Crambe cake protein replacing soybean meal protein and its effects on performance, carcass traits and meat quality of lambs

ABSTRACT The objective was to evaluate different levels of crambe cake (CC) protein (0, 250, 500, 750 and 1000 g/kg DM) as a replacement for soybean meal protein in diets for lambs and their effect on performance, carcass traits and meat quality. Forty not castrated crossbred male lambs, weighing 20 ± 3.45 kg, were settled in a completely randomized design with eight replicates per treatment. Final body weight (BW), dry matter intake per kg BW (DMI), DMI over metabolic weight and average daily gain showed linear decreasing effect (p < 0.05), meanwhile, feed efficiency presented a quadratic effect for replacement levels. There was a quadratic effect with replacement levels of CC protein for hot carcass weight; cold carcass weight; cooling loss index; muscle fat thickness; leg perimeter; leg depth and carcass finishing. Hot carcass yield, carcass length; shoulder clods weight, rib weight and the leg weight showed a linear decrease within BW. Bones and muscle weight of shoulder clods showed a decreasing linear effect. Replacing up to 500 g/kg DM of protein of soybean meal protein for CC protein does not affect performance, carcass traits and meat quality.


Introduction
Recent expanding demand for food supply is a consequence of many technological, environmental, political and social changes worldwide, once population's purchasing power has increased considerably (Shurson 2017;Ajila et al. 2012).Therefore, it is urgent to develop innovative techniques so food could be safer, more sustainable and affordable, and the whole process of food production should be more effective (Shurson 2017) with better livestock and farming practices (Romanzini et al. 2017).
Some agroindustrial by-products not edible to humans can add valuenutritive and economicalto animal nutrition and also reduce environmental impact (Jedrejek et al. 2016;Schader et al. 2015).In this way, the use of alternative feed becomes an option in feedlots to reduce costs of animal nutrition by making the system economically viable (Bezerra et al. 2016).
Crambe (Crambe abyssinica Hochst) is a plant that has been grown in tropical and subtropical regions of Africa, Asia, Europe, the United States and South America (Sokólski et al. 2020).In Brazil, its cultivation was encouraged by the MS Foundation (State of Mato Grosso do Sul) to meet the National Biodiesel Program, and research resulted in the launch of the first Brazilian crambe cultivar in 2007 (Pitol et al. 2012) to increase biodiesel production.
Crambe cake is obtained after the mechanical extraction of the oil by pressing the seeds, presenting a residual oil content higher than that contained in the crambe meal (Canova et al. 2015).The crambe cake has 92.72% dry matter, 24.67% crude protein, 29.60% ether extract, 5994 kcal/g gross energy, 39.5 g/kg nitrogen, 8.10 g/kg phosphorus and 7.70 g/kg calcium, being an alternative for the replacement of soybean meal (Brás et al. 2014).
Ruminants can convert by-products into high biological value material thanks to microbiota activity in the digestive tract (Carrera et al. 2012).Considering that the crambe cake is a good alternative food and that the intake of lambs meat has increased in recent years, although it is still lower when compared to the intake of meat from other species, the study of this food in lamb production is justified.
The objective of this study was to evaluate the effects of adding different levels of crambe cake protein to replace soybean meal protein on the productive performance, carcass traits and meat quality in lambs.

Study location and ethical considerations
The study was conducted at the experimental farm of the State University of Londrina (UEL), Brazil.All animal proceedings were conducted in accordance with ethical standards and were approved by the Animal Use Ethics Committee, following protocol N o.7748.2014.28.

