A total of 91 Creole male kids were used in this experiment in a completely randomized design with a complete 2 × 3 experimental design (feeding regimen × age at slaughter). The kids were reared at pasture from birth until weaning with their mother in a rotational grazing system. The pasture was divided into five plots grazed by 50–52 suckling goats with their progeny until weaning. Each plot was grazed for 1 week and rested for 4 weeks.

After weaning, at 84 days old ± 7 days and 9.2 kg live weight (LW) ±0.5 kg, kids were allocated by birth weight, weaning weight, pre-weaning average daily gain (ADG), and the LW of the doe, and randomly assigned to experimental groups. Animals were reared indoors in individual pens on a slatted floor. In group C0 (no concentrate, n = 45), kids received no commercial concentrate and in group C50, they received a commercial pellet in addition to their basal forage diet (50% of the diet, n = 46).

Ages of the animals at slaughter were in 4-month increments from weaning at 3 months old according to a 4-month interval of either kidding or weaning of the experimental flock: ages at slaughter were 7 months (i.e., age at weaning plus 4 months; n = 15 C0 and n = 15 C50), 11 months (n = 16 C0 and n = 17 C50), and 15 months (n = 14 C0 and 14 C50).

The basal diet was Digitaria decubens allocated ad libitum and cut daily to achieve a 28-day regrowth: 0.74 UFL (feed unit for lactation) and 79 g PDIN (ruminally degradable nitrogen) per kg DM (dry matter). The pasture was divided into 29 plots and the harvest was planned so that the whole experimental period would fall within a D. decubens regrowth age of 28 days. One plot was harvested per day. The pasture was irrigated and fertilized with 500 kg/ha/year of commercial fertilizer (30% N, 12% P₂O₅, and 18% K₂O). The commercial concentrate (1.15 UFL and 150 g PDIN per kg DM) was composed of maize (68%), soybean cake (15%), wheat bran (11%), vitamin and mineral supplement (5%), and urea (1%).

The animals had free access to fresh tap water and mineral blocks composed of sodium (32.5%), calcium (2.5%), and magnesium (2.5%). The animals were allowed to adapt to their diets and pens for 14 days before starting the measurements. The animals received the commercial concentrate first at 0600 h and then the forage was distributed in two meals: at 0800 and 1400 h. The amount of forage offered was 1.15 times higher than the voluntary intake measured the days before. Voluntary feed intake was measured from Monday to Friday during the experiment. Offered and refused forage were recorded weekly for each animal to measure voluntary dry matter intake. Samples of offered forage (two sub-samples of 200 g) and refusals (10% for each animal) were collected daily. One of the sub-samples was kept for daily dry matter determination. All the samples of the feed provided during the 2 weeks were mixed together and a new sub-sample (200 g) was used for chemical analyses. The same operation was carried out for feed refusals.

Animals were weighed the day before slaughter, and the next day while fasting just prior to slaughter. The animals were humanely slaughtered with a captive bolt pistol which hit them on the head producing immediate unconsciousness following by exsanguination. Thereafter, the digestive tract was removed, weighed in full, separated by compartment, emptied, and then re-weighed. The omental and mesenteric fat were removed and weighed. Weights of non-carcass components were recorded. Dressed carcasses were weighed within 1 h of slaughter (hot carcass weight: HC), stored at 4°C for 24 h, and then weighed (cold carcass weight: CC). Cold carcasses were rated by three experts (from 1 to 5) according to conformation and internal and external fat based on a lightweight lamb grid (13). The perirenal fat (PR) was removed and weighed. The carcass was then cut in half lengthwise and the left side was cut according to Colomer-Rocher et al. (14) into five joints (shoulder, neck, ribs, flank, and long leg). Every joint was weighed and the left shoulder was dissected into fat (subcutaneous plus intermuscular deposits, IM deposits), muscles, and bones.

The right shoulder was removed and frozen at −80°C until the end of the experiment, then cut while frozen with a MAGURIT machine (Unitcut 545 SC model), ground using a 3-mm grid (BIRO AFMG 48/52, BIRO France), and homogenized. Aliquots were freeze-dried and were finely ground in a ball grinder (Dangoumill 300, Prolabo, Paris, France) using liquid nitrogen for chemical analysis.

