Effect of a Corn Silage-Based Finishing Diet on Growth, Carcass Composition, Meat Quality, Methane Emissions and Carbon Footprint of Crossbred Angus Young Bulls

Abstract

Using locally produced forage and agro-industrial by-products can reduce dependence on imported feed and competition for human food sources, while improving meat quality. However, the overall effect of this feeding strategy on global greenhouse gas emissions must be evaluated to provide a comprehensive assessment of sustainability. This study aimed to test whether replacing the conventional concentrate finishing diet with a total mixed ration (TMR) diet based on maize silage and brewer’s spent grains (BSG) would improve meat quality without compromising productive performance, carcass composition, and the carbon footprint (CFp) of finishing beef cattle. Twenty crossbred young bulls were randomly distributed among 4 pens and randomly allocated to 2 treatments: Control—a conventional diet based on commercial concentrate and wheat straw or TMR—a maize silage-based diet with BSG, concentrate, and straw. Dry matter intake and average daily gain were 13% and 15%, respectively, lower in the TMR treatment than in the Control treatment. Daily methane emissions were 59% higher in the TMR treatment. However, life cycle assessment results revealed no differences in the CFp, and the beef from TMR treatment achieved higher meat quality. In conclusion, a maize silage-based diet offers a cost-effective alternative to conventional diets, with a lower environmental impact and improved beef quality.

1. Introduction

Beef production is a major industry worldwide that has been under strong social pressure due to its significant contribution to the emissions of greenhouse gases (GHG), particularly methane (CH₄), produced in the cattle digestive tract [1]. The beef sector is estimated to account for between 2.2% and 6% of global anthropogenic GHG emissions [2,3]. Nevertheless, the carbon footprint (CFp) varies largely with the type and intensification level of beef production systems [4]. The most effective strategies to reduce CH₄ emissions involve optimizing diets and incorporating feed additives [5,6]. In many beef production systems, calves remain with cows on natural or sown pasture until weaning and are slaughtered after a fattening period of variable duration [7]. During this finishing phase, nutrient-dense diets, primarily based on high-protein concentrates, are provided to maximize growth [7]. In Europe, including Portugal, the animal feed industry relies heavily on imported raw materials [8]. Including locally sourced raw feeds and using industrial food by-products in finishing diets can enhance sustainability by reducing reliance on imports and competition with human food. In this context, the inclusion of locally produced forage in finishing diets should be promoted [7]. Total mixed rations (TMR) diets, incorporating high-quality forages such as maize silage (MS), are becoming increasingly common. Incorporating MS into finishing diets allows for the reduction in concentrate use, which can lower production costs without compromising performance [9]. Compared to concentrate-based diets, the TMR forage-based diets can also improve the nutritional profile of beef by increasing n − 3 polyunsaturated fatty acids (PUFA), vaccenic (t11-18:1), and rumenic acids (c9,t11-18:2) while preventing the accumulation of undesirable trans fatty acids (FA) like t10-18:1 [7]. Previous studies suggest that TMR with MS can also improve meat sensory quality traits [10]. An advantage of using TMR diets is the facility to include locally produced agro-industrial by-products unsuitable for human consumption in cattle finishing diets. However, their availability is often restricted seasonally and they have a higher fibre content, which may increase CH₄ emissions, hindering the efforts to reduce net GHG emissions [7]. Increasing the digestibility of diets does not always result in lower total GHG emissions, as other sources of emissions on the farm may increase [11]. To accurately assess the net impact of dietary changes on GHG emissions, it is essential to use a systemic approach like life cycle assessment (LCA) that accounts for GHG emissions from all stages of production, including feed production, manure management, and transport, providing a comprehensive picture of the total emissions from finishing beef production systems [12]. Thus, the net effect on GHG emissions of reducing beef cattle finishing diets’ cereals and oilseeds meals by incorporating maize silage and agro-industrial by-products, such as brewers’ spent grain, needs to be evaluated. Therefore, the aim of this study was to test the hypothesis that replacing the conventional concentrate-straw finishing diet with a TMR diet based on MS and brewers’ spent grains would improve meat quality without compromising productive performance, carcass composition, CH₄ emissions, and CFp of finishing beef cattle.

2. Materials and Methods

All procedures were conducted following the European Union’s rules for the use of animals in experiments (Directive 2010/63/EU) [13]. This study was approved by the Research and Education Ethics Committee (CEIE) of the Faculty of Veterinary Medicine, University of Lisbon, Portugal (Proc. 038/2023).

2.1. Animals, Diets, and Experimental Design

Twenty young crossbred males, sired by Angus bulls and Limousin or Charolais cows and born between September and November 2021, were selected for this experiment. The young males had an average age of 13.6 months and an average live weight (LW) of 473 ± 37.6 kg. The experiment was conducted in a farm (Quinta da França, Covilhã, Portugal, 40°16′24.73″ N, 7°25′55.535″ W) where 4 pens (5.8 m × 10 m), along a 4 m wide corridor and a 2 m wide rear sleeve that directed the animals to the 50 m² outdoor weighing area, were used for this trial. The pens had a concrete floor and an automatic stainless steel water distribution system. On arrival, the young bulls were blocked by weight and farm of origin and randomly distributed among the four pens, five animals per pen. The trial began after 12 days of adaptation to the diets and environmental conditions. After the adaptation period, the LW was 500 ± 38.5 kg (mean ± sd).

The pens (experimental units) were randomly allocated to two dietary treatments: Control—a conventional diet based on commercial concentrate (NANTA II Nutrição S.A., Ovar, Portugal) and wheat straw; TMR—a forage-based diet consisting of maize silage, brewer’s spent grains, concentrate, and the same straw. The TMR was prepared daily with a vertical mixer (Mammut, Profi Mix, Gurten, Austria). Maize silage was produced on the farm in 2022 and the brewer’s spent grains were produced in November 2022 and supplied by a national brewer. The composition of the diets used is shown in

Table 1. Chemical composition of diet ingredients.

Ingredient
Chemical Composition (% DM 1)	Control Concentrate 2	TMR Concentrate 3	GF Pelleted 4	Maize Silage	Breer’s Spent Wet Grains	Straw
Dry Matter (%)	88	88.3	89	29.8	29.7	97.5
Crude Protein	14.7	13.7	16.9	7.73	29.0	-
Ether Extract	7.3	6.9	5.6	-	-	-
Crude Fibre	6.6	5.0	6	22.5	20.8	-
Starch	42.6	58.6	45	33.0	1.2	-
NDF 5	22.5	14.2	20.2	46.5	43.2	-
ADF 6	8.2	6.0	7.9	26.0	27.4	-
Calcium	0.74	1.6	1.1	-	-	-
Phosphorus	0.65	0.53	0.45	-	-	-
Sodium	0.20	0.43	0.45	-	-	-

¹—Dry matter; ²—Ground concentrate used in the control treatment; ³—Ground concentrate incorporated in the TMR diet; ⁴—Concentrate pellets used as bait in the Greenfeed apparatus; ⁵—Neutral detergent fibre; ⁶—Acid detergent fibre.

