Testing regression models for pig feeding systems

Abstract

SUMMARYRegression equations have been tested relating average daily gains in pigs to average daily intake of crude protein, lysine, and TDN for each of three growth stages. The tests were based on output contour curves representing 1 lb, 1·4 lb and 1·75 lb live-weight gain per day in the three stages. On each contour, while crude protein intake was held constant, points were selected to represent the economically optimum combination of lysine and TDN intakes, a ‘high energy-low lysine’ combination and a ‘low energy-high lysine’ combination. For the latter two points, least cost mixes of feed ingredients were computed. Four feed mixes were computed at the economically optimum combination of lysine and TDN intakes. Comparisons were made between the performance of pigs given the six different mixes to assess the validity both of the slope of the output contour and its position and to examine the influence on growth performance of the ingredient specification of diets formulated to the same nutrient levels.Using the six diets formulated on each contour, 12 feeding systems were developed all of which theoretically would take the pig from 50 to 200 lb in the same number of days. Six of these systems maintained the same diet type for each period; the others involved a change of diet type at least once during the three stages of growth. Diets formulated to meet the economically optimum combination of nutrients based largely on maize and groundnut meal stimulated inferior daily gains compared with alternative formulations. Both diets with a larger percentage of their lysine supplied from synthetic sources (the economically optimum combination based on maize and groundnut meal, and a low energy-high lysine combination) led to inferior daily gains. Pig performance was generally poorer than expectation, particularly from 50 to 150 lb live weight.No differences could be attributed to differing ingredient composition in equi-nutrient diets. Similarly, no significant differences in growth rate were found between those systems which maintained a constant diet type throughout the three stages of growth. Pigs on systems which involved a change in diet type from one stage of growth to another grew in general slower than those pigs which had a fixed-diet type throughout.Pigs on both diet combinations with a high percentage of weatings (a high cost formulation and a low energy-high lysine combination) had the best carcass results.