RESULTS

Feed composition

The chemical composition of the experimental roughage sources is presented in Table 1. Guinea grass, used as the control roughage, contained 5.8% CP, 75.4% NDF, and 44.0% ADF on a DM basis. Compared with guinea grass, both ensiled TMF rations, TMFc and TMFe, contained higher CP concentrations (8.9% and 8.8%, respectively) and lower NDF contents (62.1% and 65.9%, respectively). Ensiled TMFe had the highest ADF (46.8%) and ADL content (9.6%), whereas guinea grass and ensiled TMFc had comparable ADL values (7.8% and 8.2%, respectively). The inclusion of agro-industrial byproducts in ensiled TMFc and ensiled TMFe altered the fiber profile and nutritional density of the rations, suggesting a potential advantage in nutrient supply over conventional roughage.

One limitation of the present study is that silage fermentation characteristics, including pH, lactic acid, acetic acid, butyric acid, and NH₃–N concentrations, were not determined during the ensiling process. Therefore, although no visible spoilage, mold growth, or abnormal odor was observed, the fermentation quality of ensiled TMF could not be quantitatively evaluated. Future studies should include comprehensive silage quality assessments to better characterize the fermentation profile and preservation efficiency of cassava-based TMF.

Items | Guinea grass | TMFc | TMFe | Concentrate
Amount, % DM |  |  |  | 
Guinea grass | 100.0 | – | – | –
Bagasse | – | 30.0 | 45.0 | –
Vinasse | – | 5.0 | 5.0 | –
Urea | – | 1.6 | 1.0 | –
Rice straw | – | 10.0 | 10.0 | –
Cassava pulp | – | 53.4 | 0.0 | –
Cassava bioethanol waste | – | 0.0 | 39.0 | –
Total, kg | 100.0 | 100.0 | 100.0 | –
Chemical composition (% DM basis) |  |  |  | 
DM | 33.5 | 26.3 | 27.0 | 87.9
CP | 5.8 | 8.9 | 8.8 | 21.1
Ash | 9.1 | 7.3 | 8.3 | 8.2
EE | 1.9 | 1.0 | 1.4 | 3.3
NDF | 75.4 | 62.1 | 65.9 | 38.9
ADF | 44.0 | 44.9 | 46.8 | 34.5
ADL | 7.8 | 8.2 | 9.6 | 9.4

BW, feed intake, and apparent nutrient digestibility

Animal performance and nutrient utilization data are summarized in Table 2[31]. Dietary treatment had no significant effect on initial BW, final BW, or average daily gain (ADG; p > 0.05). Total DM intake (DMI) was significantly affected by dietary treatment (p = 0.007), with cows fed ensiled TMFe showing the highest intake (15.6 ± 0.61 kg/day), followed by cows fed ensiled TMFc (13.2 ± 0.73 kg/day) and guinea grass (12.6 ± 0.64 kg/day). This study is among the first to demonstrate that ensiled TMF formulated with cassava bioethanol waste can enhance voluntary feed intake and energy supply when used as a primary roughage source in early-lactating dairy cows.

Nutrient intake followed a similar pattern, with significantly higher intakes of DM, OM, NDF, and ADF observed in cows fed ensiled TMFe than in those fed the other treatments (p < 0.001, p < 0.001, p < 0.005, and p < 0.001, respectively). Energy intake, calculated based on digestible OM, was also significantly greater in the ensiled TMFe group (36.10 ± 1.26 Mcal metabolizable energy (ME)/day) than in the ensiled TMFc (30.83 ± 1.43 Mcal ME/day) and guinea grass groups (28.60 ± 1.56 Mcal ME/day; p < 0.001). However, apparent nutrient digestibility did not differ significantly among treatments (p > 0.05). Digestibility coefficients for DM, OM, CP, NDF, and ADF were similar across all groups, indicating that the inclusion of ensiled TMF did not negatively affect digestibility despite higher feed intake.

