All experimental procedures involving animals were conducted in accordance with the Mexican Official Standard NOM-062-ZOO-1999 for the care and use of experimental animals [11] and complied with the ARRIVE 2.0 guidelines for reporting animal research [12]. The experimental protocol was reviewed and approved by the Institutional Ethics and Research Committee of the National Institute of Forestry, Agricultural, and Livestock Research (INIFAP) under Approval No. JAG.500-00908. Throughout the experimental period, animal health, behavior, and welfare conditions were monitored daily by trained personnel to minimize stress and ensure appropriate management practices. All handling, housing, feeding, transport, and slaughter procedures were performed in accordance with national animal welfare regulations and internationally accepted ethical standards for livestock experimentation.

The study was conducted from January to March 2019 at the Mocochá Experimental Field of the INIFAP, located in Yucatán, México (21°09′ N, 89°43′ W; 8 m above sea level). The regional climate is classified as tropical subhumid (Aw) according to the Köppen classification modified by García [13]. The area has a mean annual temperature of 26.5°C and an average annual precipitation of approximately 900 mm, with the rainy season from June to November.

The experiment consisted of two complementary trials designed to evaluate the effects of tropical forage inclusion in concentrate-based diets for hair sheep under intensive production conditions. The first trial evaluated apparent in vivo digestibility, whereas the second trial assessed growth performance, carcass traits, and meat quality characteristics. In both experiments, animals were assigned to four dietary treatments in a completely randomized design. The treatments consisted of T1 = 100% CF; T2 = 70% CF + 30% Maralfalfa grass (Pennisetum sp.); T3 = 70% CF + 30% Mombaza grass (Megathyrsus maximus); and T4 = 70% CF + 30% CT-115 grass (Pennisetum purpureum) on a dry matter (DM) basis.

Apparent in vivo digestibility was determined as described by Garry et al. [14]. Sixteen adult intact Katahdin × Pelibuey rams (27.2 ± 4.3 kg body weight) were individually housed in metabolic cages (1.0 × 0.90 × 0.50 m) equipped with separate feed and water systems and independent feces collection facilities. Animals were randomly assigned to the four dietary treatments, with four replicates per treatment and one animal per cage.

Apparent in vivo digestibility (%) was calculated using the following equation:

[(nutrient intake − nutrient in feces)/nutrient intake]×100

Experimental diets were formulated on a DM basis and subsequently adjusted to an as-fed basis before feeding (Tables 1 and 2). CF was offered once daily during the morning, whereas freshly harvested forage was supplied at midday. Maralfalfa, Mombaza, and CT-115 grasses were harvested at their respective regrowth stages of 80, 60, and 80 days, chopped to approximately 4–5 cm, and offered fresh to the animals.

Before the experiment began, sheep were dewormed with ivermectin (Ivermectin®, Laboratorios Sanfer, México City, México) and adapted to the diets and housing conditions for 14 days prior to data collection.

The digestibility trial lasted 7 days, during which daily feed intake and fecal output were recorded. Composite samples consisting of 10% of the daily feces, feed offered, and feed refusals were collected, pooled, dried at 65°C for 48 h, and stored at −20°C until analysis.

DM (AOAC 950.46), crude protein (CP; AOAC 981.10; N × 6.25), and ash (AOAC 920.153) were determined using standard AOAC methods [15]. Neutral detergent fiber (NDF) and acid detergent fiber (ADF) were analyzed following the procedures described by Van Soest et al. [16].

Thirty-six intact Katahdin × Pelibuey male lambs (18.0 ± 3.3 kg initial body weight) were stratified according to body weight and assigned to the four dietary treatments (n = 9 per treatment) in a completely randomized design. Diets were formulated according to NRC [17] recommendations for growing lambs to achieve an average daily gain (ADG) of 200–250 g/day.

Animals were individually housed in shaded pens (3.0 × 1.6 m) with ad libitum access to feed and fresh water. Following deworming and a 14-day adaptation period, lambs were weighed after a 16 h fasting period at the beginning of the experiment and subsequently every 14 days throughout the 84-day feeding period. Final body weight was recorded at the end of the experiment, and ADG was calculated as the difference between final and initial body weight divided by the number of experimental days.

