Introduction

Brazil figures as the second biggest biodiesel producer in the world (^{Sawin et al., 2016}), with an annual production estimated at 3.87 billion L (^{USDA, 2016}). Crude glycerin is the main byproduct of biodiesel production, and its major component is glycerol, which has been evaluated as a feed ingredient for swine diets. Adding up to 80 and 50 g kg⁻¹ glycerol to nursery diets can potentially replace wheat (^{Zijlstra et al., 2008}), lactose, and corn (^{Groesbeck et al., 2008}; ^{Shields et al., 2011}), with positive effects on performance. Reported glycerol digestible (DE) and metabolizable energy (ME) for growing pigs are 3,344±8 and 3,207±10 kcal kg⁻¹, respectively (^{Lammers et al., 2008}). For nursery pigs, no differences on ileal digestibility of dry matter (DM) and crude protein (CP) were observed for diets in which sweet milk whey was substituted for 90 and 180 g kg⁻¹ crude glycerin, respectively, indicating that the nutritional content of this feed ingredient is well utilized by young pigs (^{Oliveira et al., 2014}). However, glycerin intake can increase urine output and urinary energy losses (^{Lammers et al., 2008}; ^{Mendoza et al., 2010}). The present study aimed to determine nutritional value of glycerin added to diets of weaned piglets and its effects on performance and urinary losses.

Material and Methods

The experimental protocol followed the ethical principles in animal research (^{CONCEA, 2016}) and was conducted according to the institutional committee on animal use (case no. 22583).

Thirty castrated male piglets (Agroceres PIC), weaned between 21 to 24 days of age and weighing an average 6.3±0.68 kg, were individually housed in metabolic cages (0.5 × 0.8 m) containing two dry feeders and a nipple drinker. Cages were in a temperature-controlled room, and temperature was kept within the thermoneutral zone throughout the trial (27-30 °C and 60-80% relative humidity).

Two basal diets were offered during the trial: phase I diet, from 1 to 12 days post weaning, and phase II diet, from 13 to 24 days post weaning. The trial was divided in a five-day adaptation period, followed by a seven-day sample collection period (phase I diet) and a 12-day sample collection period (phase II diet), summing 19 days of sample collection. Using this design, animal performance was measured in two phases, but metabolism responses were measured throughout the two phases.

Piglets were randomly assigned to one of five dietary treatments: a basal diet formulated to meet the nutritional requirements of weanling piglets (^{NRC, 2012}) was used as the control (Gly0); the other four diets were formulated as follows: a diet containing 950 g kg⁻¹ of the control diet and 50 g kg⁻¹ glycerin (Gly50); a diet containing 900 g kg⁻¹ of the control diet and 100 g kg⁻¹ glycerin (Gly100); a diet containing 850 g kg⁻¹ of the control diet and 150 g kg⁻¹ glycerin (Gly150); and a diet containing the same nutritional content of the control diet in which 100 g kg⁻¹ glycerin was added at the expense of lactose (Gly100Lac0) (Table 1). Feed and water were provided ad libitum throughout the experiment.

During sample collection periods, total feces produced by each piglet were collected once a day and stored at −15 °C. Urine was collected into plastic buckets containing 5 mL of sulfuric acid (H₂SO₄). The buckets were weighted daily, and 100 mL L⁻¹ of urine volume produced were collected and stored at −15 °C. At the end of the experiment, fecal and urinary samples from each piglet were thawed, homogenized, and a 100-mL urine and a 400-g feces subsamples were used for further analysis. Fecal subsamples were dried at 60 °C for 72 h, weighed, and ground through a 2-mm screen.

Dry matter and N content of diets and feces were analyzed according to ^{AOAC (2007)} (method 930.15 and method 984.13, respectively). Gross energy of glycerin, diets, feces, and urine were determined using isoperibol bomb calorimeter (C2000, IKAWerke GmbH & Co. KG, Staufen, Germany). Nitrogen in urine was determined according to ^{AOAC (2007)} (method 984.13).

Animal performance; total tract apparent digestibility (TTAD) coefficients of DM, CP, and gross energy (GE); total tract apparent metabolizability (TTAM) coefficient of GE, nitrogen retention (NR) coefficient, DE (kcal kg⁻¹ DM), ME (kcal kg⁻¹ DM), urine output, and fecal and urine energy content were determined according to ^{Sakomura and Rostagno (2007)}.

The DE and ME values were estimated using the substitution method: using the Gly0 diet as the reference diet, and the diets containing glycerin included at the expense of the control diet (Gly50, Gly100, Gly150) as the substitution diets. Calculations were performed according to equations proposed by ^{Campbell et al. (1983)}.

The experimental design was completely randomized, with five treatments, six replicates, and a total of 30 experimental units. Results were analyzed by ANOVA using GLM procedure of SAS (Statistical Analysis System, version 9.2.). With significant differences identified by F-test (P<0.05), polynomial regression was performed using glycerol level (g kg⁻¹) as independent variable. To test the effect of substitution of lactose with glycerol, treatments Gly0 and Gly100Lac0 were compared using paired t test, and the probability level of P<0.05 was considered as statistically significant.

