1st Page

Journal of Animal Science - Animal Nutrition

Effects of dietary sulfur concentration and forage-to-concentrate ratio on ruminal fermentation, sulfur metabolism, and short-chain fatty acid absorption in beef heifers1


This article in JAS

  1. Vol. 92 No. 2, p. 712-723
    Received: Oct 14, 2013
    Accepted: Nov 19, 2013
    Published: November 24, 2014

    2 Corresponding author(s):

  1. S. Amat*,
  2. J. J. McKinnon,
  3. G. B. Penner and
  4. S. Hendrick 2
  1. Department of Large Animal Clinical Sciences, University of Saskatchewan, Saskatoon, SK S7N 5B4, Canada
    Department of Animal and Poultry Science, University of Saskatchewan, Saskatoon, SK S7N 5A8, Canada


This study evaluated the effects of dietary S concentration and forage-to-concentrate ratio (F:C) on ruminal fermentation, S metabolism, and short-chain fatty acid (SCFA) absorption in beef heifers. Sixteen ruminally cannulated heifers (initial BW 628 ± 48 kg) were used in a randomized complete block design with a 2 × 2 factorial treatment arrangement. The main factors included F:C (4% forage vs. 51% forage, DM basis) and the S concentration, which was modified using differing sources of wheat dried distillers grains with solubles (DDGS) to achieve low- and high-S diets (LS = 0.30% vs. HS = 0.67% S on a DM basis). Elemental S was also added to increase the S content for the HS diets. Serum sulfate concentration from blood, sulfide (S2–), and SCFA concentrations from ruminal fluid, hydrogen sulfide (H2S) concentration from the ruminal gas cap, and urinary sulfate concentration were determined. Continuous rumen pH and SCFA (acetate, butyrate, and propionate) absorption were measured. There were no interactions between S concentration and F:C. The F:C did not affect DMI (P = 0.26) or ruminal S metabolite concentrations (P ≥ 0.19), but ruminal pH was lower (P < 0.01) and SCFA absorption was greater (P < 0.01) for low F:C diets. Heifers fed HS diets had less DMI (P < 0.01) but greater ruminal pH (P < 0.01), greater concentrations of ruminal H2S (P < 0.01) and serum sulfate (P < 0.01), and greater urinary sulfate concentration (P < 0.01) and output (P < 0.01) relative to heifers fed LS diets. Ruminal H2S was positively correlated with serum sulfate (r = 0.89; P < 0.01). Ruminal acetate concentration was not affected (P = 0.26) by dietary S concentration. Heifers fed the HS diet had lower (P = 0.01) ruminal propionate concentration and tended to have lower (P = 0.06) butyrate concentration than heifers fed the LS diet. Ruminal acetate was greater (P = 0.01) and butyrate was less (P < 0.01) with the high F:C diet than the low F:C diet. Both HS (P = 0.06) and low F:C (P = 0.07) diets tended to reduce urine output. Feeding HS diets reduced SCFA absorption (P < 0.05). In summary, S metabolism in beef heifers was not influenced by the F:C, but HS reduced DMI, inhibited SCFA absorption, and increased urinary S excretion.


Feeding strategies that mitigate S effects could be a solution to help prevent polioencephalomalacia (PEM) in cattle. Adverse effect of S may be influenced by the composition of the diet (Vanness et al., 2009) with greater risk for PEM when feeding high levels of S in high-concentrate diets than in high-forage diets (Kung et al., 1998). However, recent research has indicated that high-concentrate diets may mitigate the effect of dietary S. Neville et al. (2010) reported that lambs receiving a finishing diet with 60% corn dried distillers grains with solubles (DDGS) and either 0.73% or 0.87% S (DM basis) did not exhibit any clinical signs of PEM. They suggested that this was likely due to the interactive effects of dietary S, dietary grain supplementation, and thiamine supplementation in finishing rations. In addition, Amat et al. (2012) indicated that serum sulfate concentration in feedlot steers fed a backgrounding ration (59% forage) containing 0.32% S (DM) was higher than in steers fed a finishing ration (93% concentrate) that contained 0.58% S (DM). Thus, there seems to be a discrepancy as to the influence of the forage-to-concentrate ratio (F:C) with respect to the risk for S-induced PEM.

In addition to inducing PEM, an inhibitory effect of S on acetate and butyrate oxidation in colonocytes has been documented (Roediger et al., 1993; Leschelle et al., 2005). However, the relationship between high dietary S and short-chain fatty acid (SCFA) absorption from the rumen has not been studied. Because SCFA absorption is a prominent factor regulating ruminal pH (Penner et al., 2009; Aschenbach et al., 2011) and SCFA metabolism is involved in regulating concentration gradients between the rumen fluid and cytosol to promote absorption (Gäbel et al., 2001), a reduction in epithelial SCFA metabolism, induced by high S, may decrease ruminal SCFA absorption.

Our objective was to evaluate the effects of dietary S concentration and F:C on ruminal fermentation, S metabolism, and SCFA absorption in beef heifers. Our hypotheses were 1) dietary F:C will influence S metabolism and thereby mitigate the adverse effect of S and 2) high dietary S will compromise SCFA absorption across the reticulorumen.


Cattle in this experiment were cared for in accordance to the guidelines of the Canadian Council on Animal Care (1993) under University of Saskatchewan Animal Care Protocol 20100018.

