A quantitative review of on-farm feeding practices to enhance the quality of grassland-based ruminant dairy and meat products.

In the last decades, a large body of evidence has highlighted the major role of feeding management practices in improving specific nutritional, technological and sensory quality traits of ruminant products. However, results have been mostly obtained under controlled conditions, and have been rarely validated on-farm. Therefore, a quantitative review was conducted to quantify the effects of on-farm feeding management practices on carotenoids, fat-soluble vitamins, colour, fatty acids (FAs), terpenes and sensory properties in the main animal product categories (PCs): dairy products from cattle (DC), sheep (DS) and goat (DG), and meat from cattle (MC) and sheep (MS). Four feeding scenarios were selected according to the consistency of on-farm studies in the literature: (a) feeding "Fresh herbage" instead of conserved forages; (b) ban any form of silage ("Silage-free"); (c) ban maize silage ("Maize silage-free"); (d) feeding forages from permanent grasslands rich in species or plant secondary metabolites (PSMs) ("PSM-rich permanent grassland"). Feeding fresh herbage increased the concentration of carotenoids, fat-soluble vitamin, n-3 FA, rumenic acid, and branched chain FA (BCFA), and reduced the concentration of saturated FA, for all PC, with overall stronger effect for dairy products than for meat. The texture of meat and dairy products was marginally affected, whereas feeding fresh herbage decreased lactic and increased vegetal notes in DC. The "Silage-free" feeding scenario resulted in increased vaccenic acid, rumenic acid, BCFA, and C18:3n-3 in DC. The "Maize silage-free" feeding scenario lowered n-6 FA whereas increased n-3, rumenic acid and BCFA concentrations in DC. Feeding ruminants with forages from "PSM-rich permanent grasslands" increased monounsaturated FA, n-3 FA and rumenic acid and decreased n-6 FA in dairy products, and only marginally affected meat FA composition. The DC from "PSM-rich permanent grasslands" showed higher intense, spicy and animal notes. Overall, the differences between feeding management practices observed on farm were smaller than those observed under controlled trials. Several confounding factors, not controlled when operating under on-farm conditions, could be at the origin of these divergences (i.e. mixed diets, forage characteristics, animal-related factors). This review confirmed that farming practices may differently affect several quality traits of ruminant products. It also highlighted the uneven knowledge on the effect of feeding management depending on the PC: larger for milk than for meat and decreasing when moving from cattle to sheep and from sheep to goat.

What is the impact of the rearing management applied during the heifers' whole life on the toughness of five raw rib muscles in relation with carcass traits?

The aims of this study were, analysing the effects of rearing managements, carcass traits, and muscle type (M. complexus [CP], M. infraspinatus [IF], M. longissimus [LM], M. rhomboideus [RH], and M. serratus ventralis [SV]) on toughness of raw meat; developing prediction models to act on their toughness. According to our results obtained on the data of 77 heifers, the IF raw muscle was the toughest and appeared the most sensitive to a change in the rearing management. The four other raw muscles had a similar toughness within heifers from the same rearing management. The five raw muscles were less tough when the carcass was heavier and had higher dressing percentage and conformation. The 3 models explained about 40% of the variability observed. Our models showed that it is possible to improve the potential tenderness of raw meat, acting on: age of the heifer's mother, growth rate during the growth and fattening periods, slaughter age, carcass weight and temperature 24 h post-mortem.

Influence of the rearing managements and carcass traits on the sensory properties of two muscles: Longissimus thoracis and rectus abdominis.

The aim of this study was to analyse simultaneously the effect of rearing managements (RM), carcass traits, muscle type (longissimus thoracis, LT and rectus abdominis, RA) and their interactions on colour (system L*a*b*), sensory and rheological properties of 77 heifers. The data used were 46 rearing factors defining four RM applied during the heifers' whole life from and 5 carcass traits discriminating two carcass quality clusters (Low and High quality). The results showed that the RM had an impact on the carcass and meat quality traits. The redness and the overall acceptability of LT meat were more sensitive to variation of RM than RA meat. The tenderness of LT and RA were similar or not according to the RM. Higher tenderness was obtained from high carcass quality irrespective the muscle. These results show that similar carcass and meat qualities can be obtained with different RM. They also demonstrated that it is also possible to manage jointly carcass and meat qualities.

Prediction of beef carcass and meat quality traits from factors characterising the rearing management system applied during the whole life of heifers.

