Alkanes as Internal Markers to Estimate Digestibility in Horses

Fecal recoveries of n-alkanes were determined in three digestion balance experiments consisting of two periods each. Each experiment compared two diets in a switch-back design using eight mature Thoroughbred geldings. Horses were randomly assigned to two groups of four and offered one of two mixed grass/legume hays (Diets 1 and 2) in Exp. 1, mixed grass/legume hay and one of two concentrates (Diets 3 and 4) in Exp. 2, and mixed grass/legume pasture (Diets 5 and 6) in Exp. 3. In Exp. 1 and 2, horses were housed in stalls and in Exp. 3, two horses from each diet were housed in stalls and two horses remained on pasture. Balance periods lasted 11 days with d 1 to d 7 consisting of a dietary accommodation period, followed by 4 days of total collection. Results indicated that fecal recoveries of odd-chain alkanes (C25 to C33) were less than 100 % and similar between chain lengths. Estimates of DMD (DE) were similar to the total collection DMD (DTC) for Diet 1 in Exp. 1, but underestimated DTC for Diet 2 in Exp. 1 (P < .05) and Diets 3 and 4 in Exp. 2 (P < .05). For Diet 5 in Exp. 3, the DE for stall-fed horses using C25 and C33 was similar to DTC, whereas C27, C29, and C31 underestimated DTC (P < .05). For pastured horses, the DE using C29 and C31 were similar to DTC, whereas C25, C27, and C33 underestimated DTC (P < .05). For Diet 6 in Exp. 3, the DE for stalled horses calculated using C25 was similar to the DTC, whereas use of C27, C29, C31, and C33 underestimated DTC (P < .05). For pastured horses, the DE using C29 was similar to DTC, whereas all other alkanes underestimated DTC (P < .05). When DE was adjusted (DA1) using the mean recovery of each odd-chain alkane, DA1 was similar to DTC for Diet 2 in Exp. 1, Diets 3 and 4 in Exp. 2, and stalled horses offered Diets 5 and 6 in Exp. 3. The DA1 using C25 underestimated DTC for Diet 1 in Exp. 1 (P < .05). For pastured horses offered Diet 5, DA1 for C33 was not different from the DTC estimate, whereas all other DA1 for n-alkanes overestimated DTC (P < .05). For pastured horses offered Diet 6, DA1 for C29 and C31 overestimated, but were similar to the DTC, whereas the DA1 for C33 underestimated DTC and was similar to the DTC. The DA1 for C25 and C27 overestimated DTC (P < .05). When DE was adjusted for the mean recovery of all n-alkanes (DA2), all DA2 estimates for stalled horses in Exp. 1, 2, and 3 were similar. In pastured horses offered Diets 5 and 6 in Exp. 3, the DA2 overestimated DTC (P < .05). These results suggest that accurate mean estimates of DMD can be obtained by adjusting for mean recovery of each odd-chain alkane in a specific diet.