. Weekly, sub-samples of the Control and TMR diets were collected and frozen. The sub-samples were pooled monthly, and the composite samples were analysed for the chemical composition, which is shown in

Table 2. Ingredients, chemical composition, and fatty acid (FA) profile of the diets.

	Diets
	Control	TMR
Ingredients (g/kg DM)
Control concentrate	887	-
TMR concentrate	-	328
Maize silage	-	471
Brewer’s spent wet grains	-	189
Straw	113	12
Chemical composition (g/kg DM)
DM (g/kg as fed)	862	393
Crude Protein	132	137
Ether Extract	75	51
Crude Fibre	91	162
NDF 1	321	491
ADF 2	123	235
ADL 3	21	30
Sugar	34.1	18.5
Starch	396	312
Ash	54	76
Calcium (mg/g)	4.79	4.90
In Vitro Digestibility (g/kg)
DM 4	789	694
OM 5	779	709
Metabolizable Energy 6 (MJ/kg DM)	11.9	10.9
Fatty acid profile (g/kg FA)
14:0	-	0.9
16:0	300	281
18:0	60	36
c9-18:1	223	200
c11-18:1	5.4	7.5
18:2n − 6	348	416
20:0	17.9	7.6
18:3n − 3	15.3	38.5
22:0	12.0	5.6
24:0	18.5	7.7
Total FA (g/kg DM)	65.6	39.5
Feed cost (€/kg DM)	0.441	0.305

¹—Neutral detergent fibre; ²—Acid detergent fibre; ³—Lignin; ⁴—Dry matter; ⁵—Organic matter; ⁶—Calculated using the formula by [14].

.

Animals were fed ad libitum, once daily, at approximately 8:00 a.m., allowing for a 10% refusal rate. Orts in each pen were recorded daily. Animals were weighed every 2 weeks to quantify growth performance. The trial lasted for 12 weeks, after which the animals were slaughtered as described in point 2.3. Feed costs were computed using raw materials market prices in 2022.

2.2. Gas Emissions Measurement

Methane and carbon dioxide (CO₂) emissions of animals were recorded using GreenFeed equipment (GF) (GreenFeed—Large Animals, C-Lock Inc., Rapid City, SD, USA). The pelleted compound feed (Rações Zêzere S.A., Ferreira do Zêzere, Portugal) used as bait in the GF is described in Table 1.

The equipment was automatically calibrated every three days for CH₄ (zero and span) using the manufacturer’s built-in automatic calibration system. According to the manufacturer’s instructions and previously described [7]. Every 15 days, the air filter was changed and cleaned. The GF was available in each pen with free access for seven days. After this time, the GF equipment was moved to the subsequent pen. So, throughout the experimental period, the animals in each pen had access to the GF unit for three 7-day periods separated by 4 weeks. The GF was instructed to deliver eight doses of bait a maximum of five times per day, with no minimum time between each visit. Each dose weighed an average of 38 g. The homogeneity of the animals’ ages and LW attempts to prevent social dominance during visits to the GF. Metallic alley barriers were also installed at the entrance to the GF to ensure that only one animal was visited at a time.

2.3. Slaughter and Meat Sample Collection

The animals were transported and slaughtered at a commercial abattoir (OVIGER—P.T.C. Carnes e Derivados S.A., Castelo Branco, Portugal) located 48 km from the farm.

Animals were stunned using a penetrating captive bolt and slaughtered by exsanguination following the European standards for the protection of animals at the time of killing [15]. After weighing, carcass conformation and fattening were evaluated after dressing using the SEUROP grading scheme [16]. The carcasses were split in half along the vertebral column and chilled in a cold room at 0–4 °C. At 48 h post-slaughter, longissimus lumborum (LL) muscle samples and the “Rib Primal 103” [17] portion were taken from the left half-carcasses. The 9th, 10th, and 11th ribs were obtained from the “Rib Primal 103” using a saw [18]. The sections were vacuum-packed and frozen at −20 °C until they were dissected.

2.4. Prediction of Carcass Composition and Carcass Weight, Energy, and Protein Gain

The 9th, 10th, and 11th rib cuts were weighed and dissected to obtain the lean, fat, and bone weights to allow the prediction of the carcass proportions of adipose tissue, lean, and bone using the equations proposed by [18]:

Proportion of fat = 3.06 + 0.82X

(1)

Proportion of lean = 15.56 + 0.81X

(2)

Proportion of bone = 4.30 + 0.61X

(3)

where X is the percentage of the respective separable tissue in the rib cut.

The fat and lean tissues from each carcass were ground three times in a 9.5 mm meat grinder (Braher Internacional, S.A., Andoain, Spain) and a homogeneous representative sub-sample was taken.

The proportions of lean body mass (LBM), protein, and water in the carcasses were calculated according to the equations of the Agricultural Research Council [19]:

LBM = Carcass weight − Fat weight in carcass

(4)

Protein = LBM × 0.216

(5)

Water = LBM × 0.729

(6)

The energy content of the carcasses was estimated by the sum of the energy value of protein (23.6 kJ/g) and the energy value of fat (39.3 kJ/g) in the carcass. The carcass weight (CW) gain and carcass energy gain were estimated using linear regression models.

A linear regression model was fitted to estimate carcass energy (CE) content from CW yielding. Carcass protein (CPr) gain was estimated through the slope of CW gain derived from a linear regression. These equations are presented in sub-chapter Section 3.2.

2.5. Analytical Procedures

2.5.1. Feeds

The diets used in the experiment were analysed for dry matter (DM), crude protein, ether extract, crude fibre, ash and minerals as described by Santos-Silva et al. [7]. Starch and sugar were determined according to the method described by Clegg [20]. Neutral detergent fibre (NDF), acid detergent fibre (ADF) and acid detergent lignin (ADL) were determined according to the method [21]. The in vitro digestibility was determined according to [22]. Metabolizable energy (ME) was calculated using the formula by [14]. Direct transesterification of dietary lipids was adapted according to the method from [23] using nonadecanoic acid (19:0) as the internal standard and fatty acid methyl ester (FAME) analysed by gas chromatography.

2.5.2. Muscles, Subcutaneous and Inter-Muscular Adipose Tissue

For chemical analysis and pH determination, 200 g of LL was taken from the 9th, 10th, and 11th rib cuts, stripped of visible adipose and connective tissues, minced in a Moulinex food processor (Group SEB Portugal Lda, Lisbon, Portugal) for 3 × 5 s, then vacuum-packed and stored at −20 °C. The analyses were performed in duplicate. Briefly, dry matter was determined according to [24,25], ether extract content was determined by the Soxhlet method [26], using a Soxhlet extractor (VELP Scientific SER 148/6 Solvent Extractor, Usmate, Italy) and pH at 1:10 in 0.1 M KCl solution according to [27].