Animals, dietary treatment, experimental design and general procedures
Forty noncastrated crossbreed male lambs (Dorper × Santa Ines × Texel), with 20 ± 3.45 kg of body weight, were housed in individual stalls with slatted floor, equipped with feeders, mineral supplementation troughs and drinking fountains.The animals were distributed in a completely randomized design with five treatments (0, 250, 500, 750 and 1000 g/kg DM of crambe cake protein replacing soybean meal protein in concentrate ration) and eight replicates per treatment.
All diets were formulated to be isonitrogenous and meet nutrient requirements for growing and finishing lambs (NRC 2007).Roughage: concentrate ratio was 30:70, using corn silage as forage (Table 1).For diet formulation, chemical compositions of the ingredients, was determined at UEL's Laboratory of Animal Nutrition, according to the methodologies described by AOAC (2016).Total digestible nutrient (TDN) content of the feed used for diet balance was estimated by the equation proposed by Patterson et al. (2001): TDN = [88.9-(0.779 × ADF%)].
There was an initial period of 10 days for the lambs to adapt to their diets and management routines and 60 days for assay.Animals were fed twice a day (08 am and 04 pm), and the amount offered was adjusted daily, so leftovers were around 10% of total feed supplied.Mineral supplementation and water were provided ad libitum.At the end of the study, leftover samples were collected from each lamb and stored (freezer 4°C) for later analysis to evaluate nutrient intake.
The animals were weighed every 15 days and were slaughtered when they reached an average weight of 30 kg, considering all animals in the experiment.Dry matter intake (DMI) was calculated according to the body weight (% LW/day) and the metabolic weight of DMI (g/kg 0.75 ), as well as average daily gain (ADG) and feed efficiency (FE).For the FE calculation, daily weight gain was divided by average DMI.
Offered and leftover feed samples were dried (forced-air oven for 72 h at 55°C) and then ground (1 mm; Willis type knife mills).According the AOAC (2016) standard procedures, feed samples were analysed for dry matter (105°C for 24 h), organic matter (600°C in muffle oven for 4 h), crude protein (Kjeldahl procedure) and ether extract (Soxhlet procedure).NDF and ADF were determined according to Detmann et al. (2012).

Slaughtering procedure, carcass characteristics and meat quality
After 12 h fasting animals were slaughtered in a slaughter house with municipal inspection.They were stunned by electronarcosis, followed immediately by bleeding, skinning and evisceration.
Fasting body weight was recorded immediately before slaughter and hot carcass weight (HCW) was recorded immediately after slaughter.Carcasses were then chilled (at 2°C for 24 h) and then weighed to determine cold carcass weights (CCW).Hot carcass yield (HCY) and cold carcass yield (CCY) and cooling loss index (CLI) were calculated according to methodology described by Osório (2003).
Carcass composition was evaluated by the methodology of Luchiari Filho (2000) and the carcasses were classified according to muscle related to the size of the skeletal tissues, as described by Cezar and Souza (2007).Both were evaluated using photographic standards.The chilled carcasses were split into halves and then into commercial cuts, to measure the carcass length (CL), leg length, leg perimeter, leg depth, arm length and arm perimeter as described by Osório (2003).From the left half carcass, pieces were separated, in cuts of the national market: neck, loin, shoulder, rib and leg, which were weighed separately (McManus et al. 2010).
Loin area was obtained according to Müller (1980) by the exposure of the Longissimus lumborum muscle after a transverse cut in the carcass.The marbling rate was evaluated subjectively using photographic standards of the American Meat Science Association -AMSA (2012).
The Longissimus lumborum muscles of each animal were removed, after keeping the carcass at 2°C for 24 h, to estimate the shear force (SF), determine colour, marbling and proximate analysis.
The animals' left shoulders were separated and frozen to be dissected and evaluated for muscle-bone-fat ratio, and each tissue component was weighed to calculate the relative weight in relation to the total sample.
The determination of pH was performed in samples of 5 g of Longissimus lumborum collected 24 h post mortem.The reading was performed using a portable potentiometer with insertion electrode (Testo® 205).
Meat tenderness was evaluated by obtaining the shear force using the AMETEK Brookfield® CT3 Texture Analyser with a 3mm shear probe blade.To obtain samples of the Longissimus lumborum muscle, a cylindrical steel sampler was used to take two portions per animal, which were baked until internal temperature of 71°C and from each portion three subsamples were taken, sheared once, totalling six readings per animal (Whipple et al. 1990).The color (L, a and b) was evaluated at 30 min using a Konica Minolta®, CR-10 model portable colorimeter expressed in the CIELAB color system.A D65 illuminant was used at and observation angle of 10°and with a measuring area of 8.0 mm.

