The purpose of this thesis was to investigate, in a series of 5 experiments, the basis of
the diet selection of sheep, with the aim of shedding some light on the mles that
ruminants might use when selecting their diet. This was done by investigating the
effects of offering foods with rapidly degradable nitrogen and rapidly fermentable
carbohydrates, within the rumen, on the food intake and diet selection of sheep.
Experiment 1 investigated whether sheep select between foods that differ in their
effective rumen degradable protein (eRDP) content to achieve an estimated ‘optimal’
eRDP: fermentable metabolisable energy (fME) ratio. The foods offered were
presented in a meal form to prevent chemical changes to the food ingredients that
may occur during the pelleting process. However, as the food ingredients separated
during the course o f the experiment and the composition of the selected diet could
not be sufficiently predicted, the results of the experiment were unclear.
Experiments 2 and 3 re-asserted the hypothesis that diets selected by sheep would be
those that meet the eRDP requirements of the animals and at the same time avoid
excess consumption of rumen degradable protein (RDP). The sheep were given
choices between pelleted foods differing in their urea content, which was used as an
eRDP source. However, sheep selected a diet in favor of the urea-supplemented food
regardless of the basal food to which the urea was added. This resulted in an excess
consumption of RDP. The results did not support the hypothesis that sheep would
avoid excess RDP when given a choice and suggest that eRDP may not be a relevant
dimension in diet selection in the conditions of these experiments. It was thought that
urea may have had other properties, such as a high buffering potential, that the sheep
were selecting for when selecting a diet.
Experiment 4 tested the hypothesis that the preference for an eRDP adequate food
supplemented with urea would be reduced by the addition of a buffer (sodium
bicarbonate (SB)) to both foods offered as a choice or offering ad libitum access to
hay. In both cases urea may no longer be needed as a buffer and the preference for
the urea-supplemented food would then be expected to be reduced or even disappear.
Insufficient amounts of hay were consumed to reduce the preference for the ureasupplemented food. Supplementing both foods with SB significantly reduced the
preference for the urea-supplemented food. It was proposed that the effect of SB
addition on the diet selection of sheep was carried though its buffering properties.
Therefore it could be hypothesized that the sheep in this experiment, and those in the
previous two, were prepared to over consume RDP, when supplied by urea, due to
the buffering potential of the urea supplemented food. When this excess consumption
was made unnecessary by adding SB, sheep avoided an excess intake of RDP. The
idea that sheep might give greater weighting to the benefits of maintaining their
rumen within a desired physiological range rather than avoid an excess consumption
of RDP, led into the second line of research of this thesis.
The final experiment investigated whether a large rather than small change within the
rumen environment, due to the consumption of a ‘disruptive’ food (i.e. food that
would lower rumen pH), would alter subsequent diet selection. Sheep were offered
high or low energy density (ED) foods with or without SB for a period of 2 hours.
All sheep were then offered a choice between the high and low ED foods
unsupplemented for a further 6 hours. SB supplementation did not alter subsequent
diet selection. The magnitude of change in the pH of the rumen may not have been
sufficient to alter subsequent diet selection. However, sheep initially offered the high
ED food subsequently selected more of the low ED food during the 6 hours when a
choice was offered compared to sheep initially offered the low ED food. The results
from this experiment agree with the hypothesis that changes within the rumen of
large, rather than small, magnitude due to the consumption of a disruptive food alters
subsequent diet selection.
Taken together, the results from these experiments are in agreement with the
hypothesis that one of the objectives of the diet selection of ruminants is to maintain
the rumen conditions within a certain physiological range. Therefore, in the absence
of long forage, ruminants would be expected to select foods with the greater
buffering potential firstly to prevent further disruption to the rumen environment and
secondly to return the conditions in the rumen to within the accepted range as soon as
possible. This need appeared to override other diet selection objectives such as
avoiding an excess intake of RDP. However, the basis of protein selection by
ruminants still remains unclear. Further research is required to strengthen the
argument that ruminants select a diet to avoid both deficient and abundant
concentrations of eRDP.