20 J
.Th. Schonewille et al. Livestock Production Science 63 2000 17 –26
and the supernatant was stored in plastic tubes at stuffs and faeces was estimated by atomic absorption
2 188C. An aliquot of the supernatant from the spectroscopy
Perkin-Elmer 3110;
Perkin-Elmer, rumen liquid samples taken at 07:45, 09:00, 11:00,
Norwalk, CT, USA. Calcium and K in feedstuffs 13:00, 15:00 and 17:00 h was centrifuged at 208C at
and K in supernatant of rumen fluid were estimated 30,000 g for 30 min and the supernatant was stored
by atomic absorption spectroscopy and Na in feed- in plastic tubes at 2 188C. From the rumen liquid
stuffs by atomic emission spectroscopy Perkin samples taken at 07:45, 09:00, 11:00, 13:00, 15:00
Elmer 3110. Magnesium in plasma, urine, superna- and 17:00 h, 2 ml non-centrifuged fluid was de-
tant of rumen fluid, and chromium III in rumen proteinised according to the method of Bergmeyer
fluid were measured directly by atomic absorption 1970. After 10 min, the de-proteinised rumen
spectroscopy. The accuracy of each assay run was liquid samples were centrifuged at room temperature
monitored using a commercial reference sample hay at 2700 g for 15 min and the supernatant was
powder, CRM 129; Community Bureau of Refer- collected and stored in plastic tubes at 2 188C until
ence, Brussels, Belgium and laboratory reference analysis of volatile fatty acids VFAs.
samples, and was found to be within 5 deviation from the target values. The combined within and
2.4. Chemical analyses between-run precision of the determinations was
3.0 coefficient of variation. Samples of the feedstuffs were subjected to the
VFAs in rumen fluid were determined gas- Weende analysis. Nitrogen contents were determined
chromatographically Perkin Elmer Autoscan with a by the macro-Kjeldahl method IDF, 1986; a factor
2 m glass column Chromosorb WAW DMCS; 17 of 6.25 was used to convert g of N into crude
neopentyl glycol adipate, 1 H PO . Pivalic acid
3 4
protein. Ether extracts of the feedstuffs were pre- 0.045, w v was added to the de-proteinised
pared according to the AOAC 1984; the solvent ruminal fluid as an internal standard.
was evaporated and the crude-fat residue weighed. The crude fibre contents of the feedstuffs were
2.5. Statistical analyses estimated using the Fibertec System M2 Tecator,
Stockholm, Sweden. The NDF, ADF and ADL All data were checked for normal distribution
content of the feedstuffs were estimated according to using the Kolmogorov–Smirnov test and then were
the methods described by Goering and Van Soest subjected to analysis of variance ANOVA with
1970. To determine the starch content of the animal, experimental period and dietary treatment as
feedstuffs, they were enzymatically treated with factors Wilkinson, 1990. Rumen-liquid mineral
amyloglucosidase from
Aspergilles niger
EC data were subjected to repeated measurement analy-
3.2.1.3 to hydrolyse all starch to glucose Keppler ses with animal, experimental period and dietary
and Decker, 1974. Subsequently, glucose was mea- treatment as factors Wilkinson, 1990. When a
sured enzymatically with a test combination Boeh- dietary factor had a statistically significant influence,
ringer Mannheim Diagnostica, Mannheim, Germany Bonferroni’s test was used to identify the diets that
and a
computerised autoanalyser
Beckman had different effects on the variable involved.
Synchron CX Systems; Beckman, Mijdrecht, The
Throughout, the level of statistical significance was Netherlands. The free glucose content of the feed-
pre-set at P , 0.05. stuffs was measured directly. Starch was calculated
as total glucose minus free glucose. For the ex- perimental feedstuffs, i.e., the control mix and the
3. Results
two types of cornmeal, the in-vitro rate of fermen- tation was determined according to the method
3.1. In-vitro fermentation of the dietary variables described by Cone et al. 1996. Prior to the de-
termination of the selected minerals in feedstuffs and To check the fermentation profiles of the dietary
faeces, the samples were ashed 4808C for 6 h and variables, the control mix, native and popped cor-
dissolved in 15 ml 4 M HCl. Magnesium in feed- nmeal were incubated with a rumen culture and total
J .Th. Schonewille et al. Livestock Production Science 63 2000 17 –26
21
gas production was measured Cone et al., 1996. 3.3. Magnesium balance and plasma Mg
The high rate and subsequent decline in the rate of gas production during the first hour of the incuba-
Magnesium intake was similar for all five dietary tions Fig. 1 probably represents the degradation of
treatments Table 3. Upon ANOVA, faecal excre- the water-soluble fraction of the experimental feed-
tion of Mg was found to be significantly affected by stuffs Cone et al., 1997. After 5 h of incubation,
the factor diet P 5 0.013, but significant differences the incubation containing the popped cornmeal had
between specific rations could not be identified by the highest rate of gas production. After 10 h, the
Bonferroni’s t-test. Consequently, Mg absorption and rate had fallen to approximately 20 of the peak
urinary excretion were similar for the treatments. value, but now the control mix and native cornmeal
The dietary treatments did not affect P 5 0.205 produced their peak values. After 20 h the rate of
plasma Mg. For all treatments combined, mean total gas production had returned to baseline values
plasma Mg concentration was 0.80 mmol l S.E. for the three feedstuffs.
