area and one where firm conclusions are
not possible. In our laboratory, it has been
observed consistently that, despite the
ready availability of water, tube-fed birds
were rarely seen to be drinking. Further-
more, because 90% of water consumption
by poultry is believed to be a direct conse-
quence of eating, it could be that the lack
of access to food has a marked effect on
the stimulus to drink. Whether atypically
low and variable water intakes explain
erratic and slow food passage rates, and
consequently residue clearance, must
remain uncertain, but some rather simple
evaluations in our laboratory (McNab and
Blair, 1988) have resulted in the routine
administration of water (50 ml per bird, by
tube) during the balance period about 24 h
after feeding. This practice also provides
the opportunity to palpate the crop and
mix any food residues still there with the
water. Only rarely (e.g. with blood meal)
has it led to losses of food from the crop.
However, it does seem reasonable to expect
that it will change the relationship between
the amounts of food given and the
clearance rates. More work is required to
clarify the effects of water consumption on
ME values over a range of ingredients and
diets.
Endogenous Energy Loss (EEL)
Knowledge of the EEL is a prerequisite for
the determination of TME and it patently
has an effect on the value attributed to
AME. That there are both difficulties and
uncertainties associated with its quantifica-
tion is undisputed. Three methods have
been used to derive EEL: starving birds;
giving birds a completely metabolizable
energy source (e.g. glucose); or by extra-
polating to zero intake a line relating
energy excretion to its consumption.
Starvation has been the most widely
used means of deriving EEL. However,
during starvation, individual birds void
quite variable amounts of energy. Values
ranging from 33 to 82 kJ 24 h^1 (Farrell,
1978) and from 25 to 69 kJ 24 h^1 (Sibbald
1979) have been reported for the second
24 h period of 48 h of starvation (24 h +
24 h assay). In our laboratory, a somewhat
wider range has been found, presumably a
consequence of the greater maintenance
demands associated with the 48 h + 48 h
assay. Individual values ranged from 47 to
238 kJ 48 h^1 (24–119 kJ 24 h^1 ) and the
average coefficient of variation within an
experiment (six replicates) was 36.8%.
No consistent variation can be ex-
plained by bird weight or by body weight
changes and, although there have been
quite reasonable claims that the size of the
EEL is related to the temperature of the
environment, we have not been able to
observe any differences in the EEL from
birds fed 50 g of glucose and housed at
either 5 or 35°C. In our laboratory, with a
48 h + 48 h bioassay and using the birds
approximately every 4 weeks, we have
found a substantial reduction in EEL and
conspicuously less between-bird variation
when droppings are collected from birds
fed glucose (50 g) rather than from
starved counterparts. There are some sound
scientific reasons to believe that body com-
position and basal metabolic rate influence
EEL, but the significance of these in ME
evaluations has still to be clarified.
Collaborative studies with a type 1
bioassay, designed to establish a standard
method throughout Europe, have shown its
reliability as a means of deriving the AMEN
values of diets (Bourdillon et al., 1990).
However, when these data were used to
derive AMENvalues of two of the com-
ponents of the diets, wheat and soybean
meals, by a replacement approach, the
results proved conspicuously less robust.
This example illustrates the difficulty, if
not impossibility, of deriving meaningful
ME values for raw materials, as opposed to
diets, using type 1 assays, even when great
care is taken by experienced people.
The simplicity and speed of type 3
assays mean that many more can be carried
out in a given time, and this leads to an
increase in the amount of information
available to practising nutritionists. For the
evaluation of ingredients, which can be
assayed directly rather than after dietary
substitution/replacement, they must be theRapid Metabolizable Energy Assays 313