Items | Guinea grass | TMFc | TMFe | SEM | p-value
Initial weight, kg | 406 ± 19.53 | 442 ± 34.57 | 440 ± 32.36 | 16.21 | 0.321
Final weight, kg | 448 ± 25.38 | 508 ± 48.48 | 529 ± 18.90 | 20.09 | 0.052
ADG, kg/day | 0.58 ± 0.05 | 0.92 ± 0.28 | 0.89 ± 0.11 | 0.10 | 0.433
Roughage DMI, kg/day | 5.4 ± 0.11 | 6.7 ± 1.00 | 8.4 ± 0.91 | 0.53 | 0.060
% BW | 1.3 ± 0.09 | 1.5 ± 0.30 | 1.7 ± 0.18 | 0.12 | 0.198
g/kg BW⁰·⁷⁵ | 57.5 ± 3.28 | 67.7 ± 12.62 | 79.7 ± 8.47 | 5.38 | 0.152
Concentrate DMI, kg/day | 7.2 ± 0.58 | 6.5 ± 0.66 | 7.2 ± 0.85 | 0.39 | 0.625
% BW | 1.7 ± 0.14 | 1.4 ± 0.15 | 1.5 ± 0.19 | 0.09 | 0.443
g/kg BW⁰·⁷⁵ | 76.8 ± 5.98 | 64.8 ± 6.74 | 68.7 ± 8.57 | 4.06 | 0.537
Total DMI, kg/day | 12.6 ± 0.64ᵇ | 13.2 ± 0.73ᵇ | 15.6 ± 0.61ᵃ | 0.50 | 0.007
% BW | 3.0 ± 0.20 | 2.9 ± 0.34 | 3.1 ± 0.13 | 0.14 | 0.623
g/kg BW⁰·⁷⁵ | 134.3 ± 7.84 | 132.5 ± 13.32 | 148.3 ± 5.58 | 5.41 | 0.328
Nutrient intake, kg/day |  |  |  |  | 
DM | 7.78 ± 0.46ᶜ | 8.33 ± 0.32ᵇ | 9.77 ± 0.36ᵃ | 0.32 | <0.001
OM | 7.52 ± 0.41ᶜ | 8.10 ± 0.38ᵇ | 9.53 ± 0.33ᵃ | 0.32 | <0.001
CP | 1.23 ± 0.10 | 1.26 ± 0.09 | 1.41 ± 0.21 | 0.07 | 0.430
NDF | 3.78 ± 0.14ᵇ | 3.78 ± 0.27ᵇ | 4.68 ± 0.20ᵃ | 0.16 | <0.005
ADF | 2.77 ± 0.13ᶜ | 3.01 ± 0.21ᵇ | 3.62 ± 0.09ᵃ | 0.13 | <0.001
Energy intake (Mcal ME/day) | 28.60 ± 1.56ᶜ | 30.83 ± 1.43ᵇ | 36.10 ± 1.26ᵃ | 1.20 | <0.001
Apparent nutrient digestibility (%) |  |  |  |  | 
DM | 61.8 ± 1.26 | 63.5 ± 3.10 | 61.6 ± 2.69 | 1.23 | 0.856
OM | 65.5 ± 1.02 | 67.1 ± 2.88 | 61.6 ± 2.69 | 1.28 | 0.759
CP | 66.7 ± 1.10 | 65.9 ± 4.40 | 62.1 ± 5.24 | 1.88 | 0.975
NDF | 55.3 ± 1.22 | 55.9 ± 2.49 | 54.8 ± 1.82 | 0.98 | 0.983
ADF | 57.3 ± 0.80 | 58.0 ± 1.35 | 55.2 ± 1.86 | 0.72 | 0.873

Rumen fermentation characteristics and blood metabolites

Rumen fermentation characteristics and BUN concentrations are presented in Table 3. Ruminal temperature and pH remained within physiological ranges across all treatments and were not significantly affected by dietary treatment (p > 0.05). Mean ruminal temperature ranged from 38.3 ± 0.54°C to 39.2 ± 0.12°C, whereas mean pH ranged from 6.9 ± 0.09 to 7.1 ± 0.11. At 0 h after feeding, no significant differences were observed among treatments for BUN, NH₃–N, propionate, or acetate concentrations (p > 0.05), indicating that baseline rumen fermentation characteristics were generally comparable among treatments before feeding.

At 4 h after feeding, BUN concentration was highest in cows fed ensiled TMFc (21.2 ± 1.89 mg/dL), which was significantly greater than that in cows fed guinea grass (15.0 ± 1.37 mg/dL; p < 0.05), whereas cows fed ensiled TMFe showed intermediate values. However, mean BUN concentration did not differ significantly among treatments (p = 0.387). The NH₃–N concentration at 4 h after feeding was also highest in the ensiled TMFc group (14.24 ± 0.35 mg%), which was significantly greater than that in the ensiled TMFe and guinea grass groups (p = 0.001), suggesting increased ruminal nitrogen availability with cassava pulp inclusion.

The VFA profile showed no significant differences in mean acetate or butyrate concentrations among treatments. However, mean propionate concentration was significantly higher in the guinea grass group (27.41 ± 1.89 mM) than in the ensiled TMFc (22.36 ± 0.77 mM) and ensiled TMFe groups (22.23 ± 1.67 mM; p = 0.044). Butyrate concentration at 0 h after feeding was significantly higher in cows fed ensiled TMFc than in those fed the other treatments (p = 0.013). Overall, ensiled TMF diets, particularly ensiled TMFc, influenced nitrogen metabolism and selected fermentation end-products without impairing rumen stability.

