Comparative digestibility of polysaccharide-complexed zinc and zinc sulfate in diets for gestating and lactating sows

Abstract

We hypothesized that the digestibility of a zinc polysaccharide complex is greater than zinc sulfate when sows consume high fiber diets containing corn dried distillers grains with solubles (DDGS). Gilts and sows (n = 32) were blocked according to parity and assigned randomly to one of four dietary treatments (n = 8 sows per treatments). Dietary treatments consisted of: 1) Control (ConZnSO₄)—corn–soybean meal-based diet + 100 ppm supplemental Zn from ZnSO₄; 2) Control PSZn (ConPSZn)—corn–soybean meal-based diet + 100 ppm supplemental Zn from Zn polysaccharide complex; 3) DDGS/ZnSO₄—corn–soybean meal–40% DDGS gestation diet and a 30% DDGS lactation diet, with each containing 100 ppm supplemental Zn from ZnSO₄; 4) DDGS/PSZn—corn–soybean meal–40% DDGS gestation diet and a 30% DDGS lactation diet, with each containing 100 ppm supplemental Zn from Zn polysaccharide complex. A fifth dietary treatment was imposed using a subset of sows (n = 20) to determine basal Zn losses in gestating and lactating sows fed corn–soybean meal-based diets containing no supplemental Zn. Nutrient balance experiments were conducted in both gestation and lactation to evaluate the digestibility of Zn sources of the four dietary treatments and to determine basal Zn losses when no supplemental Zn was provided. The statistical model included fixed effects of diet, Zn source, and their interaction, and random effects of parity. Estimated endogenous losses of Zn were used to adjust apparent total tract digestibility (ATTD) to true total tract digestibility (TTTD) of Zn in the four dietary treatment balance periods. There were no differences in Zn concentrations of urine, plasma, colostrum, or milk samples among treatments at any time of the experiment (P > 0.05). Gestating sows fed DDGS/PSZn had improved (P < 0.05) ATTD, TTTD, and overall retention of Zn compared with both Control treatments, with the DDGS/ZnSO₄ treatment responses being intermediate. Lactating sows consuming diets without DDGS and supplemented with Zn polysaccharide complex had the greatest (P < 0.05) ATTD, TTTD, and retention of Zn, which were opposite to responses observed in gestation. Furthermore, ATTD, TTTD, and Zn retention for lactating sows consuming DDGS/PSZn were less (P < 0.05) than all other treatments. Overall, zinc digestibility of ZnSO₄ and PSZn appears to be differentially influenced by the stage of the reproductive cycle and presence of dietary fiber from DDGS.

Introduction

The current zinc (Zn) requirements for gestating and lactating sows are 100 mg/kg of diet or 210 mg/d for gestating sows and 597 mg/d for lactating sows (NRC, 2012). However, the most recent studies to determine the dietary Zn requirements of sows were conducted about 40 yr ago (Hedges et al., 1976; Kirchgessner et al., 1981) with genetic lines that were far less productive than those currently used in the swine industry (Stalder, 2018). Consequently, industry nutritionists have increased the safety margin above current requirement estimates due to uncertainty of actual Zn requirements and factors that can modify the availability of zinc (Flohr et al., 2015; Brugger and Windisch, 2017). This practice increases diet cost and leads to increased Zn excretion in manure which can have negative environmental consequences over time. Excessive accumulation of Zn in manure can be detrimental to the quality of soil and water (Jongbloed and Lenis, 1998) and must be minimized to achieve sustainable pork production systems. Inclusion of grain coproducts such as dried distillers grains with solubles (DDGS) or wheat middlings in swine diets during the last two decades further exacerbates the uncertainty regarding dietary Zn utilization. Many dietary components such as phytate, fiber, Ca, and Cu may strongly inhibit Zn absorption and negatively influence Zn utilization (Baker and Ammerman, 1995; Lönnerdal, 2000; Solomons, 2001). Source of supplemental Zn may also influence digestibility and absorption of Zn for swine. Previous investigations of Zn utilization in the nursery and growing-finishing pigs have revealed that organic Zn sources have greater digestibility, improved absorption, and decreased excretion when compared with inorganic Zn sources (Lee et al., 2001; Acda and Chae, 2002; Carlson et al., 2004; Liu et al., 2014). Currently, no studies have been published regarding the nutritional value of a Zn polysaccharide complex compared with an inorganic Zn source in diets for gestating and lactating sows. The objectives of this study were to determine and compare excretion, retention, apparent total tract digestibility (ATTD), and true total tract digestibility (TTTD) of a Zn polysaccharide complex and Zn sulfate (ZnSO₄) in gestation and lactation diets for sows with and without a dietary antagonist, DDGS. A second objective was to estimate basal daily Zn losses of sows fed diets without supplemental Zn.

Materials and Methods

This experiment was conducted at the University of Minnesota West Central Research and Outreach Center in Morris, Minnesota. The experimental protocol was reviewed and approved by the University of Minnesota Institutional Animal Care and Use Committee (IACUC# 1706-34924A).

Animals, housing, and dietary treatments

The experiment was conducted in two similar rooms in the same facility containing 16 farrowing stalls per room. Sows were housed individually in farrowing stalls (1.52 × 2.13 m) from about day 80 of gestation through lactation and weaning of litters. Farrowing stalls were equipped with one stainless steel dry feeder and one nipple waterer on a perforated floor over a deep (2.4 m) manure collection pit. All heaters and ventilation fans were operated by an independent controller in each room that maintained room temperature between 15.5 and 18 °C during gestation and lactation, and 21 °C during the week of farrowing. One infrared heat lamp (125 W bulb) was placed in the creep area of each farrowing stall as a supplemental heat source for piglets during lactation. On day 80 ± 0.8 of gestation, 32 sows (Topigs Norsvin, Burnsville, MN) were randomly assigned to one of four dietary treatments based on parity in a 2 × 2 factorial arrangement. Eight sows were assigned to each treatment and remained on their assigned dietary treatment over the entire experimental period, except during the low Zn balance feeding and collection period.

Dietary treatments consisted of: 1) Control—corn–soybean meal-based diet + 100 ppm supplemental Zn from ZnSO₄ (ConZnSO₄); 2) Control polysaccharide Zn—corn–soybean meal-based diet + 100 ppm supplemental Zn from polysaccharide-complexed Zn (ConPSZn); 3) High fiber control—corn–soybean meal–40% DDGS gestation diet and a 30% DDGS lactation diet, with each containing 100 ppm supplemental Zn from ZnSO₄ (DDGS/ZnSO₄); and 4) High fiber polysaccharide Zn—corn–soybean meal–40% DDGS gestation diet and a 30% DDGS lactation diet, with each containing 100 ppm supplemental Zn from polysaccharide-complexed Zn (DDGS/PSZn). Polysaccharide-complexed Zn was supplemented to dietary treatments using SQM Zinc (AAFCO 57.29 metal polysaccharide complex; QualiTech, Inc., Chaska, MN). Corn DDGS (Glacial Lakes Energy LLC, Watertown, SD) with low vomitoxin concentration (<1.2 ppm) was utilized in experimental diets. A fifth dietary treatment (LowZn) was imposed on a subset of sows (n = 20) to determine baseline Zn losses in gestating and lactating sows fed corn–soybean meal-based diets containing no supplemental Zn or DDGS.

Sows remained on their initially assigned dietary treatments through the completion of gestation and lactation balance periods (Figure 1). Immediately following gestation and lactation balance periods, 20 sows were selected randomly and fed the LowZn diet (gestation and lactation) during a 5-d adaptation period and a 3-d collection period. After completion of LowZn collection periods, sows returned to their originally assigned gestation and lactation dietary treatments until parturition or weaning, respectively.

Figure 1.

Experimental timeline of gestation and lactation balance periods.

