Feeding urea ammonia treated rice straw in the tropics. II. Assumptions on nutritive value and their validity for least cost ration formulation

Abstract

Field application of new feeding methods, such as urea ammonia treatment of straws requires an ex-ante assessment of their economic feasibility. Animal experiments are expensive for such an analysis, therefore cheaper and quicker methods are required. One such method is least cost ration formulation (LCRF), which extrapolates from a restricted number of feeding trials to a large number of feed combinations. The assumptions underlying LCRF can, however, be challenged and this paper discusses their validity in nutritional terms. The first part reviews literature on nutritive quality of rice straw, in relation to nutrient requirements and animal performance on rice straw based diets. The second part calculates the expected performances ( P calc) on straw based rations, using measured dry matter intake (DMI) and estimated total digestible nitrogen (TDN) and crude protein (CP) values. It then compares P calc with real performance ( P real) as found in the literature. Most dose response trials indicate a rather linear response to supplementation, with exceptions particularly at low levels of supplementation. The pooled values of P calc relate linearly with P real in several feeding trials. The P calc only comes close to P real when calculated on a TDN basis, when energy requirements for liveweight gain of light tropical animals are taken to be higher than those of NRC and when digestibility of energy in straws is assumed to be low. This suggests a low utilisation of digestible energy from either treated or untreated straw for reasons that cannot easily be established. Total CP supply cannot be limiting because the P calc on the basis of CP also considerably overestimates P real. The available literature values do not allow specification of CP into fractions according to rumen degradability. At production levels around maintenance, the correlation between P real and P calc decreases, possibly because at those levels the variability of maintenance requirements exerts a bigger effect on the total performance. Also, at those levels, straw constitutes a larger part of the ration, and associative effects are more likely. The sparse literature values for milk production also indicate differences between P real and P calc. The conclusions are that, in the absence of better methods, LCRF can be useful for ranking purposes but absolute values need to be interpreted with care, and P calc tends to overestimate P real. Future performance trials should measure digestible organic matter intake, substitution rates and maintenance requirements by dose response curves in order to improve their predictive value, and to allow interpretation of differences between P real and P calc. Differences between P real and P calc can then be investigated and weighed against accuracy and cost of performance trials.