Items | Guinea grass | TMFc | TMFe | SEM | p-value
Temperature, °C |  |  |  |  | 
0 h post-feeding | 37.9 ± 0.83 | 38.9 ± 0.44 | 38.0 ± 0.24 | 0.32 | 0.265
4 h post-feeding | 38.3 ± 0.43 | 39.4 ± 0.29 | 38.9 ± 0.60 | 0.26 | 0.068
Mean | 38.3 ± 0.54 | 39.2 ± 0.12 | 38.3 ± 0.23 | 0.22 | 0.209
pH |  |  |  |  | 
0 h post-feeding | 7.3 ± 0.11 | 7.4 ± 0.05 | 7.3 ± 0.04 | 0.04 | 0.314
4 h post-feeding | 6.7 ± 0.13 | 6.6 ± 0.15 | 6.9 ± 0.30 | 0.10 | 0.660
Mean | 7.0 ± 0.12 | 6.9 ± 0.09 | 7.1 ± 0.11 | 0.06 | 0.825
BUN, mg/dL |  |  |  |  | 
0 h post-feeding | 13.5 ± 1.77 | 14.5 ± 1.36 | 14.6 ± 2.06 | 0.94 | 0.904
4 h post-feeding | 15.0 ± 1.37ᵇ | 21.2 ± 1.89ᵃ | 18.8 ± 1.96ᵃᵇ | 1.14 | 0.047
Mean | 14.3 ± 1.50 | 17.8 ± 1.58 | 16.7 ± 1.97 | 0.98 | 0.387
NH₃–N, mg% |  |  |  |  | 
0 h post-feeding | 10.70 ± 0.45 | 10.42 ± 0.43 | 10.76 ± 0.44 | 0.24 | 0.790
4 h post-feeding | 12.75 ± 0.30ᵇ | 14.24 ± 0.35ᵃ | 13.17 ± 0.41ᵇ | 0.25 | 0.001
Mean | 11.72 ± 0.22 | 12.33 ± 0.23 | 12.12 ± 0.26 | 0.14 | 0.137
Volatile fatty acids, mM |  |  |  |  | 
Acetate |  |  |  |  | 
0 h post-feeding | 78.61 ± 1.40 | 74.94 ± 2.06 | 74.69 ± 1.52 | 1.03 | 0.305
4 h post-feeding | 88.27 ± 1.17 | 87.23 ± 1.37 | 86.19 ± 2.51 | 0.88 | 0.747
Mean | 83.44 ± 1.22 | 81.09 ± 1.63 | 80.14 ± 1.63 | 0.88 | 0.407
Propionate |  |  |  |  | 
0 h post-feeding | 25.91 ± 1.82 | 19.84 ± 1.59 | 20.31 ± 1.95 | 1.19 | 0.062
4 h post-feeding | 28.91 ± 2.17 | 24.88 ± 0.22 | 24.33 ± 1.26 | 0.98 | 0.089
Mean | 27.41 ± 1.89ᵃ | 22.36 ± 0.77ᵇ | 22.23 ± 1.67ᵇ | 1.03 | 0.044
Butyrate |  |  |  |  | 
0 h post-feeding | 7.87 ± 0.56ᵇ | 8.86 ± 0.71ᵃ | 7.74 ± 0.49ᵇ | 0.35 | 0.013
4 h post-feeding | 11.81 ± 1.09 | 10.76 ± 0.62 | 9.52 ± 1.05 | 0.55 | 0.322
Mean | 9.84 ± 0.76 | 9.81 ± 0.46 | 8.82 ± 0.59 | 0.36 | 0.320

Milk yield and composition

Milk production and composition data are summarized in Table 4. Although the difference was not statistically significant, cows fed ensiled TMFc tended to produce higher daily milk yield (15.5 ± 0.82 kg/day; p = 0.069). This tendency may indicate differences in nutrient partitioning and energy utilization between cassava-derived byproducts. Milk composition, including fat, protein, lactose, solids-not-fat, and total solids, was not significantly affected by dietary treatment (p > 0.05). Milk fat content averaged 3.51 ± 0.37%, 3.51 ± 0.25%, and 3.57 ± 0.34% in the guinea grass, ensiled TMFc, and ensiled TMFe groups, respectively, whereas milk protein content ranged from 3.06 ± 0.07% to 3.14 ± 0.04%.