Intramuscular lipids of freeze-dried LL samples were extracted according to the method of [28], but where chloroform was replaced by dichloromethane. The extracted intramuscular lipids and intermuscular fat (ImF) samples taken from the 9th, 10th, and 11th rib cuts were transesterified according to the methods detailed in [29]. For FAME identification, commercial standard mixtures (FAME mix containing 37 components, supplied by Supelco Analytical Products, Darmstadt, Germany) and published chromatograms [30,31] were used to compare FAME retention times. Additional FAME identification was performed when necessary, using electron impact mass spectrometry with a Shimadzu GC-MS QP2010 Plus (Shimadzu, Kyoto, Japan).

The colour of subcutaneous fat (ScF), ImF, and LL meat was assessed seven days post-mortem using a CR-400 chroma meter (Konica Minolta Inc., Tokyo, Japan) set up with a 10 mm diameter aperture, a D65 illuminant, and a 2° observer. The chroma meter was calibrated daily using a white standard plate (D65:Y84.9, x:0.3199, y:0.3359). Measurements were recorded in the CIELAB system, where L* is lightness, a* is redness, and b* is yellowness. Colour saturation (C*) and hue angle (H*) were calculated as (a*² + b*²)^1/2 and tan⁻¹(b*/a*) × (180/π), respectively [32].

Meat cooking loss and Warner–Bratzler shear force were determined as described in detail by Santos-Siva et al. [7].

2.6. Sensory Panel

The sensory analysis was carried out by a panel of 9 members from INIAV-Fonte Boa, selected and trained according to [33], who assessed the (1) tenderness, (2) juiciness, (3) flavour, and (4) overall perception of the meat in seven sessions. For each session, six frozen LL samples were randomly selected: three from each diet, with the same ageing time, allowed to thaw for 24 h at 2 °C, and prepared as described by [7]. For the sensory analysis, a defined scale from 1 to 6 was used, with values between: 1—extremely tough, dry, soft, or unacceptable; and 6—extremely tender, juicy, and intense, or acceptable.

2.7. Carbon Footprint Calculation

A partial life cycle assessment (LCA) approach was used to account for additional direct and indirect emissions in addition to the measurements of CH₄ and CO₂. The LCA is a method that considers all direct and indirect potential environmental impacts that are due to a given production system [34]. As it only uses global warming potential as an impact indicator, the LCA approach used here can be considered a CFp metric. In the CFp assessment, the functional units used were 1 kg live weight (LW) gain, 1 kg carcass weight (CW) gain, and 1 kg of CPr gain, all of which occurred during the finishing period only. The functional unit used was cradle to farm gate, but the temporal boundaries of the analysis included only the finishing period of the young bulls. Therefore, the pre-fattening period of the animals, the transport, slaughter process, carcass deboning, and meat processing up to the point of consumption were not included in the LCA calculations. This choice was made because those stages are not affected by the treatments studied here. The CFp was calculated individually for each animal using the online tool developed by Terraprima (https://www.terraprima.pt/en/area-de-actividade/10 (accessed on 5 January 2025)). This tool calculates the contribution to climate change measured in kg CO₂eq using a classification model based on the IPCC 4th Assessment Report [35]. The CFp calculation included emissions from rumen digestive fermentation, manure in the pens, and emissions from the life cycle of feed ingredients production. The GHG emissions from the digestive tract were acquired using the GF measurements described previously. The emissions from excrement during confinement were estimated, individually for each bull, by using equation 10.24 (Tier 2) from [36], refined in 2019, considering specifically the duration of the trial, the gross energy intake, and the digestibility of the diets. For maize silage, the emission factor was calculated using field data for fertilizer applications and operations in the farm where the trial took place, applying emission factors from [35,37]. For both concentrate feeds and straw, the emission factors were calculated using the Ecoalim V1 database from Agribalyse [38]. For brewer’s grains, we considered an average of beer production emission factors present in the CarbonCloud Database [39] of 0.676 kg CO₂eq/L of beer. Then we considered that for each 100 L of beer, 20 kg of brewer’s grains are produced and made an economic allocation, using 2.15 €/L of beer and 0.08 €/kg of brewer’s grains as the economic value, in order to calculate the brewer’s grains final emission factor. The resulting emission factors were 0.037 kg CO₂eq/kg of fresh weight silage, 0.045 kg CO₂eq/kg of fresh weight straw, and 0.025 kg CO₂eq/kg of fresh brewer’s grains. For the concentrate feeds in the Control diet, the calculated emission factor was 0.574 kg CO₂eq/kg and 0.542 kg CO₂eq/kg in the TMR diet.

2.8. Statistical Analysis

Statistical analysis was performed using Proc MIXED (SAS Institute Inc., Cary, NC, USA). The variables were analysed considering the pen as the experimental unit and the diet as the single fixed effect, after checking data for homogeneity of variance. The animals within the pens were treated as subsampling using the statement “repeated/sub = pen(diet) type = cs”. The meat sensory data were analysed with a similar model but including the panellist as a block. Average daily gain was obtained through the slope of the random regression mixed model of the LW at several time points, using the ARH(1) covariance structures for the repeated measurements within animals and the pen(treatment) as a random factor. In all cases, the covariance structure (CS and ARH(1)) was selected according to the model fit statistics.

A linear regression between final live weight (FLW) and CW was fitted (CW = 0.618 kg ± 0.038 × FLW − 38.48 kg, R² = 0.935, RSD = 8.88, n = 20) and used to estimate the carcass weight gain using the initial live weight. Carcass composition and energy content data were computed as explained in Section 2.5.

3. Results

3.1. Feed Intake, Growth Performance, and Carcass Traits

The feed intake, growth performance, carcass traits, and feeding costs during the finishing period are presented in

Table 3. Effects of the diet on growth performance, carcass characteristics, and feed costs of crossbred Angus young bulls.

	Diets		SEM 1	p-Value
	Control	TMR
DM 2 intake (kg/d)	10.6	9.2	0.191	0.035
ME 3 intake (MJ/d)	126	100	0.38	0.001
Initial LW 4	501	499	12.5	0.920
Slaughter LW (kg)	634	610	12.5	0.302
LW gain (kg)	133	111	8.1	0.191
Average daily gain (g/d)	1693	1442	78	0.024
Cold carcass weight (kg)	354	338	9.0	0.345
Dressing percentage (%)	55.7	55.4	0.513	0.757
Feed efficiency	0.16	0.15	0.009	0.717
Feed costs (€/d)	4.67	2.80	0.072	0.003
Feed costs (€/kg LW gain)	2.86	2.05	0.137	0.053

¹—Standard error of the means; ²—Dry matter; ³—Metabolizable energy; ⁴—Live weight.

