represented mean values of 61.2 and
73.5%, respectively. The ED of DM was
closely related with their a (r = 0.856;
P < 0.001) and u (r = –0.920; P < 0.001)
fractions, whereas for CP, the ED was
related with both a and b fractions (r = 0.607
and –0.534, respectively; P < 0.01) and
more closely with kd (r = 0.704; P < 0.001).
Correlation coefficients between the
degradation characteristics of DM and CP
and chemical composition including
hemicellulose and cellulose contents
(estimated respectively as NDF-ADF and
ADF-ADL) are presented in Tables IV and
For DM, the degradation characteristics
were mainly correlated with the contents of
fibre (ADF or NDF), but a high positive
correlation was also observed with the
lucerne CP concentration. The content of
HCEL: hemicellulose; CEL: cellulose; for other abbreviations see Tables II and III.
*, **, ***: significant at P < 0.05, P < 0.01 and P <0.001, respectively.
ADF showed the closest correlation with
the ED of DM and the different fractions,
except b, which was mainly correlated with
ADL (r = –0.701; P < 0.001). However, the
effects of ADF should be mainly due to the
cellulose content, since this carbohydrate
showed a similar correlation coefficient
with the ED of DM than ADF. The
degradation rate of DM was related to the NDIN
proportion (r = –0.505; P < 0.05) and, at a
lower level, to the hemicellulose content
(P = –0.455; P < 0.05). The best prediction
of ED of DM (DMED) was derived from
the ADF content (Fig. 1). The range of
application of this equation was only from
275 to 380 g of ADF·kg–1 DM.
The CP solubility was related to the a and
b fractions of CP (r = 0.838 and r = –0.821,
respectively; P < 0.001) and, as a
consequence, to the ED of CP (r = 0.626; P < 0.01).
However, the most close correlation for ED
was observed with NDIN (r = –0.665;
P < 0.01), which also displayed the highest
correlation with kd (r = –0.692; P < 0.001).
This last trait was also affected (P < 0.05)
by the content of both hemicellulose
(r = –0.536) and cellulose (r = 0.488).
Conversely, only this last type of carbohydrate
was related to the undegradable fraction
(r = 0.681; P < 0.001). The best prediction
of ED of CP (CPED) included NDIN and
CP as the first and second predictive
variables (Fig. 2), although its determination
coefficient only reached a moderate level
(R2 = 0.592), mainly as a consequence of
the important deviations observed for the
granulated samples. When these samples
were excluded from the regression, CPED
was only dependent of NDIN (Fig. 3) and a
higher accuracy was observed (R2 = 0.814).
The range of application of NDIN
proportions in both equations was large: from 14
to 33% of total N.
The moderate variability observed in the
study sample for most tested chemical
parameters should be, in part, a consequence
of the similar harvesting stage used for the
different samples. Dehydration of lucerne
leads to an increase in the proportions of
ADIN and specially NDIN and to a
decrease in CP solubility [
]. Therefore, the
high variability observed for the proportion
of NDIN (CV = 26.3%) or the values of CP
solubility (CV = 16.4%) will be related
with the intensity of heating effects derived
from the forage management and industrial
treatments employed for the different
plants. On the contrary, ADIN proportions,
which were slightly higher than the usual
values of fresh lucerne [
5, 7, 17
] and its
moderate variation (CV = 9.68%) showed a
limited and relatively uniform effect of
dehydration on this trait. The proportion of
ADIN has been used to assess overheating
]. Therefore, heat damage was not
suspected in the present samples.
Considered globally, the correlation co
efficients recorded between degradative
characteristics of DM and the contents of
CP or fibre related parameters (Tab. IV)
showed that the ruminal availability of
dehydrated lucerne is logically dependent on
its quality. In this way, the CP content can
be used as an index of its level of rumen
fermentation, but a better evaluation can be
derived from the ADF contents. On the
contrary to what was expected, there were no
relations between the lignin content and the
undegradable fraction or the ED of DM.
Jung and Allen [
] indicated that the
relations between lignin and degradability or
digestibility are consistent when forages
with different maturity stages or from
different species are considered. On the
contrary, these relations are usually not
found for forages of the same species with a
similar maturity stage.