At the end of the feeding trial, lambs were transported to a commercial slaughter facility, fasted for 16 h, and slaughtered in accordance with NOM-033-SAG/ZOO regulations for humane slaughter [18]. Slaughter live weight (SLW) and hot carcass weight (HCW) were recorded to calculate carcass yield (CY) using the following equation:

CY (%) = (HCW/SLW) × 100

After chilling for 24 h at 4°C, carcasses were sectioned between the 12th and 13th ribs to measure backfat thickness (BF) using a stainless-steel ruler, as described by Elizalde et al. [19]. The Longissimus dorsi muscle (LDM) area was determined using a transparent grid following the Ohio State University protocol [20].

After 24 h of chilling, the LDM was exposed between the 12^th and 13^th ribs. Muscle pH was measured using a portable pH meter (HI99163 Meat pH Meter, Hanna Instruments, Woonsocket, RI, USA) previously calibrated with standard buffer solutions at pH 4.0 and 7.0. Measurements were obtained by inserting the probe approximately 2 cm into the geometric center of the muscle.

Immediately after pH determination, muscle temperature was measured at the same location using a digital thermometer (Taylor Precision Products Inc., Oak Brook, IL, USA).

The LDM was subsequently excised, and 2.5-cm-thick steaks were collected for the determination of drip loss (DL) at 24 and 48 h postmortem, as described by Lunesu et al. [21]. Three replicate samples (100 g each) from each carcass were placed in polyethylene bags, suspended in a refrigerated chamber at 4°C, and reweighed after each storage period. DL was expressed as the percentage of weight loss during storage. Outliers exceeding ±1.5 interquartile ranges were identified and excluded before statistical analysis.

DL (%) was calculated as:

where W i represents the initial sample weight (g), and W f represents the final sample weight after drainage (g).

For proximate analysis, 150 g samples of LDM were vacuum-packed, stored at −20°C, thawed at 4°C, trimmed of visible fat, minced, and homogenized for 15 s. Moisture (AOAC 950.46), ash (AOAC 920.153), CP by the Kjeldahl method (AOAC 981.10; N × 6.25), and intramuscular fat by Soxhlet extraction (AOAC 991.36) were analyzed according to AOAC methods [15].

All analyses were performed in technical duplicate using a Kjeldahl system (Kjeltec™ 8200, FOSS, Hillerød, Denmark) and a Soxhlet extraction system (Thermo Scientific™ SE400, Thermo Fisher Scientific, Waltham, MA, USA).

Data obtained for apparent in vivo digestibility, growth performance, carcass traits, chemical composition, pH, and temperature were analyzed using the following general linear model:

where Yijrepresents the response variable of the j-th animal receiving the i-th dietary treatment, μ is the overall mean, 𝜏i is the fixed effect of dietary treatment, and 𝜀ij is the residual error.

Before analysis, data normality was evaluated using the Shapiro–Wilk test, whereas homogeneity of variances was assessed using Levene’s test. When significant treatment effects were detected, means were separated using Tukey’s honestly significant difference test to control the overall Type I error rate.

Statistical analyses were performed using SAS software version 9.1 (SAS Institute Inc., Cary, NC, USA) [22]. Results are presented as mean ± standard error. Statistical significance was declared at p < 0.05.

A post hoc power analysis (1 − β = 0.80; α = 0.05) indicated that the experimental design was sufficient to detect minimum differences of approximately 15% for digestibility-related variables. Experimental data were recorded daily using standardized forms and subsequently entered into electronic spreadsheets. Data entries were independently verified, and no missing values were identified.

DL data were analyzed using a mixed-effects model to account for repeated measurements within carcasses. The statistical model used was:

where Yijk represents the DL value of the k-th replicate from the j-th animal receiving the i-th dietary treatment, μ is the overall mean, 𝜏i is the fixed effect of diet, aj(i) is the random effect of animal nested within treatment, and 𝜀ijk is the residual error.

Time points (24 and 48 h postmortem) were considered repeated measures within animals, and the covariance structure was selected according to model fit criteria. Analyses were performed using PROC MIXED in SAS software version 9.1 [22]. Least squares means were compared using Tukey’s test, and significance was declared at p < 0.05.

Dietary effects on feed intake and nutrient digestibility are presented in Table 3. Significant differences in concentrate intake were observed among treatments (p < 0.05). Lambs fed the concentrate-only diet (CF) showed the highest concentrate intake, which did not differ significantly from CF + Maralfalfa or CF + CT-115, but was significantly greater than CF + Mombaza. Regarding forage intake, lambs receiving CF + Mombaza exhibited the highest forage consumption (p < 0.01), followed by CF + Maralfalfa and CF + CT-115.