Experimental diets were formulated to represent ingredient composition typically used in commercial U.S. pork production systems and were based on corn, soybean meal, and DDGS (Tables 1 and 2). Previous production data from sows in this herd were used in the NRC (2012) model to estimate the nutrient requirements. The model output was then used to formulate diets that met or exceeded nutrient requirements during gestation and lactation, except for Zn. All diets were mixed in the same feed mill on-site and were offered to sows in meal form.

Table 1.

Ingredient and nutrient composition of gestation diets (as-fed basis)¹

	Control		DDGS
Items	ZnSO4	PSZn	ZnSO4	PSZn	LowZn
Ingredient composition, %
Corn	89.29	89.29	54.52	54.52	89.29
DDGS	—	—	40.00	40.00	—
Soybean meal, 46.5% CP	7.71	7.71	2.88	2.88	7.71
Monocalcium phosphate, 21%	1.01	1.01	0.28	0.28	1.01
Limestone	0.99	0.99	1.32	1.32	0.99
Salt	0.50	0.50	0.50	0.50	0.50
ZnSO4 premix2	0.50	—	0.50	—	—
PSZn premix3	—	0.50	—	0.50	—
Zn-free premix4	—	—	—	—	0.50
Total	100.00	100.00	100.00	100.00	100.00
Calculated nutrient composition
ME, kcal/kg	3,319	3,319	3,334	3,334	3,319
Crude protein, %	11.00	11.00	16.83	16.83	11.00
Crude fiber, %	2.43	2.43	4.49	4.49	2.43
Crude fat, %	3.72	3.72	5.22	5.22	3.72
NDF, %	9.36	9.36	17.73	17.73	9.36
SID Lys, %	0.38	0.38	0.38	0.38	0.38
SID Trp, %	0.09	0.09	0.10	0.10	0.09
SID Thr, %	0.33	0.33	0.45	0.45	0.33
SID Met + Cys, %	0.38	0.38	0.51	0.51	0.38
Calcium, %	0.60	0.60	0.60	0.60	0.60
Phosphorus, total %	0.52	0.52	0.51	0.51	0.52
STTD Phosphorus, %	0.26	0.26	0.26	0.26	0.26
Zinc added, ppm	100	100	100	100	0
Zinc total, ppm	123	123	146	146	23
Analyzed composition
Moisture, %	13.2	13.1	11.2	11.7	13.2
Crude protein, %	9.5	9.2	17.5	17.1	9.6
Crude fat, %	1.9	2.0	3.4	3.4	1.8
Crude fiber, %	1.7	1.7	3.6	3.5	1.8
Ash, %	4.4	4.4	5.0	4.7	4.5
Zinc total, ppm5	160.0	112.0	130.0	150.0	21.6

¹Titanium dioxide used as indigestible marker added at 0.50% of the diet.

²Contained the following nutrients per kilogram of premix: vitamin A, 3,742,468 IU; vitamin D₃, 748,493 IU; vitamin E, 40,827 IU; vitamin K, 220 mg; riboflavin, 716 mg; niacin, 1,918 mg; pantothenic acid, 2,315 mg; vitamin B₁₂, 7 mg; iodine, 66 mg; selenium, 66 mg; zinc (as zinc sulfate monohydrate), 44,092 mg; iron, 35,274 mg; manganese, 11,023 mg; copper, 8,818 mg.

³Premix as above but SQM Zn (QualiTech, Chaska, MN) was the source of zinc.

⁴Premix as above but containing no supplemental zinc.

⁵Average of two dietary sample analyses.

Table 2.

Ingredient and nutrient composition of lactation diets (as-fed basis)¹

	Control		DDGS
Items	ZnSO4	PSZn	ZnSO4	PSZn	LowZn
Ingredient composition, %
Corn	75.62	75.62	52.35	52.35	75.62
DDGS	—	—	30.00	30.00	—
Soybean meal, 46.5% CP	20.89	20.89	14.33	14.33	20.89
l-Lysine HCl	0.10	0.10	0.19	0.19	0.10
Monocalcium phosphate, 21% P	1.44	1.44	0.91	0.91	1.44
Limestone	0.95	0.95	1.22	1.22	0.95
Salt	0.50	0.50	0.50	0.50	0.50
ZnSO4 premix2	0.50	—	0.50	—	—
PSZn premix3	—	0.50	—	0.50	—
Zn-free premix4	—	—	—	—	0.50
Total	100.00	100.00	100.00	100.00	100.00
Calculated nutrient composition:
ME, kcal/kg	3,297	3,297	3,308	3,308	3,297
Crude protein, %	16.08	16.08	19.41	19.41	16.08
Crude fiber, %	2.69	2.69	4.18	4.18	2.69
Crude fat, %	3.58	3.58	4.73	4.73	3.58
NDF, %	9.39	9.39	15.64	15.64	9.39
SID Lys, %	0.78	0.78	0.78	0.78	0.78
SID Trp, %	0.16	0.16	0.15	0.15	0.16
SID Thr, %	0.50	0.50	0.55	0.55	0.50
SID Met + Cys, %	0.49	0.49	0.57	0.57	0.49
Calcium, %	0.70	0.70	0.70	0.70	0.70
Phosphorus, total %	0.66	0.66	0.64	0.64	0.66
STTD phosphorus, %	0.35	0.35	0.35	0.35	0.35
Zinc added, ppm	100	100	100	100	0
Zinc total, ppm	128	128	146	145	28
Analyzed composition:
Moisture, %	12.7	12.7	11.5	11.6	12.3
Crude protein, %	15.7	14.8	19.6	19.3	14.0
Crude fat, %	1.7	1.7	2.8	2.9	1.8
Crude fiber, %	2.1	2.0	3.3	3.5	1.9
Ash, %	5.4	5.5	6.0	5.9	5.1
Zinc total, ppm5	129.9	134.2	136.0	123.7	29.3

¹Titanium dioxide used as indigestible marker added at 0.50% of the diet.

²Contained the following nutrients per kilogram of premix: vitamin A, 3,742,468 IU; vitamin D₃, 748,493 IU; vitamin E, 40,827 IU; vitamin K, 220 mg; riboflavin, 716 mg; niacin, 1,918 mg; pantothenic acid, 2,315 mg; vitamin B₁₂, 7 mg; iodine, 66 mg; selenium, 66 mg; zinc (as zinc sulfate monohydrate), 44,092 mg; iron, 35,274 mg; manganese, 11,023 mg; copper, 8,818 mg.

³Premix as above but SQM Zn (QualiTech, Chaska, MN) was the source of zinc.

⁴Premix as above but containing no supplemental zinc.

⁵Average of two dietary sample analyses.

Sow and piglet performance

Sows were identified individually using ear tags and weighed at the initiation of the experiment, within 24 h of expected farrowing date, within 24 h after parturition, and at weaning. Daily feed intake and instances of illness were monitored and recorded daily. Sows were limit-fed 2.2 kg of feed once daily during gestation. Immediately following parturition and throughout lactation, sows were offered an increasing amount of feed twice daily to achieve ad libitum intake by day 4 of lactation. Instances of feed wastage were monitored and recorded if necessary. Sows were allowed ad libitum access to water throughout the experiment, which was supplied by one nipple drinker in each farrowing stall.

Measurements of sow performance at farrowing included: total number of piglets born, born alive, stillborn, mummified, and weaned per litter. Litter sizes were standardized as close to 11 piglets per sow as possible by cross-fostering piglets within the treatment group within 24 h of birth. Piglets were processed according to the standard operating procedures of the farm which included administering injectable iron, clipping needle teeth, docking tails, and castration of males within 24 h of birth. Before cross-fostering, all piglets were ear-tagged individually and weighed within 24 h of birth. Piglets were weighed again at weaning (20.2 ± 1.6 d of age) to determine the total body weight (BW) gain during the nursing period.