Animal Care and Experimental Design

Sixteen mixed-breed and ovariectomized heifers with ruminal cannulas (initial BW 628 ± 48 kg) were housed in individual pens (9 m2) equipped with a feed bunk and water bowl and rubber floor mats at the University of Saskatchewan Livestock Research Barn. The experiment was conducted as a randomized complete block design using a 2 × 2 factorial treatment arrangement with the main effects of dietary S content and F:C. Heifers were grouped by initial BW into 4 blocks: block 1 (564 ± 11 kg), block 2 (612 ± 23 kg), block 3 (652 ± 6 kg), and block 4 (683 ± 19 kg). Within each block, heifers were randomly assigned to 1 of 4 treatments that differed in F:C and dietary S concentration. The ingredient and chemical composition of the experimental treatments is shown in Table 1. The F:C was modified by altering the proportion of barley silage (4% vs. 51%, DM basis), whereas the S content was modified using 2 different wheat DDGS containing 0.55% or 1.07% S (DM basis) at approximately 38% of diet DM to achieve low- and high-S diets (0.3% [low S; LS] vs. 0.67% S [high S; HS] on DM basis). To achieve HS diets with 0.67% S (DM), elemental S (catalog number 13803, Sigma-Aldrich, Oakville, ON, Canada) was added to the HS supplements.

View Full Table | Close Full ViewTable 1.

Ingredient composition and chemical analysis of treatment diets

Treatment diets1
Diet composition, % DM
Barley silage 3.7 3.7 50.6 51.0
Barley grain 53.0 53.0
DDGS-TG2 37.8 39.5
DDGS-NWT3 37.4 39.1
Supplement 1 5.6
Supplement 2 5.6
Supplement 3 9.9
Supplement 4 9.9
Supplement composition, % DM
Barley grain 34.4 36.7 57.7 58.7
Canola oil 3.2 3.2 3.9 3.9
Limestone 35.1 35.2 15.0 15.0
Ionophore/thiamine premix4 8.5 8.6 12.8 12.8
Trace mineral salt5 4.9 4.9 2.8 2.8
LS1066 11.4 11.4 6.9 6.9
Elemental sulfur 2.5 1.0
Chemical composition (mean ± SD), % DM
CP 20.8 ± 0.35 20.7 ± 0.19 21.8 ± 0.35 21.8 ± 0.29
Starch 34.1 ± 0.85 34.1 ± 0.73 16.4 ± 1.17 17.0 ± 0.85
ADF 9.2 ± 0.70 9.4 ± 0.51 20.4 ± 1.15 20.6 ± 0.20
NDF 25.6 ± 0.54 27.4 ± 0.65 36.5 ± 0.37 38.4 ± 1.25
Calcium 0.67 ± 0.17 0.76 ± 0.02 1.14 ± 0.26 0.96 ± 0.04
Phosphorus 0.56 ± 0.03 0.56 ± 0.00 0.57 ± 0.04 0.57 ± 0.02
Sodium 0.27 ± 0.05 0.26 ± 0.00 0.44 ± 0.07 0.41 ± 0.01
Potassium 0.87 ± 0.01 0.87 ± 0.00 1.59 ± 0.05 1.59 ± 0.05
Sulfur7 0.59 ± 0.05 0.29 ± 0.04 0.64 ± 0.03 0.33 ± 0.02
1Treatment diets: LF:C-HS, low forage-to-concentrate ratio (F:C), high S; LF:C-LS, low F:C, low S; HF:C-HS, high F:C, high S; HF:C-LS, high F:C, low S.
2DDGS-TG: high-S-containing wheat dried distillers grains with solubles (1.07% S DM basis).
3DDGS-NWT: low-S-containing wheat dried distillers grains with solubles (0.55% S DM basis).
4Ionophore/thiamine premix of supplements 1 and 2 contains 95% barley and 3% Rumensin premix (as monensin sodium at 200 g/kg; Elanco, Guelph, ON) and 2% thiamine (DM basis); ionophore/thiamine premix of supplements 3 and 4 contains 98% barley and 1% Rumensin premix and 1% thiamine (DM basis).
5Trace mineral salt: 95% NaCl, 12,000 mg/kg Zn, 10,000 mg/kg Mn, 4,000 mg/kg Cu, 400 mg/kg I, 60 mg/kg Co, and 30 mg/kg Se.
6University of Saskatchewan vitamin A & D supplement contains 440,500 IU vitamin A and 88,000 IU vitamin D3/kg.
7Sulfur contents of barley silage and barley grain were 0.23% and 0.13% (DM), respectively. Sulfur content of supplements 1, 2, 3, and 4 were 1.93%, 0.10%, 1.03%, and 0.11% (DM basis), respectively.

Heifers were fed their individual diets in equal portions at 0800 and 1600 h. Diets were formulated to meet or exceed the NRC (2000) nutrient requirements for CP, energy, and trace minerals (Table 1). The trial included a 31-d adaptation period and a 37-d experimental period. During the adaptation period, a 7-step dietary transition protocol was used to allow each heifer on the low F:C treatment to adapt, whereas a 4-step dietary transition protocol was used for heifers fed the high F:C treatments. During this adaptation period, low-S-containing DDGS and pellets containing no elemental S were fed. There were 4 d between each step and 7 d between the final step and the trial diets. Orts were weighed and recorded daily before the morning feeding and then discarded.

Barley silage (AC Rosser) used in the study was grown at the University of Saskatchewan farm, harvested, and stored in a bunker silo. Barley grain (605 g/L) was purchased from commercial sources and dry rolled (Ross Kamp Champion, Waterloo, IA). High-S-containing wheat DDGS (1.07% S) was supplied by Terra Grain Fuels (Belle Plaine, SK), whereas low-S-containing wheat DDGS (0.55% S) was purchased from North West Terminal Ltd. (Unity, SK). Thiamine was added to all diets at 20 mg/kg (DM) in each of the supplements.

Heifers were examined twice daily for any signs of neurological disease, including head pressing against the fence, hypersensitivity to sound or touch, teeth grinding, staggering, and blindness. One heifer on the high F:C and high-S diet from block 1 was diagnosed with a bladder infection not related to dietary treatment after 2 wk of the experiment. This heifer was treated and recovered; however, no urine, blood, ruminal fluid, or gas cap samples were collected on sampling d 21.

Sample Collection and Laboratory Analysis

Sampling Regimen.

