Residual frying oil in the diets of sheep: intake, digestibility, nitrogen balance and ruminal parameters

Study location

The experiment was carried out at the ruminant metabolism sector of the farm school of the Londrina State University, Paraná, Brazil. Analyses of the collected materials were carried out at the Animal Nutrition Laboratory of the Londrina State University, Brazil. All protocols for the research have been approved by the Animal Ethics Committee at the same institution under register number CEUA 46/12 process 28732/11.

Animals housing and feeding

Inclusion levels of 0, 20, 40, 60, and 80 g/kg DM base of RFO in diets for lambs were evaluated. Diets were formulated to be isonitrogenous (Table 1) and were increased by 10% to the requirements according to the NRC [4]. All diets were kept at a forage:concentrate ratio of 60:40, where the source of forage was sorghum silage and the concentrate ration sources were, corn, wheat bran, soybean meal, a vitamin and mineral mixture, and RFO (Table 1).

Table 1

Food and chemical compositions in dry matter basis of diets containing different levels of residual frying oil for lambs

The residual frying oil was obtained from the company BEM AMBIENTAL enterprise which captures oils residuals in Londrina-PR and region. The residual frying oil was stored in dark plastic drums with 20 L capacity, without the addition of antioxidant. The collected product is a mixture of oil from different vegetable sources and distincts origins such as restaurants, coffee shops, snack bar and bakeries. It had dark brown color with a strong odor of rancid due to oxidation process.

Five castrated lambs of undefined breed with initial body weights of 36.8±3.3 kg were used. The animals were kept in cages with individual feeders and waterers for the metabolism trial. For total feces collection, collector bags were used. The animals were weighed at the beginning and end of each evaluation period for adjustment of diets. Each trial period lasted 15 days, the first ten days were for diet adaptation. Experimental animals were submitted to the pre-experimental period (15 days) for adaptation to the metabolism cage, routine handling and feeding. Diets were given as a whole mixture twice a day at 7 am and 4 pm to minimize feed loss by animals.

Sampling and analysis of feed, orts, and feces

The weighing of food offered and orts was performed daily, and data collected regarding the feeding portions for each animal during each trial period were later analyzed. Offered food and orts, as well as feces and urine, were sampled from the 11th up to 14th day of each trial period to determine intake, nutrient digestibility and nitrogen balance. Solid samples (supplied diet, orts and feces) were placed in plastic bags, identified, and frozen at a temperature of −18°C immediately after collection. At the end of the experimental period, the samples were defrosted and dried in forced air ovens at a temperature of 55°C for 72 hours. Subsequently, they were milled in a Willey-type mill with a sieve of 1 mm.

Chemical composition was identified using the methodology described by Mizubuti et al [5] for determining the content of DM, crude protein (CP), ether extract (EE), ash, neutral detergent fiber (NDF), acid detergent fiber, and lignin. Total carbohydrates (TCH) contents were calculated according to the equation proposed by Sniffen et al [6]: %TCH = 100 − (%CP+%EE+% ash). The nonfibrous carbohydrate (NFC) content was calculated using the difference between %TCH and %NDF.

The direct method was used to determine the apparent digestibility (AD) of nutrients by evaluating the nutrient content in feed samples and feces. The AD was obtained using the equation described by Coelho da Silva and Leão [7]:

Where, Nf = nutrient in feed (g), No = nutrient in orts (g), and Nfc = nutrient in feces (g).

The total digestible nutrient (TDN) content of feedstuffs and diets were estimated according to the equation described by Sniffen et al [6]:

Where, DCP = digestible crude protein, DEE = digestible ether extract, and DTCH = digestible total carbohydrates.

Sampling and analysis of urine

For urine collection, 25 mL of 1:1 of 50% (wt/wt) solution of 95% HCl and water was placed in the collection containers daily between 11 and 14 days in each experimental period to prevent fermentation and ammonia losses in the urine by volatilization The total volume of urine excreted per animal was measured at intervals of 24 hours, and an aliquot corresponding to 10% of the total volume per animal was stored at −15°C for later composition analysis. Nitrogen balance was calculated as the difference between the total nitrogen intake and the total nitrogen excreted in feces and urine [8], which were expressed in g/d, g/kg BW and g/kg of N_intake, respectively.

Sampling and analysis of rumen liquor

The rumen fluid was collected on the last day (15th day) of each experimental at 3 and 6 hours after de morning feeding [9] using an esophageal silicone probe connected to a vacuum pump. Approximately 100 mL of rumen fluid were collected from each animal. Afterwards, samples were filtered through three layers of cheesecloth, and three aliquots of 20 mL were taken from each sample. These samples were packed in pots containing sixteen drops of sulfuric acid (50% wt/wt) solution of 98% sulfuric acid and water and were stored at −18°C for determination of the ruminal ammonia concentration (NH₃-N). NH₃-N concentrations in the rumen fluid samples were determined by distillation with potassium hydroxide (2N), according to a technique described by Preston [10].

Statistical analysis

The experimental design was a Latin square design (5×5) with five levels of RFO, five periods of sampling and five animals (experimental units). Data were submitted to a normality test for distribution of errors, to a homogeneity of variance test, and subsequently, to an analysis of variance (α = 0.05). The statistical package R [11] was used to study the mean values by regression analysis, using “F” test (α = 0.05).

Nutrients intake

There was a linear decreasing effect on DM, organic matter (OM), CP, NDF, TCH, and NFC intake when expressed in g/d/animal (Table 2).