Zinc balance during gestation and lactation

Two separate nutrient balance trials were conducted on days 89 ± 0.8 and 97 ± 0.8 of gestation to evaluate digestibility of dietary Zn among treatments. Sows were fed their respective experimental diets containing 0.50% titanium dioxide. Although sows began consumption of experimental diets at about day 80 of gestation, titanium dioxide was not included in diets until about day 86 of gestation. Sows were housed in individual stalls for sample collection and were allowed a 9-d adaptation period and a subsequent 3-d collection period. Upon completion of the collection period (about day 92 of gestation), five sows per initial treatment (n = 20 total) were fed the LowZn dietary treatment and allowed a 5-d acclimation period before a second 3-d collection period.

During lactation, nutrient balance experiments were conducted on days 7 ± 2.0 and 15 ± 1.5 of lactation as described for gestation with minor differences. Sows were fed their respective experimental diets containing 0.50% titanium dioxide on day 110 of gestation, allowing an 11-d adaptation period which preceded a 3-d collection period to determine Zn digestibility of experimental lactation diets. Digestibility of a LowZn lactation diet was determined using 20 sows as described for the gestation period.

Fresh fecal samples were collected once daily at 1500 hours, placed in plastic bags, and stored separately at −20 °C for all balance periods. At the end of each collection period, fecal samples were pooled by sow within the balance period and weighed. To determine moisture content, samples were dried in a forced draft oven at 55 °C and weighed once daily for about 3 d or until samples maintained a constant weight. Once dried, feces were ground through a 1-mm screen and stored in Whirl-Pak bags until subsequent Zn and titanium analysis.

Urinary catheters were inserted into the urethra of each sow 1 d before each nutrient balance period. To begin this process, the entire vulva area of the sow was washed with an antiseptic (betadine). Next, sterile urinary catheters (LubriCath, 2-way, 30 mL balloon, 18 French; Bard Medical Canada Inc., Oakville, ON, Canada) were lubricated with a sterile lubricant and inserted flaccidly into the urethra. The lubricant was laced with lidocaine to prevent spasms in the urethra upon placement. While the sow was standing, the tip of the catheter was guided along the base of the vagina by a technician’s finger until it entered the urethra. The technician used sterile gloves for this procedure. Once the catheter tip was fully inserted into the urethra, the balloon was inflated with 30 mL of saline solution to retain the catheter in the bladder. Catheters were connected to polyvinyl tubing that dispensed urine into a closed vessel which allowed the sow to urinate as necessary. Urine was collected in a vessel that contained 20 mL of sulfuric acid to maintain urine pH ≤ 5. Sulfuric acid was used to preserve urinary nitrogen, prevent microbial contamination of the urine sample, and subsequently minimized the occurrence of ascending urinary tract infections. Catheters remained in place for 48 h and urine collection vessels were emptied and subsampled to obtain 5% aliquots as necessary. At the end of the collection period, the inflatable cuff was deflated and the catheter was extracted from the bladder. Two urine samples (50 mL each) from each sow during each balance period were frozen and stored at −20 °C for later Zn analysis. Sow body temperature, behavior, and occurrence of vaginal discharge were evaluated twice daily within each balance period to monitor for urinary tract infections. Any urine collections with instances of catheter complications such as infection or misplacement were not sampled or stored for later analysis.

Blood samples were collected on day 3 of each balance period via jugular venipuncture into heparinized Vacutainer tubes. Blood samples were placed immediately on ice after collection and then centrifuged at 1,400 × g to obtain plasma. Plasma samples were frozen and stored at −80 °C for later analysis.

Colostrum samples (50 mL) were collected within 12 h of parturition from all functional teats by hand stripping teats and collecting equal quantities of colostrum from each teat to obtain a representative sample. Exogenous oxytocin was not used for the collection of colostrum. Milk was collected on the final day of each lactation balance period by administering 10 IU of oxytocin IM to each sow before using a collection process similar to that described for colostrum. Immediately after collection, all colostrum and milk samples were frozen and stored at −20 °C.

Digestibility determinations and milk output

Values for ATTD, TTTD, and basal endogenous losses of Zn were calculated using the equations described by Stein et al. (2007). The equations are as follows:

$${\displaystyle \begin{array}{ll}{\displaystyle \mathrm{A}\mathrm{T}\mathrm{T}\mathrm{D},\mathrm{\%}}& =\left[1-\left(\mathrm{Z}{\mathrm{n}}_{\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}}/\mathrm{Z}{\mathrm{n}}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{t}}\right)\mathrm{x}\left({\mathrm{M}}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{t}}/{\mathrm{M}}_{\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}}\right)\right]\times 100;\\ & \mathrm{Z}{\mathrm{n}}_{\mathrm{e}\mathrm{n}\mathrm{d}}=\mathrm{Z}{\mathrm{n}}_{\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}}\times \left({\mathrm{M}}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{t}}/{\mathrm{M}}_{\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}}\right)\end{array}}$$

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{T}\mathrm{T}\mathrm{T}\mathrm{D},\mathrm{\%}=\mathrm{A}\mathrm{T}\mathrm{T}\mathrm{D}+[\left(\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{Z}{\mathrm{n}}_{\mathrm{e}\mathrm{n}\mathrm{d}}/\mathrm{Z}{\mathrm{n}}_{\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{t}}\right)\times 100];}\end{array}}$$

where Zn_feces represents fecal Zn concentration in mg/kg dry matter (DM), Zn_Diet represents dietary Zn concentration in mg/kg DM, M_diet represents dietary indigestible marker (titanium) in mg/kg DM, M_feces represents fecal indigestible marker (titanium) in mg/kg DM, and Zn_end represents estimated endogenous Zn losses in mg/kg DM.

Milk energy output was estimated using an equation from NRC (2012): Milk energy output, kcal/d = (4.92 × litter gain, g/d) – (90 × number of pigs). Estimated energy density of milk (Hurley, 2015) was used to estimate daily milk output: Milk output, g/d = Milk energy output/Milk energy density. Energy density of milk corresponded to the stage of lactation during which the balance experiments were conducted: Milk energy, kcal/g = 1.29 (day 7) or 1.17 (day 15).

Sample analysis

About 2 kg of each experimental diet was collected at mixing and stored in a freezer at −20 °C. Two randomly selected samples from each phase and treatment were sent to the University of Missouri Agricultural Experiment Station Chemical Laboratories (Columbia, MO) for proximate analysis and determination of Zn and titanium concentration. Feces, urine, plasma, colostrum, and milk samples were also analyzed for zinc concentration. Standard procedures (AOAC International, 2006) were followed for analysis of moisture (method 934.01), ash (method 942.05), ether extract (method 920.39), crude fiber (method 978.10), crude protein (method 990.03), and zinc (method 985.01) concentrations. Experimental diets and fecal samples were analyzed for titanium concentration according to procedures described by Myers et al. (2004).

Statistical analysis

Data were evaluated for the presence of outliers, normal distribution, and heterogeneous variance among treatments. Outliers were deemed to be any value greater than or less than two standard deviations from the mean and were removed from the final analysis to achieve normal distribution of data and equal variances among treatments. Considering all Zn balances conducted in this experiment (n = 262), 20 balances were deemed to be outliers and removed from the dataset. Removed balances were evenly distributed across dietary treatments (Gestation: ConZnSO₄, 2; ConPSZn, 0; DDGS/ZnSO₄, 0; DDGS/PSZn, 0; LowZn, 7; Lactation: ConZnSO₄, 3; ConPSZn, 2; DDGS/ZnSO₄, 1; DDGS/PSZn, 1; LowZn, 4). Commonly, these balance removals were a direct result of sows not consuming feed, which did not allow for digestibility calculations.