. The average daily DM intake (DMI) was 13.2% lower in TMR (p = 0.035) compared to the Control group. Consistently, the average daily gain (ADG) was significantly higher in Control, with a value of 1693 g/day, compared to 1442 g/day in the TMR treatment (p = 0.024). Despite that, the difference of 22 kg of LW gain observed between treatments did not reach statistical significance (p = 0.19). No statistical differences were detected for the initial weight (p = 0.92), slaughter weight (p = 0.41), cold carcass weight (p = 0.41), dressing (p = 0.76), and feed efficiency (p = 0.72).

The feed costs per day were approximately 40% lower in TMR than in Control (p = 0.003) and the feed cost per kg LW gain was still 28% lower in TMR than in Control, although not reaching statistical significance (p = 0.054).

3.2. Carcass Grading and Composition

The results of carcass classification according to the SEUROP system showed that in the TMR treatment, 9 out of 10 animals were classified as “R” (Good) and only one carcass was classified as “U” (Very good). In the Control treatment, 50% of the carcasses were classified as ‘R’ and 50% as ‘U’. Regarding the level of fat cover, all the carcasses were classified as class 2 (Slight) in both treatments.

The carcass composition results are presented in

Table 4. Effects of the diet on physical composition of ninth-tenth-eleventh-rib cuts and on total % separable fat, separable lean and separable bone, chemical constituents, and energy content (MJ) in carcasses of crossbred Angus young bulls.

	Diets		SEM	p-Value
	Control	TMR
9th–11th Ribs Dissection 1 (%)
Separable fat	22.3	21.4	2.16	0.795
Separable lean	57.8	57.7	1.75	0.987
Separable bone	17.3	17.8	0.13	0.102
Carcass estimations 2 (%)
Separable fat	21.4	20.6	1.78	0.795
Separable lean	62.4	62.3	1.42	0.987
Separable bone	14.9	15.1	0.08	0.102
Carcass (kg)
Weight	354	338	9.0	0.345
Lean Body Mass	278	268	12.9	0.638
Protein	60.1	57.9	2.79	0.638
Bone	61.2	60.4	1.65	0.776
Water	202	196	9.4	0.638
Fat	75.2	69.6	4.87	0.499
Carcass energy (MJ)	4375	4102	138	0.298
Carcass weight (CW) gain 3 (kg)	82.4	68.0	5.03	0.191
Carcass energy gain 3,4 (MJ)	973	809	59.7	0.191
Carcass protein (CPr) gain 3 (kg)	15.6	12.9	0.95	0.191

¹ Observed values. ² Estimated by the equations generated by [18]; ³ Estimated values of predicted linear regression equations; ⁴ adjusted to a carcass fat percentage of 21%.

. The proportion of separable fat, lean, and bone in the rib cuts was not significantly different between the diets (p > 0.5) and averaged 21.9%, 57.8%, and 17.6%, respectively. This resulted in estimates for the separable fat, lean, and bone in the carcass of 21.0%, 62.4%, and 15.0%, respectively. Applying this estimation to CW and using the mean chemical composition of lean and fat, we estimated that the carcass had an average of 273 kg of lean body mass, 59 kg of protein, 61 kg of bone, 199 kg of water, and 72.4 kg of fat, none differing significantly among treatments.

A linear regression model was fitted to estimate CE content from CW, yielding:

CE = (9.31 ± 2.96) MJ × CW + 1020

(7)

with R² = 0.355, RSD = 437 and n = 20. However, this model showed poor fit due to the large variation in estimated carcass fat content (15.8% to 27.9%). Fitting a linear regression model with both CW and carcass fat (%) yielded a model with a well-fitted (R² = 0.995, RSD = 43 MJ) and a slope of 11.85 MJ per kg of carcass gain adjusted to carcass fat percentage of 21%. Diet and its interactions with regressors were tested and had no effects (p > 0.79) and were removed from the model. The value of 11.85 MJ/kg carcass was used to estimate the energy value of carcass gain. Protein gain was estimated through the slope 0.189 kg/kg of CW gain derived from a linear regression relating CPr to CW:

CPr = (0.189 ± 0.019) kg × CW − 6.43

(8)

with R² = 0.850, RSD = 2.76 and n = 20.

The estimated gross energy content of carcass (4375 MJ in Control and 4102 MJ in TMR), carcass energy gain (973 MJ in Control vs. 809 MJ in TMR), and protein gain (15.6 kg in Control and 12.9 kg in TMR) were numerically higher for the Control than in the TMR treatment, but not significantly different (p = 0.191).

3.3. CH₄ and CO₂ Production

The results for CH₄ and CO₂ production for the control and TMR treatments assessed by GF equipment are reported in

Table 5. Effect of diet on digestive methane (CH₄) and carbon dioxide (CO₂) production and energy loss estimation during finishing period (12 weeks) of crossed Angus young bulls.

	Diets		SEM	p-Value
	Control	TMR
CH4 (g/d)	121	192	7.2	0.020
CH4 (g/kg DMI)	11.5	21.0	0.64	0.009
CH4 (g/kg LW gain)	78	150	4.66	0.008
CH4 (g/kg CW gain)	126	243	7.54	0.008
CH4 (kg/kg CPr gain)	33.3	64.3	1.99	0.008
CH4 Energy losses
CH4 (MJ/d)	6.08	9.65	0.361	0.020
CH4 (% GE Intake)	3.0	5.7	0.20	0.011
CO2 (kg/d)	8.94	8.88	0.05	0.511
CO2 (kg/kg DMI)	0.84	0.97	0.018	0.039
CO2 (kg/kg LW gain)	5.72	7.07	0.498	0.196
CH4/CO2 (g/kg)	13.6	21.7	0.34	0.004

. All 20 animals visited the GF with valid emission measurements for all animals. The animals spent an average of 4 min 30 s in the GF during each visit and visited it between 1 and 7 times a day. The digestive CH₄ emissions were higher (p < 0.05) for TMR than for Control treatments, regardless of the unit as expressed. Daily CH₄ emissions were higher in the TMR group compared to the Control group (192 vs. 121 g/day, p = 0.02). Similar results were observed when CH₄ emissions are expressed per kg of DMI (21.0 g for TMR vs. 11.5 g for Control, p = 0.009) per kg of LW gain (150 g for TMR vs. 79 g for Control, p = 0.008) and per kg of CPr gain (64.3 kg for TMR vs. 33.3 kg for Control, p = 0.008). These daily CH₄ emissions correspond to a loss of 9.65 MJ/d (5.7% of GE intake) for the TMR-fed animals and of 6.08 MJ/d (3.0% of GE intake) for the Control-fed animals.