Blood, ruminal fluid, ruminal gas cap, and urine samples were collected on d 1, 7, 10, 14, 21, 28, and 35 of the experimental period. Blood and ruminal samples were collected simultaneously at 0600, 1200, 1800, and 2400 h, ruminal gas cap samples were collected at 1200 h, and urine samples were collected over a 24-h period. A staggered sampling approach was used such that samplings from blocks 2, 3, and 4 were started 2, 10, and 12 d after the first block.

Feed and Water Sampling and Analysis.

Feed samples were collected for each batch of barley, supplement pellets, and wheat DDGS. Silage samples were collected every week to determine DM content, and the diets were adjusted, if necessary. All feed samples were dried in a forced-air oven at 55°C for 72 h and then ground through a hammer mill fitted with a 1-mm screen (Christy & Norris 8” Lab Mill, Christy Turner Ltd., Chelmsford, UK). Feed samples were analyzed for CP, starch, ADF, NDF, Ca, P, Na, K, and S using wet chemistry methods (Cumberland Valley Analytical Services Inc., Maugansville, MD; Cumberland Valley Analytical Services, 2013). Water samples from water bowls were collected weekly, composited on an equal-volume (20 mL) basis, and stored at –20°C for sulfate testing. Water sulfate was determined using inductively coupled plasma atomic emission spectroscopy by the Saskatchewan Research Council Analytical Laboratory (Saskatoon, SK, Canada).

Ruminal Gas Cap and Ruminal Fluid Sampling and Analysis.

To maintain the integrity of the ruminal gas cap, a cannula plug was modified to allow simultaneous ruminal gas cap and fluid sampling from cannulated cattle without opening the cannula (Fig. 1). Cattle were also fitted with a ruminal cannula (9C; Bar Diamond Inc., Parma, ID), which has previously been shown to minimize gas leakage from the cannula (Beauchemin et al., 2012). This modified cannula plug was placed in each heifer 24 h before ruminal gas cap sampling. At 1200 h, the ruminal gas cap was sampled before sampling the ruminal fluid. Ruminal gas cap sampling and measurements were adapted from Neville et al. (2010) with the following modifications. Two 120-mL ruminal gas cap samples were drawn into 140-mL syringes. Ruminal H2S concentration was measured via precision H2S gas detector tubes attached to a calibrated gas detection pump (model AP-20S, Sensidyne, Clearwater, FL) within 1 h after sampling. Two types of H2S detector tubes differing in the range of measurement (Sensidyne numbers 120SM and 120SF) were used for the samples collected from the HS cattle and LS cattle, respectively. The concentration of H2S was recorded from the detector tube by the same individual. Duplicate measurements were taken for each heifer, and the average of the 2 samples was used for data analysis.

Figure 1.
Figure 1.

Ruminal fluid and ruminal gas cap sampling apparatus. A, cannula plug (1EZ Easy-out Stopper, Bar Diamond Inc., Parma, ID); B, ruminal fluid collection tube (Bar Diamond, reference 33055, DESC-1); C, 102-mm 14-gauge needle; D, 8-cm-long (4.8-mm-diam.) tubing; E, plastic tubing clamp; F, female luer lock catheter end; G, Argyle intermittent infusion plug (Covidien, Saint-Laurent, QB, Canada, reference 115006); H, 140-mL syringe used for ruminal fluid collection; I, 140-mL syringe used for rumen gas cap collection; J, 3-way stopcock (Smith Medical, Markham, ON, Canada, reference MX5311l); K, 15-cm-long (4.8-mm-diam.) tubing used for sucking the gas from the rumen; L, 15-cm-long (4.8-mm-diam.) tubing used for sucking the gas from the syringe to the pump; M, gas detection pump (model AP-20S, Sensidyne, Clearwater, FL); N, precision hydrogen sulfide gas detector tube (Sensidyne, number 120SM or 120SF); O, 6-cm-long (4.8-mm-diam.) tubing with female luer lock catheter end used for connecting detector tube with the tubing of syringe. See online version for figure in color.


Following ruminal gas cap sampling, ruminal fluid was sampled using a 140-mL syringe. The first 10 to 15 mL of ruminal fluid was discarded, and then 100 mL of ruminal fluid were collected and transferred to a 250-mL plastic beaker. A 30-mL ruminal fluid sample was collected for the measurement of ruminal S2–. Rumen S2– was measured between 2 and 3 h after sampling according to Khan et al. (1980) with the following modifications. An Orion Dual Star pH/Ion Selective Electrodes (ISE) meter (Thermo Scientific, Mississauga, ON, Canada, catalog number 100–240V) was used to analyze the ruminal S2–. Ruminal fluid (10 mL) was placed in a 50-mL beaker and mixed with an equal amount of sulfide antioxidant buffer (SAOB). The SAOB buffer contained 198.7 mM ascorbic acid, 180 mM Na2 EDTA, and 2 M NaOH. After the stabilization of the S2– ion by the SAOB, the S2– ion was measured with a sulfide ion selective electrode (Orion ion silver/sulfide electrode BN, Thermo Scientific, catalog number 9616 BNWP). A 10-mL sample of ruminal fluid was also collected and preserved with 2 mL of 25% (wt/vol) metaphosphoric acid. The ruminal fluid was stored at –20°C until analyzed for SCFA concentration as described by Khorasani et al. (1996).

Blood Sampling and Analysis.

Blood samples were collected from the jugular vein of all heifers via temporary vinyl catheters (0.86 mm i.d. × 1.32 mm o.d.; Scientific Commodities Inc., Lake Havasu City, AZ) that were inserted 1 d before sampling. Each catheter was flushed with 10 mL of heparinized saline (20 IU of heparin/mL of saline). Blood (10 mL) was collected with tubes containing no anticoagulant (BD Vacutainer, Mississauga, ON, Canada, reference 367820) and centrifuged (15 min at 2,526 × g at room temperature). Serum was transferred to 4-mL cryovials and stored at –80°C for sulfate analysis. Serum samples at 1200 and 2400 h sampling points were analyzed for sulfate with high-performance ion chromatography (HPIC) according to Russo and Karmarker (1998) at the Veterinary Diagnostic Laboratory of Colorado State University (Fort Collins, CO).