Statistical analysis
All data were subjected to statistical analyses as a completely randomized design with five treatments and eight replicates per treatment.The results were evaluated by analysis of variance and when significant, polynomial regression was used, considering 5% the level of significance, through the statistical package SAS (SAS Inst.Inc., Cary, NC).When we observed quadratic effect of the replacement levels of soybean meal protein by crambe cake protein on studied characteristics, we derived the equations to obtain the maximum (maximum point = max.p)or minimum level (minimum point = min.p)that affected these characteristics.

Results
Final body weight, dry matter intake (% LW/day and DMI g/ kg 0,75 ) and average daily gain (ADG) presented a linear decreasing effect (P < 0.05) as a function of the levels of crambe cake protein.However, the feed efficiency (FE) presented a quadratic effect with a maximum point at 377.3 g/kg DM crambe cake protein to replace the soybean meal protein (Table 2).
However, hot carcass yield (HCY) as a function of body weight and carcass length, presented a linear decreasing effect depending on the levels of crambe cake protein.The other variables did not show any effect with the levels of crambe cake replacing soybean meal protein (Table 3).
The cuts shoulder, rib and leg had a linear decreasing effect (P < 0.05) with increasing levels of replace of crambe cake protein (Table 3).Tissue composition of the shoulder clods (Table 3), bone and muscle weight presented a linear decreasing effect (P < 0.05), with increasing levels of crambe cake protein in the diet, similar to the linear decreasing effect occurring in the total weight of the shoulder (Table 3).
Meat quality, a and b color parameters, pH and proximate composition were not influenced (P > 0.05) by the inclusion levels of crambe cake protein.However, the L colour presented a linear increase effect (P < 0.05) as a function of the levels of crambe cake protein (Table 4).
h Ŷ = 8.72678 + 0.01595x -0.00028213× 2 (R 2 = 0.90) (P max = 24.32%). i general, the better is feed efficiency, the better is animal performance presenting a higher the final body weight.However, in this study, it was observed that there was a reduction in the final weight as the levels of crambe cake protein in the diet were increased.
The increasing levels of ether extract in the diets (Table 1) may have contributed to a reduction in dry matter intake, since it impairs food intake and limits the oxidation of fatty acids by the rumen microbiota (Palmquist and Mattos 2006).Medeiros et al. (2015) stated that the lipid content should be restricted to up to 60 g/kg DM of the diet, because when it exceeds this level, there may be negative effects with regard to feed efficiency, performance and carcass attributes.Levels above 70 g/kg DM of ether extract affect fibre degradation because it forms a hydrophobic layer preventing the adherence of bacteria.
Another factor that may have contributed to decline in dry matter intake was the selection of food by the animals that consumed more than 500 g/kg DM crambe cake protein, indicating that there was a reduction effect on the palatability probably due to the glucosinolates present in crambe (content not determined this study) that cause a reduction in the intake of non-fibre carbohydrates (starch), promoting negative effects on ruminants (Canova et al. 2015).According to Tripathi and Mishra (2007), the content of glucosinolates and their metabolites varies depending on the oil extraction process, being higher for those foods in which its oil is extracted with solvents.On the other hand, the dry matter intake and the daily weight gain obtained in the present study were higher than those indicated by NRC (2007) for 20 kg sheep.
The animals from treatments with 250 and 500 g/kg DM of crambe cake protein presented a similar feed efficiency (0.20), indicating that they had higher meat production per kg of DM ingested when the level of soybean meal protein replacement by crambe cake protein was 377.3 g/kg DM.
The linear reduction observed in hot carcass yield (HCW) shows that the crambe cake protein influences carcass composition, reducing commercial cuts weights (Table 3).Ovine carcass may represent 40-50% or more of body weight, varying as a function of intrinsic factors related to the animal itself, such as age, sex, genetics, morphology, birth weight and slaughter weight (Lima Júnior et al. 2016); and also by extrinsic factors, such as feeding, management, homogeneity of weighing and pre-slaughter fasting (Guerrero et al. 2013).
