agricultural production guidelines
dairying in kwazulu-natal
Dairying in KwaZulu-Natal
Dairying 5.11 1995
ANALYSIS OF FEEDS
T J Dugmore
Cedara Agricultural Development Institute
Sampling feeds for analysis
procedures for analysing farm feeds
The nutritive value of a
feed is determined by the contribution it is able to make to the nutritional requirements
(energy, protein, and mineral) of the class of animal. Digestibility and metabolizable
energy (ME) are crucial measures of feed quality. Unfortunately, the determination of the
digestible nutrients and the energy available to livestock from a feed can be made only
from animal feeding trials (in vivo trials), which are time consuming and expensive
procedures and require specialized facilities. Mineral content can be determined directly
by chemical analysis in the laboratory.
The farmer who needs to
know the energy value of his feed has one of the following options:
he may estimate the
digestibility of the feed from experience (thumb sucking)
he may use standard
energy values from tables of the chemical composition and nutritive value of feeds
he may send a sample of
the feed to a laboratory.
If he chooses the last
option, then the sample should be analyzed for mineral, fat, fibre, and protein content,
and submitted to special in vitro, digestibility analyses to determine dry matter
loss with time. Comparison of the results so obtained, with those obtained from feeds of
known digestibility, enable the nutritive results of the feed to be derived.
It is with the systems of
chemical analysis used, and the prediction of feed energy values, that the remainder of
this leaflet will be concerned.
The original system for
analysing feeds, the Weende system of proximate analyses, was developed in the 1860s by
Henneberg & Stohman at the Weende Experimental Station in Germany. Their method was to
separate feeds into the nutrient components needed by the animal, that is water, crude
protein (CP), crude fat or ether extract (EE), ash or mineral matter, crude fibre (CF),
the indigestible fraction of the carbohydrates present, and nitrogen free extract (NFE)
the readily digestible carbohydrate fraction.
Later research has,
however, shown that a substantial amount of CF is digested by dairy cattle, whereas much
of the NFE in some feeds, especially those derived from tropical plant species, is not
digested. This problem prompted van Soest, in 1963, to develop his system of detergent
fibre analysis, which determines lignin, the totally indigestible fraction of ruminant
feeds, as a separate entity.
The objective of the van
Soest detergent system is the fractionation of foods of plant origin, relative to their
nutritive availability and fibre content. The first fraction corresponds to the cellular
contents and is composed of lipids, soluble carbohydrates, most of the protein, and other
water soluble material. These cell contents are almost completely digestible. The second
fraction corresponds to the cell wall, or the total fibre, fraction of the plant cell, and
is known as neutral detergent fibre (NDF). Acid detergent fibre (ADF) separates the NDF
into an acid soluble (hemicellulose) and insoluble (lignocellulose) fractions. Acid
detergent fibre has become widely used as a quick method for determining fibre in feeds,
often substituting for crude fibre, and used on much the same basis as crude fibre in
Acid detergent fibre is
considered to be the analysis of choice by many nutritionists for estimating the
digestibility of feeds, although the use of ADF as a predictor of digestibility is not
founded on any solid theoretical basis, but rather on statistical association. Neutral
detergent fibre has been found to be closely related to the voluntary intake of feeds, as
it measures total fibre, or bulk, in feeds, and this is the primary determinant of the
amount which an animal will consume voluntarily.
There are numerous measures
of fibre available, but not all measure the same fibre fraction in the plant, and
therefore will give different answers (see Figure 1). This is important, and must be taken
into consideration when using fibre in dairy ration calculations, i.e. use the
appropriate fibre measure for its appropriate function.
Comparative analyses of the different fibre fractions in kikuyu and ryegrass
SAMPLING FEEDS FOR ANALYSIS
Samples taken for chemical
analysis should always be taken in an unbiased fashion i.e. at random. Always take
as many sub-samples as possible to avoid possible sampling bias. Several pointers to
taking a good sample are outlined below:
Hand cut, or hand pluck,
grass samples, avoiding contamination by soil or dung.
Walk across the pasture
taking small samples at a given number of steps, to give a combined sample of 10 to 30
sub-samples. A kilogram of fresh material should be sufficient.
For silages, take several
samples from different points on the silage face, dig into the face to get a fresh sample.
Ideally, use a soil auger to get samples, at various depths, from different points in the
If a dry matter
determination is required for the pasture or silage, seal fresh sample in a plastic bag
and weigh accurately. Send fresh sample mass to the laboratory, along with the sample.
Refrigerate the sample during storage, if possible.
For hays, take samples
from the inside of 20 bales, at different depths, and combine into a single sample.
For bagged feeds, take a
sample from the top, centre and bottom of several bags.
RECOMMENDED PROCEDURES FOR
ANALYSING FARM FEEDS
Water, although a major
nutrient, does not contribute protein, energy or minerals to a feed. Feed samples are
therefore dried by the laboratories to prevent the moisture present in the sample from
diluting the concentration of the other nutrients. Because the amount of water present in
feeds varies, the dilution effect will be variable. Feed composition is therefore
expressed on a dry matter basis by laboratories, to avoid this dilution effect and to
allow comparison of different feeds on an equal basis.
