Given the drawbacks associated with the use of antibiotics as feed additives and the imminent banning of its use in the European Union, the aim of this project was to develop an extract that could be used as an alternative feed additive in poultry production. The desired extract preferably had to be rich in antibacterial activity to control proliferation of undesired microorganisms, and antioxidant activity to boost the immune system of the poultry. A number of trial extraction procedures were employed on dried leaf material samples to identify the best extraction method. In preliminary extraction studies, direct extraction was performed on leaf samples from the Lowveld National Botanical Gardens (LNBG) and from University of Pretoria Botanical Garden (UP). The principle aim of preliminary studies was to identify the solvents that extracted high antibacterial and antioxidant activity while also extracting large quantities of material. The secondary objective was to test for differences in activities between samples collected from LNBG and UP. Five extractants of varying polarities; acetone, ethanol, ethylacetate, dichloromethane and hexane were used. Antibacterial activity of all extracts was quantified by a serial dilution microplate technique while bioautography was used in qualitative analysis of the antibacterial active compounds. ATCC strains of Staphylococcus aureus, Pseudomonas aeruginosa, Escherichia coli and Enterococcus faecalis were used as test organisms. Qualitative antioxidant activity was determined by using a DPPH assay on TLC plates. Results from preliminary extraction studies showed larger quantities of material were present in extracts from the LNBG sample than in the UP sample. Two major antioxidant compounds (Rf values of 0.85 and 0.35 in EMW solvent system) were seen on DPPH sprayed TLC plates, while bioautography showed the presence of a number antibacterial active compounds in the acetone, ethanol and ethylacetate extracts with Rr values ranging between 0.85 and 0.56 on TLC plates developed in the EMW solvent system. MIC values of the extracts tallied with the results from bioautography. The acetone, ethanol and ethyl acetate extracts had the highest antibacterial activity while the hexane extracts had the lowest activity with average MIC value of 0.55 mg/ml for both the LNBG and UP samples. MIC values as low as 0.04 mg/ml were measured in the acetone and ethylacetate extracts of the LNBG sample against S. aureus and E. faecalis. Based on results from preliminary extraction studies, hexane was identified as a possible pretreatment solvent for application in enrichment procedures, acetone and ethanol were chosen as the main extractants and only the LNBG sample was used for future work. Enrichment procedures were employed along two pathways; the first pathway involved the use of hexane "wash" as a pretreatment procedure prior to extraction with acetone or ethanol. The second pathway involved the use of various mixtures of acetone in water and ethanol in water as extractants. The rationale of using these various ratios was an attempt to identify solvent mixtures that would selectively extract the bioactive components or otherwise selectively remove inactive material. A serial dilution microplate method was used to determine Minimum Inhibitory Concentrations (MICs) and the Trolox Equivalent Antioxidant Capacity (TEAC) assay was used to quantify antioxidant activity of all extracts. The optimal extract was the one developed by pretreatment with a single direct extraction with hexane prior to extraction with acetone. It had a TEAC value of 2.3, an increase in TEAC value of 283% compared to that of the crude acetone extract. The average MIC of the crude acetone extract against ATCC stains of S. aureus, Ps. aeruginosa, E. coli and E. faecalis had dropped from 0.15 mg/ml to 0.08 mg/ml in the optimal extract (an improvement in antibacterial activity of 87.5%). Since the optimal extract is intended for commercial application in poultry production, its antibacterial activity was tested against Campylobacter jejuni, Clostridium perfringens, Salmonella enteritidis, E. coli and multi drug resistant E. coli isolated from chickens. Its in vitro toxicity was ascertained using the brine shrimp assay and the MTT cytotoxicity assay on monkey kidney cells. The optimal extract was effective against Campylobacter jejuni and Clostridium perfringens with MIC values ranging from 40 µ/ml to 80 µ/ml. It was also active against multi-resistant strains of E. coli and Salmonella enteritidis (MIC values of 125 µ/ml for both strains). LC50 results from the brine shrimp assay and the MTT cytotoxicity assay on monkey kidney cells gave values of 863 µ/ml and 226 µ/ml respectively indicating low toxicity. These results meant that though in some cases the MICs of the optimal extract were higher than befitting of typical antibiotics, due to its relatively low toxicity, large quantities of the extract may possibly be feed to achieve the desired activity without causing any toxicity in the poultry. The major antioxidant compound was isolated by silica gel column chromatography. The isolated compound was identified by nuclear magnetic resonance and mass spectroscopy as combretastatin BS (2', 3', 4-trihydroxyl, 3, S, 4'-trimethoxybibenzyl), previously isolated from the seeds of C. kraussii and also from C. woodii leaves. Famakin (2002) showed this compound to be the major antibacterial compound in C. woodii leaves. Combretastatin BS (CBS) demonstrated in vitro cytotoxicity in the MTT assay on monkey kidney cells with an LC50 value of 1 0 µ/ml. In vitro cytotoxicity of CBS could be due to its antimitotic activity. The TEAC value of 7.9 found in this study means that combretastatin BS has about 8 times the antioxidant capacity of vitamin E. This is the first report of the antioxidant activity of any of the combretastatins. Tolerance of broiler chickens to the optimal extract was assessed at clinically inferred doses of 2 mg/kg, Smg/kg and 10 mg/kg . After 21 days of infeed-dosing with the optimal extract, none of the chickens died or showed any behavioral signs of toxicity. There were no statistically significant differences in weight gain between broilers fed the optimal extract and the positive and negative control. There was also no positive correlation between weight gain and amount of the optimal extract incorporated in feed. Although the optimal extract did not result in significant growth promotion relative to the positive and negative control, 2 mg/kg dose regimens showed the best Feed Conversion Ratio (FCR), with a 6.2% improvement compared to the negative control. The positive control was the only other feed regimen to provide a positive FCR with an improvement of 1.73% compared to the negative control. Because purchase of feed could represent up to 80% of costs of broiler production, this is an important finding. If these results can be confirmed, the product may therefore have commercial value. Repetition of the experiment with lower doses of the optimal extract on poultry challenged with bacterial infections is required to confirm the commercial applicability of this product. Copyright / Dissertation (MSc (Paraclinical Science))--University of Pretoria, 2004. / Paraclinical Sciences / unrestricted