[In vitro simulation of rabbit cecal fermentation in a semi- continuous flow fermentor. I. Role of food substrate pretreatment].

A Rusitec semi-continuous flow fermentor was used to study the influence of enzyme pretreatment of food substrates on the fermentation profile over a 2-week period following inoculation with rabbit caecal contents. Three types of substrate were examined: 1) homogenized commercial rabbit feed; 2) the solid remains of this feed after digestion with alpha-amylase for 24 h; and 3) substrate 2 digested for 4 h with pepsin (double enzyme treatment). One of a pair of nylon pouches containing 15 g substrate was replaced each day, thus producing a uniform 48-h fermentation. Fermentation of the untreated feed (1) for 5-6 days produced a fermentation profile quite different from that obtained in vivo in the rabbit caecum: propionic acid accounted for over 35% of total volatile fatty acid (VFA), and butyric acid for about 15%. Amylase digestion (2) gave a stable ferment profile closer to the in vivo profile, except that propionic and butyric acids were similar at 15% of total VFA. Digestion with both amylase and pepsin (3) produced a stable fermentation profile very close to the in vivo profile: C2 > 60%, C3 < 11% and 17% < C4 < 21%. The rate at which membrane constituents (acid detergent fibre, ADF) were lost in 48 h was similar to the digestibility coefficient measured in vivo by others for the same basic feed. Lastly, there was a high percentage (about 5%) of volatile C5 fatty acids; this could be due to the discontinuous fermentor input of one pouch per 24 h. Thus, feed pretreated with both amylase and pepsin simulates, in vitro, rabbit caecal fermentation in a semi-continuous Rusitec type fermentor.