Crude protein (N x 6,25)
values are necessary to balance such rations for dairy cows as dairy meals, complete
rations, etc. Crude protein values for pastures are also useful in determining
whether protein may be either limiting animal production owing to insufficiency, or
inhibiting production and fertility through excessively high levels in the herbage. Crude
protein on its own is becoming of limited value in the programming of rations for the high
yielding dairy cow (> 35l/day). Future analytical procedures, which are presently being
developed, will include measures of bound protein and accurate estimates of protein
Fibre determination is
recommended only in situations where fibre may be a limiting factor in the diet, e.g.
complete dairy meals and feedlot rations. NDF is considered to be the method of choice
because it is highly correlated with chewing time in ruminants. NDF values are not
considered appropriate to estimate the energy content of feeds. Although NDF is correlated
with intake, it is not considered precise enough to predict intake accurately. It has been
suggested that dairy cows consume 0,9% of their live-mass as forage NDF drymatter per day.
Optimum NDF levels in the
diet differ for different forage species, and within species, owing to analytical effects.
This is especially true when forages have been chemically treated. Until there are
improved methods for characterising NDF, the best level in the diet should be judged by a
competent nutritionist, instead of some ideal ration content.
To establish mineral
imbalances or shortages, samples should be analysed for the macro-minerals. Mineral
deficiencies can be positively diagnosed only by using liver biopsies etc., and
need to be confirmed by using response trials.
Protein has been found to
be positively correlated with digestibility through its decline with age in the plant, but
factors such as the luxury uptake of N by the plant complicate the prediction of
digestibility using crude protein values. Similarly, fibre is negatively associated with
digestibility through its increase in the plant as it ages, and the decline in
digestibility of the plant with age, rather than through being a measure of the
indigestible fraction of the plant. Temperate plant species decline in nutritive value by
0,5% unit of digestibility per day, as opposed to tropical species, which decline by 0,1%
unit of digestibility per day.
Because it is important to
know the digestibility of a feed, it was decided to test several of the known regression
equations against feeds of known in vivo digestibility values. The in vivo
data comprised the results of 40 digestion trials on Cedara, using steers, supplemented
with other digestion trial data for southern and East Africa, as well as those of tropical
areas similar to the local situation. These feeds were classified into the following
categories: (i) dry roughages and forages, e.g. hay, (ii) pasture and forages fed
fresh, and (iii) silages.
The results of the study,
and others, indicate that the chemical composition of a feedstuff cannot be used to
predict the energy value of a feed with accuracy, except in the cases of temperate
forages, hays and crop residues. There appears to be no broadly based equation for
tropical forages, where each species requires its own specific equation, while some
species, such as kikuyu, cannot be predicted from chemical composition. The best fitting
equations, derived from the Cedara data, excluding kikuyu and silages, were;
ME (MJ/kg) = 11,95 - 0,112
(CF%) + 0,219 (EE%)
ME (MJ/kg) = 13,24 - 0,136 (CF%)
Similar equations reported
using ADF values are:
ME (MJ/kg) = 16,654 - 0,24
ME (MJ/kg) = 11,8 - 0,105 (ADF%)
Predictive errors are
typically in the range of ± 4 to 6 digestibility units. In practice, this could mean an
error of 3 to 4 l in the estimated milk yields of cows consuming these feeds, resulting in
incorrect concentrate allocations. The reason for this lack of accuracy is that
digestibility is affected by many aspects of forage composition, but only one or two of
these are measured by these techniques.
The method of choice is
considered to be the in vitro system. In vitro digestibility is the most
accurate method of predicting the digestibility or energy value of a feed. The method
requires rumen fluid from fistulated sheep, an expensive and impracticable measure for
commercial laboratories. The in vitro digestibility determination is an empirical
method and does not entirely simulate ruminant digestion, which is a dynamic process.
Therefore, the in vitro results must be converted to estimated in vivo
results by the use of an appropriate regression equation, e.g.
ME (MJ/kg DM) = 0,84 + 0,14
in vitro DOM%
ME (MJ/kg DM) = -0,376 + 0,166 in vitro DOM% for fresh grass.
ME (MJ/kg DM) = -1,69 + 0,182 in vitro DOM% for hay
The conversion to estimated
in vivo digestibility or energy values is necessary because animal requirements are
expressed on this basis.
The inaccuracy of the
chemical fraction in predicting the digestibility of roughages constitutes a major problem
in assessing a particular farmer's feed quality and, unless an in vitro service is
available, we are left with no recourse but to use tables of known values of feedstuffs,
adapted to the particular situation by a competent nutritionist.
Predicting the energy
value of concentrates
The following equations
have been derived to predict the energy values of concentrate or grain mixtures. The
reliability of these equations has been shown to be superior to those of roughages, with a
predictive error of 0,36 ME/kg DM.
ME (MJ/kg DM) =
13,50 + 0,263 EE% - 0,133 ash % - 0,136 ADF%
ME (MJ/kg DM) = 11,78 + 0,0654 CP% + 0,0665 EE%2 - (0,0414 EE% x CF%) - 0,118
ME (MJ/kg as-fed) = 10,25 + 0,0654 CP% + 0,0764 EE%2 - (0,0476 EE% x CF%) -
Analysis of silages
The products of the silage
fermentation process include volatile compounds, i.e. volatile fatty acids and
ammonia. The method of drying silage samples is therefore critical to ensure that these
volatile components are not driven off by heating. It was recommended that the appropriate
analyses for silages, in addition to any other analysis required, are crude protein, dry
matter, ammonia nitrogen, pH, lactic acid, and volatile fatty acids, all indicators of the
quality of the fermentation process.
A quote from Prof. John
Owen of the University College of North Wales at Bangor is appropriate. "There is
little sense in using a computer to formulate rations when we use blunderbuss methods to
assess the feeding value of the fresh or conserved roughage".
PREVIOUS CHAPTER | CONTENTS
| NEXT CHAPTER >> ]