results for rumen and duodenal bacteria.
Microbial nitrogen flows estimated from
(^15) N procedures have been shown to be
lower than estimates based on purines
(Perez et al., 1996), but this finding has
been ascribed to an inadequate isotope
distribution or to difficulties in obtaining
representative samples of microbial pools.
The presence in the diet of fish meal,
which contains relatively high concentra-
tions of nucleic acids, could lead to
erroneous predictions. With most other
feeds, the intestinal passage of dietary
nucleic acids appears to be insignificant. A
further concern regarding the use of
nucleic acids as markers is the finding that
nucleic acid to protein ratios could differ
in bacteria and protozoa and also among
the fluid and particle-associated pools in
the rumen. Since the rumen microbial
biomass is largely associated with the
particulate fraction in the rumen, it is
recommended that the particulate fraction,
or mixed rumen bacteria pooled over time
after feeding, is used for measuring purine
to nitrogen ratios of bacteria leaving the
rumen. Although fluid-associated bacteria
are much easier to isolate than bacteria
associated with the particulate fraction,
their use in determining purine to nitrogen
ratios could lead to the underestimation of
bacterial protein yields. The ruminal
protozoal purine to nitrogen ratios are
usually about half those of ruminal bacteria
(Broderick and Merchen, 1992), and will
lead to the underestimation of the proto-
zoal protein contribution if total bacterial
plus protozoal flow is estimated from bac-
terial purine to nitrogen ratios.
Non-invasive methods based on the
urinary excretion of purine derivatives
assume that duodenal nucleic acids are
largely of microbial origin and are recovered
proportionally in the urine mainly as
allantoin and, to a lesser extent, as hypo-
xanthine, xanthine and uric acid after
intestinal digestion. Reliable estimates of
duodenal flow were obtained using this
technique, provided the minor non-
allantoin purine derivatives were also taken
into account in the calculations. Dietary
nucleic acids contribute little to purine
derivative excretion as they are degraded
extensively in the rumen. Estimation of
microbial protein reaching the duodenum,
based on urinary purine derivative assays,
did not differ significantly from estimations
based on purine nitrogen in duodenal
digesta (Perez et al., 1996).
Digestibility
Conventionally, the digestibility of a forage
is the difference between the quantities con-
sumed and excreted in the faeces, and
involves the measurement of feed intake and
total faecal excretion. Apparent digestibility
of herbage or feeds can be estimated by
using an internal marker, or incorporating an
external marker into the diet at a known
concentration and collecting faecal samples
when a steady-state has been reached. The
advantage of the use of markers in digesti-
bility studies is that total quantitative faecal
collection is replaced by random grab
sampling, which saves costs and labour,
especially in studies with large herbivores.
A combination of rumen evacuation-
derived rate of digestion values and of two-
compartmental passage models gave good
estimates of rumen NDF digestibility
(0.94–0.96 of the actual digestibility)
(Huhtanen et al., 1995). Partial digestibility
coefficients have been estimated by the
dual-marker reconstitution method of
Faichney (1980) using Cr-EDTA and
ytterbium as markers, or the single marker
concentration method of Schneider and
Flatt (1975), using ytterbium. The two
methods gave similar results, but the recon-
stitution method required more sample
preparation time and three times the num-
ber of atomic absorption spectrophotometer
analyses than the single indicator method
(Remillard et al., 1990).
Digestibility can also be derived from
Equation 12.5.
Dry matter digestibility (g kg^1 )
= 1000(1 Mh/Mf) (12.5)
where Mh= the marker concentration in
herbage and Mf= the marker concentration
in faeces.
Use of Markers 267