Samples of muscle (supraspinatus, SE) and intermuscular (IM) adipose tissues of the shoulder and samples of perirenal (PR) fat were frozen at −80°C until analysis.

DM concentrations of both allocated and refused forage were determined by drying samples to a constant weight at 65°C in a ventilated drying oven. The samples were then ground (0.75 mm) prior to analysis. The forage and carcass samples were analyzed according to AFNOR procedures (15): dry matter (DM, AFNOR NF V18-109), ash (AFNOR NF V18-101), crude protein (CP, N x 6.25, AFNOR NF V18-120), and ether extract (EE, AFNOR NF V18-117). Neutral detergent fiber (NDF), acid detergent fiber (ADF), and acid detergent lignin (ADL) were determined according to Van Soest et al. (16).

The DM of PR, IM, and SE were determined after freeze-drying for 48 h. The lipid extract concentrations of PR, IM, and SE were determined by ether-petroleum extraction. The fatty acids (FA) extraction was carried out according to the method described by Rule (17), but using tricosanoic acid as an internal standard in the extraction solvent mixture. Lipid extraction and methylation were described by Bas et al. (18). Samples were then injected with auto sampler CP-8410 into a Varian CP-3900 gas liquid chromatograph (GLC, Varian, Les Ulis, France) on a DB-wax-fused silica capillary column (60 m × 0.25 mm i.d. × 0.25 μm film thickness: JW, Folsom, CA). For the GLC procedure, the split/split less injector type 1177, and the flame ionization detector were held at 250°C. The oven temperature increased from 120 to 195°C at a rate of 4°C/min, and then held for 60 min at constant temperature. The injector was in splitless mode for 1.0 min and then in split mode until the end of the run with a split ratio of 30:1. The column flow rate was 1.2 mL/min of He. FA were identified by comparison to reference standards (Fatty Acid Methyl Ester, FAME, Sigma, St. Louis, MO; Interchim, Montlucon, France) and analyzed under similar conditions. Under these conditions, some positional isomers, notably trans-octadecenoic acid, could not be separated from each other (19).

Several evaluations were made on the cold carcasses. The color and ultimate pH of the rib eye area of the fourth rib on the left side was evaluated, respectively with Minolta chromameter CR-300 calibrated to a white standard, using the L^*, a^*, b^* scale, and a Bioblock Scientific IP67 pH probe (Fischer scientific, Illkirch, France) calibrated to pH 4 and 7 using buffer standards.

Empty body weight (EBW) was calculated by subtracting the weight of the gut content from the slaughter weight. The carcass yield was the weight of cold carcass related to EBW.

The FA were summed by families according to their structure: even straight-chain saturated FA (ESFA) = C10:0 + C12:0 + C14:0 + C16:0 + C18:0 + C20:0 + C22:0 + C24:0; odd-numbered straight-chain FA (odd FA) = C13:0 + C15:0 + C17:0 + C17:1n−8 + C19:0 + C21:0; methyl-branched chain FA of the iso and anteiso forms, (isoFA and anteisoFA) = isoC14:0 + isoC15:0 + anteisoC15:0 + isoC16:0 + isoC17:0 + anteisoC17:0; straight-chain monounsaturated FA (MUFA) = C14:1 + ∑C16:1 (n−9 and n−7) +∑C18:1 (n−7, n−9, trans-10) + C17:1n−8 + C20:1 + C22:1 + C24:1; n−3 FA = C18:3n−3 + C20:3n−3 + C20:5n−3 + C22:5n−3 + C22:6n−3; n−6 FA= C18:2n−6 + C18:3n−6 + C20:2n−6 + C20:3n−6 + C20:4n−6 + C22:4n−6; and polyunsaturated FA (PUFA) = n−3 FA + n−6 FA.

Data were analyzed using PROC GLM (Version 9, SAS Inst., Inc., Cary, NC, 1999) with diet, age at slaughter, and interaction diet^*age as the main effects in the model. The statistical unit was the animal. Initial LW and slaughter weight were used as co-variables for ADG and carcass traits. Co-variables were kept in the model when significant. Carcass quality scores and cut weights were studied with carcass weight used as a co-variable, which was only kept in the model when significant.