The CO₂ production values were similar for the two treatments when represented as kg/day, averaging 8.91 kg/d. The same was observed when CO₂ emissions were expressed as kg/kg LW gain (5.72 in Control and 7.07 in TMR, p = 0.196), but when expressed per kg of DMI, they were higher (p = 0.039) in TMR (0.97 kg/kg DMI) when compared to the Control treatment (0.84 kg/kg DMI). The CH₄/CO₂, molar ratio (g/kg) was larger (p = 0.004) in TMR (21.7 g/kg) than for the Control (13.6 g/kg) treatment.

3.4. Meat Quality

The effects of diet on the chemical, physical, and sensory traits of the LL muscle evaluated at 7 days post-mortem are presented in

Table 6. Effects of the diet on chemical composition and on physical and sensorial properties of m. Longissimus lumborum of crossbred Angus young bulls.

	Diets		SEM 1	p-Value
	Control	TMR
Dry matter (%)	27.6	26.7	0.26	0.128
Fat (%)	3.08	2.88	0.186	0.524
pH	5.75	5.73	0.032	0.657
Cooking loss (%)	32.7	31.7	0.42	0.236
Shear force (N/cm2)	57.1	45.2	4.22	0.186
Colour parameters
L*	40.2	42.5	0.36	0.046
a*	23.0	24.7	0.79	0.272
b*	13.6	15.1	0.48	0.165
C*	26.7	28.8	0.90	0.231
H*	30.5	31.4	0.31	0.167
Sensory traits
Tenderness	3.48	4.16	0.107	0.046
Juiciness	3.03	3.45	0.060	0.039
Flavour	2.43	2.49	0.044	0.484
Overall acceptability	3.84	4.27	0.082	0.065

¹—Standard error of the means; L*—lightness; a*—redness; b*—yellowness; C*—chroma; H*—hue.

. There were no significant effects of treatments on muscle DM, intramuscular fat, pH, cooking losses, and shear force (p > 0.05). There were no significant effects of diet on the meat colour parameters (p > 0.05), except for the L* parameter that was higher for TMR than for the Control treatment (42.5 vs. 40.2, p = 0.046). The colour parameters of ScF and ImF did not differ between treatments. The colour parameters averaged L*, 75.5 ± 0.71; a*, 2.46 ± 0.683; b*, 8.61 ± 0.713 for ScF and L*, 80.5 ± 0.47; a*, 0.32 ± 0.224; b* 7.84 ± 0.250 for ImF.

The tasting panel graded the meat from TMR-fed animals as being more tender (4.16 vs. 3.48, p = 0.045) and with higher juiciness (3.45 vs. 3.03, p = 0.039) than that from Control, but with similar flavour (≈2.46, p = 0.484). The overall acceptability of the meat tended (p = 0.065) to be higher for TMR treatment (4.27) than for Control (3.84).

3.5. Fatty Acid Composition of Lean Meat and Adipose Tissue

The FA composition of LL expressed as mg/100 g of meat is presented in

Table 7. Effects of the diet on fatty acid (mg/100 g of meat) of m. Longissimus lumborum (LL) of crossbred Angus young bulls.

	Diet		SEM	p Value
	Control	TMR
Total FA	2273	1899	134	0.188
FA profile
16:0	698	678	24.5	0.617
c9-16:1 1	42.6	30.7	6.08	0.299
18:0	534	522	10.7	0.499
c9-18:1	559	397	64.6	0.219
18:2n − 6	114	58	7.4	0.034
18:3n − 3	5.1	3.6	0.68	0.265
20:4n − 6	17.6	13.2	0.78	0.058
20:5n − 3	1.7	1.0	0.15	0.083
Biohydrogenation intermediates (BHI)
t10–18:1	34.0	8.6	3.50	0.036
t11–18:1	17.8	13.2	1.23	0.118
c9,t11-18:2	5.6	3.5	0.63	0.150
Partial sums
SFA	1327	1272	37.7	0.410
MUFA	745	508	82.3	0.179
cis-MUFA	659	463	76.2	0.211
trans-MUFA	86.4	45.3	6.33	0.044
PUFA	147	83	9.87	0.045
n − 6 PUFA 2	138	76	8.9	0.039
n − 3 PUFA 3	8.9	7.0	0.97	0.313
BHI 4	117	63	9.9	0.061
FA Ratios
n − 6/n − 3 PUFA	18.6	11.5	0.81	0.025
SCD-17 5	22.6	19.1	1.45	0.236
t10-/t11-18:1	1.8	0.6	0.03	0.001

¹ Coelutes with minor amounts of anteiso-17:0; ² n − 6 PUFA = 18:2n − 6 + 18:3n − 6 + 20:2n − 6 + 20:3n − 6 + 20:4n − 6 + 22:4n − 6; ³ n − 3 PUFA = 18:3n − 3 + 20:5n − 3 + 22:5n − 3; ⁴ BHI = Sum of biohydrogenation intermediates presented individually in the table; ⁵ SCD17, Stearoyl-CoA activity index computed as c9-17:1/(c9-17:1 + 17:0) × 100; Values with different superscripts are significantly different (p < 0.05).

and detailed in Table S1 of the Supplementary Material. The total FA concentration was similar in Control and in TMR (2273 vs. 1899 mg FA/100 g meat, p = 0.188). Most of the FA, including the most abundant (i.e., 16:0, 18:0, and c9-18:1), did not present differences between treatments. However, despite the lack of statistical difference observed for c9-18:1, the difference between means is large (559 mg vs. 397 mg), and the variability of values is also large (Coefficient of variation = 78%). The clear differences observed between diets were restricted to the 18:2n − 6 and several biohydrogenation intermediates (BHI). The 18:2n − 6 proportion in meat was 49% higher in the Control than in the TMR treatments. Moreover, the 20:4n − 6 and 20:5n − 3 also tended (p < 0.1) to be higher in the Control than in the TMR treatment. Several BHI were present in higher proportions (i.e., the coeluted peak of t6-18:1, t7-18:1, and t8-18:1, the t10-18:1, t12-18:1, c13-18:1, and c15-18:1; p < 0.05) or tended to be higher (i.e., t9-18:1, c12-18:1, and the coeluted peak of c14-18:1 and t16-18:1; p < 0.1) in the Control than in the TMR treatment.

The individual FA results are translated into the partial sums of FA, where only the trans-MUFA, PUFA, and n − 6 PUFA are higher (p < 0.05), and the BHI tended (p = 0.061) to be higher in the Control than in the TMR treatment. The n − 6/n − 3 PUFA ratio was 38% lower (p = 0.025) in the TMR than in the Control treatment. The t10-18:1/t11-18:1 ratio for Control meat was 200% higher (p = 0.001) than that of the TMR meat. No differences were observed for the SCD-17 (p = 0.236), a proxy of stearoyl-CoA activity.