Urine Sampling and Analysis.

Total urine was collected each sampling day for determination of urine output and sulfate concentration. Urine was collected with bladder catheters (Bardex Foley Catheter, 75 mL capacity balloon; C. R. Bard Inc., Covington, GA). Bladder catheters were inserted 24 h before the first 24-h collection and remained in place throughout the trial. During collection, the heifers were haltered with enough space for their normal activities. Total urine was collected in 25-L Nalgene jugs with sterile Nalgene tubing, and the weight was recorded daily. Urine was subsampled and transferred to a 15-mL centrifuge tube and frozen at –20°C for sulfate analysis. Urine sulfate was analyzed with HPIC according to Magee et al. (2004) at the Veterinary Diagnostic Laboratory of Colorado State University (Fort Collins, CO).

Indwelling Ruminal pH Measurement.

On d 3, 12, 23, and 32 indwelling ruminal pH probes (model LRCpH; Dascor, Escondido, CA) were placed in the ventral sac and programmed to record pH at 1-min intervals over a 24-h duration (Penner et al., 2006). After 24 h, the ruminal pH probes were removed from the ruminal, the data were recorded, and the probe was restandardized in pH buffers 4 and 7. The millivolt data were combined with the preinsertion and postmeasurement standardization data to calculate pH on the basis of the slope and y-intercept values. Ruminal pH data were used to summarize minimum, mean, and maximum pH. The duration (min/d) and pH area (min/d × pH) below the threshold of 5.5 were also calculated to summarize the severity of ruminal pH depression.

Measurement of SCFA Absorption.

The temporarily isolated and washed reticulorumen (WRR) technique was performed on all heifers to measure SCFA absorption across the reticulorumen epithelium (Care et al., 1984; Zhang et al., 2013). The WRR procedure was performed on 2 heifers from each block at 1300 h on d 36, and the remaining 2 heifers from the same block were subjected to the WRR at 1300 h on the following day (d 37). Briefly, the reticulorumen digesta were completely removed from the rumen and stored in an insulated covered container until the end of experiment when the digesta were returned to the reticulorumen. The reticulorumen was washed twice using 10 L of preheated tap water (38°C). After washing, the excess water was removed using a wet/dry vacuum. Then, 20 L of preheated (38°C) washing buffer (Table 2) was used to wash (5 to 6 L/wash) the rumen. Washing buffer was poured into the rumen and agitated manually to dislodge material from the rumen epithelium. After each wash, the buffer was removed using the wet/dry vacuum, and the procedure was repeated until the effluent was free of digesta. After the reticulorumen was completely washed, the reticulorumen was isolated from the rest of the gastrointestinal tract. The esophagus was temporarily occluded using a saliva collection device (University of Leipzig, Leipzig, Germany). The saliva collection device was connected to a vacuum pump (model N86KT45P, KNF Neuberger Inc., Trenton, NJ) and thus allowed the continuous removal of saliva. A Foley catheter (75-mL volume) was used to occlude the omasal orifice. After occluding the esophagus and omasal orifices, the reticulorumen was washed again with washing buffer (5 L) to remove any saliva that was secreted during placement of the occluding devices. Fifteen liters of incubation buffer containing a fluid marker (2 mM Co-EDTA; Table 2), warmed to 38°C and adjusted to pH 6.2 was poured into the washed reticulorumen and continually gassed with CO2. Samples were collected at 0 (before infusion), 5, and 45 min after the incubation buffer was introduced into the reticulorumen. Two buffer samples (15 and 30 mL) were collected at each sampling time and were immediately stored at –20°C for SCFA and Co analysis, respectively. Short-chain fatty acid concentration was analyzed using GLC by the Ruminant Nutrition Laboratory of Agriculture and Agri-Food Canada (Lethbridge, AB, Canada). Cobalt analysis was analyzed using inductively coupled plasma mass spectrometry (Thermo Jarrel Ash-Corporation, Franklin, MA) by the Toxicology Laboratory of Prairie Diagnostic Services Inc. (Saskatoon, SK, Canada). The osmolality of the experimental buffer (before infusion) was measured in duplicate using an osmometer (model 3250, Advanced Instruments Inc., Norwood, MA). The initial and final volumes were calculated by the concentrations of Co at the 5- and 45-min sampling points relative to the actual Co concentration in the 0-min sample. Rates of SCFA absorption were calculated according to Gäbel et al. (1991).

View Full Table | Close Full ViewTable 2.

Chemical composition of buffers used for washing the reticulorumen and for measurement of short-chain fatty acid (SCFA) absorption across the reticulorumen epithelia

Chemical Washing buffer,1 mM Incubation buffer,1,2 mM
NaCl 105 5
KCl 5
CaCl2 2
MgCl2 2
Na acetate 30
K acetate 35
Na propionate 20 35
Na butyrate 8
Butyric acid 7
NaHCO3 25 25
l-Lactic acid 5
Na2S 0.2
Acetic acid 10
1The pH of the buffer was adjusted to 6.2.
2Osmolality ± SD of the experimental buffer was 279 ± 9.7 mOsmol/kg.

Statistical Analysis

All data, except those of SCFA, were analyzed as a randomized complete block design (RCBD) with repeated measurements accounting for a 2 × 2 factorial treatment arrangement using the Mixed Model procedure of SAS (version 9.2; SAS Inst. Inc., Cary, NC). Because no significant block effect was observed, block was removed from the model, and data were reanalyzed with the fixed effects of F:C, S, F:C × S, day, F:C × day, S × day, and F:C × S × day. The covariance error structure for each model was chosen on the basis of that producing the lowest Akaike information criterion value. Short-chain fatty acid data were first analyzed as RCBD with 2 × 2 factorial treatments. After observing no significant block effect, the block was removed from the model. The model used for SCFA analysis included fixed effects of F:C, S, and F:C × S. Correlations between ruminal H2S and serum sulfate were evaluated using the Proc Reg procedure of SAS. The Satterthwaite method was applied to determine degrees of freedom. Means were compared using Tukey’s multiple comparison test. Results were considered significant when P ≤ 0.05, and trends are discussed when 0.05 < P ≤ 0.10.