In this study, four prediction models were developed by logistic regression using individual data from 96 heifers. Carcass and sensory rectus abdominis quality clusters were identified then predicted using the rearing factors data. The obtained models from rearing factors applied during the fattening period were compared to those characterising the heifers' whole life. The highest prediction power of carcass and meat quality clusters were obtained from the models considering the whole life, with success rates of 62.8% and 54.9%, respectively. Rearing factors applied during both pre-weaning and fattening periods influenced carcass and meat quality. According to models, carcass traits were improved when heifer's mother was older for first calving, calves ingested concentrates during pasture preceding weaning and heifers were slaughtered older. Meat traits were improved by the genetic of heifers' parents (i.e., calving ease and early muscularity) and when heifers were slaughtered older. A management of carcass and meat quality traits is possible at different periods of the heifers' life.

Prediction of beef carcass and meat traits from rearing factors in young bulls and cull cows.

The aim of this study was to predict the beef carcass and LM (thoracis part) characteristics and the sensory properties of the LM from rearing factors applied during the fattening period. Individual data from 995 animals (688 young bulls and 307 cull cows) in 15 experiments were used to establish prediction models. The data concerned rearing factors (13 variables), carcass characteristics (5 variables), LM characteristics (2 variables), and LM sensory properties (3 variables). In this study, 8 prediction models were established: dressing percentage and the proportions of fat tissue and muscle in the carcass to characterize the beef carcass; cross-sectional area of fibers (mean fiber area) and isocitrate dehydrogenase activity to characterize the LM; and, finally, overall tenderness, juiciness, and flavor intensity scores to characterize the LM sensory properties. A random effect was considered in each model: the breed for the prediction models for the carcass and LM characteristics and the trained taste panel for the prediction of the meat sensory properties. To evaluate the quality of prediction models, 3 criteria were measured: robustness, accuracy, and precision. The model was robust when the root mean square errors of prediction of calibration and validation sub-data sets were near to one another. Except for the mean fiber area model, the obtained predicted models were robust. The prediction models were considered to have a high accuracy when the mean prediction error (MPE) was ≤0.10 and to have a high precision when the was the closest to 1. The prediction of the characteristics of the carcass from the rearing factors had a high precision ( > 0.70) and a high prediction accuracy (MPE < 0.10), except for the fat percentage model ( = 0.67, MPE = 0.16). However, the predictions of the LM characteristics and LM sensory properties from the rearing factors were not sufficiently precise ( < 0.50) and accurate (MPE > 0.10). Only the flavor intensity of the beef score could be satisfactorily predicted from the rearing factors with high precision ( = 0.72) and accuracy (MPE = 0.10). All the prediction models displayed different effects of the rearing factors according to animal categories (young bulls or cull cows). In consequence, these prediction models display the necessary adaption of rearing factors during the fattening period according to animal categories to optimize the carcass traits according to animal categories.

Establishment of ruminal bacterial community in dairy calves from birth to weaning is sequential.

AIM: Establishment of ruminal bacterial community in dairy calves. METHODS AND RESULTS: Rumen bacterial community was analysed on 6 calves bred according to commercial practices from day one to weaning at day 83 of age, using 454 16S rRNA-based pyrosequencing. Samples taken at day 1 did not produce amplicons. Analysis of data revealed a three-stage implantation process with a progressive but important shift of composition. At day 2, the bacterial community was mainly composed of Proteobacteria (70%) and Bacteroidetes (14%), and Pasteurellaceae was the dominant family (58%). The bacterial community abruptly changed between days 2 and 3, and until day 12, dominant genera were Bacteroides (21%), Prevotella (11%), Fusobacterium (5%) and Streptococcus (4%). From 15 to 83 days, when solid food intake rapidly increased, Prevotella became dominant (42%) and many genera strongly decreased or were no longer detected. A limited number of bacteria genera correlated with feed intake, rumen volatile fatty acids and enzymatic activities. CONCLUSION: The ruminal bacterial community is established before intake of solid food, but solid food arrival in turn shapes this community. SIGNIFICANCE AND IMPACT OF THE STUDY: This study provides insight into the establishment of calves' rumen bacterial community and suggests a strong effect of diet.

Repeated acidosis challenges and live yeast supplementation shape rumen microbiota and fermentations and modulate inflammatory status in sheep.

