Thesis (MScAgric (Animal Sciences))--University of Stellenbosch, 2005. / The demand for protein for human and animal nutrition in South Africa is increasing and it will continuously increase. The effect of replacing soybean oilcake meal as protein source for broilers and full-fat soybean meal for weaner pigs, with different levels of sweet lupins (Lupinus angustifolius), canola oilcake and full-fat canola was examined. A basal diet with soybean oilcake as protein source was mixed with a diet using either sweet lupins or canola oilcake or full-fat canola in ratios of 100%, 67% and 33% respectively. In the broiler trial the test diets were fed for a period of six weeks with or without the enzyme Vegpro (Alltech). Pigs were fed the test diets, with and without Roxazyme® enzyme, ad libitum from 8.5 to 25 kg live weight. The fatty acid content of the fat pads of the broilers raised on the different diets was determined. The pigs were kept in the trial up to the grower- finisher phase. The fatty acid content of the carcass fat and muscle of pigs raised on the different diets was determined. The inclusion of enzymes had no effect on the growth, feed intake or feed conversion ratio of broilers fed the test diets. The provision of external dietary enzymes to the weaner pig diets failed to improve either dry matter intake or growth rate, but improved the feed conversion ratio. Broiler weights at six weeks of age were significantly higher for the control diet compared to the 20% lupin diet. There was no significant difference in the feed intake as the lupin content of the diets increased. The feed conversion ratio did not differ significantly between the control diet and the 6.6% lupin diet but became significantly poorer as the lupin content increased to 13.2% and 20% of the test diet. There were no significant differences in production performance of the control diet and the canola oilcake containing diet. The broiler weights at six weeks decreased significantly with each increase in the canola oilcake content of the diets. The feed intake of the 20% canola oilcake diet at week six was significantly less than the intake of the control diet, but not significantly less than the 6.6% and 13.2% canola oilcake diets. The feed conversion ratio of the control diet was significantly better than the 13.2% and 20% canola oilcake diets. No significant differences were found in week six between the 6.6% full-fat canola diet and the control diet for broiler weights and feed intake. The feed conversion ratio of the broilers fed the 13.2% and 20% full-fat canola diets was significantly poorer than the control diet. The final body weights of the weaner piglets fed the control diet were significantly higher than the final body weights of weaner piglets fed the lupin containing diets. The average daily gain of the weaner piglets fed the control diet was significantly higher than the gain of the weaner piglets fed the lupin containing diets. No significant differences in the feed intake and feed conversion ratio between the different lupin inclusion levels were detected. There were no significant differences in body weight, feed intake, average daily gain and the feed conversion ratio, between the various full-fat canola containing diets and the control diet. A significant difference in body weight was found between the weaner piglets fed the 20% canola oilcake diet and the weaner piglets fed the control diet. There were no significant differences in feed intake between the various inclusion levels of canola oilcake. The control and 6.6% canola oilcake containing diets had significantly higher average daily gains than the 20% canola oilcake containing diet. In the final trial the finisher pigs fed the test diet containing 25.00% lupins, had a final body weight significantly lighter than the final body weight of the finisher pigs fed the other test diets. The finisher pigs fed the test diet containing 25.00% lupins, also had a significantly reduced average daily gain and feed intake. The feed intake of the 25.00% canola oilcake diet was significantly lees than the feed intake of the 16.75% full-fat canola containing diet. The feed conversion ratio of the 25.00% lupin containing diet was significantly poorer than the feed conversion ratio of the 16.75% lupin containing diet, 8.25% lupin containing diet, 16.75% full-fat canola containing diet, 25.00% canola oilcake containing diet and the control diet. In a choice feeding trial growing pigs were offered four diets with four different protein sources: sweet lupins (25% inclusion level), canola oilcake (25% inclusion level), full-fat canola (25% inclusion level) and soybean oilcake (25% inclusion level), while their daily intakes were recorded. In a second choice feeding trial pigs were offered ten different diets with increasing levels (6.6%, 13.2%, 20%) of either sweet lupins, canola oilcake meal or full-fat canola meal. The pigs consumed significantly more of the soybean oilcake containing diet compared to diets containing the alternative protein sources. Pigs consumed significantly less of the full-fat canola diet compared to the sweet lupin and canola oilcake diets. Ten different canola cultivars were collected from two different locations in the Mediterranean rainfall area of South Africa namely the Western Cape (Swartland) and Southern Cape (Rûens) grain producing areas. The sinapine and glucosinolate content of various canola cultivars was compared and the influence of locality on the sinapine and glucosinolate content of the canola cultivars was determined. There were no significant differences (P < 0.05) in sinapine content when the canola produced in the Western and Southern Cape were compared. Varola 54 and Rainbow cultivars had significantly higher (P < 0.05) sinapine contents compared to the Varola 50 cultivar. Samples of lupins, field peas, faba beans and narbon beans were collected and analysed for amino acids, alkaloids, non-starch polysaccharides, tannin and starch. The digestible energy value of these alternative protein sources for pigs was determined. Significant differences were found in the amino acid content of the various crops. The alkaloid content of the lupins varied significantly between the sweet and bitter lupin varieties. Sweet L. angustifolius cultivars contained ca 50mg/kg and the bitter L. angustifolius cultivars ca 15000mg/kg alkaloids. The mean alkaloid content of L. albus cultivars was ca 1300mg/kg. The faba beans, narbon beans and peas had significantly higher values for tannins and starch, compared to lupins.