The FA profile of intermuscular adipose tissue, expressed as mg/g of total FA, is presented in

Table 8. Effects of the diet on fatty acid (FA) (mg/g of total FA) of intermuscular adipose tissue from m. Longissimus lumborum (LL) of crossbred Angus young bulls.

	Diet		SEM	p Value
	Control	TMR
Total FA (mg/g DM)	863	815	29.4	0.366
FA profile
16:0	265	307	1.5	0.002
c9–16:1 1	26.0	26.4	0.42	0.530
18:0	210	224	2.4	0.051
c9-18:1	312	287	2.4	0.017
18:2n − 6	29.4	18.8	0.14	0.001
18:3n − 3	1.4	1.3	0.04	0.137
Biohydrogenation intermediates (BHI)
t10–18:1	28.0	12.1	1.03	0.008
t11–18:1	9.8	11.3	0.46	0.144
c9,t11-18:2	4.3	3.5	0.102	0.029
Other CLA 2	0.6	0.2	0.05	0.030
Partial Sums
SFA	532	589	1.38	0.001
MUFA
cis-MUFA	369	340	2.4	0.013
trans-MUFA	51.5	35.8	0.99	0.008
PUFA	31.3	20.4	0.18	0.001
BHI	62.0	44.0	0.93	0.005
FA ratios
n − 6/n − 3 PUFA	20.1	14.8	0.43	0.009
SCD-17 3	30.5	27.6	0.34	0.027
t10-/t11-18:1	3.16	1.12	0.153	0.011

¹ Coelutes with minor amounts of anteiso-17:0; ² Sum of t10,c12-18:2 + t11,c13-18:2 and other minor t,t conjugated isomers of linoleic acid; ³ SCD-17, Stearoyl-CoA activity index computed as c9-17:1/(c9-17:1 + 17:0) × 100; Values with different superscripts are significantly different (p < 0.05).

and detailed in Table S2 of the Supplementary Material. Most of the FA were affected by treatments. Compared to Control the TMR treatment increased iso-15:0, anteiso-15:0, 16:0, 17:0, and 18:0, and decreased the c9-17:1, c9-18:1, c11-18:1, 18:2n − 6, c9-20:1, 20:2n − 6, and most of BHI, notably the t10-18:1 and c9,t11-/t7,c9-18:2. The 18:3n − 3 and the t11-18:1 did not differ between treatments.

These effects of individual FA are reflected in the higher concentration of SFA (589 vs. 532, p = 0.001) and lower concentration of cis-MUFA (369 vs. 340, p = 0.013) and trans-MUFA (52 vs. 36, p = 0.008), PUFA (31 vs. 20, p = 0.001), and BHI (62 vs. 44, p = 0.005) in the intermuscular adipose tissue of TMR-fed animals. The n − 6/n − 3 PUFA, SCD-17, and t10-18:1/t11-18:1 ratios are higher in the Control than in the TMR treatment.

3.6. Carbon Footprint

The effects of diet on various emission components of the CFp of the finishing period of young bulls are presented in

Table 9. Effects of the diet on emissions components of partial life cycle analysis for the finishing period (12 weeks) of crossbred Angus young bulls.

	Diets		SEM 1	p-Value
	Control	TMR
kg CO2eq/kg LW 2 gain
Digestive tract	1.86	3.60	0.202	0.026
Feeds	3.70	2.00	0.166	0.018
Manure	0.29	0.41	0.022	0.059
Finishing period carbon footprint	5.9	6.0	0.34	0.770
kg CO2eq/kg CW 3 gain
Finishing period carbon footprint	9.5	10.0	0.80	0.716
kg CO2eq/kg CPr 4 gain
Finishing period carbon footprint	50.3	52.8	4.21	0.716

¹—Standard error of the means; ²—Live weight; ³—Carcass weight; ⁴—Carcass protein.

. When the various components of the footprint are expressed in kg of CO₂eq per kg of LW gain, the CFp associated with digestive tract fermentation was 48% higher in animals fed the TMR diet compared to the Control diet (p = 0.026). In contrast, emissions related to feed production were about 46% lower in the TMR diet compared to the Control diet (p = 0.018). For manure management, emissions tended to be higher (p = 0.059) in the TMR than in the Control. Despite these variations, the total CFp during the finishing period was similar between treatments, with values of 5.9 kg CO₂eq/kg LW gain for the Control and 6.0 kg CO₂eq/kg LW gain for the TMR treatment (p = 0.770). When the footprint used a functional unit of kg of carcass gain and kg of protein gain, the results were also the same for both treatments. The average CFp was 9.8 kg CO₂eq/kg CW gain and 51.6 kg CO₂eq/kg of CPr gain, for both treatments.

4. Discussion

The main aim of the trial was to test the hypothesis that replacing the conventional beef finishing diets based heavily on cereal- and oilseeds-derived feedstuffs concentrates with a TMR finishing diet based on maize silage and brewer’s spent wet grains (BSWG) would improve meat quality and reduce costs, without compromising the productive performance and the CFp of beef.

4.1. Productive Performance

The DM intake of the TMR diet was lower than that of the Control, despite the much larger intake as a fresh diet. The lower rumen digestibility and passage rate of forage-based diets, together with their bulkiness, might restrict the DM voluntary intake due to the intake regulatory mechanism linked to rumen fill [40]. Avilés et al. [10] and Cooke et al. [41] also observed higher DMI in Limousine and Retinta young bulls and in Charolais crossbred heifers, respectively, fed a conventional concentrate diet than those fed a maize silage-based TMR. On the other hand, the TMR diet contained 18.9% of DM as BSWG, which are difficult to preserve due to their high fermentation rates and spoilage within a few days of production, developing flavours that might reduce the palatability [42]. Thus, we cannot exclude the possibility that the inclusion of the BSWG in the TMR diet had a negative effect on the intake, health, and growth of animals [43].

The ADG was 16% lower in the TMR compared to the Control treatment, in full accordance with lower DMI and ME intake and similar feed efficiency. The ADG presented here are in line with those obtained in a similar trial with Limousine and Retinta young bulls that also compared a concentrate diet with a TMR diet [10]. The slaughter weight, cold carcass weight, LW gain, and carcass gain all follow the same consistent and expected trend of being lower for TMR than for Control, although the power of the experiment was not enough to validate these effects. The ADG was obtained through a random regression method, which is more powerful than the simple mixed model applied to the other variables. We did not find differences in the dressing percentage, which is surprising considering the bulkiness of the TMR diet compared to the Control. In fact, increasing both the DMI and forage proportion in diets is expected to increase the weight of the digestive tract and thus decrease the dressing percentage [7,44]. Probably the higher DMI in animals fed Control compensated for the higher bulkiness of the TMR diet, resulting in a similar dressing percentage.