This study aimed to investigate the impact of repeated acidosis challenges (ACs) and the effect of live yeast supplementation (Saccharomyces cerevisiae I-1077, SC) on rumen fermentation, microbial ecosystem and inflammatory response. The experimental design involved two groups (SC, n=6; Control, n=6) of rumen fistulated wethers that were successively exposed to three ACs of 5 days each, preceded and followed by resting periods (RPs) of 23 days. AC diets consisted of 60% wheat-based concentrate and 40% hay, whereas RPs diets consisted of 20% concentrate and 80% hay. ACs induced changes in rumen fermentative parameters (pH, lactate and volatile fatty-acid concentrations and proportions) as well as in microbiota composition and diversity. The first challenge drove the fermentation pattern towards propionate. During successive challenges, rumen pH measures worsened in the control group and the fermentation profile was characterised by a higher butyrate proportion and changes in the microbiota. The first AC induced a strong release of rumen histamine and lipopolysaccharide that triggered the increase of acute-phase proteins in the plasma. This inflammatory status was maintained during all AC repetitions. Our study suggests that the response of sheep to an acidosis diet is greatly influenced by the feeding history of individuals. In live yeast-supplemented animals, the first AC was as drastic as in control sheep. However, during subsequent challenges, yeast supplementation contributed to stabilise fermentative parameters, promoted protozoal numbers and decreased lactate producing bacteria. At the systemic level, yeast helped normalising the inflammatory status of the animals.

Modification of activities of the ruminal ecosystem and its bacterial and protozoan composition during repeated dietary changes in cows.

Dietary change alters the ruminal ecosystem and can be regarded as a disturbance. Studying the response to a disturbance can help us understand the behavior of the ecosystem. Our work is concerned with the response of the ruminal ecosystem (composition and activities) during the application of repeated dietary disturbances to 6 dry Holstein cows. For 2 mo, the cows received a hay-based diet [experimental period (EP) 0], followed by 3 EP of successive changes (EP 1, 2, and 3) comprised of 2 parts: the first (10 d) with a corn silage-based diet and the second (25 d) with a hay-based diet. The measurements and samplings were done on the last days of EP 0 and of each part of EP 1 through 3, with the results of EP 0 used as covariables in the statistical models. The physicochemical measurements (pH and redox potential) and the fermentation variables (VFA, ammonia) were determined hourly between the morning and evening meals (n = 8 measurements/d). Samples of ruminal contents were taken 3 h after the morning meal to determine enzymatic activity [amylase, carboxymethyl cellulase (CMCase), and xylanase], to count the main protozoan genera and to quantify the bacteria by quantitative PCR, and to determine its structure by the capillary electrophoresis single-strand conformation polymorphism process. The pH fell for the corn silage-based diet with the EP (P < 0.05) but not for the hay-based diet. The VFA concentration decreased for both diets with the EP (P < 0.001), with the primary changes in acetate and propionate. The ammonia concentration increased for the corn silage-based diet with the EP (P < 0.05), whereas for the hay-based diet the highest value was observed for EP 2 (P < 0.05). The total quantity of bacteria decreased between EP 1 and 3 (P < 0.05) for both diets. The structure of the bacterial community was not affected by the disturbances for the corn silage-based diet, whereas for the hay-based diet large differences were evident between EP 1 and 3 (P < 0.05) and 2 and 3 (P < 0.01). The number of protozoa increased over the EP, with a more marked effect for the corn silage-based diet (diet × EP interaction, P < 0.05). The specific amylase, CMCase, and xylanase activities decreased over the EP for both diets (P < 0.05). The dietary changes applied in our experiment involved strong modifications of the ruminal ecosystem and alterations of ruminal fermentation and enzymatic activities. These alterations were reinforced with the repetition of the dietary changes.

Establishment of ruminal enzyme activities and fermentation capacity in dairy calves from birth through weaning.