The addition of 50% of concentrate to the diet significantly increased DMI, ADG, feed conversion ratio (FCR), and carcass weights, yields, and scores but not the color (P < 0.01, Table 2). A significant effect of age was observed for DMI and FCR but not for ADG (P < 0.05). The FCR was affected only by the interaction diet × age (P < 0.05). The ADG increased by 80% in the C50 compared with the C0 group (P < 0.05). The differences between the C50 and the C0 groups in carcass traits were high: addition of concentrate accounted for 75–80% increases in the weight of hot and cold carcasses, prime cuts, and muscle and 50% increases for conformation scores. Finally, the dressing percentage increased with the addition of concentrate, although to a lesser extent (10%). The non-carcass components parameters were significantly affected by the diet except for the proportion of red organs (related to EBW) and retail cuts (related to carcass weight; P < 0.05, Table 2). In addition, abdominal fat tissues and neck proportions increased by 30% with age at slaughter (P < 0.05).

The intermuscular fats were ~2-fold greater in the C50 compared to C0 group (P < 0.05). The proportions of muscle ranged from 70 to 77% within the groups. The muscle/bone ratio was significantly affected by age at slaughter and diet (P < 0.05) but no interaction between these two effects was observed.

With the exception of traits related to collagen and glycogen analyses, a significant diet effect was observed on almost all parameters (P < 0.05, Table 3). The pH values, water losses, and “L” and b” parameters were higher in C0 compared with C50 kids (P < 0.05). The shoulder DM and ether extract concentrations increased in the C50 group and with age while ash declined (P < 0.05). Interactions were observed for the color parameter “a” measured in the longissimus dorsi and the shoulder CP concentration. A significant effect of age was observed for collagen solubility (%) which decreased with time (P < 0.05).

Perineal and intramuscular adipose tissue DM concentrations were affected by diet (P < 0.01, Table 4). The DM concentrations of intermuscular adipose tissue were affected by diet (P < 0.01) and age at slaughter (P < 0.05) whereas only age at slaughter had an effect on DM concentrations of supraspinatus muscle tissue (P < 0.01). In addition, the DM concentrations of intermuscular adipose tissue increased more sharply with age at slaughter in the C50 group, resulting in a diet × age interaction (P < 0.01). The age at slaughter had a significant effect on lipid concentrations of perineal and intermuscular adipose tissues but no effect of diet was observed (P < 0.01 and 0.05, respectively).

The FA compositions of perineal, intermuscular, and supraspinatus adipose tissues were significantly affected by diet and/or age at slaughter (P < 0.05, Table 4). Proportions of saturated fatty acid (SFA), odd chain FA, iso and anteiso FA, and n−3 FA in perineal adipose, intermuscular adipose, and supraspinatus muscle tissues were higher in the C0 group, with the exception of SFA in supraspinatus muscle tissue (P < 0.01). Furthermore, this effect was more pronounced with increasing slaughter age. In contrast, proportions of monounsaturated FA (MUFA) and n−6 and n−6/n−3 FA were higher in the C50 group in perineal and intramuscular adipose tissue (P < 0.05), and in supraspinatus muscle tissue (P < 0.01). Proportions of polyunsaturated FA were higher in perirenal adipose tissue of C50 goats (P < 0.01) with no effect of age at slaughter, whereas no effect was observed in intermuscular adipose tissue and supraspinatus muscle tissue. The composition of the supraspinatus muscle tissue in the major FA C16:0 and C18:0 was affected by diet and age at slaughter whereas only diet influenced C18:1n−9 (P < 0.05); nonetheless no effect of diet × age at slaughter was observed. Effects of diet, age at slaughter, and the interaction diet × age at slaughter were observed in the proportions of C20:0, C17:1n−8, C20:3n−6, C20:4n−6, and C20:5n−3 FA in supraspinatus muscle tissue (P < 0.05, Table 4). With the exception of C20:4n−6 FA, these FA were significantly higher in the C0 group. In addition, the differences were more important as the age at slaughter increased. In the present study, C18:1cis-11, C18:1trans-11, and other positional isomers of C18:1trans FA could not be quantified separately because they co-eluted on the column used (19).