Dry Matter and Sulfur Intake

Actual S concentration in LS and HS diets averaged 0.31% and 0.62%, respectively (Table 1). The actual mean S concentration in HS diets was 0.05% lower than the formulated value (0.67%). This discrepancy was a result of the slightly lower S content of the supplements to which elemental S (average of 0.12%, DM) was added during pelleting.

No interaction was observed between dietary S and F:C for DMI (P = 0.87; Fig. 2). Dry matter intake was reduced by 22.5% with high dietary S (P < 0.001) but was not affected by F:C (P = 0.26). Uwituze et al. (2011) observed lower DMI in steers fed a corn-DDGS-based finishing diet containing high S (0.65%, DM) relative to steers fed low S (0.42% DM). Likewise, for steers fed a finishing diet containing 48% corn and 40% DDGS, Richter et al. (2012) observed steers fed diets containing 0.5% to 0.6% S tended to have lower feed intake than those fed diets with 0.2% to 0.3% S. In contrast, Boila and Golfman (1991) did not observe reduced feed intake in Holstein steers fed a finishing diet containing 0.39% S compared to control steers fed a diet with 0.19% S. This might be because the concentration of S (0.39%) in that diet was under the maximum tolerable level (0.4%) as set by NRC (2000) and therefore not high enough to exert an effect on feed intake.

Figure 2.
Figure 2.

The effects of dietary sulfur concentration (high S [HS] vs. low S [LS]) and forage-to-concentrate ratio (low F:C [LF:C] vs. high F:C [HF:C]) on DMI (kg/d) of beef heifers. Effects of dietary S (P < 0.001), F:C (P = 0.26), their interaction (P = 0.87), week of sampling (P = 0.002), week × S interaction (P = 0.66), week × F:C interaction (P = 0.15), and week × S × F:C interaction (P = 0.34). Error bars represents SEM.


The adverse impact of high dietary S on feed intake has been proposed to be due to suppressed ruminal motility. High dietary S results in elevated ruminal H2S production, which inhibits rumen motility (Bird, 1972). In the present study, HS diets yielded greater (P < 0.01) ruminal H2S concentrations relative to LS diets, further supporting this theory (Table 3). A negative correlation between ruminal H2S concentration and DMI has also been shown by Sarturi et al. (2011) and Uwituze et al. (2011). Reduced DMI in heifers fed HS diets in the current study could also be associated with the palatability of the wheat DDGS. The wheat DDGS offered to HS heifers contained almost twice the S of the low-S-containing DDGS (1.07% vs. 0.55% of DM; Table 1). Elevated S content of the high-S-containing DDGS is in part due to sulfuric acid added during the fermentation process at the particular plant of origin, which may reduce its palatability. A negative impact of S concentration in corn DDGS on feed intake of feedlot steers was observed by Sarturi et al. (2011). They compared the effect of 2 different S-containing (0.82% vs. 1.16%) DDGS sources on feed intake and found high-S-containing DDGS resulted in lower DMI relative to low-S-containing DDGS.

View Full Table | Close Full ViewTable 3.

The effect of dietary sulfur concentration and forage-to-concentrate ratio (F:C) on ruminal pH, ruminal short-chain fatty acid (SCFA) concentrations, and sulfur metabolites in beef heifers

Dietary treatments1
Item LF:C-HS LF:C-LS HF:C-HS HF:C-LS SEM F:C S F:C × S D2 F:C × D S × D F:C ×S × D
Ruminal pH
    Minimum pH 5.53 5.49 6.09 5.81 0.087 <0.01 0.08 0.17 0.86 0.02 0.89 0.43
     6.11 5.92 6.43 6.17 0.060 <0.01 <0.01 0.60 0.39 <0.01 0.04 0.18
    Maximum pH 6.72 6.54 6.84 6.64 0.053 0.05 <0.01 0.90 0.65 0.02 0.68 0.21
    Duration pH < 5.53 60.0 268.7 3.0 7.9 60.2 0.02 0.10 0.11 0.01 0.02 0.05 0.1
    Area pH < 5.54 7.12 49.30 0.48 0.34 12.17 0.04 0.11 0.11 0.01 0.01 0.01 0.01
Total SCFA,5 mM 150.0 156.3 143.8 169.5 7.86 0.77 0.05 0.18 0.85 0.35 0.51 0.79
Acetate, mM 54.4 54.1 61.1 68.9 3.21 0.01 0.26 0.24 0.95 0.03 0.72 0.79
Propionate, mM 27.3 30.5 21.7 30.2 1.89 0.14 0.01 0.19 0.39 0.26 0.05 0.36
Butyrate, mM 17.8 19.5 10.6 13.9 1.21 <0.01 0.06 0.53 0.02 0.46 0.40 0.88
H2S, g/m3 3.46 0.44 3.44 0.96 0.260 0.34 <0.01 0.31 0.22 0.19 0.42 0.71
S2–, µM 2.98 1.29 1.41 0.51 0.27 <0.01 <0.01 0.19 0.01 0.11 0.34 0.22
Serum sulfate, mg/L 174.0 141.7 169.0 135.2 3.67 0.13 < 0.01 0.81 0.11 0.48 0.69 0.90
1Dietary treatments: LF:C-HS, low F:C, high S; LF:C-LS, low F:C, low S; HF:C-HS, high F:C, high S; HF:C-LS, high F:C, low S.
2Day of sampling effect.
3Duration < 5.5 = min pH below 5.5 over 24 h.
4Area < 5.5 = min × pH below 5.5 over 24 h.
5Total SCFA = acetate + propionate + isobutyrate + butyrate + isovalerate + valerate + isocaproate + caproate.