Total DMI was significantly greater (p < 0.01) in all forage-supplemented diets compared with the CF diet, indicating that forage inclusion increased overall nutrient intake without reducing total feed consumption. Significant differences (p < 0.01) were also observed for all digestibility variables. Lambs fed the CF diet showed the greatest digestibility coefficients for DM, CP, ash, NDF, and ADF. Diets containing CF + Maralfalfa or CF + Mombaza presented intermediate digestibility values. Acid detergent fiber digestibility in the CF + Maralfalfa treatment was statistically similar to that of CF, whereas CF + Mombaza exhibited lower ADF digestibility than CF (p < 0.01). The lowest digestibility coefficients were consistently observed in lambs fed CF + CT-115, particularly for ADF digestibility (p < 0.01), highlighting the influence of forage source on nutrient utilization.

Growth performance and carcass characteristics are presented in Table 4. No significant differences were detected among treatments for growth performance or carcass variables (p > 0.05). ADG ranged from 0.240 to 0.265 kg/day, whereas SLW varied from 43.50 to 44.87 kg. Likewise, HCW ranged from 21.09 to 21.82 kg, and CY varied from 48.32% to 48.61%.

Backfat thickness values were similar among treatments (p > 0.05), with measurements close to 1.1 mm in CF, CF + Maralfalfa, and CF + Mombaza treatments, whereas slightly lower values were observed in the CF + CT-115 group. The largest LDM area was recorded in lambs fed CF + Mombaza, whereas the smallest area was observed in the CF + Maralfalfa treatment; however, these differences were not statistically significant (p > 0.05).

Meat physical traits did not differ significantly among treatments (p ≥ 0.05); therefore, descriptive statistics were used for reporting (Table 5). Median ultimate pH values at 24 h postmortem ranged from 5.5 to 5.8 and were comparable across diets, with lower variability in CF and CF + Mombaza and slightly greater dispersion in CF + Maralfalfa and CF + CT-115 (Figure 1A). Meat temperature during chilling ranged from 5.0°C to 6.2°C, with CF + CT-115 showing the lowest median and CF + Mombaza the highest median (Figure 1B).

DL medians at 24 h ranged from 0.85% to 1.10%. The CF + Mombaza group showed the widest variability, ranging from 0.5% to 1.5%, whereas CF displayed the narrowest interquartile distribution, despite the presence of a low outlier. CF + Maralfalfa and CF + CT-115 exhibited intermediate ranges, with slightly higher medians than CF and CF + Mombaza (Figure 2A).

At 48 h, DL medians increased to 1.2%–1.8%. CF showed the broadest variability, ranging from 0.6% to 2.6%, whereas CF + Mombaza had the narrowest range, from 1.0% to 2.0%. Similar to the 24 h results, CF + Maralfalfa and CF + CT-115 presented intermediate variability, with comparable medians across treatments (Figure 2B). Although the range of values was wide, no statistically significant differences were detected among treatments (p > 0.05).

The chemical composition of lamb meat is presented in Table 6. Moisture content was significantly greater (p < 0.05) in lambs fed CF + Mombaza compared with those fed the CF diet, whereas intermediate values were observed in CF + Maralfalfa and CF + CT-115 treatments.

Crude protein and ash contents did not differ significantly among treatments (p > 0.05), with mean values of approximately 20.52% and 1.02% on a DM basis, respectively. Total fat content was highest in lambs fed the CF diet and lowest in those receiving CF + Maralfalfa (p < 0.05). Intermediate fat values were observed in CF + Mombaza and CF + CT-115 treatments.

The inclusion of both CF and forage in sheep diets has been widely reported to enhance ruminal function and improve nutrient utilization, often resulting in greater DMI. This trend was partially observed in the present study, where forage-based diets stimulated feed intake. Similar findings have been reported by Obeidat et al. [23] and Liu et al. [24], who demonstrated that varying forage proportions in total mixed rations stabilize rumen fermentation and support higher intake in lambs. Mohammed et al. [25] also reported comparable DMI in lambs fed Mombaza hay versus other forages. Conversely, reduced DMI has been reported when less digestible or more fibrous forages, such as Maralfalfa, are used [26]. This finding was recently supported by de Azevedo et al. [27], who showed that forage intake increases as forage availability and quality improve, underscoring the direct role of these factors in determining consumption in small ruminants. Collectively, these findings indicate that DMI responses are shaped by forage characteristics, the animal’s physiological status, and the overall dietary profile.