Table 2

Means, regression equations, regression coefficients, and variance coefficients for nutrient intake of diets for lambs with different levels of residual frying oil

A reduction in the intake of these components was most likely due to the high intake of EE that increased linearly (p<0.05) when expressed in the same unit. In this study, the levels of lipids in the diet with greater levels of RFO (40, 60, and 80 g/kg DM) exceeded the maximum level of EE recommended by the Palmquist and Mattos [12]: That is, 50 g/kg in the total diet. According to these authors, levels above 50 g/kg in the diet may negatively affect nutrient intake, either by affecting the regulatory mechanisms that control food intake or by limiting the capacity of the ruminants to oxidize fatty acids. Another aspect that could explain a reduction in intake (g/d) of the components NDF, TCH, and NFC is a reduction in the levels of these components in the diet (Table 1).

The CP intake, when expressed in g/d for RFO levels of 60 and 80 g/kg of DM, was not able to fulfill the requirements proposed by the NRC [4], which is 160 g for lambs that are eight months old and weigh 40 kg. Dry matter intake was also lower than that estimated by the NRC [4] when expressed in kg/d (1.39) and g/kg BW (32.5) for the same category at all levels of inclusion of RFO. The NDF intake, when expressed in g/kg BW, was within the limits needed for the rumen to function as reported by Van Soest [13]: That is, 8 to 12 g/kg BW.

The intake of all the nutritional components, when expressed in g/kg BW and g/kg BW^0.75, was not affected by the different levels of RFO in the diet (Table 2). This could be due to the experimental design because all animals were fed all diets at different moments. Therefore, it could be that the body weight of the animals influenced the nutrient intake, keeping them at balance. Another aspect that may have contributed to this effect is the variation in intake behavior of the animals according to their individual acceptance of the diets. According to Forbes [14], acceptability is not only a characteristic of the food but should also be observed as a function of the animal, with likes and dislikes for certain foods.

The different levels of RFO were not able to promote an effect (Table 2) on the TDN intake, regardless of the form expressed, even with the significant increase in EE intake, which is the more energetic fractional component of this variable. A similar effect was observed by Nunes et al [15], who evaluated the nutrient intake of lambs from diets containing different levels of palm kernel cake. The recommended intake of this component, as recommended by the NRC [4], is 860 g/d.

Nutrients digestibility

The digestibility of DM, OM, CP, NDF, TCH, and NFC was not influenced by the addition of RFO on the diet (Table 3). These results corroborate the results obtained by Bateman II and Jenkins [16] and Pimentel et al [17], who did not observe any effects of lipid inclusion in the ration on apparent digestibility of these nutrients, suggesting that significant amounts of unprotected lipids may be added to the rations because they do not affect nutrient digestibility.

Table 3

Nutrients digestibility of diets of lambs containing different levels of residual oil frying

However, the EE digestibility exhibited crescent linear behavior (Table 3). This can be explained by higher fat digestibility. Another possibility would be the presence of higher levels of EE in the diet, which can cause higher digestibility of this fraction. According to the NRC [17], the digestibility of fatty acid, a component of EE, can be influenced by dry matter intake, by the amount consumed, and by the characteristics of dietary lipids. The degree of unsaturation is perhaps the feature that most influences lipid digestion. Therefore, an optimal source of lipids for ruminants would be one that does not interfere with the digestibility of other nutrients and that produces high intestinal digestibility.

Nitrogen balance

Urine excretion and nitrogen retention were not influenced by the levels of RFO (Table 4). All diets resulted in positive nitrogen retention, with an average 6.4 g N retained per day, the equivalent of 22.24% of nitrogen intake. In this work, the nitrogen lost in urine 368.7 g/kg N_intake was higher than that found by Zeoula et al [9] of 309.0 g/kg N_intake, a difference that can be explained by the higher content of dietary CP: 12.8% vs 10.9%. However, our results were similar to those reported by Alves et al [18], who found a mean value of 376.5 g/kg N_intake, which is similar to the CP concentration in the diets.

Table 4

Intake, excretion, nitrogen balance and ruminal ammonia nitrogen of lambs fed diets with different levels of residual frying oil

A negative linear effect was observed for the nitrogen intake and nitrogen excretion in the feces when expressed in g/d (Table 4). The reduction in nitrogen excretion in feces is due to the reduction in the intake of this nutrient (Table 2). However, when expressed as a function of nitrogen intake, nitrogen excretion was not affected by the levels of RFO. The N excretion in the feces was higher than in the urine. Zeoula et al [9], used diets with 10.9% CP and found lower N excretion through the urine (309 g/kg N_intake) in relation to the fecal excretion (352 g/kg N_intake), similar to what occurred in this work. They provided diets that contained different levels of cassava meal to replace corn and soybean meal as a protein source for sheep. According to Hristov et al [19], improved use of ammonia in the rumen can transfer nitrogen excretion from the urine to the feces.

Ruminal ammonia nitrogen

The concentration of ruminal ammonia nitrogen (NH₃-N) was not affected by the diets containing RFO; additionally, there was no interaction between diet and the collection time (Table 4). The ruminal ammonia concentration may be indicative of the fermentation of amino acids and may indicate high levels of rumen-degraded protein and reduced levels of NFC in the rumen. Thus, no change was observed in the concentrations of NH₃-N, showing that lipids in the diet from the addition of RFO does not promote a depressant effect on the population of Gram-positive bacteria, which ferment obligatory amino acids to meet their energy and protein needs [20]. The concentrations of NH₃-N were higher than the range considered optimal (15 to 29 mg/dL) for achieving maximum microbial fermentation in ruminants without negatively influencing fiber degradation [21].