1. 1. INTRODUCTION Acacia erioloba, A. nilotica and Z. mucronata tree species are ubiquitous in the semi-arid regions of the North-West province but their utility as protein sources for ruminants is a function of their chemical composition, which is in turn influenced by the growth environment as well as grazing/browsing pressure. Of major concern is the high concentration of phenolic compounds such as tannins, which are produced by these plants for various purposes. The concentration and biological activity of phenolic compounds is known to vary with factors such as tree species, growth environment and grazing/browsing pressure, among other factors. This study, therefore, investigated the influence of harvesting sites, which vary in terms of grazing/browsing pressure, on the concentration and in vitro ruminal biological activity of phenolic compounds and how this may affect the potential protein value of tree eaves. We hypothesized that higher grazing/browsing pressure represents a quantifiable biotic stress factor induce higher biosynthesis of phenolic compounds with higher biological activities. Acceptance of this hypothesis would be evidence of the existence of an inductive phenolic biosynthesis mechanism in the three tree species investigated.. Mnisi C.M ¥ and Mlambo V. Department of Animal Science, North West University Mmabatho 2735, South Africa ¥ Corresponding author: Kenny.Mnisi@nwu.ac.zaKenny.Mnisi@nwu.ac.za. Influence of harvesting site on the chemical composition and potential protein value of Acacia erioloba, A. nilotica and Ziziphus mucronata leaves for ruminants 2. METHODOLOGY 2.1.Study site The harvesting sites included, Molelwane (25°85’S, 25°63'33" E), which is a private farm with a very low number of ruminants that do not rely entirely on the rangeland, and Masuthle (25° 53’ S, 31° 49’ E), which is a communal grazing area where a large number of herbivores have unlimited access to the communal rangeland. While the grazing/browsing pressure in Masuthle is greater than in Molelwane, the soil types and rainfall patterns are similar. The two sites were purposefully selected based on proximity to each other, availability of tree species of interest, and variation in terms of grazing/browsing pressure. 2.2. Chemical analyses and buffer N solubility Total N was determined by the standard macro-Kjeldahl method (AOAC 2005, method no. 984.13). NDF and ADF were determined using the ANKOM 2000 Fibre Analyzer (ANKOM Technology, New York). Minerals were analysed following the guidelines provided by AgriLASA (1998). Soluble phenolics were determined using the Folin-Ciocalteau method (Makkar, 2000). Soluble condensed tannins were analysed using butanol-HCl (95:5 v/v) (Porter et al., 1986). Buffer-soluble N was determined according to a modified Licitra et al. (1996) method, in which ANKOM F57 (ANKOM Technology) filter bags were used for the filtration process (Cudjoe & Mlambo, 2014). N solubility index was then calculated as the ratio of buffer-soluble N to total N. 2.3. In vitro ruminal N degradability In vitro ruminal N degradability was determined using the ANKOM Daisy II incubator (ANKOM Technology Corporation, Fairport, NY) in which dried and milled samples were weighed into F57 filter bags and heat-sealed. Rumen fluid was obtained from a fistulated Bonsmara heifer and incubation was done under anaerobic conditions at 39 °C. Bags were withdrawn after 12, 24, and 36 h of incubation, washed, dried and analysed for residual N. 2.4. In vitro ruminal gas production tannin bioassay The condensed tannin bioassay (ivCTB) was done with the aid of a tannin-binding compound, polyethylene glycol (PEG), using an automated ANKOM RF gas production system (ANKOM Technology Corporation, Fairport, NY). Leaf samples (1 g) were weighed into individual ANKOM RF bottles to which the ANKOM buffer solution (90 mL per bottle) with (400 mg PEG/90 mL buffer) or without PEG was added. Anaerobic incubation was carried out at 39°C using rumen fluid from a Bonsmara heifer. Headspace gas pressure was automatically measured at 10 min intervals over a 96 h period. Tannin bioactivity was determined as the % increase in cumulative gas production upon PEG treatment. 2.5. Statistical analyses Chemical composition, buffer N solubility and in vitro ruminal degradability of N data were analysed based on a 3 (tree species)× 2 (harvesting sites) factorial treatment design in a CRD while the ivCTB data were analysed based on a 3 (trees species) × 2 (harvesting site) × 2 (PEG) factorial treatment design also in a CRD. Both analyses employed the general linear models (GLM) procedures of SAS (2010). Statistical significance was declared at p ≤ 0.05. Lsmeans were separated using the probability of difference option. 3. RESULTS Leaves of Z. mucronata had the highest CP content (177.67 g/kg DM), followed by A. erioloba (150.6 g/kg DM) and A. nilotica (127.6 g/kg DM. Leaves harvested from Masuthle had higher soluble phenolics (SPh) (44.6 g TAE/kg DM) compared to those harvested from Molelwane (29.8 g TAE/kg DM). In Molelwane, leaves from A. nilotica had a higher (P 0.05). There were no site differences (P >0.05) in soluble condensed tannins content for all tree species. There was no difference in SI between leaves of A. erioloba and Z. mucronata. Leaves of A. nilotica harvested from Molelwane had the same (P >0.05) N solubility index (SI) as leaves of A. erioloba and A. nilotica harvested from Masuthle. However, leaves of Z. mucronata and A. erioloba had higher SI content when compared to A. nilotica with the exception of A. erioloba harvested in Molelwane. The largest PEG effect was seen in leaves of A. nilotica (448%) harvested in Masuthle after 36 h of incubation. Z. mucronata (25%) leaves harvested from Molelwane had the least PEG effect at 48 h. In both sites, there was no significant association (P >0.05) between SPh content and PEG effect at 24, 36 and 48 h of incubation. In Molelwane, CT had a significant (P <0.05) positive correlation with PEG effect at 24 (r = 0.501) and 48 h (r = 0.751) but not at 36 h. In Masuthle, there was a positive association (P <0.05) between CT and PEG effect at 24 (r = 0.588) and 36 h (r = 0.630). 4. DISCUSSION & CONCLUSIONS Higher browsing pressure in Masuthle communal rangeland resulted in leaves with higher levels of condensed tannins, which had higher in vitro ruminal biological activity, compared to those harvested from Molelwane. Leaves harvested from Molelwane had higher potential nutritional value than those from Masuthle because they had high levels of N and minerals but low levels of phenolics. Buffer-soluble N was poorly correlated with in vitro ruminal N degradability of leaf substrates harvested from Masuthle. The leaves of A. nilotica were rich in condensed tannins, which depressed rumen microbial activity and therefore should be used with caution or at least with a tannin-inactivating agent when used to supplement low-protein fibrous diets. Leaves of Z. mucronata and A. erioloba, whose condensed tannins had the lowest in vitro ruminal biological activity but had high N content, can be used as protein supplements in fibrous basal diets in order to maintain animals throughout dry periods. 5. REFERENCES Cudjoe, N., Mlambo, V., 2014. J. Anim. Physiol. Anim. Nutr. 98, 722 - 730. Licitra, G., Hernandez, T.M., Van Soest, P.J., 1996. Standardization of procedures for nitrogen fractionation of ruminant feeds. Anim. Feed Sc. Technol. 57, 347 - 358. Makkar, H.PS., 2000. A Laboratory Manual. Animal Production and Health Sub- Programme. Pp. 1 - 99..