Faría-Mármol et al. [
] indicated that the
differences in protein value due to lucerne
quality are mainly derived from its potential
to synthesise microbial protein in the rumen,
as a consequence of the high CP degradability
of lucerne and of the moderate intestinal
digestibility of the undegraded protein (about
50–55% for dried lucerne). The interest in
using dehydrated lucerne of a high quality
in high productive ruminants is not only
derived from its higher level of rumen
fermentation, but also from its lower effect
on rumen fill, which leads to a higher intake
and, therefore, to a higher total microbial
The values of ED of CP were higher than
those reported by the NRC [
et al. [
] and Kowalski et al. [
] (41, 60,
and 64%, respectively). These
disagreements can be associated with factors related
to sample characteristics and also to
differences in methodology. Thus, most of the
values cited above were calculated for kp
values higher than those used in our work.
On the contrary, our values agreed with
those of Faría-Mármol et al. [
with determined kp values similar to those
observed in our work. The outflow rate of
particles from the rumen is conditioned
mainly by the rumen fill, which is basically
associated with fibrous particles. When
good quality lucerne is the only forage of a
diet consumed at a restricted level, as in this
work or that of Faria-Marmol et al. [
fill will be low. Therefore, rumen pressure,
rumination activity, rumen motility
(contractile and propulsive movements) and
consequently the evacuation through the
reticulo-omasal orifice will also be low. As
a consequence of the low kp values, our
results of ED are logically higher than those
of other works or those recorded in tables
of nutritive value, which have been
calculated using a fixed and higher kp value. On
average, 73.5% of total CP in dehydrated
lucerne was apparently degraded in the
rumen. Nevertheless, when these values were
corrected for microbial contamination
using the equation proposed by Rodriguez et
] (which was obtained using the same
methodology and with a diet with the same
intake level and forage to concentrate
ratio), the results showed that 79.1% of the
CP of these forages is diverted to microbial
ruminal fermentation. The underestimation
of ED, produced by the microbial
contamination, varied from 5.7 to 8.4% and
averaged 7.1%. The high values of corrected
ED were associated with low proportions of
undegradable CP. So, based on the
prediction of Rodriguez et al. [
], 51.4% on
average of fraction u was of microbial origin.
Consequently, the corrected u values
represented only 7.2% of total CP on average.
These authors showed that the microbial
CP in the undegradable CP fraction is
related positively with the cellulose content
and negatively with the N concentration of
feed, which agrees with the correlation
coefficients observed between fraction u and
the chemical composition of the samples.
The lack of a relation between ADIN and
this fraction reinforced the previous
indication on the lack of heat damage.
The fractional degradation rate of CP
was negatively related with the values of
NDIN and hemicellulose. This same
behaviour was observed for DM degradation,
since a close correlation was recorded
between both kd values (r = 0.813; P < 0.001).
The degradation rate of insoluble CP in
dehydrated lucerne is, therefore, associated in
great part to the progression of degradation
in the feed, which seems to be diminished
by the protein-hemicellulose complexes.
Solubility in buffer of CP was not only
positively related to the soluble CP (a), but
also negatively to the potentially degradable
CP (b), since there was a close correlation
between both fractions (r = 0.911; P < 0.001)
as a consequence of the moderate value and
variation of the undegradable fraction.
Nevertheless, the correlation of this trait
with the ED of CP was insufficient to be
considered as a valuable index. Similarly,
the accuracy of the prediction of the ED of
CP based on NDIN and CP only reached an
intermediate level, mainly as a
consequence of the deviation of granulated
lucerne samples, which could have been
mixed with other raw materials (straws,
urea, ...) in the industrial process. However,
NDIN allowed an accurate prediction of
CPED of dehydrated lucerne presented in
the long form.
Using fresh and dried lucerne samples,
Faría-Mármol et al. [
] observed a close
and negative relationship between ED of
CP and the fraction of CP digested in the
intestine, which is a consequence of the
change of the digestion site. Since the ED of
DM and CP were not related to the same
parameters, it can be stated that the reduction
of ED of CP always enhances the feed
protein value, as a consequence of a reduction
of N losses [
] and an increase of the CP
digested in the intestine [
Therefore, the chemical ideotype de
rived from the present results, for a high
protein value in dehydrated lucerne
combines low contents of fibre (especially
lignocellulose) and high proportions of
Financial support was provided by C.I.C.Y.T.
Project AGF 93-0549-CO2.