Experimental data were analyzed using the PROC GLIMMIX procedure of SAS (Version 9.4, SAS Institute Inc., Cary, NC) with a Gaussian distribution. Sow was considered the experimental unit. The statistical model included fixed effects of diet, Zn source, and their interaction, and random effects of parity. Treatment means were separated using the PDIFF option with the Tukey–Kramer adjustment for multiple comparisons.

Chi-square analysis was used to determine the influence of treatments on categorical response variables such as piglet mortality over the entire experiment. All data are reported as least square means and considered statistically significant at P < 0.05 with P < 0.10 considered a trend.

Results

Digestibility of Zn during gestation

Sows consumed an identical amount of feed during the gestation balance period, regardless of dietary treatment (Table 3). Estimated daily fecal excretion of DM was greater (P < 0.05) for sows consuming diets containing DDGS compared with that of sows consuming corn–soybean meal-based diets because of lower overall DM digestibility of DDGS-containing diets. There was an interaction (P < 0.05) between diet and Zn source on concentrations of fecal Zn, where feeding DDGS/PSZn had lower fecal Zn concentrations than all other treatments (P < 0.05). Although daily fecal excretion increased (P < 0.05) for sows consuming DDGS diets, daily fecal Zn and Ti excretion decreased (P < 0.05), compared with sows consuming Control diets. Consequently, ATTD, TTTD, and retention of Zn were greatest (P < 0.05) for sows consuming DDGS/PSZn compared with both Control treatments, with DDGS/ZnSO₄ being intermediate. Therefore, despite lower DM digestibility for diets containing DDGS, ATTD, and TTTD of Zn improved (P < 0.05) for sows consuming DDGS diets containing greater dietary fiber content compared with those fed conventional corn and soybean meal diets.

Table 3.

Daily Zn balance, ATTD, and TTTD of Zn for sows during gestation

	Control		DDGS			P-value
Item	ZnSO4	PSZn	ZnSO4	PSZn	SE	Diet	Zn source	Diet * Zn source
No. of sows assigned	8	8	8	8	—	—	—	—
ADFI, kg/d	2.22	2.22	2.22	2.22	—	—	—	—
Diet DM, %	86.8	86.9	88.7	88.3	—	—	—	—
ADFI, kg/d DM	1.93	1.93	1.97	1.96	—	—	—	—
Diet Zn, mg/kg DM	184	129	146	170	—	—	—	—
Diet Ti, mg/kg DM	3,443	3,314	3,087	3,115	—	—	—	—
Zn consumed, mg/d DM	356	249	289	333	—	—	—	—
Ti consumed, mg/d DM	6,646	6,401	6,090	6,112	—	—	—	—
Fecal excretion1, kg/d DM	0.33ab	0.31b	0.38a	0.38a	0.02	<0.01	0.62	0.63
Fecal Zn, mg/kg DM	1,401a	1,220b	911c	822d	37.8	<0.01	<0.01	0.02
Fecal Ti, mg/kg DM	20,682a	22,275a	16,029b	16,194b	1,136.2	<0.01	0.33	0.41
Fecal Zn excreted2, mg/d	462a	385ab	352b	320b	33.2	<0.01	0.03	0.36
Fecal Ti excreted3, mg/d	6,646a	6,401b	6,090d	6,112c	1.6	<0.01	<0.01	<0.01
DM digestibility4, %	82.9xy	83.9x	80.5y	80.4y	1.2	<0.01	0.66	0.59
ATTD Zn5, %	−31.0b	−52.9b,y	−21.9ab,x	4.1a	11.5	<0.01	0.82	0.01
TTTD Zn6, %	−3.2b	−13.1b	13.1ab	34.4a	11.5	<0.01	0.54	0.09
Urine excretion7, kg/d	13.1	12.2	12.2	8.1	4.0	0.53	0.54	0.70
Urine Zn concentration, mg/kg	0.3	0.8	0.5	0.6	0.2	0.99	0.19	0.32
Urine Zn excreted8, mg/d	3	3	4	4	0.6	0.10	0.74	0.80
Total Zn excreted9, mg/d	465a	387ab	356b	324b	33.2	<0.01	0.03	0.36
Zn retained10, mg/d	−110b	−138b	−68ab	9a	33.2	<0.01	0.34	0.04

¹Fecal excretion = DM_intake * (Ti_diet/ Ti_feces).

²Fecal Zn excreted = Zn_feces * Output_feces.

³Fecal Ti excreted = Ti_feces * Output_feces.

⁴DM digestibility, % = [1 −(Ti_diet/Ti_feces)] * 100.

⁵ATTD Zn, % = [1 – (Zn_feces/Zn_diet) x (M_diet/M_feces)] * 100.

⁶TTTD Zn, % = ATTD + [(Zn_end/Zn_diet) * 100].

⁷Represents urine output of the following sows per treatment: ConZnSO₄ = 5; ConPSZn = 3; DDGS/ZnSO₄ = 6; and DDGS/PSZn = 6.

⁸Urine Zn excreted = (Zn_urine * Urine_output).

⁹Total Zn excreted = (Fecal Zn_excreted + Urine Zn_excreted).

¹⁰Zn retained = (Zn_intake – Zn_excreted).

^a–dMeans within a row with different superscripts differ (P < 0.05).

^xyMeans within a row with different superscripts differ (P < 0.10).

Supplemental PSZn improved (P < 0.05) digestibility of Zn in gestating sows fed DDGS diets that contain higher concentrations of dietary fiber, but not in diets based on corn and soybean meal. Improved digestibility of Zn for diets with DDGS and PSZn reduced (P < 0.05) fecal Zn excretion, and ultimately improved Zn retention, when compared with corn and soybean meal diets. Additionally, total Zn excretion was greater (P < 0.05) for Control diets compared with DDGS-containing diets, resulting in improved Zn retention for sows fed DDGS diets. Overall, diet and Zn source interactions affected Zn retention so that PSZn supplemented to DDGS diets allowed positive Zn retention in sows, which was not observed for the other dietary treatments.

Digestibility of Zn during lactation

Before initiation of feeding lactation diets, four sows were removed from the experiment that were confirmed not pregnant. The distribution of sows removed from the lactation balance trial were: ConZnSO₄ = 2 of 8; ConPSZn = 0 of 8; DDGS/ZnSO₄ = 2 of 8; DDGS/PSZn = 0 of 8. An additional gilt that was not included in the gestation balance periods, but received the DDGS/PSZn dietary treatment during gestation, was added to the lactation balance trial. Therefore, the distribution of sows per treatment in the lactation balance trial were: ConZnSO₄, n = 6; ConPSZn, n = 8; DDGS/ZnSO₄, n = 6; and DDGS/PSZn, n = 9.

Immediately after farrowing, sows were allowed an amount of feed that they would completely consume each day to best represent ad libitum feeding. However, there were differences in average daily feed intake (ADFI) among treatments (Table 4). Sows assigned to DDGS treatments consumed less (P < 0.05) feed compared with sows assigned to Control treatments. Consequently, sows consuming Control diets consumed a greater amount of daily Zn (P < 0.05) and tended to consume greater amounts of Ti (P = 0.07). Calculated daily fecal excretion of DM was not different across treatments. As expected, and similar to responses in gestation, DM digestibility was reduced by DDGS inclusion in diets (P < 0.05). In addition, sows consuming DDGS diets with high dietary fiber excreted less Ti (P < 0.05) and tended to excrete less Zn (P = 0.09), than sows consuming lactation diets based on corn and soybean meal, which were similar to gestation responses. Diet and Zn source interactions significantly affected ATTD, TTTD, excretion, and overall retention of Zn (P < 0.01). Sows consuming diets without DDGS and supplemented with PSZn exhibited the highest ATTD, TTTD, and retention of Zn, which was opposite to responses observed during gestation. However, when PSZn was supplemented to diets with DDGS, ATTD, TTTD, and Zn retention were less (P < 0.05) than all other treatments. Despite slight differences in feed intake of sows in lactation, there were no differences in estimated milk yield or secretion of Zn in milk.

