Prediction of intake of tropical forages in dairy cattle based on studies from fibre digestion kinetics

PhD Thesis

Two experiments were conducted simultaneously to estimate the physical fill of tropical forages and the maximum intake capacity of mature non-pregnant crossbred heifers kept under zero grazing system. The experiments were carried out at Magadu Dairy Research Farm, Department of Animal Science and Production, Sokoine University of Agriculture. The objectives were to describe the tropical forages in terms of their rumen degradability characteristics, digestion and passage rates of fibre and their resultant rumen pool sizes in order to estimate the physical fill and potential intake of tropical forages. The study was planned to test the hypothesis that in the tropics where ruminants are fed forages of high fibre content and low digestibility, the physical fill is the major factor in regulation of feed intake. Five forages [Signal grass (Brachiaria brizantha) hay (BH), Maize (Zea mays) silage (MS), Lucerne (Medicago saliva) hay (LH), Lucerne grass hay (LGH) and urea-treated rice (Oiyza saliva) straw (UTRS)] were fed to 10 mature non-pregnant rumen fistulated dairy heifers in two 5x5 Latin square designs. In experiment 1 (Expt.l) the forages were fed as a sole diet with only minerals and vitamins supplementation to form five diets. Diet 1 = BH + minerals and vitamins (BH), Diet 2 = MS + minerals and vitamins (MS), Diet 3 = LH + minerals and vitamins (LH), Diet 4 = LGH + minerals and vitamins (LGH), Diet 5 = UTRS + minerals and vitamins (UTRS). In experiment 2 (Expt.2) the forages were supplemented with urea (20 g d'*/100 kg body weight) and 200 g d’1 fishmeal (FM). Diet 1 = BH + minerals and vitamins (UNBH), Diet 2 = BH + minerals and vitamins + urea + FM (SBH), Diet 3 = MS + minerals and vitamins + urea + FM (SMS), Diet 4 = LH + minerals and vitamins + urea + FM (SLH) and Diet 5 = UTRS + minerals and vitamins + urea + FM (SUTRS). Parameters measured were similar in both experiments. These were voluntary feed intake (VFI) and faecal output by total collection method for seven days, whereas rumen pH and NH3 were measured at an interval of 2 h for 24 h. In situ degradability was carried out in a standardprocedure and in both experiments. Rumen pool sizes were measured by rumen evacuation technique (RET) during the last seven days of the experiments. Fractionation of the NDF into digestible (DNDF) and indigestible (INDF) was done using the in situ long time (30 days) incubation of feeds, rumen content and faeces. The parameters were used together with rumen pool sizes to derive rate of digestion and passage of NDF, DNDF and INDF. Passage rate of NDF was also estimated using chromium mordanted fibre (CR-MF). The chemical composition, the degradability characteristics, digestion and passage rates and the obtained rumen pool sizes varied substantially between the tropical forages studied. The crude protein (CP) ranged from 64-151 g kg’1 DM, NDF ranged from 615 - 770 g kg’1 DM and ash content ranged from 93 - 180 g kg’1 DM. There was a marked difference (P<0.0001) between forages in the rate and extent at which these forages were degraded in the standard procedure and in experiments 1 and 2 for DM, N and NDF. The lag times were short (0 - 9 h) and not different (P>0.05). In the standard procedure the NDF had b value of 47, 58, 70, 72 and 76 % DM for LH, LGH, BH, UTRS and MS respectively. The rate constant c for NDF when forages were incubated in the standard diet was 2.9, 2.9, 3.6, 3.7 and 8.8 % h’1 for BH, UTRS, LGH, MS and LH respectively. The rate constant c for NDF when forage were incubated in diets for expt.l was 2.0, 3.2, 3.8, 5.0 and 6.9 % h'1 for MS, UTRS, BH, LGH and LH respectively, whereas that of expt.2 was 2.6, 2.9, 3.0, 3.6, and 9.1 % h’1 SMS, SBH, UNBH, SUTRS and SLH respectively.iv The forage dry matter intake (DMI) were different (P<0.01) and the animals consumed 4.2, 5.1, 6.3. 