The cold carcass weight observed in the treatments with 250, 500 and 750 g/kg DM of crambe cake protein was related to the lowest cooling loss index, which is related to lower carcass shortening and liquid loss, probably due to a better level of finishing and higher subcutaneous fat thickness presented by the animals.
According to Rozanski et al. (2017), CLI in lamb meat is around 2.5%, ranging from 1% to 7%.When evaluating the mean cooling loss index obtained in this experiment (3.13%), it was observed that the carcasses from all treatments presented values higher than those indicated by Martins et al. (2000), and the lowest CLI (3.32%) from animals fed 500 g/kg DM crambe cake protein is related to the greater fat thickening.According to Carneiro et al. (2019), CLI may vary due to fat covering consistency, sex, slaughter weight, cold chamber temperature and humidity.
As stated in Gallo et al. (2014), values between 3 and 5 mm MFT are suitable for satisfactory protection of ovine carcasses.However, values found in this study were inferior to those recommended and could be influenced by low NDF content and the high values of EE in diets, causing low fermentation of total carbohydrates and production of acetate (Ribeiro et al. 2016).Animal age influences fat deposition; similar results were reported by Pires et al. (2006), who slaughtered 30 kg lambs and obtained values below 3 mm.These results may have influenced the increase in CLI.The results of CL may be related to the linear decreasing effect of DM intake and the lower observed nutrient intake.
The results obtained for leg perimeter (LPER) and leg depth (LD) are related to the intrinsic and extrinsic effects found in both hot and cold carcass weights.Anyhow, the effect observed in the finishing can be related to the muscular fat thickness.The treatments that presented the best finishing score were those which presented the greatest muscle fat thickness (Table 3).Moreover, factors such as sex, breed, age and rearing system can influence this variable (Cordão et al. 2012).
The results obtained for the cuts from shoulder, rib and leg may be associated with the lower weight of the animals at slaughter, especially in the treatments with 750 and 1000 g/kg DM crambe cake protein.
Growth curves of bone, muscle and adipose tissues show different patterns according to weight gain and age (Santos et al. 2010).Thus, when analysing the animal development regarding tissue composition of the shoulder, we must consider the developmental aspects of the tissues together (bone: muscle: fat ratio).
The values observed for the meat color are within the range indicated by Warris (2003), in which lamb meat values for L at 30 min may vary from 30.03 to 49.47.The pH reduction rate as well as the final pH of the meat after 24-48 h is very variable, ranging from 5.4 to 5.7 (Rego et al. 2017).Issakowics et al. (2017) when replacing soybean meal protein for increasing levels of crambe cake protein (0, 220, 440 and 660 g/kg DM) noticed a reduction in L values (34.10; 34.0; 33.7 and 32.5, respectively), so colour negatively affected consumer's preference without pH effect.
The results obtained in the proximate composition may be due to the diets containing similar levels of protein and energy.The proximate composition of lamb meat is intrinsically related to its sensory aspects and can be influenced by several factors such as species, age, breed, sex, nutrition and slaughter weight.
Therefore, lambs fed 500 g/kg DM of crambe cake (55 g EE/ kg DM) shall maintain carcass quality, presenting greater life shelf and meat quality.Therefore, to gather the best meat and fat deposition characteristics of the meat carcass, the best content is 427.0 g/kg DM of crambe cake.

Conclusion
Crambe cake can be used in up to 500 g/kg DM as a protein source to replace soybean meal protein in concentrate diets for lambs, resulting in a satisfactory productive performance without affecting carcass traits and meat quality.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Table 1 .
Chemical composition of ingredients and chemical analysis of experimental diets (g/kg DM).

Table 2 .
Performance of lambs fed with different levels of Crambe cake replacing soybean meal protein in the diets.

Table 3 .
Carcass traits and tecidual composition of lambs fed with different levels of protein of Crambe cake replacing soybean meal protein in the diets.