The objectives of this study were to characterize the establishment of ruminal fermentation and enzymatic activities in dairy calves from birth to weaning (d 83). Six Holstein calves, immediately separated from their mother at birth, were fed colostrum for 3 d after birth, and thereafter milk replacer, starter pelleted concentrate, and hay until d 83 of age. Ruminal samples were collected from each calf every day for the first 10 d, and additionally at d 12, 15, 19, 22, 26, 29, 33, 36, 40, 43, 47, 50, 55, 62, 69, and 83. Ruminal samples were collected 1h after milk feeding with a stomach tube. The pH and redox potential (E(h)) were immediately measured. Samples were kept for further determination of ammonia nitrogen (NH(3)-N) and volatile fatty acid (VFA) concentrations, and xylanase, amylase, urease, and protease activities. Ruminal pH averaged 6.69, 5.82, and 6.34, from d 1 to 9, d 10 to 40, and d 43 to 83 of age, respectively. At first day of life, the ruminal E(h) value was positive (+224 mV). From d 2 to 9, d 10 to 40, and d 43 to 83 of age, ruminal E(h) averaged -164, -115, and -141 mV, respectively. From d 1 to 3, d 4 to 22, and d 26 to 83 of age, NH(3)-N concentration averaged 60.1, 179.8, and 58.2 mg/L, respectively. No VFA were detected in ruminal samples collected on d 1 of life of calves. From d 2 to 10 and d 12 to 83 of age, ruminal total VFA concentration averaged 19.5 and 84.4mM, respectively. Neither ruminal xylanase or amylase activities were observed at d 1 of age. From d 5 to 15 and d 19 to 83 of age, the xylanase activity averaged 182.2 and 62.4 µmol of sugar released per hour per gram of ruminal content dry matter (DM), respectively. From d 5 to 83 of age, the amylase activity reached 35.4 µmol of sugar released per hour per gram of ruminal content DM. The ruminal ureolytic activity was observed with an average value of 6.9 µg of NH(3)-N released per minute per gram of ruminal content DM over the 83-d experimental period. From d 1 to 4 and d 5 to 83 of age, the proteolytic activity was 8.2 and 27.9 optical density units per hour per gram of ruminal content DM, respectively. The fermentative and enzymatic activities were rapidly established in the rumen from d 2 after birth. Most parameters did not evolve further after 1 mo of age.

Changes over time in the bacterial communities associated with fluid and food particles and the ruminal parameters in the bovine rumen before and after a dietary change.

This work aimed to study the changes over time in the bacterial communities associated with the fluid and food particle fractions of the cow rumen following a change in diet. Four cannulated cows were fed a hay-based diet for 21 days and were then switched to a corn-silage-based diet for 33 days. The bacterial communities were regularly characterized by capillary electrophoresis - single-strand conformation polymorphism (CE-SSCP) and qPCR, and the main ruminal parameters were determined. The dietary change led to slight reductions in the diversity index, bacterial concentration, pH, and NH(3)-N concentration, and to an increase in the redox potential and volatile fatty acid concentrations. CE-SSCP profiles were not significantly affected by the dietary change but did change over time, with frequent fluctuations in both fluid and food particle fractions before and after the dietary change. The food particle fraction had a higher diversity index of bacterial community (+1.2 points, P < 0.001) and slightly more total bacteria than the fluid fraction of the rumen.

Rapid adaptation of the bacterial community in the growing rabbit caecum after a change in dietary fibre supply.

This work aimed to study the response of the growing rabbit caecal ecosystem (bacterial community and caecal environmental parameters) after a switch from a control to a low-fibre diet (LFD). A group of 160 rabbits were fed ad libitum a control diet (ADF: 20.4%) from weaning (36 days). At 49 days of age (day 0), 75 rabbits were switched to a LFD group (ADF: 10.7%), whereas 85 others (control group) remained on the control diet, for 39 days. Caecal contents were regularly sampled throughout the trial (60 rabbits per group). The bacterial community structure was characterized using CE-SSCP (capillary electrophoresis single strand conformation polymorphism) and total bacteria were quantified using real-time PCR. Redox potential (Eh), pH, NH(3)-N, volatile fatty acid (VFA) were measured in the caecum to characterize environmental parameters. The reduction of fibre in the diet modified the CE-SSCP profiles (P < 0.001) but not the diversity index (5.6 ± 0.8, ns). The number of 16S rRNA gene copies of total bacteria decreased (P < 0.01) in LFD rabbits compared with controls. In LFD rabbits, the caecal environment was less acid (+0.2 units; P < 0.01), more reductive (-11 mV; P < 0.05) and drier (+3.4 g 100 per g; P < 0.001), with an increase in NH3-N (+77%; P < 0.001) and a decrease in total VFA concentration (-17%; P < 0.001). We found significant correlations between the bacterial community, the quantity of bacteria and the caecal traits of the caecal ecosystem. Indeed, in both groups, the caecal traits barely constrained the total inertia of the CE-SSCP profile set (less than 14%), whereas total bacteria were positively related to total VFA, acetic acid and butyric acid levels, and Eh, and negatively related to pH. All the microbial and environmental modifications had occurred by day 2 and remained stable thereafter. These results suggest that the bacterial community in the growing rabbit caecum is able to adapt quickly after a change to in the dietary fibre supply to reach a new steady-state equilibrium.