Table 4.

Daily Zn balance, ATTD, and TTTD of Zn for sows during lactation

	Control		DDGS			P-value
Item	ZnSO4	PSZn	ZnSO4	PSZn	SE	Diet	Zn source	Diet * Zn source
No. of sows	6	8	6	9	—	—	—	—
ADFI, kg/d	6.71a	5.68ab	4.47b	4.83b	0.64	<0.01	0.43	0.09
Diet DM, %	87.3	87.3	88.4	88.5	—	—	—	—
ADFI, kg/d DM	5.85a	4.96ab	3.96b	4.27b	0.56	<0.01	0.43	0.10
Diet Zn, mg/kg DM	149	154	154	140	—	—	—	—
Diet Ti, mg/kg DM	3,425	3,139	3,368	3,518	—	—	—	—
Zn consumed, mg/d DM	811a	713ab	618ab	583b	60.7	0.01	0.28	0.61
Ti consumed, mg/d DM	18,662a	14,554b	13,549b	14,649ab	1,179	0.07	0.27	<0.01
Fecal excretion1, kg/d DM	0.79	0.66	0.73	0.82	0.1	0.42	0.79	0.10
Fecal Zn, mg/kg DM	987a	889b	759c	798c	26.4	<0.01	0.07	<0.01
Fecal Ti, mg/kg DM	25,788a	23,700b	18,369c	18,394c	296.1	<0.01	<0.01	<0.01
Fecal Zn excreted2, mg/d	772a	600ab	547b	653ab	80.6	0.09	0.52	<0.01
Fecal Ti excreted3, mg/d	19,965a,x	15,656ab,y	13,329b	15,028b	1,878	<0.01	0.29	0.01
DM digestibility4, %	86.7a	86.7a	81.6b,x	80.8b,y	0.2	<0.01	0.14	0.07
ATTD Zn5, %	11.6b	22.9a	10.0b	−8.6c	2.6	<0.01	0.03	<0.01
TTTD Zn6, %	28.3b	39.0a	26.1b	9.1c	2.6	<0.01	0.05	<0.01
Urine excretion7, kg/d	17.2	10.5	13.3	16.0	5.8	0.88	0.74	0.42
Urine Zn concentration, mg/kg	0.5	0.7	0.8	0.6	0.2	0.65	0.92	0.32
Urine Zn excreted8, mg/d	6	8	9	7	1.7	0.60	0.86	0.23
Milk yield, kg/d	9.2	9.1	9.8	10.3	0.7	0.24	0.83	0.71
Milk Zn concentration, mg/kg	5.2	5.7	5.0	5.4	0.4	0.48	0.33	0.88
Milk Zn secreted9, mg/d	52	52a	49	54	5.7	0.99	0.63	0.64
Total Zn excreted10, mg/d	830a	660ab	606b	714ab	80.6	0.10	0.55	<0.01
Zn retained11, mg/d	42b	101	4b	−115c	19.1	<0.01	0.02	<0.01

¹Fecal excretion = DM_intake * (Ti_diet/ Ti_feces).

²Fecal Zn excreted = Zn_feces * Output_feces.

³Fecal Ti excreted = Ti_feces * Output_feces.

⁴DM digestibility, % = [1 − (Ti_diet/Ti_feces)] * 100.

⁵ATTD Zn, % = [1 – (Zn_feces/Zn_diet) × (M_diet/M_feces)] * 100.

⁶TTTD Zn, % = ATTD + [(Zn_end/Zn_diet) * 100].

⁷Represents urine output of the following sows per treatment: ConZnSO₄ = 4; ConPSZn = 2; DDGS/ZnSO₄ = 5; and DDGS/PSZn = 5.

⁸Urine Zn excreted = (Zn_urine * Urine_output).

⁹Milk Zn secreted = (Zn_milk * Milk_output).

¹⁰Total Zn excreted = (Fecal Zn_excreted + Urine Zn_excreted).

¹¹Zn retained = (Zn_intake – Zn_excreted).

^a–cMeans within a row with different superscripts differ (P < 0.05).

^x,yMeans within a row with different superscripts differ (P < 0.10).

Sow health and performance

There were no differences in parity or BW among sows fed any of the four dietary treatments at the initiation of the experiment (Table 5). Furthermore, no major health challenges were experienced by sows or pigs during this study. Sows were monitored daily for elevated body temperatures by measuring and recording rectal temperatures after farrowing and twice daily within balance periods. Two sows experienced transient fevers, for 1 to 3 d and consumed little or no feed, which was likely due to mild urinary tract infections from catheter placements. These sows were removed from the lactation LowZn balance period.

Table 5.

Effects of Zn source and diet on farrowing performance of sows

	Control		DDGS			P-value
Item	ZnSO4	PSZn	ZnSO4	PSZn	SE	Diet	Zn source	Diet * Zn source
Parity	3.2	2.7	2.2	2.8	0.6	0.43	0.82	0.41
Lactation length, d	21.2	19.2	19.8	20.7	0.6	0.94	0.36	0.03
Bodyweight, kg
D80 gestation	244.4	244.4	239.5	245.8	12.5	0.80	0.68	0.67
Pre-farrow1	249.0	250.7	243.2	258.8	12.1	0.88	0.30	0.38
Post-farrow2	234.3	237.3	237.3	241.6	11.4	0.54	0.53	0.91
Weaning	239.0	236.8	225.5	226.7	13.3	0.10	0.94	0.80
Farrowing Performance
Total pigs born	14.2	14.0	16.2	14.2	1.2	0.38	0.41	0.49
Pigs born alive	11.5	11.6	15.0	13.2	1.3	0.07	0.54	0.51
Pigs weaned	10.8	10.4	13.4	11.9	1.0	0.03	0.31	0.56
Piglet birth wt., kg	1.31ab	1.29b	1.24b	1.41a	0.1	0.42	0.03	<0.01
Litter birth wt., kg3	18.1	17.7	20.2	20.4	1.7	0.17	0.96	0.87
Litter start wt., kg4	14.4b	15.1b	19.0a	18.8a	1.4	<0.01	0.84	0.76
Piglet weaning wt., kg	7.1a	6.3b	5.6c	6.7ab	0.3	<0.01	0.26	<0.01
Litter weaning wt., kg5	69.8xy	64.5y	75.1xy	82.6x	5.3	0.04	0.83	0.24
Piglet gain, g/d	251	244	212	242	12.4	0.11	0.37	0.14
Litter gain, g/d	2,615	2,582	2,831	2,919	207.2	0.20	0.90	0.77

¹Sows were weighed 1 d prior to expected farrowing date.

²Sows were weighed within 24 h of completing farrowing.

³Total litter birth weight before cross-fostering.

⁴Total litter weight of live pigs after cross-fostering.

⁵Total litter weaning weight after cross-fostering.

^a,bMeans within a row with different superscripts differ (P < 0.05).

^x,yMeans within a row with different superscripts differ (P < 0.10).

Sows had a stillborn piglet rate of 9.7%, with low (2.7%) overall preweaning piglet mortality. Distribution of stillborn piglets in relation to the number of pigs born within each treatment were: ConZnSO₄, 15 of 84 (17.8%); ConPSZn, 15 of 113 (13.3%); DDGS/ZnSO₄, 4 of 97 (4.1%); and DDGS/PSZn, 8 of 137 (5.8%; χ ² = 13.75; df = 3; P ≤ 0.01). The number of piglet deaths prior to weaning was not different among treatments (χ ² = 0.41; df = 3; P = 0.94) and was distributed among treatments as follows: ConZnSO₄, 3 of 84 (3.6%); ConPSZn, 3 of 113 (2.6%); DDGS/ZnSO₄, 3 of 97 (3.1%); and DDGS/PSZn, 3 of 137 (2.2%).