Results

No significant differences (P>0.05) on average daily gain (ADG), average daily feed intake (ADFI), and feed conversion ratio were observed among treatments (Table 2). A linear effect of glycerin on urinary production was observed in Phase I (P<0.0001; R² = 0.656) and Phase II (P<0.0001; R² = 0.431) (Table 2). Piglets fed the Gly100Lac0 diet showed an increase in urinary production compared with piglets fed Gly0, during Phase I (P<0.001) and Phase II (P<0.002). There were no significant effects (P>0.05) on DM TTAD, CP TTAD, GE TTAD, GE TTAM, and NR coefficients when all glycerin levels were compared. However, in the paired t test analysis, Gly100Lac0 resulted in lower DM TTAD (P = 0.030) and GE TTAM (P = 0.019) compared with Gly0. In addition, Gly100Lac0 increased GE content in feces (P = 0.043) and urine (P = 0.001) compared with Gly0.

No effects of glycerin inclusion levels on DE, ME, and GE losses in feces (Table 3) were observed (P>0.05); however, urinary gross energy increased quadratically with increased levels of glycerin inclusion (P = 0.001; R² = 0.870). In the paired t test analysis, piglets fed Gly100Lac0 had higher GE losses in feces (P = 0.043) and urine (P = 0.001) compared with piglets fed Gly0.

Estimated DE and ME values for glycerin based on the substitution method calculation were, respectively: 3528 and 3378 kcal kg⁻¹ DM in diets containing 50 g kg⁻¹ glycerin; 4118 and 3469 kcal kg⁻¹ DM in diets containing 100 g kg⁻¹ glycerin; and 3918 and 2869 kcal kg⁻¹ DM in diets containing 150 g kg⁻¹ glycerin (Table 4).

Discussion

The ADG and ADFI observed in this study were higher than the estimated expected ADG and ADFI for piglets at a relatively stress-free environment (^{NRC, 2012}), indicating that the nutrient dilution resultant from glycerin inclusion did not impair performance. Other studies also reported no effects on performance in nursery piglets fed diets containing 120 and 140 g kg⁻¹ glycerin (^{Diaz-Huepa et al., 2015}; ^{Gallego et al., 2016}), whereas several authors found improved performance results in nursery pigs fed up to 100 g kg⁻¹ crude glycerol (^{Groesbeck et al., 2008}; ^{Zijlstra et al., 2008}; ^{Shields et al., 2011}; ^{Rocha et al., 2016}). The replacement of 100 g kg⁻¹ of lactose with glycerin in diets of weaned piglets resulted in lower DM TTAD and GE TTAM coefficients and higher GE losses in urine and feces; however, piglet performance was not negatively affected, indicating that glycerin may be a viable alternative to replace up to 100 g kg⁻¹ of lactose in diets of weaned piglets.

The inclusion levels tested had no negative effects on digestibility and metabolizability of the diets; however, increases in urine output and energy excretion were observed as glycerin inclusion levels increased. The results indicated that glycerin levels higher than 100 g kg⁻¹ could limit energy utilization of the diet and agree with ^{Lammers et al. (2008)}, who reported that 11-kg pigs may not be able to metabolize diets containing more than 100 g kg⁻¹ of crude glycerol. Similarly, ^{Oliveira et al. (2014)} reported that inclusion of 90 and 180 g kg⁻¹ of glycerin to nursery diets increased urinary glycerol excretion but had no effect on apparent ileal digestibility of the diets. It is suggested that the metabolic pathways of glycerol utilization may become saturated when pigs are fed high levels of glycerin, and the higher urinary excretion of glycerol may explain the higher gross energy content of urine. The glycerol present in the urine may also exert a diuretic effect, which may increase water intake and urine production (^{Mendoza et al., 2010}). Additionally, the higher urinary production observed in piglets fed glycerin in the present study may be a consequence of the higher sodium content of those diets due to sodium present in glycerol that was not considered during diet formulation. Diets with high sodium content can result in increased water intake and urinary production, but will not necessarily impair pig performance, provided that animals have ad libitum access to non-saline water (^{Chittavong et al., 2013}).

According to ^{Campbell et al. (1983)}, the substitution method estimates can be impaired when using lower levels of inclusion, due to an amplification of analytical variability. Thus, at 50 g kg⁻¹ inclusion level, the variability is multiplied by 20, reducing the reliability of the results. As increased gross energy of urine at 150 g kg⁻¹ inclusion leads to a numerical reduction of glycerin estimated ME content, we considered the DE and ME estimates obtained at 100 g kg⁻¹ inclusion, which were 4118 and 3469 kcal kg⁻¹ DM. The glycerin DE and ME values estimated by the substitution method were intermediate in relation to those reported in the literature for weaned piglets. Using regression analysis, ^{Carvalho et al. (2012)} reported estimated DE and ME values of glycerin as 5070 and 4556 kcal kg⁻¹, respectively; ^{Lammers et al. (2008)} reported 4401 kcal kg⁻¹ DE and 3463 kcal kg⁻¹ ME for glycerol; whereas ^{Diaz-Huepa et al. (2015)} reported 3534 kcal kg⁻¹ DE and 3279 kcal kg⁻¹ ME for glycerin. The glycerin origin and purity can affect its composition and nutritional value (^{Carvalho et al., 2013}), which may explain the variability across studies. Further investigation of glycerol excretion in the urine of pigs fed glycerin is needed, as it may impact ME content of this ingredient, and the increased urine output can also increase environmental impact of swine production.

Conclusions

The nutritional value of glycerin allows its inclusion at up to 100 g kg⁻¹ in diets of weaned piglets without impairing animal performance and metabolism. Considering animal performance, glycerin can replace 100 g kg⁻¹ of lactose in diets of weaned piglets, but this substitution causes lower energy metabolizability and higher urinary production and energy losses.