Effects of varying proportions of concentrates on ruminal-reducing power and bacterial community structure in dry dairy cows fed hay-based diets.

The objectives of this study were to evaluate the influence of diet composition on ruminal parameters, more particularly redox potential (Eh). Four Holstein dry dairy cows, fitted with ruminal cannulas, were allocated in a 4 × 4 Latin square design. They were given four experimental hay-based diets D0, D25, D42 and D56 consisting of 0%, 25%, 42% and 56% of ground wheat and barley concentrate mixture, respectively. They were fed at a daily feeding rate of 8.0 kg DM per cow during a 24-day experimental period (a 21-day diet adaptation, three consecutive days of measurement and sampling). The physicochemical parameters, such as pH and Eh, were measured and Clark's exponent (rH) was calculated from 1 h before feeding to 8 h after feeding at 1-h interval. Samples of ruminal fluid were taken at 0, 1, 2, 4, 6 and 8 h after feeding for the determination of volatile fatty acid (VFA) and ammonia N (NH3-N) concentrations. Ruminal bacterial populations were also studied by means of capillary electrophoresis single-strand conformation polymorphism (CE-SSCP) technique to focus on the structure of the ruminal microbiota and the diversity index was calculated. Mean ruminal Eh and rH were not modified by the concentrate-to-forage ratio and averaged - 210 mV and 6.30, respectively, across diets. The pH decreased slightly by 0.10 pH unit between treatments D0 and D56 with an average of 6.58. Nevertheless, the time during which physicochemical parameters remained at nadir value after feeding varied with diets: 2 and 7 h for D0 and 6 and 5 h for D56, respectively for pH and Eh. Moreover, fermentative parameters were altered by treatments: total VFA and NH3-N were greater in D56 (72.2 mM and 17.5 mg/100 ml) compared with D0 (65.2 mM and 14.2 mg/100 ml). However, neither the structure of bacterial populations of the rumen nor the diversity index (Shannon) was altered by treatments.

Spatial and temporal variations of the bacterial community in the bovine digestive tract.

AIMS: Improved knowledge of the bacterial community of the digestive tract is required to enhance the efficiency of digestion in herbivores. This work aimed to study spatial and temporal variations of the bacterial communities in the bovine digestive tract and their correlation with gut environmental parameters. METHODS AND RESULTS: Rumen content and faeces of five cows were sampled for 3 weeks. In addition, reticulum content was sampled during the third week. Bacterial communities were assessed by studying capillary electrophoresis single-stranded conformation polymorphism (CE-SSCP) profiles of 16S rRNA genes. The bacterial community structure differed between the forestomach and faecal contents. The abundance of several operational taxonomic units changed from week to week. Bacterial community structure of the rumen was correlated to propionic acid and NH(3)-N concentrations. CONCLUSIONS: The bacterial community of the bovine digestive tract varied in space and time. SIGNIFICANCE AND IMPACT OF THE STUDY: The study of the bacterial communities of the digestive tract in herbivores should be widened from the rumen to the large intestine. The amplitude and origin of the temporal variation of the ruminal bacterial community need to be better understood to improve the control of the fermentative activity in herbivores.

How does live yeast differ from sodium bicarbonate to stabilize ruminal pH in high-yielding dairy cows?