There were no differences among treatments in the total number of pigs born or number of pigs born alive per litter, but there were slight differences in piglet birth weight and number of piglets weaned among treatments. However, it is not likely that these slight differences were associated with dietary treatment, but rather were due to numerical differences in the number of piglets born per litter. Sows assigned to the DDGS/ZnSO₄ treatment produced piglets with lower (P < 0.05) weaning weights than sows assigned to all other treatments, but there were no overall differences in piglet and litter daily gain among treatments.

Colostrum, milk, and plasma Zn concentrations

Composition, including Zn concentrations, of colostrum and milk were not different (P > 0.05) among diets, Zn source, or their interactions at any time of the trial with one exception (Table 6). Diet and Zn source interacted to influence crude fat concentration of colostrum. Similarly, there were no differences (P > 0.05) in Zn concentrations of plasma among dietary treatments at any time of the trial (Table 7).

Table 6.

Colostrum and milk composition by treatment (as-is basis)

	Control		DDGS				P-value
Item	ZnSO4	PSZn	ZnSO4	PSZn	LowZn	SE	Diet	Zn source	Diet * Zn source
Colostrum1
Moisture, %	77.3	75.3	77.0	77.8	—	1.0	0.25	0.53	0.13
Crude protein, %	14.1	15.0	12.0	14.1	—	1.1	0.19	0.18	0.59
Crude fat, %	4.3	5.8	6.3	3.7	—	1.0	0.99	0.55	0.04
Zn, mg/kg	12.2	13.7	13.0	14.8	—	1.3	0.47	0.23	0.90
Milk
Lactation2
Moisture, %	79.1	79.9	79.9	79.8	—	0.9	0.66	0.70	0.57
Crude protein, %	5.3	5.7	5.0	5.3	—	1.0	0.14	0.13	0.81
Crude fat, %	9.7	8.6	9.5	9.1	—	1.1	0.90	0.47	0.77
Zn, mg/kg	5.2	5.7	5.0	5.4	—	0.4	0.48	0.33	0.88
LowZn Balance3
Moisture, %	—	—	—	—	80.5	0.9	—	—	—
Crude protein, %	—	—	—	—	4.9	0.1	—	—	—
Crude fat, %	—	—	—	—	8.9	1.1	—	—	—
Zn, mg/kg	—	—	—	—	4.3	0.2	—	—	—

¹Samples collected within 12 h after parturition.

²Samples collected on day 10 of lactation.

³Samples collected on day 17 of lactation.

Table 7.

Zinc concentration of plasma in sows during gestation and lactation

	Control		DDGS				P-value
Item	ZnSO4	PSZn	ZnSO4	PSZn	LowZn	SE	Diet	Zn source	Diet * Zn source
Plasma Zn, mcg/dL
Initial	69.6	67.0	66.8	72.1	—	2.9	0.67	0.62	0.16
Gestation1	70.0	75.0	75.3	73.3	—	3.6	0.54	0.63	0.25
Gest. LowZn2	—	—	—	—	57.9	2.2	—	—	—
Pre-farrow3	78.3	77.8	79.8	83.5	—	4.7	0.17	0.56	0.42
Lactation4	86.0	76.7	83.2	79.7	—	3.5	0.98	0.08	0.41
Lact. LowZn5	—	—	—	—	49.9	2.4	—	—	—
Weaning6	86.7	97.6	89.8	90.3	—	5.7	0.68	0.28	0.32

¹Samples collected on day 3 of gestation balance period.

²Samples collected on day 3 of LowZn balance period in gestation.

³Samples collected 24 h prior to expected farrowing date.

⁴Samples collected on day 3 of lactation balance period.

⁵Samples collected on day 3 of LowZn balance period in lactation.

⁶Samples collected on about day 21 of lactation.

Baseline Zn losses

When gestating sows were fed a diet without supplemental Zn, daily fecal excretion of Zn and total daily excretion of Zn exceeded the amount of Zn consumed, resulting in negative ATTD (−107.4%) and daily loss of Zn (−44.6 mg/d; Table 8). In contrast, daily Zn intake of lactating sows exceeded total Zn excretion, resulting in positive ATTD of Zn at 25.6%. After correcting for excretion of Zn in urine and milk, there was positive retention of Zn (11.4 mg/d) during lactation.

Table 8.

Daily Zn balance, ATTD, and TTTD of Zn for sows during gestation and lactation when fed diets not supplemented with Zn

	Gestation		Lactation
Item	LowZn	SE	LowZn	SE
No. of sows	20	—	18	—
ADFI, kg/d	2.22	—	8.09	1.1
Diet dry matter (DM), %	86.8	—	87.7	—
ADFI, kg/d DM	1.93	—	7.10	0.9
Diet Zn, mg/kg DM	25	—	33	—
Diet Ti, mg/kg DM	3,422	—	3,158	—
Zn consumed, mg/d DM	487	—	231	7.8
Ti consumed, mg/d DM	6,601	—	21,890	740.0
Fecal excretion1, kg/d DM	0.26	<0.01	0.88	0.03
Fecal Zn, mg/kg DM	372	45.8	203	8.4
Fecal Ti, mg/kg DM	26,028	871.5	25,705	623.5
Fecal Zn excreted2, mg/d	90	6.7	171	6.1
Fecal Ti excreted3, mg/d	6,601	0.1	21,890	740.0
DM digestibility4, %	86.6	0.5	87.6	0.3
ATTD Zn5, %	−107.4	31.2	25.6	1.3
TTTD Zn6, %	131.5	4.7	110.4	0.2
Urine excretion, kg/d	20.3	5.6	14.2	1.9
Urine Zn concentration, mg/kg	0.3	0.1	0.3	<0.1
Urine Zn excreted7, mg/d	3	0.3	4	0.3
Milk yield, kg/d	—	—	10.5	0.4
Milk Zn concentration, mg/kg	—	—	4.3	0.2
Milk Zn secreted8, mg/d	—	—	45	2.6
Total Zn excreted9, mg/d	92	6.7	220	6.1
Zn retained10, mg/d	−454	6.7	11	2.8

¹Fecal excretion = DM_intake * (Ti_diet/ Ti_feces).

²Fecal Zn excreted = Zn_feces * Output_feces.

³Fecal Ti excreted = Ti_feces * Output_feces.

⁴DM digestibility, % = [1 − (Ti_diet/Ti_feces)] * 100.

⁵ATTD Zn, % = [1 – (Zn_feces/Zn_diet) × (M_diet/M_feces)] * 100.

⁶TTTD Zn, % = ATTD + [(Zn_end/Zn_diet)*100].

⁷Urine Zn excreted = (Zn_urine * Urine_output).

⁸Milk Zn secreted = (Zn_milk * Milk_output).

⁹Total Zn excreted = (Fecal Zn_excreted + Urine Zn_excreted).

¹⁰Zn retained = (Zn_intake – Zn_excreted).

Discussion

To reduce potential environmental impacts of sow feeding programs, precise adjustments of dietary Zn supplementation must occur. However, the impact of nutritional factors that affect zinc utilization must be well recognized to adjust feeding programs accordingly. Namely, the impact of zinc source (inorganic vs. organic) and the presence of antagonists (such as dietary fiber) commonly force nutritionists to incorporate safety margins when supplementing dietary zinc. Furthermore, no studies have been published regarding the nutritional value of a Zn polysaccharide complex compared with an inorganic Zn source in diets for gestating and lactating sows fed diets with additional dietary fiber. Therefore, we determined and compared the ATTD and TTTD of Zn from sows fed diets containing a supplemental Zn polysaccharide complex and Zn sulfate (ZnSO₄).

A priori, we did not expect negative estimates for Zn digestibility in the gestation or lactation phases of this study. A definitive explanation for these unexpected results is not readily apparent. Three possible reasons for negative digestibility estimates relate to the accuracy and precision of dietary Zn analyses, use of adaptation periods of sufficient length, and presence of phytate in experimental diets.