Despite the decrease in DMI, the inclusion of maize silage in beef cattle diets can be desirable, reducing the amount of concentrate, reducing feed costs, and preventing losses due to rumen acidosis [45]. Among forages, maize silage is notably rich in starch and has a high quality of its fibrous fraction [46]. In fact, the feed costs and feed cost per gain were reduced in our experiment. In Portugal, the compound feed industry is heavily dependent on imported cereals and oilseed-derived feedstuffs and thus on their international price fluctuations [7]. Partial replacement of concentrate by on-farm or locally produced maize silage warrants an improved predictability of feed costs. Our data are consistent with previous studies of our team [7,47] that found that the costs of feeding diets based on forages and agro-industrial by-products that replace cereals are competitive with commercial concentrates. Nevertheless, our results only relate to the direct costs of feed ingredients, excluding costs associated with feed management such as labour, energy, and equipment.

4.2. Carcass Composition

We estimated the carcass muscle, bone, and fat composition using a classic approach that uses the dissection of the 9th, 10th, and 11th rib sections [18]. Despite genetic and management development of beef production systems, the method continues to be used, and the lean and fat values obtained from the dissection of the 9th, 10th, and 11th rib sections are still strongly correlated with the overall carcass composition, particularly for fat and lean content, with correlation coefficients of 0.87 and 0.80, respectively [48]. This approach allows us to estimate the carcass composition and hence the carcass weight, protein, and energy gain. Naturally, estimations need to be interpreted with caution. In fact, previous studies have suggested that breed differences could explain the differences in the percentage of lean meat, fat, and bone and stated that it was necessary to make small adjustments according to the breed analysed [49,50]. In the present experiment, crossbred Angus × Limousin or Charolais young bulls were used, and in fact, the proportions of lean and fat tissue observed were higher than those reported in dissections of Holstein steers [51]. Duckett et al. [52] reported that crossbred Angus steers fed on pastures presented carcasses with higher lean percentage and less fat than those fed concentrate diets. In our experiment, we did not detect differences between TMR and Control. The variability of estimated carcass fat was large and independent of the treatments, with values ranging from 15.8% to 27.9%. As no effect of treatments on carcass composition was detected, the estimates of CW, energy, and protein gain follow closely that of LW gain.

4.3. Methane Emissions

Individual CH₄ production was measured using a GF unit. This involves the animals voluntarily visiting the equipment during the period when it is available. This variation in the number of records obtained from each animal throughout the trial is due to differences in behaviour towards the equipment and the dominance of some animals over others. In average daily CH₄ production, a high individual variation was observed, suggesting a genetic variability between individuals [53], with average values ranging between 89 and 157 g/d for the Control and between 153 and 234 g/day for the TMR diet. This may be related to differences in rumen microbiota, particularly the abundance of the Methanobrevibacter genus, resulting in variations in the expression of methyl-coenzyme M reductase, which catalyses methanogenesis [54]. In fact, several studies have demonstrated correlations between host genotype and CH₄ emissions, which may explain part of the variation observed between individuals [54]. Another possible explanation is the individual feed passage rate, a biological factor that appears to be inversely proportional to individual CH₄ production [55]. The values obtained in the present experiment were consistent with previous studies on beef cattle that obtained CH₄ emissions of 22 g/kg DMI [56] and 20 g/kg DMI [57] for TMR diets, and of 11 g/kg DMI [7] for concentrate diets.

The emission of CH₄ by animals represents a net loss of feed energy, meaning it cannot be utilised by ruminants for productive purposes [58]. Our data indicate that CH₄ energy losses are 3% of GE intake for Control and 5.7% of GE intake for TMR treatment. These results are consistent with the values reported by Johnson et al. [59]: CH₄ energy losses of around 5.8 to 6.5% of GE intake for all categories and classes of cattle, except for the single situation of a feedlot with high concentrate content, where the typical CH₄ loss can drop to approximately 3%. Thus, as expected, we confirm that a maize silage-based TMR diet, due to its larger contribution of fibre, results in more CH₄ emission and energy losses than a concentrate-based diet.

The GF equipment also provides estimates of animals’ CO₂ emissions, despite having no participation in computing the GHG of animals, as the C involved is part of biogenic C recycling, it reflects the oxidation of organic matter and thus the energy yield of the animal [58]. The evaluation of the energy balance by indirect calorimetry also requires information on O₂ consumption, N balance, and retained energy [60]. Nevertheless, the use of the production of CO₂ to estimate energy metabolism and CH₄ in cattle has been explored [61]. The CO₂ produced averaged did not differ between diets, but due to reduced DMI, animals fed TMR produced more CO₂ yield (kg/kg DMI) than those fed the Control diet. Higher CO₂ yields in TMR treatment might be due to lower dietary fat content [62] and putative decreased retained energy. The ratio CH₄:CO₂ (g/kg) is almost double in TMR treatment than in Control, confirming the usefulness of the CH₄:CO₂ ratio as a proxy of relative CH₄ as proposed in the Sniffer methodological approach [62].

4.4. Meat Quality

The quality of meat produced is another important factor that should be taken into consideration when evaluating both feeding systems. We could not detect relevant differences in the composition of meat and its physical quality traits, probably because of a lack of statistical power, but the sensorial analysis clearly pointed to a better eating quality of meat from animals fed TMR compared to those fed Control. In fact, the numerically lower shear force of TMR meat compared to Control is very consistent with the significantly higher score for tenderness given by the panellists. Meat colour is also an important meat quality trait, and consumers value the bright red colour as an indicator of freshness and healthiness [63]. The L* was higher in TMR meat, which indicates slightly higher lightness, and it has also been recognised that for a given b*, the colour acceptability increased linearly with increasing L* [64]. This suggests that the colour of TMR beef would be slightly more acceptable to the consumers. Diet had no significant effect on adipose tissue colour, which was white in both diets, contrary to what has been observed in several studies, which state that the fat of animals fed concentrates is whiter and that of animals fed TMR is more yellowish [41,65,66].

In addition to tenderness, the panellists also score the TMR meat better for juiciness and overall acceptability traits. The higher juiciness for beef from animals fed TMR diets compared to concentrate-based diets was also reported by [10,67].