6.5 and 6.7 kg d'1 of MS, BH, UTRS. LH and LGH respectively in expt.l. Supplemented forage diets in expt.2 were not different (P>0.05) for all other nutrients intake except N and water. Forage DMI in expt.2 was 4.4, 4.8, 4.8, 5.9 and 6.4 kg d'1 for SMS, UNBH, SBH, SLH and SUTRS respectively. Water intake was 10, 22, 23. 30 and 31 / d’1 for MS, BH. UTRS, LH and LGH respectively in expt.l and 14. 21, 23, 26 and 30 / d'1 for SMS. UNBH, SBH, SLH and SUTRS respectively in expt.2. The rumen pool sizes of NDF (kg) in expt.l were different (P<0.01) and were higher in animals fed LGH (5.2) and lowest in MS (3.8) compared to those fed BH (4.8) and UTRS (4.5). In expt.2 the NDF pool sizes (kg) were also different (P<0.002) highest in animals fed SBH (4.9) and lowest in animals fed SMS (3.3) compared to those fed UNBH (4.0) with similar values in SLH and SUTRS (4.3). The passage rates (A>) measured using CR-MF and RET showed high variability between forages and between the two methods in both experiments. The passage rates estimated by RET for NDF differed (P<0.03) between the forage diets and were 1.78, 1.63, 1.54, 1.53 and 1.36 % h’1 for UTRS. LGH, BH, LH and MS respectively in expt.l and 2.02, 1.50, 1.41, 1.38 and 1.32 % h" for SUTRS, UNBH, SBH, SMS and SBH respectively in expt.2 (P < 0.05). The kp measured using CR-MF was higher than kp derived from RET and varied from 2.26 - 3.24 % h’1 in expt.l (P>0.42) and 2.22 - 3.86 % h'1 in expt.2 (P<0.003). There were differences in Fill (day) when calculated using kp obtained from RET - total NDF in expt.l (P<0.045) and expt.2 (P<0.046). In expt.l the Fill (day) values were 1.84, 1.77, 1.55,V 1.41 and 1.35 for MS, LH, BH, LGH and UTRS respectively. Fill (day) values were 1.91, 1.65, 1.64. 1.55 and 1.09 for SBH. UNBH, SLH, SMS and SUTRS respectively in expt.2. Fill values were not different when calculated using kp from CR-MF in both experiments. Fill values calculated using an assumed passage rate of 2 % h’1 as a constant value were also not different in expt.l (P>0.35) and expt.2 (P>0.09). The predicted NDF intake (PNDFI) across forages were over-estimated when fill calculated from passage rate estimated from CR-MF was used and under-estimated when passage rate obtained from RET was used. A similar trend to that obtained in PNDFI was observed for predicted DM intake (PDMI) for forages in expt.l and 2. The accuracy of PDMI from PNDFI from passage rates obtained from other NDF fractions-RET was generally poor for both experiments, except for RET-total NDF. The best prediction was obtained from RET-total NDF (P<0.0001) with /?’= 0.70 (RMSE = 0.83) in expt.l and /?’ = 0.75 (RMSE = 0.71) in expt.2. It was concluded that the most important limitation to tropical forage intake is the physical fill of NDF in the rumen. 'ITie results have demonstrated that intake of tropical forages can adequately be described from physical fill based on degradability characteristics, rumen pool size and passage rate of NDF measured by RET. The use of NDF parameters as predictors of VFI is recommended as a realistic estimate of forage DMI because NDF was found to be distinct from microbial and endogenous materials. It can further be concluded that PDMI of tropical forages fed to cross-bred dairy heifers can be calibrated in the equation DMI (kg d’1) = 1.19 + 1.10 PDMI (RMSE = 0.83) in expt.l and DMI (kg d1) = 0.98 +1.16 PDMI (RMSE = 0.71) in expt.2.

Danish International Development Agency (DANIDA) through ENRECA (Enhancement of the Research Capability) Project