First, the assessment of Zn digestibility using the indirect method relies heavily on accurate laboratory analysis of dietary Zn. Dietary Zn concentration has a more dramatic influence on Zn digestibility estimates than fecal Zn concentrations because one diet is fed to many animals. An incorrect determination of dietary Zn concentration will influence the digestibility estimates of many more animals than an incorrect determination of fecal Zn concentration. Although slight variation within laboratory analysis is expected, we initially observed very high and inconsistent variation among analyzed Zn concentrations of the same diet subsamples analyzed at different times. Recently, Jones et al. (2017) also observed high variation of Zn analysis within the same laboratory and estimated that at least 34 feed samples would need to be analyzed to have 95% confidence that analyzed Zn concentrations of a diet are within 4 ppm of the actual value. However, analyzing numerous feed samples (34 samples) would be very costly and impractical. To achieve accurate sample analysis results within 15 ppm of the true Zn concentration, Jones et al. (2017) recommended that two to five feed samples be analyzed. Based on our experience, we suspect that procedures employed by laboratory personnel to select the subsample for analysis and the fineness and uniformity of grinding the subsample in the laboratory also influence analyzed Zn concentrations. To control the observed variation, diet samples were homogenized and ground (1.0 mm screen) prior to submission of feed samples to the laboratory. As a result, we observed reduced variation of Zn concentration of dietary treatments.

Second, there is no widely accepted standard adaptation period or number of sample collection days for swine nutrient balance experiments (Jacobs et al., 2017). Furthermore, literature specific to diet acclimation periods for gestating sows is very limited. Gestating sows typically express superior digestibility of nutrients due to limited feed intake, increased BW, higher degradation rate of dietary fiber in the hindgut, and decreased rate of passage compared with growing-finishing pigs (Noblet and Shi, 1993, 1994; Le Goff and Noblet, 2001). These differences may influence the length of diet acclimation time necessary for sows in a balance experiment utilizing an indigestible marker, when compared with acclimation periods necessary for piglets or growing/finishing pigs that are fed amounts near ad libitum intake. Specifically, the decreased rate of passage for sows may increase the adaptation period necessary to observe a consistent plateau of indigestible marker in feces.

Finally, phytate, present in most swine feed ingredients, strongly impairs Zn absorption (Lönnerdal, 2000), but addition of the feed enzyme, phytase, can release trace elements that may be bound in feedstuffs, such as corn, wheat, soybean meal, wheat bran, and wheat middlings (Jongbloed et al., 2004; Yu et al., 2018). Dietary treatments in this experiment did not contain phytase. Therefore, the observed improvement in Zn digestibility during gestation from feeding DDGS diets may be due to reduced phytate concentrations resulting from DDGS in diets. Researchers have previously reported significant increases in ATTD of P in diets containing DDGS compared with diets without DDGS (Almeida and Stein, 2010). The improved ATTD of P in DDGS occurs when part of the phytate present in corn is hydrolyzed during ethanol and DDGS production (Stein and Shurson, 2009). We suspect that this process may similarly affect ATTD of Zn when DDGS is included in the diet. However, it is puzzling that this same effect on Zn digestibility was not observed during lactation.

Gestation

High fiber diets often reduce DM digestibility in adult sows (Le Goff et al., 2002a; Holt et al., 2006). As a result, we expected that including DDGS in the experimental diets would have an antagonistic effect on Zn digestibility in this experiment, but the opposite response was observed. The more negative ATTD, TTTD, and retention of Zn in Control diets compared with DDGS diets were not expected. Daily excretion of Zn is the sum of indigestible dietary Zn, endogenous losses of Zn, and urinary Zn excretion (a minor contributor). For reasons that are not entirely clear, daily Zn excretion exceeded dietary intake resulting in the calculated negative digestibilities and retentions observed.

Previously, scientists not only observed that the average mean retention time of feed in the gastrointestinal tract of sows was 81 h but also reported great variability with ranges from 2.8 d (68 h) to 3.9 d (95 h) (Le Goff et al., 2002b). Jo and Kim (2017) suggested that an adaptation phase of 7 d is adequate to achieve a plateau of titanium excretion for sows. Although sows consumed experimental diets for 7 d (168 h) before collection periods, sows did not begin consumption of dietary treatments with the indigestible marker until 3 d (72 h) prior to the collection period. A subset of daily fecal samples (n = 12) was analyzed for titanium concentration separately by collection day before pooling within balance period (data not shown). These data confirmed that consistent analysis of the indigestible marker was not observed until day 2 of the collection period or after 5 d (120 h) of adaptation to the indigestible marker. Collection periods within the current study were only 72 h. As a result, one may hypothesize that the adaptation and collection periods for gestating sows in this experiment were not long enough for the indigestible marker to thoroughly mix with digesta and plateau before passing through the gastrointestinal tract, causing the observed negative digestibility of Zn among treatments. We speculate that sows fed diets with higher dietary fiber, supplied by DDGS, increased retention time of feed in the gastrointestinal tract, which likely increased fermentation of dietary fiber in the cecum (Williams et al., 2007; Jha and Berrocoso, 2015). Increased retention time may explain the observed increase of Zn digestibility within DDGS gestation treatments. To clarify this issue, further investigation of rate of passage for digesta and indigestible markers in meal-fed sows with restricted intake is necessary.

Total Zn excretion was reduced (P < 0.05) for diets containing organic Zn supplemented as PSZn, which is in agreement with reduced Zn excretion reported by Nitrayova et al. (2012) when they compared responses from feeding a different organic Zn source to an inorganic Zn source. In contrast, Van Riet et al. (2016) reported that Zn source did not influence fecal Zn excretion nor absorption of Zn in gestating sows. Digestibility and bioavailability of various Zn sources among several studies (Hill et al., 1986; Wedekind et al., 1992, 1994; Close, 2003) are not consistent, yet studies conducted by researchers studying poultry and growing-finishing pigs suggest that digestibility and bioavailability of organic Zn sources are generally equal to that of ZnSO₄ (Cao et al., 2000; Lebel et al., 2014). Although this may be true for growing-finishing pigs, it may not be so for gestating and lactating sows. The improved ATTD, TTTD, and retention of Zn for sows consuming DDGS–corn–soybean meal-based diets containing polysaccharide-complexed Zn in the current study confirm that digestibility of Zn from organic Zn may actually be superior to inorganic Zn for diets fed to gestating sows containing DDGS. Furthermore, we suspect that reduced phytate due to DDGS inclusion in diets may have improved Zn digestibility in gestation diets.

The primary route of excretion for Zn is through feces (Poulsen and Larsen, 1995). Therefore, as expected, there were no effects of diet, Zn source, or their interaction on urinary Zn excretion among treatments. Urine collected from sows that exhibited signs of infection, contained blood in their urine, had a catheter that did not remain inserted, or had disconnected tubing from sample buckets were removed from the balance. The distribution of sows with successfully placed and functional catheters from the gestation balance trial were: ConZnSO₄ = 5 of 8; ConPSZn = 3 of 8; DDGS/ZnSO₄ = 6 of 8; DDGS/PSZn = 6 of 8. The distribution of sows with successfully placed and functional catheters from the lactation balance trial were: ConZnSO₄ = 4 of 6; ConPSZn = 2 of 8; DDGS/ZnSO₄ = 5 of 6; DDGS/PSZn = 5 of 9. Generally, sows maintained good health status and only 3.7% of urine samples from catheterized sows were discarded and not analyzed for Zn content. Therefore, within dietary treatment, the average urine Zn excreted from sows with successfully placed catheters was used to calculate total Zn excretion and retention for all sows.