Nutrient digestibility in the present experiment ranged from 76% to 80%, values considered high for forage-based diets and likely reflecting CF’s contribution to the ration. The lower ADF digestibility observed in lambs fed CT-115 compared with those fed Maralfalfa may be explained by the higher NDF and ADF contents of CT-115, which indicate a greater proportion of structural carbohydrates and potential lignification. These factors are known to limit ruminal fiber degradation and reduce apparent digestibility. However, the increased feed intake observed in forage-based diets may partially offset lower digestibility, resulting in sufficient nutrient availability to sustain growth despite potential reductions in feed efficiency.

These findings are consistent with Muhammad et al. [28], who reported improved digestibility with increasing concentrate inclusion, but contrast with the lower digestibility values of 51%–69% reported by Partida et al. [26] and Ibrahim et al. [29]. Overall, the results support the general agreement that increasing the proportion of forage, particularly with more fibrous forages, reduces DM and organic matter digestibility [30].

The ADG observed in this study was slightly higher than previously reported for Pelibuey lambs. Previous studies reported ADG values of 195–205 g/day in animals fed diets containing agricultural by-products, crop residues, or fresh Mombaza grass, supplemented with concentrate [31, 32]. The improved performance observed in the present study may also be attributed to genotype effects, as Katahdin crosses are recognized for greater growth potential than Pelibuey and Blackbelly lambs [33, 34]. The lower ADG reported by Mohammed et al. [25] in older sheep further supports the role of age and physiological status in modulating growth performance.

No significant differences were observed in SLW, HCW, CY, or BF thickness among treatments. These findings support the concept that carcass quality is more closely related to slaughter weight than to diet composition or forage type when nutrient supply is adequate. Wang et al. [35] reported minimal effects of the feeding system on lamb carcass parameters. Guim et al. [36] also observed negligible effects of different roughage-to-concentrate ratios on carcass traits in Santa Inês lambs. The similar slaughter weights across treatments in the present study likely explain the absence of significant differences in carcass characteristics.

Meat quality parameters were within normal physiological ranges. Ultimate pH values of 5.5–5.8 confirmed adequate postmortem glycolysis and the absence of pale, soft, exudative meat or dark, firm, dry meat defects. This finding is consistent with Holman et al. [37] and Ge et al. [38], who reported similar pH declines in lamb longissimus muscles during rigor mortis. Meat temperature during chilling showed some variation among carcasses but did not differ among dietary treatments. This finding agrees with Muela et al. [39], who demonstrated that conventional chilling effectively maintains appropriate temperature and pH dynamics regardless of feeding regimen.

DL at 24 and 48 h postmortem was not significantly affected by diet (p > 0.05). At 24 h, CF-fed lambs showed the least variability, whereas at 48 h they displayed a wider range of values, reflecting normal temporal changes in muscle structure. The more uniform DL at 24 h may be related to consistent early pH decline, while greater variation at 48 h corresponds to progressive alterations in myofibrillar proteins and weakening of the extracellular network during storage [40, 41]. In contrast, CF + Mombaza consistently showed narrower ranges at both time points, suggesting a more stable water-holding capacity. These observations agree with Orzuna et al. [42], who reported similar variability patterns at 24 h. Liu et al. [43] also demonstrated that diet composition and forage inclusion can modulate the dynamics of water-holding capacity during postmortem aging. Overall, these findings indicate that although dietary treatments did not significantly alter DL, forage inclusion, particularly Mombaza, may contribute to greater stability of water-holding capacity over time.

It should be noted that ruminal fermentation parameters and methane emissions were not measured in this study. Therefore, the observed effects of forage inclusion on intake, digestibility, and meat quality should be interpreted based on productive and carcass responses rather than direct rumen metabolic indicators.

Chemical composition analysis showed higher intramuscular fat and lower moisture in CF-fed lambs, consistent with the greater energy density of cereal- and oilseed-based diets that promote lipid deposition at the expense of water content [44, 45]. In contrast, forage-based treatments produced leaner carcasses with higher moisture, likely due to their higher fiber and lower energy content, which favor protein accretion and greater water retention [46]. These results emphasize the influence of diet composition on meat characteristics, with practical implications for eating quality. Higher intramuscular fat is associated with improved juiciness and flavor [47], whereas leaner meat with greater moisture content aligns with consumer preferences for healthier, lower-fat products [48].

Although productive performance, carcass traits, and meat quality were evaluated, rumen fermentation parameters and methane emissions were not measured in the present study and should be addressed in future research.