Although the HS diets reduced DMI, daily S intake was greater (P < 0.01) for heifers fed HS diets relative to those fed LS diets (53.0 vs. 34.5 g/d, respectively; Fig. 3). As expected, no difference (P = 0.53) in daily S intake was observed between heifers fed high and low F:C diets (44.4 and 43.0 g/d, respectively). Water S intake was not measured or estimated; however, the sulfate content of the water used in this trial (110 mg/L) was low (Olkowski, 2009).

Figure 3.
Figure 3.

The effects of dietary sulfur concentration (high S [HS] vs. low S [LS]) and forage-to-concentrate ratio (low F:C [LF:C] vs. high F:C [HF:C]) on S intake (g/d) of beef heifers. Effects of dietary S (P < 0.001), F:C (P = 0.53), their interaction (P = 0.52), week of sampling (P = 0.003), week × S interaction (P = 0.97), week × F:C interaction (P = 0.12), and week × S × F:C interaction (P = 0.36). Error bars represents SEM.


Ruminal pH

There were no interactions (P > 0.11) between dietary S and F:C for any ruminal pH variables (Table 3). Mean ruminal pH was influenced by both dietary S (P = 0.01) and F:C (P < 0.01). Heifers fed low F:C diets exhibited lower (P < 0.01) mean ruminal pH relative to those fed high F:C diets (6.02 vs. 6.30, respectively). Heifers fed HS diets had higher (P < 0.01) maximum rumen pH and tended to have higher (P = 0.08) minimum ruminal pH. Both minimum (P < 0.01) and maximum (P = 0.05) rumen pH were lower for cattle fed low F:C diets. Dietary S tended to decrease the duration (min/d; P = 0.10) that ruminal pH was below the threshold of 5.5; however, duration (P = 0.02) and area (P = 0.04) were increased by low F:C diets. Despite the impact of day of study on DMI, there was no effect of time on minimum, mean, and maximum ruminal pH (P > 0.39).

An increase in rumen pH for steers exposed to high dietary S (0.65% vs. 0.42%) was observed by Uwituze et al. (2011). They speculated that the increase in ruminal pH was due to reduced feed intake, which is further supported by the data from the present study, as heifers fed HS diets had reduced feed intake. Feed intake is a primary factor influencing ruminal pH (Nordlund, 2003), primarily through ruminal SCFA production. Although SCFA concentration is not truly representative of production, Uwituze et al. (2011) reported that increasing the dietary S concentration decreased SCFA concentration, further supporting the notion that the effect of dietary S is mediated via changes in DMI.

The duration and area that ruminal pH is below the threshold value of 5.5 can be used as an indicator of ruminal acidosis (Penner et al., 2007). The mean ruminal pH in all treatments was above 6.0; however, heifers fed low F:C and LS diets experienced pH values below pH 5.5 for 2.7 and 2.3 h/d, respectively. Heifers fed high F:C and HS diets did not experience ruminal acidosis.

Ruminal H2S and Sulfide and Serum Sulfate Concentration

Results of ruminal H2S and S2– and serum sulfate concentrations are shown in Table 3. All three S metabolites were greater for heifers fed HS than for those fed LS (P < 0.01), but only ruminal S2– was influenced by F:C (P < 0.01). There was no significant interaction (P > 0.19) between dietary S and F:C for these S metabolites.

High-S diets produced about a 3.9-fold greater ruminal H2S gas concentration relative to LS diets (P < 0.01), with LS heifers having ruminal H2S concentrations lower than 0.75 g/m3 Gould et al. (1997) found similar ruminal H2S concentrations in cattle fed low-S-containing finishing diets. The mean ruminal H2S concentration for heifers fed HS diets was 3.45 g/m3, with values exceeding 1.50 g/m3 being potentially toxic and those exceeding 3.00 g/m3 increasing the risk for the development of PEM (Gould et al., 1997). Despite the high ruminal H2S observed in HS heifers in the present study, there were no clinical signs of PEM observed. Gross and histopathological examination revealed that there were no gross or microscopic changes indicative of PEM in the brain of HS heifers (Amat et al., 2013a). Likewise, Neville et al. (2010) reported that lambs exposed to elevated dietary S (0.65% or 0.83%, DM) exhibited relatively high ruminal H2S gas ranging from 3.00 to 12 g/m3 but did not observe any clinical signs of PEM. The same authors also did not observe any clinical signs of PEM in cattle fed finishing diets containing 20% to 60% DDGS with S concentrations ranging from 0.6% to 0.9% (DM), despite significantly higher ruminal H2S concentrations (Neville et al., 2012). Taken together, ruminal H2S does not seem to be a good indicator for assessing the risk of PEM in cattle. Ruminal H2S concentrations did not vary (P = 0.34) between low F:C and high F:C diets. Hydrogen sulfide generation in the rumen is pH dependent. When ruminal pH is low, more H+ is expected to be available to protonate S2– and form H2S (Gould, 1998). Although heifers fed low F:C diets had lower ruminal pH, mean pH only differed by 0.28 units from LS heifers. This pH difference may not be large enough to affect ruminal H2S concentration.

Ruminal S2– concentration was increased (P < 0.01) by low F:C and did not differ with sampling time (P = 0.11, data not shown) within the day but was different between sampling days (P = 0.01). It was expected that high dietary S would increase ruminal S2–, as all S compounds in the rumen are metabolized to sulfide (Lewis, 1954; Olkowski, 1992). Felix and Loerch (2011) found ruminal S2– concentrations were greater for steers fed 0.43% S compared to those fed 0.25% S (DM).

Unlike the other S metabolites evaluated, ruminal S2– was increased by the high-grain diet (P < 0.05). This may be due to reduced ruminal pH, which provides more favorable conditions for sulfate-reducing bacteria to produce more sulfides in the rumen (Gould, 1998).