 Alvir M.R. , González J. , Caja G. , Gálvez J.F. , Relación entre la solubilidad y la degradación ruminal de la proteína en alimentos concentrados , Invest. Agr.: Prod. Sanid. Anim . 2 ( 1987 ) 43 - 52 .
 AOAC, Official Methods of Analysis (15th ed .), Association of Official Analytical Chemists, Arlington, VA , 1990 .
 Broderick G.A. , Alfalfa silage or hay versus corn silage as the sole forage for lactating dairy cows , J. Dairy Sci . 68 ( 1985 ) 3262 - 3271 .
 Faría-Mármol J. , González J. , Rodríguez C.A. , Alvir M.R. , Effect of diet forage to concentrate ratio on rumen degradability and postruminal availability of protein from fresh and dried lucerne , Anim. Sci . 74 ( 2002 ) 337 - 345 .
 Goering H.K. , A laboratory assessment on the frequency of overheating in commercial dehydrated alfalfa samples , J. Anim. Sci . 43 ( 1976 ) 869 - 872 .
 González J. , Rodríguez C.A. , Andrés S.G. , Alvir M.R. , Rumen degradability and microbial contamination of fish meal and meat meal measured by the in situ technique , Anim. Feed Sci. Technol . 73 ( 1998 ) 71 - 84 .
 González J. , Faría-Mármol J. , Rodríguez C.A. , Alvir M.R., Effects of stage of harvest on the protein value of fresh lucerne for ruminants , Reprod. Nutr . Dev. 41 ( 2001 ) 381 - 392 .
 Jung H.G. , Alleng M.S. , Characteristics of plant cell walls affecting intake and digestibility of forages by ruminants , J. Anim. Sci . 73 ( 1995 ) 2774 - 2790 .
 Kowalski Z.M. , Pisulewski P.M.P. , Peyraud J.L. , Kaminski J., The effect of drier outflow temperature on rumen protein degradability and intestinal digestibility of rumen-undegraded protein of dehydrated grass and lucerne , Ann. Zootech. 44 ( Suppl .) ( 1995 ) 88 .
 Michalet-Doreau B. , Vérité R. , Chapoutot P. , Méthodologie de mesure de la dégradabilité in sacco de l' azote des aliments dans le rumen , Bull. Tech. CRZV Theix 69 ( 1987 ) 5 - 7 .
 NRC , Nutrients Requirements of Dairy Cattle (6th ed.), National Academy Press, Washington, DC, 1989 .
 Ørskov E.R. , McDonald I. , The estimation of protein degradability in the rumen from incubation measurements weighted according to rate of passage , J. Agric. Sci. Camb . 92 ( 1979 ) 499 - 503 .
 Repetto J.L. , González J. , Cajarville C. , Effect of dehydration on ruminal degradability of lucerne , Ann. Zootech. 49 ( 2000 ) 113 - 118 .
 Robertson J.B., Van Soest P.J., The detergent system of analysis and its application to human foods , in: James W.P.T., Theander O . (Eds.), The Analysis of Dietary Fibre in Food, Marcel Dekker , New York, 1981 , pp. 123 - 158 .
 Rodríguez C.A. , González J. , Alvir M.R. , Cajarville C. , Underestimation of in situ effective degradability of N due to microbial contamination , in: Lobley G.E., White A. , MacRae C. (Eds.), Book of abstracts of the VIIIth International Symposium on Protein Metabolism and Nutrition , Aberdeen, United Kingdom, 1999 , p. 67 .
 Rodríguez C.A. , González J. , Alvir M.R. , Repetto J.L. , Microbial nitrogen contamination of in sacco ruminal incubated feeds , in: Lobley G.E., White A. , MacRae C. (Eds.), Book of abstracts of the VIIIth International Symposium on Protein Metabolism and Nutrition , Aberdeen, United Kingdom, 1999 , p. 68 .
 Sanderson M.A. , Weding W.F. , Nitrogen in the detergent fibre fractions of temperate legumes and grasses , Grass Forage Sci . 44 ( 1989 ) 159 - 168 .
 SAS, SAS/STAT User's Guide (version 6 , 4th ed.), Statistical Analysis System Institute Inc ., Cary, NC , 1990 .
 Vérité R. , Michalet-Doreau B. , Chapoutot P. , Peyraud J.L. , Poncet C. , Révision du système des protéines digestibles dans l'intestin (PDI) , Bull. Tech. CRZV Theix 70 ( 1987 ) 19 - 34 .