Assessment of mathematical models to describe the rate of passage of enzyme-treated or sugar beet pulp-substituted lucerne silage in equids

A 3 × 3 Latin-square design experiment was used to assess the ability of mathematical models to describe the rate of passage of enzyme-treated or sugar beet pulp (SB) substituted lucerne silage in equids. Three experimental diets were evaluated: a lucerne silage control (WS), enzyme-treated lucerne (WE) and lucerne substituted with 300 g/kg DM of SB (WSB). Each diet was fed to give a total DM intake of 17.5 g/kg of liveweight per day. The control diet, WE, and the basal silage in the WSB diet were labelled with ytterbium chloride (Yb). Labelled feeds (60 g) were offered as an oral dose on the first day of each recording period prior to the morning meal. Faecal sampling was initiated immediately after marker administration and continued for 120 h. Faecal excretion data were subjected to the algebraic models of Faichney (1975) and Theilmans et al. (1978), and to the compartmental models of Grovum and Williams (1973), Dhanoa et al. (1985) and Pond et al. (1988). The models of Grovum and Williams (1973), Dhanoa et al. (1985) and the G2G1 model of Pond et al. (1988) failed to converge with 6 out of the 9 faecal excretion data collected; thus were rejected from further analysis. For the remaining models, similar total tract MRT values were recorded for the algebraic and compartmental models. Furthermore, enzyme treatment or SB substitution did not affect the total tract MRT of lucerne silage in ponies, with values ranging from 25 to 27.4 h. Of the compartmental models, the time-dependent G3G1 model of Pond et al. (1988) yielded the highest R2 values and was chosen to compare differences between the marked feeds. However, results indicated that further studies, aimed at developing equine-specific models, are required to conclusively determine digesta MRT within the different segments of the equid gut. © 2009 Elsevier B.V. All rights reserved.