It is apparent that overall ruminal sulfide concentration was relatively low compared with other studies (Gould et al., 1991; Suttle, 2010; Felix and Loerch, 2011). The minimum sulfide level required by microorganisms in the rumen is reported to be about 0.03 mM (Suttle, 2010). Our maximum ruminal sulfide concentrations were lower than this critical value and were also much lower than S levels in cattle fed a diet containing 0.25% S (Felix and Loerch, 2011). The reason for these lower ruminal S2– concentrations could be associated with the time of sulfide measurement after sample collection. We measured ruminal samples for S2– with a device similar to that of Gould et al. (1991) and Felix and Loerch (2011); however, in the present study, ruminal S2– was measured 2 to 3 h after sample collection, whereas others measured S2– within 2 min after collection. Ruminal S2– has been reported to be unstable and can easily be converted to H2S gas or incorporated into S-containing AA by ruminal microorganisms; therefore, in the current study, S2– may have been lost during the time between collection and analysis.

Sulfate is the end product of sulfide metabolism; thus, it was hypothesized that measuring serum sulfate would be an effective means of identifying cattle at risk of PEM. Serum sulfate concentration was 24% greater (P < 0.01) for HS heifers compared with LS heifers. The relationship between serum sulfate and S intake in the present study is in agreement with our previous study in which serum sulfate levels in cattle fed DDGS-based backgrounding or finishing diets reflected dietary S intake (Amat et al., 2012). It has also been reported that serum sulfate levels in cattle increased with increased S intake from feed (Weir and Rendig, 1954) or water (Hansard and Mohammed, 1969). Thus, serum sulfate level in cattle reflects the dietary S intake.

Surprisingly, serum sulfate levels were not affected by the F:C (P = 0.13). This is not in agreement with the results of our previous study (Amat et al., 2012), where steers fed finishing diets based on 40% corn DDGS or wheat DDGS (average 0.58% S) had lower serum sulfate levels compared to steers fed backgrounding diets based on 17% corn DDGS or wheat DDGS (average 0.32% S). One potential explanation for such a discrepancy with regard to the effect of diet on serum sulfate between these 2 studies may be due to the different mineral concentrations in the diet resulting from different inclusion rates of DDGS. In the present study, DDGS levels in both types of diets were similar (∼38%). However, DDGS in the finishing diet in Amat et al. (2012) was 23% higher than in the backgrounding diet. Thus, the mineral concentrations in the finishing diets would presumably be higher than in the backgrounding diet because DDGS have higher concentrations of minerals such as Mg, Mo, Cu, Se, and Zn than barley grain or barley silage (Amat et al., 2013b). Elevating mineral concentrations in the diet may increase the potential for the formation of biologically unavailable complexes such as CuS, thiomolybdate, MgSO4, or other unknown complexes (Cammack et al., 2010; Gooneratne et al., 2011). As a result, more of the dietary S would be affected by these interactions and become biologically less available, which may subsequently result in lower serum sulfate levels.

Serum sulfate was positively correlated with the ruminal H2S (r = 0.89; P < 0.01). Because PEM was not observed in the present study, it was not possible to correlate serum sulfate level with the occurrence of PEM. Future work should focus on establishing a correlation between serum sulfate and PEM.

Ruminal SCFA Concentration

There was no significant interaction (P > 0.18) between dietary S and F:C for the total SCFA concentration or the concentrations of acetate, propionate, and butyrate. The total SCFA concentration was lower (P = 0.05) for heifers fed the HS diet than for those fed the LS diet but was not affected by F:C (P = 0.77; Table 3). Acetate concentration was not affected (P = 0.26) by dietary S, but it was greater (P = 0.01) for heifers fed the high F:C diet compared to the low F:C diet. Heifers fed the HS diet had a lower (P = 0.01) propionate concentration and tended to have a lower (P = 0.06) butyrate concentration than heifers fed the LS diet. The F:C ratio did not affect the propionate concentration (P = 0.14), but the butyrate concentration was greater for the low F:C diet than for the high F:C diet (P < 0.01).

Dietary S concentration may decrease ruminal SCFA concentration (Uwituze et al., 2011). In the present study, the total ruminal SCFA was reduced with the greater dietary S concentration. The effect of dietary S concentration on ruminal acetate concentration seems to be controversial. Like us, Sarturi et al. (2011) observed no difference in ruminal acetate concentration between steers fed a finishing diet containing 40% DDGS with 0.54% vs. 0.40% S. In contrast, greater ruminal acetate concentration was observed in steers fed 0.65% vs. 0.42% S in diets based on dry rolled corn (Uwituze et al., 2011), but no effect of high dietary S concentration was noted when diets were based on steam-flaked corn. Thus, the effect of dietary S concentration on ruminal acetate concentration may be influenced by the nature of the diet. In our study, we observed lower propionate and butyrate concentrations in the rumen when the HS diet was fed. Such a detrimental impact of dietary S concentration on ruminal propionate and butyrate concentrations was also observed by other studies (Uwituze et al., 2011; Sarturi et al., 2011).

Urine and Urinary Sulfate Output

Daily urine output tended to be reduced by HS (P = 0.06) and low F:C (P = 0.07; Table 4). In contrast, Neville et al. (2011) reported that the urinary output in lambs increased linearly with increasing DDGS and S intake. They ascribed this increase to increased water intake and proposed that the increased water intake resulted from elevated S intake. Although water intake was not measured in the present study, it can be seen from the urinary output results that dietary S did not have a major impact on water intake. On the basis of our results obtained from heifers fed diets with similar levels of DDGS but HS or LS, it can be concluded that dietary S does not seem to increase urinary output.

View Full Table | Close Full ViewTable 4.