Stage of gestation may be an important factor to consider when evaluating the digestibility and retention of Zn sources. This balance experiment was conducted from about day 89 to 91 of gestation. Often, swine producers begin to “bump feed” sows in late gestation beginning around day 90 until farrowing, to improve individual piglet birth weight (Goncalves et al., 2016). Feed intake was limited during the gestation balance trials, but not in lactation. Increased feed intake in lactation or the ability to consume feed more than once per day, a practice some producers implement in late gestation, may decrease digestibility and utilization of nutrients such as Zn (Cunningham et al., 1962; Shi and Noblet, 1993) because passage rate of digesta increases. Therefore, Zn utilization in late gestation under varying commercial practices must still be evaluated in the future.

Lactation

The observed reduction in Zn digestibility and increased fecal output from diets containing DDGS are in agreement with conclusions from previously investigated impacts of high dietary fiber on nutrient digestibility (Wenk, 2001; Agyekum and Nyachoti, 2017). ATTD and TTTD in lactating sows fed supplemental PSZn in corn–soybean meal-based diets were superior to all other dietary treatments, similar to observations from an experiment evaluating organic and inorganic Zn sources in growing pigs by Liu et al. (2014). In contrast, ZnSO₄ was more digestible than PSZn for lactating sows consuming diets containing DDGS.

We speculate there may be a threshold of dietary fiber inclusion that decreases the digestibility of organic Zn so much that the initial positive effects of PSZn in corn and soybean meal diets become negatively affected when dietary fiber concentration exceeds some threshold. Further investigation is necessary to confirm this speculation. Nonetheless, it appears that organic Zn inclusion as PSZn in corn and soybean meal-based diets had the greatest ATTD, TTTD, and overall retention of Zn for lactating sows.

Baseline Zn losses

Although we did not feed a completely Zn-free diet to truly measure endogenous losses of Zn, the negative retention of Zn observed in the gestation balance period may represent daily losses of Zn during gestation of sows fed corn–soybean meal diets with no supplemental Zn. Currently, the NRC (2012) suggests a daily Zn requirement of 210 mg/d for gestating sows. Gestating sows fed diets without supplemental Zn consumed 47.9 mg/d, yet excreted 92.5 mg/d. However, we could not determine the proportion of Zn loss from endogenous losses and the proportion that can be attributable to undigestible Zn in the feed ingredients used in the experimental diets.

We assume that baseline losses of Zn also occurred during lactation, however, these losses did not exceed dietary intake. The NRC (2012) suggests a daily Zn intake of 597 mg/d for lactating sows. Increased daily feed intake of lactating sows compared with gestating sows resulted in a greater daily Zn intake of 231.5 mg/d. Sows excreted 220.1 mg of Zn, therefore exhibiting slightly positive daily Zn retention (11.4 mg/d). Similar to gestation, we were unable to estimate the proportion of the excreted Zn that is due to poorly digestible dietary Zn and the proportion that is attributable to endogenous losses.

Daily Zn requirements for gestating and lactating sows have not been evaluated for modern sows. Without accurate estimates of Zn requirements for modern sows, nutritionists may be overfeeding or underfeeding Zn in practical situations. Overfeeding Zn increases feed costs and could potentially have detrimental effects on the environment through excess Zn concentrations of slurry that are spread on cropland. Our intent with feeding sows a diet without supplemental Zn was to quantify basal Zn losses in reproducing sows as an approach to estimate the maintenance Zn requirement of sows. Because we could not formulate a practical Zn-free diet, we used an un-supplemented diet that might suggest the magnitude of endogenous Zn losses if we would have fed a Zn-free diet. We expected negative retention of Zn for both gestating and lactating sows. The negative Zn retention for gestating sows suggests that Zn provided in typical corn–soybean meal-based diets is not sufficient to satisfy daily Zn requirements. Therefore, supplementation of Zn to these diets clearly is required to satisfy the Zn requirements of gestating sows. However, the level of Zn supplementation necessary to satisfy requirements was not an objective of this study and could not be determined in this experiment. In contrast to gestation, sows fed a corn–soybean meal-based diet in lactation with no supplemental Zn exhibited a slightly positive Zn retention. This observation suggests that the practical feed ingredients (corn and soybean meal) used in this experiment satisfied the sows’ requirement for Zn. Presumably, a much lower level of Zn supplementation would be required for lactating sows compared with gestating sows based on our findings. The important point of this work is that lower levels of Zn supplementation may be possible in lactating sow diets compared with diets for gestating sows if one is pressured by environmental concerns over Zn entering the environment from swine production.

Colostrum, milk, and plasma zinc concentrations

Mineral requirements of sows are met by providing adequate amounts of bioavailable mineral sources in the diet relative to nutritional demands during gestation and lactation. If the daily requirement exceeds dietary intake, body stores may be mobilized to satisfy any deficit until body reserves become depleted. Depleted body mineral reserves lead to reduced health and performance of sows (Mahan, 1990). Zinc requirements change throughout the reproductive cycle of sows, and as a result, many indicators of Zn status fluctuate during gestation and lactation (Van Riet et al., 2015). Several biomarkers such as: plasma, liver, and milk concentrations of Zn; metallothionein concentrations in blood; quantities of enzymes such as alkaline phosphatase and superoxide dismutase present in tissues; and concentrations of Zn in bone, kidney, and pancreas tissues may reflect Zn status of sows (McDowell, 2003). Unfortunately, there is no single, widely accepted biomarker for assessing Zn status in sows (Wood, 2000). Therefore, an assessment of a multitude of easily accessible indicators of Zn status, such as plasma and milk concentrations, and growth performance of piglets should be considered to assess the Zn status of sows.

The lack of differences in Zn concentrations among plasma, colostrum, and milk Zn samples, regardless of dietary treatment, confirms that short-term Zn status was not affected by Zn source or diet composition in this study. Sows may have mobilized Zn to maintain plasma, colostrum, and milk Zn status so that short-term Zn status did not appear to be deficient. Overall, sows seemed to maintain a relatively similar Zn status throughout collection periods within the experiment. Previous evaluation of Zn sources among sow diets have reported increased Zn concentrations in milk for sows consuming organic minerals (Acda and Chae, 2002), but this response was not observed in the current experiment.

The inability to distinguish differences in Zn concentrations of plasma among dietary treatments suggests a similar response to that of piglets and nursery pigs when fed inorganic or organic Zn sources (Case and Carlson, 2002; Schlegel et al., 2013). The lack of differences in plasma was not surprising because blood sampling occurred at only one time point within each balance period.

In conclusion, ATTD and TTTD of Zn, regardless of supplemental Zn source, improved when gestating sows consumed diets containing DDGS, that increased dietary fiber intake. Supplementing PSZn to DDGS diets improved ATTD and TTTD of Zn in gestation when compared with sows fed conventional corn–soybean meal-based diets. In contrast, lactating sows consuming diets without DDGS and supplemented with PSZn had the greatest ATTD and TTTD of Zn. Furthermore, of the DDGS-containing diets, supplemental PSZn fed to lactating sows decreased ATTD and TTTD of Zn. There appears to be greater flexibility in selecting the levels of dietary Zn supplementation for lactating sows compared with gestating sows if concerns over Zn contamination of the environment arise.

Glossary

Abbreviations

ADFI

average daily feed intake

ATTD

apparent total tract digestibility

BW

body weight

Con

control

CP

crude protein

DDGS

dried distillers grains with solubles

DM

dry matter

LowZn

low zinc treatment

M

indigestible marker

NDF

neutral detergent fiber

PSZn

polysaccharide-complexed zinc

SID

standardized ileal digestible

STTD

standardized total tract digestible

TTTD

true total tract digestibility

Conflict of interest statement

J.P.H., L.J.J, P.E.U., and G.C.S. declare no conflict of interest. J.E.G. was employed by the financial sponsor of this experiment, QualiTech, and served in a consultative role in experimental design, data analysis, and manuscript preparation.