The objectives of this study were to evaluate the capacity of 2 dietary feed additives, sodium bicarbonate and live yeast Saccharomyces cerevisiae (strain Sc 47), in optimizing ruminal pH in dairy cows and to determine their modes of action. Three early lactating Holstein cows, fitted with ruminal cannulas, were allocated in a 3 x 3 Latin square design. They were given a total mixed ration as control diet (CD) at a daily feeding rate of 28.0 kg of dry matter (DM)/cow supplemented with 150 g/d of sodium bicarbonate (SBD) or 5 g/d of live yeast (YD) during a 21-d experimental period (14 d of diet adaptation, 4 consecutive days of measurement and sampling and 3 d of transition). The pH and redox potential (E(h)) were measured from 1 h before feeding to 8 h after feeding at 1-h intervals, and samples of ruminal fluid were taken at 0, 2, 4, 6, and 8 h after feeding for the determination of volatile fatty acids and lactate concentrations. Total tract apparent digestibility of the diet was also determined. Ruminal pH fluctuated between 6.53 at feeding and 5.57 at 5 h postfeeding. Mean pH was greater with SBD (6.21) and YD (6.14) compared with CD (5.94), showing that both additives had a pH stabilization effect. The E(h) varied from -88 mV at 1 h before feeding to -165 mV at 1 h after feeding. Mean E(h) and Clark's Exponent (rH) were lower with YD (-149 mV and 7.31, respectively) than with SBD (-137 mV and 7.85, respectively) and CD (-115 mV and 8.05, respectively), indicating that the yeast strengthened the reducing power of the milieu. Total volatile fatty acids were greater in SBD (95.3 mM) and YD (99.4 mM) compared with CD (85.3 mM). Acetate concentration was greater in SBD (60.8 mM) and YD (59.1 mM) compared with CD (53.2 mM). Propionate concentration was greater in YD (25.8 mM) than in SBD (20.0 mM) and CD (18.0 mM). Butyrate remained constant between diets. Mean total lactate concentrations were 16.5, 12.2, and 5.4 mM for CD, SBD, and YD, respectively, with a 67% decrease with YD. Total tract organic matter digestibility was greater for YD (66.6%) compared with SBD (61.7%) and CD (62.2%). The neutral detergent fiber digestibility was greater with YD (41.6%) compared with SBD (34.3%) and CD (29.6%), whereas acid detergent fiber digestibility was greatest in YD (32.3%), intermediate in SBD (24.4%), and lowest in CD (18.1%). By inducing a lower ruminal E(h) and rH, live yeast prevented accumulation of lactate and allowed better fiber digestion, whereas sodium bicarbonate seemed to act only as an exogenous buffer.

Variations of trans octadecenoic acid in milk fat induced by feeding different starch-based diets to cows.

The impact of starch sources differing in their velocities of ruminal degradation on the milk fat of dairy cows was studied. The animals received diets containing a slowly degradable (potatoes) or rapidly degradable (wheat) starch concentrate (40% of the dry matter) in a total mixed diet. Milk fat was the only animal performance factor affected: Cows produced significantly less milk fat when fed the wheat diet than the potato diet (-3.3 g/kg, -122 g/d; P < 0.05). With the wheat diet, milk fat was poorer in short-chain FA and richer in unsaturated long-chain FA, especially in trans octadecenoic acid (4.4 vs. 2.7% of the total FA, P < 0.05). A very large increase in the isomer trans-10 18:1 (+1.46% of the total FA) was observed. Because no difference in volatile FA concentrations in the rumen was revealed, the increase in trans octadecenoic acids, and particularly the isomer trans-10 18:1, was associated with the larger postprandial drop in ruminal pH with wheat. Similar concentrate levels and FA profiles in both diets indicated that the decrease in milk fat was due to changes in the ruminal environment. Quicker degradation of wheat starch, and hence a greater drop in pH with this diet associated with the absence of any effect on volatile FA, strengthen the hypothesis developed in the literature of enzyme inhibition via increased levels of trans octadecenoic acids, especially the trans-10 isomer. Hence, milk fat can be decreased with rapidly degradable starch sources and not only with high levels of concentrates in the diet or added fat. More detailed work is necessary to elucidate the microorganisms involved and to determine whether metabolic pathways similar to those reported for high-concentrate diets are involved.

Kinetics of ruminal degradation of wheat and potato starches in total mixed rations.

Wheat and potato are rich in starch but their starches differ in their rate of ruminal degradation. Kinetics of in sacco disappearance and profiles of ruminal fermentation were studied for these two concentrates in total mixed rations based on grass silage or corn silage. Wheat starch was more rapidly (34%/h) degraded by rumen microorganisms than potato starch (5%/h). The differences in starch degradation in sacco were found again in the VFA concentrations, mainly in grass silage-based diets. Overall ruminal pH, total VFA concentration, and proportions of acetate, propionate, and butyrate are more variable for wheat during the kinetic (amplitude and quickness) than for potato in grass silage-based diets. In these diets, risks of acidosis were more elevated with wheat than with potato but the VFA concentrations were also higher. These differences of fermentation profile were so reduced in corn silage-based diets that, in this case, wheat can be substituted by potato without any effect on digestion and no risk of acidosis.