grinding of hay markedly increased the rate
of DM and NDF disappearance or
decreased the lag time, i.e. the time
between the introduction of the bag into
the rumen and the beginning of degrada-
tion. The influence of grinding on the rate
and extent of sample degradation in the
rumen is contradictory, probably because
grinding does not affect the particle size of
feeds to the same extent. Thus a given
increase in particle size causes the same
increase in in sacco N degradability
(Michalet-Doreau and Cerneau, 1991).
After milling at the same screen mesh
size, particle size distribution is different
between forages (Emanuele and Staples,
1988) and between concentrates (Michalet-
Doreau and Cerneau, 1991). For the
0.8 mm screen, the mean particle size of
pea was 5, barley 52, maize 186, soybean
meal 292, lucerne hay 321 and cocksfoot
hay 348 μm (Michalet-Doreau and Cerneau,
1991). The chemical composition of
different sized particles varies. In this last
experiment, the smallest particles of
lucerne hay contained a higher proportion
of N than the coarser particles, probably
because the leaves highest in N are the
most fragile parts of this plant and so are
ground most finely. Similarly Gerson et al.
(1988) observed that when meadow grass
hay was chopped and subsequently wet
sieved, the proportion of lipid (cuticle)
increased as the size of sieve decreased
and, conversely, as sieve size increased the
proportion of lignified (stem) material
increased. These variations of particle size
distribution with grinding of a specific
feedstuff have implications for the in situ
technique. There is a linear relationship
between the proportion of N in the smallest
particles, measured by wet sieving and
weighing, and N particulate losses through
the pore bags (Michalet-Doreau and Ould-
Bah, 1992). On 20 feeds with 0.8 mm
grinding fineness, N losses were a mean of
10%, ranging from 4% for soybean to 32%
for oat. For starch, the proportion of
particles lost through the bag pores without
being degraded is also important and very
variable. With immature maize, oven dried
and ground at 3 mm screen, the starch
losses represented 11% of starch initially
introduced in the bags, and ranged from 5
to 25% according to stage of maturity and
genotype. However, this proportion was
much greater when samples were freeze-
dried (Philippeau and Michalet-Doreau,
1997). In agreement with these results,
Lopez et al.(1995) determined that for
1 mm ground forages, greater DM losses
occurred after freeze-drying than after
oven-drying below 60°C.
Particle size also has an influence on
the accessibility of dietary components to
the microorganism enzymes. Emanuele and
Staples (1988) studied the in situdegrada-
tion kinetics of different sized particles of
the same feed. The solubilized fraction or
the lag time significantly increased with an
increase in particle size. Similarly, Nocek
and Kohn (1988) noted that chopping and
then sieving lucerne or timothy hay
through a 3.2 mm screen resulted in a shift
in the proportion of insoluble and non-
degradable DM towards a larger particle
size. In a trial using meadow grass hay
(Gerson et al., 1988), in vitrogas produc-
tion was 30% lower for the large particles,
but the surface area was 10% that of the
small particles. External surface area per
unit of mass and an intact cuticle could be
determinants of ruminal degradation
kinetics.
Another aspect of the effect of mastica-
tion on feeds is the increase in feed wetting,
and thus solubilization rate and accessi-
bility of feed components to micro-
organisms. In order to facilitate the attack
of feeds by rumen microorganisms, some
researchers pre-soak samples in water or in
buffer. The time of pre-soaking varied
between 1 and 15 min. Little information is
available to study the advantages of pre-
soaking bags before their incubation. This
treatment should nevertheless enable the
degradation rates of insoluble nutrient
components to be determined more easily.
In conclusion, the aim of sample
preparation is to simulate the resultant
particle distribution that would be
achieved by mastication. However, there
may be considerable differences in particle
sizes between feeds ground using the sameIn Sacco Methods 235