The health benefits of n − 3 long-chain PUFA are widely recognized and there are official recommendations for the minimal daily intake for adults based on cardiovascular disease risk [68]. Lean meat from forage-fed animals can be a relevant dietary source of n − 3 long-chain PUFA, particularly where the consumption of marine fish and seafood is low [69,70]. Due to the extensive biohydrogenation in the rumen, ruminant meat is also a source of rumenic acid (c9,t11-18:2), the major conjugated linoleic acid (CLA) isomer, and of its main precursor vaccenic acid (t11-18:1). Both biohydrogenation intermediates have numerous health-promoting effects, including anti-carcinogenic properties, as reviewed by Vahmani et al. [70]. One of our hypotheses was that the FA composition of meat tissues would be improved with the incorporation of forage in the diet, particularly that of the n − 3 PUFA, t11-18:1, and c9,t11-18:2. In fact, increasing forages in ruminant diets often led to an increase in n − 3 PUFA on meat, as we observed in a previous experiment when we compared a concentrated-based finishing diet with a biodiverse haylage-based TMR diet [7]. Moreover, the increased incorporation of forage in ruminants tends to prevent the establishment of trans-10 shifted rumen biohydrogenation pathways (hereafter trans-10 shift) and consequent production of the undesirable t10-18:1 isomer instead of the healthier t11-18:1 and c9,t11-18:2 [69,71]. In the present experiment, the FA composition of lean meat was very similar between treatments and the expected increase in n − 3 PUFA in meat lipids was not observed, although the meat from TMR-fed animals presented much lower content of 18:2n − 6 and of several BHI, particularly of t10-18:1, than in the Control treatment. In the present experiment, the forage used in the TMR was maize silage. When compared to most fresh and preserved forages, maize silage is notable for presenting a low 18:3n − 3 and high 18:2n − 6 content [72]. Thus, the concentration of 18:3n − 3 in the TMR diet compared to Control (38.5 vs. 15.3 g/kg FA) was mild compared to what was found in [7] (152 g/kg FA in TMR vs. 23.3 g/kg in Control). This, together with the lower fat content of the TMR diet and its lower DMI, resulted in approximately similar 18:3n − 3 intake (14 g/d in TMR vs. 11 g/d in Control), explaining the absence of response on n − 3 PUFA in meat.

The FA profile of intermuscular adipose tissue is more informative regarding the dietary effects on lipid metabolism of animals. Although both diets contained high amounts of starch and 18:2n − 6, the Control diet contained more starch (396 vs. 312 g/kg DM) and fat (75 vs. 51 g/kg DM) and less NDF (321 vs. 491 g/kg DM) than the TMR diet. The high-fat intake in the Control treatment is consistent with augmented BHI and reduced de novo synthesis of 16:0 [73]. In ruminants fed pasture/forage-based low-starch diets, commonly, the t11-18:1 is the overwhelming BHI formed and thus the t10-18:1/t11-18:1 ratio is well below 1 [69,71]. However, both high dietary content of starch and 18:2n − 6 favour the establishment of the trans-10 shift in the rumen [71]. In fact, the t10-18:1/t11-18:1 ratio in adipose tissue was close to 1.1 for TMR and 3.2 for Control. The trans-10 shift was much more exacerbated in the Control than in TMR, due to a large increase in t10-18:1. Despite the t11-18:1 remaining similar in both diets, the c9,t11-18:2 was higher in the Control. The c9,t11-18:2 present in the tissues is mainly derived from the endogenous desaturation of t11-18:1, which originates in the rumen, by the stearoyl-CoA desaturase (SCD) [74]. Concentrate-fed ruminants showed higher SCD activity in ImF, increasing the conversion of t11-18:1 into c9,t11-18:2 [69]. This probably explains the increased c9,t11-18:2, as well as other SCD products like c9-18:1, in Control and is also consistent with augmented SCD-17, a proxy of SCD activity [69]. Moreover, the contribution of conjugated isomer t7,c9-18:2 to the reported increased c9,t11-18:2 cannot be excluded, as in our chromatographic conditions, both isomers coelute. The t7,c9-18:2 is also formed by SCD through the desaturation of t7-18:1 and is commonly increased in ruminants with concentrate feeding [75].

Overall, the FA profile of both lean and fat meat tissues of TMR-fed animals was only slightly improved, through the mitigation of the t10-shift, lower trans-MUFA, and lower n − 3/n − 6 PUFA ratio. Despite that, the PUFA content was reduced, mostly due to diminished 18:2n − 6 content as the supply of n − 3 PUFA remained quite low in both diets, below 10 mg/100 g of lean meat.

4.5. Carbon Footprint

The evaluation of GHG emissions of a feeding system should not be based only on direct digestive tract emissions from animals, and other GHG sources must be considered using an LCA perspective. To address this, we calculated the overall CFp over the 12-week test period only, as the contribution of the other production and consumption stages is common to both treatments. Most of the emission sources were calculated using emission factors, except for rumen emissions, which were measured. It is important to note that the assessment only involved young bulls. It did not consider any pre-test emissions, such as those from the breeding herd and calves prior to reaching the finishing stage, nor those from the dam. Moreover, as this earlier cow-calf stage has a large contribution in the whole production cycle (accounting for about 80% of total emissions) [12,76,77], there would be a dilution of the finishing stage and low sensitivity of the analysis to detect the effect under study. The CFp results were expressed using the LW gain, CW gain, and CPr gain as functional units. The LCA results show that the large differences observed for digestive CH₄ emissions between treatments were compensated for by the also large differences of GHG emissions attributed to the feeds and to a lesser extent of manure produced. Thus, a similar CFp was obtained for both treatments (≈6.0 kg CO₂eq/kg LW gain, ≈9.8 kg CO₂eq/kg CW gain, and ≈51.6 kg CO₂eq/kg CPr gain). These confirm the previous results obtained by our team, where Limousine × Alentejana crossbred young bulls were fed for 64 days either a concentrate-based diet or a biodiverse forage-based TMR, and the CFp was 6.6 kg CO₂eq/kg LW gain for both treatments [7]. Berton et al. [78] reported a higher value of 8.3 kg CO₂eq/kg LW gain for bulls of French breeds and crosses. In addition to the differences in animal genetics and feeding management, these authors used published databases to estimate the rumen emissions, whereas we used individual data on rumen digestive emissions and intake, as well as detailed feeding information.

The CFp results clearly demonstrate the trade-off between concentrate-based diets and forage-based diets in beef finishing, as the GHG contribution of the compound feed production chain can be large enough to compensate for the larger digestive tract CH₄ yield of forage-based diets.

5. Conclusions

The crossbred young bulls finished with a TMR diet presented less 13% DMI and 15% of ADG and produced 59% more daily CH₄ compared to those fed the Control diet. However, when assessed through a partial LCA, these differences were fully offset by the lower feed costs and lower GHG emissions associated with the lower compound feed incorporation in diets. The carcass composition remained unaltered, but the eating quality of the meat from bulls finished with a TMR diet improved, showing better tenderness, juiciness, and overall acceptability scores. The FA composition of lean meat was fairly similar between diets, but with less t10-18:1 and hence total trans-MUFA, which is an improvement regarding the consumer’s health.

These results point out that replacing conventional concentrate-based diets with maize silage-based TMR diets could improve the sustainability of beef cattle finishing systems commonly used in south-western Europe, by decreasing costs and reducing reliance on imported grains and off-farm inputs, while maintaining CFp and improving beef quality. This offers a practical pathway for aligning intensive beef production with regional and EU sustainability goals.