0.014; n 5 5. 3.2. Feed intake and body weight
3.4. Rumen liquid: Mg and K concentration, pH and VFAs
Experimental rations were consumed completely throughout the experiment. Mean body weight at the
Mg and K concentrations in rumen liquid, before end of the experiment was 746 kg S.E. 16.0; n 5 5,
the morning feed and also after feeding, were not which was almost identical P 5 0.803; paired t-test
significantly influenced by the dietary treatments to pre-experimental values 744 kg, S.E. 14.0; n 5
Table 4. Post-prandial mean Mg values were 5.
highest after feeding the rations with either 20
Fig. 1. In-vitro gas production rates after incubation of a rumen culture with either the control mix h, native cornmeal d or popped cornmeal s.
22 J
.Th. Schonewille et al. Livestock Production Science 63 2000 17 –26 Table 3
Magnesium balance after the feeding of the experimental rations Control
Native cornmeal Popped cornmeal
Pooled S.E.M. P
10 starch 20 starch
10 starch 20 starch
of dm of dm
of dm of dm
a
Intake g d 12.6
12.3 12.1
12.4 12.2
nd nd
Faeces g d 11.9
11.8 11.5
11.5 11.4
0.094 0.013
Absorption g d
0.7 0.5
0.6 0.9
0.8 0.094
0.206 of intake
5.6 4.1
5.3 6.7
6.2 0.767
0.213 Urine g d
0.3 0.3
0.3 0.4
0.3 0.073
0.759 Balance g d
0.4 0.2
0.3 0.5
0.5 0.072
0.197
a
nd 5 Not determined because the cows were offered a restricted amount of feed. Table 4
Concentrations of Mg and K in rumen liquid mmol l, total volatile fatty acids VFAs in mmol l and ruminal pH in cows fed the experimental rations
Control Native cornmeal
Popped cornmeal Pooled S.E.M.
P 10 starch
20 starch 10 starch
20 starch of dm
of dm of dm
of dm Ruminal Mg concentration
07.45 0.32
0.28 0.45
0.56 0.35
0.169 0.789
a
Post prandial 0.49
0.60 0.96
0.49 1.03
0.194 0.191
Ruminal K concentration 07.45
23.3 23.5
25.3 23.9
25.0 0.771
0.309 Post prandial
42.0 41.1
42.0 38.3
41.6 0.993
0.102 Ruminal VFA concentration
07.45 93.7
97.6 95.5
95.8 97.1
2.980 0.898
Post prandial 105.0
100.9 102.1
102.3 113.8
3.594 0.141
Ruminal pH 07.45
6.73 6.67
6.71 6.71
6.65 0.045
0.716
b b
c
Post prandial 6.49
6.52 6.52
6.49 6.34
0.040 0.038
a
Post-prandial values are geometrical means for five values; i.e., these values can be considered as an estimation of the area under the curve Wolever and Jenkins, 1986. Cows were given a restricted amount of feed at 08:00 and 17:00 h.
b,c
Values in the same row with different superscript were borderline significantly different P , 0.10, Bonferroni test.
native or popped cornmeal. The ruminal pH de- mean post-prandial concentrations of total VFAs
creased significantly P , 0.05 after feeding, irre- were highest after feeding the ration with 20
spective of dietary treatment. Post-prandial ruminal popped cornmeal, but the increase versus that for the
pH was borderline significantly lower after feeding other treatments was not statistically significant.
of the ration with 20 popped cornmeal instead of Post-prandial concentrations mM of total VFAs,
native cornmeal. acetate, and propionate, and the acetate-to-propionate
Before the morning feed, concentrations of total ratio were similar P . 0.100 for the five rations; for
VFAs were similar for all dietary treatments. After all treatments combined n 5 5, the mean values
feeding, total VFA concentrations were significantly were 104.8 S.E. 2.341, 72.5 S.E. 1.648, 15.1
P , 0.05 increased when either the control ration or S.E. 0.356 and 4.8 S.E. 0.047, respectively. Upon
the ration with 20 popped cornmeal was fed. The ANOVA, post-prandial concentrations of butyrate in
J .Th. Schonewille et al. Livestock Production Science 63 2000 17 –26
23
ruminal fluid were found to be significantly affected 1987 is not yet fully understood. A possibility is
by the factor diet P 5 0.044, but Bonferroni’s t-test that the addition of starch to the ration may stimulate
did not identify specific rations that produced differ- Mg absorption by increasing the rate of ruminal
ent values. The butyrate concentration for all treat- fermentation and decreasing ruminal pH. A lowering
ments combined was 13.8 mM S.E. 0.432 of ruminal pH renders Mg more soluble and thus
more available for transport across the epithelium of 3.5. Rumen volume and passage rate
the rumen which is the major site of Mg absorption in ruminants Rogers and Van ‘t Klooster, 1969;
Rumen volume, absolute and fractional outflow of Tomas and Potter, 1976. The discrepancy between
the liquid phase were not affected P . 0.807 by the outcome of our study and that of Wilson et al.
any dietary treatment, the combined values for all 1969 might be related to a difference in starch-
treatments n 5 5 being 70 l S.E. 1.4, 5.9 l h S.E. induced change of ruminal pH. Feed intake in our
0.07 and 8.7 h S.E. 0.215, respectively. study was approximately three-times lower than that
in the study of Wilson et al. 1969. The level of feed intake is negatively associated with the buffer-
4. Discussion ing capacity of ruminal fluid Robinson et al., 1986.