Effects of dietary sulfur concentration and forage-to-concentrate ratio (F:C) on daily urine and urinary sulfate output of beef heifers

Dietary treatments1
Item LF:C-HS LF:C-LS HF:C-HS HF:C-LS SEM F:C S F:C × S D2 F:C × D S × D F:C ×S × D
Urine output, L/d 9.6 13.2 13.2 15.0 1.29 0.07 0.06 0.52 1.00 0.20 0.62 0.70
Urinary sulfate, g/L 8.49 3.38 6.58 3.17 0.51 0.06 <0.01 0.11 0.07 0.19 0.42 0.30
Urinary sulfate output, g/d 80.6 44.6 86.4 47.8 5.78 0.49 <0.01 0.82 0.96 0.29 0.59 0.70
1Dietary treatments: LF:C-HS, low F:C, high S; LF:C-LS, low F:C, low S; HF:C-HS, high F:C, high S; HF:C-LS, high F:C, low S.
2Day of sampling effect.

Urinary sulfate concentration in HS heifers was 131% greater and urinary sulfate output was 180% greater than those of LS heifers. Heifers fed low F:C diets tended to have greater (P = 0.06) urine sulfate concentrations compared with those fed high F:C diets, but urinary sulfate output was not influenced by F:C (P = 0.49). Sulfur excretion from the body is primarily accomplished through urine (Underwood and Smitasiri, 1999), with sulfate being the major form (Suttle, 2010). Doyle and Moir (1979) observed a similar influence of dietary S on urinary S excretion in ruminants as in the present study. Neville et al. (2011) reported that lambs fed 0.84% S excreted 480% more S in the urine relative to those fed 0.22% dietary S. Our results support the mechanism suggested by Neville et al. (2011), who proposed that ruminants on high-S diets have the ability to avoid S toxicity by increasing urinary S excretion.

Short-Chain Fatty Acid Absorption

Fractional rates of acetate, propionate, and butyrate absorption were affected by both dietary S (P < 0.05) and F:C (P < 0.01; Table 5). High-S diets led to lower fractional rates of acetate, propionate, and butyrate absorption by 40%, 39%, and 36%, respectively, compared to LS heifers. On the other hand, low F:C diets led to greater (P < 0.01) fractional rates of SCFA absorption. The overall fractional rate for total SCFA absorption with the low F:C diet was 54% greater relative to high F:C diets.

View Full Table | Close Full ViewTable 5.

The effect of dietary sulfur concentration and forage-to-concentrate ratio (F:C) on short-chain fatty acid absorption in beef heifers

Dietary treatments1
Fractional rate of absorption, %/h
    Acetate 32.0 50.8 14.2 25.6 6.1 <0.01 0.04 0.56
    Propionate 44.6 67.2 24.6 37.9 4.3 <0.01 <0.01 0.31
    Butyrate 46.6 67.2 32.9 48.4 4.8 <0.01 <0.01 0.61
Absolute disappearance rate, mmol/h
    Acetate 267 468 152 273 70 0.06 0.05 0.58
    Propionate 193 319 130 195 30 0.01 0.01 0.33
    Butyrate 85 136 71 101 14 0.11 0.02 0.45
1Dietary treatments: LF:C-HS, low F:C high S; LF:C-LS, low F:C low S; HF:C-HS, high F:C high S; HF:C-LS, high F:C low S.

Similar effects of dietary S and F:C on absolute disappearance rates (mmol/h) of SCFA were observed with the exception that there was no significant effect of F:C on butyrate disappearance (P = 0.11; Table 5). A stimulatory effect of grain on absorption rates of SCFA in cattle has been reported (Dirksen et al., 1985; Gäbel et al., 1991), but not all studies have observed this response (Penner et al., 2009). An increase in the absorptive surface area of the ruminal epithelium (Aschenbach et al., 2011; Penner et al., 2011), increased cellular activity of ruminal epithelial cells (Etschmann et al., 2009), and reduced ruminal pH (Dijkstra et al., 1993) caused by the high-grain diet could contribute to the increased SCFA absorption rates.

A novel finding in the current study was that dietary S inhibited SCFA absorption across the reticulorumen in cattle. Although the relationship between high dietary S and SCFA absorption in ruminants has not been studied, adverse effects of sulfur on colonic acetate and butyrate oxidation have been identified (Roediger et al., 1997; Leschelle et al., 2005; Blachier et al., 2009). Moreover, sulfide-induced epithelial damage is used as a strategy to experimentally induce ulcerative colitis in monogastric animal models (Pitcher and Cummings, 1996). Thus, it could be expected that high dietary S for ruminants may induce similar pathogenesis in the ruminal epithelium. Moreover, sulfides are thought to inhibit butyrate oxidation in epithelial cells by inhibiting acyl-CoA dehydrogenation of activated fatty acids. This is an important rate-limiting step in the oxidation of SCFA (Bremer and Osmundsen, 1984; Moore et al., 1997). Inhibition of short-chain acyl-CoA by sulfides is proposed to be caused by forming butyrate CoA persulfides (Shaw and Engel, 1987; Babidge et al., 1998) or by inactivating the electron transfer flavoprotein that is specific to the activity of short-chain acyl-CoA dehydrogenase (Babidge et al., 1998). Inhibited activity of short-chain acyl-CoA dehydrogenase will result in impaired β-oxidation of butyrate, which subsequently causes an energy deficiency state in epithelial cells (Roediger et al., 1993). In addition, Blachier et al. (2009) proposed that because the inhibitory effect of sulfide on acetate oxidation is also identified, the inhibition of cytochrome c oxidation by sulfide rather than the inhibition of short-chain acyl-CoA dehydrogenase might be a major factor associated with SCFA oxidation inhibition in epithelial cells. The rate of SCFA metabolism in epithelial cells is one of the important factors influencing the absorption of SCFA (Stevens and Settler, 1966; Gäbel et al., 2001; Penner et al., 2011). Intraepithelial SCFA metabolism, particularly butyrate metabolism, not only provides the majority of the energy to epithelial cells (Bergman, 1990) but also facilitates the absorption of SCFA by maintaining the concentration gradient between the cytosol and the lumen (Penner et al., 2011). The adverse impact of dietary S on SCFA absorption could be one of the contributors to poor animal performance in cattle associated with excessive S intake. Future study is needed to elucidate the mechanism underlying the negative impact of high dietary S on SCFA absorption.

An alternative explanation for the effect of HS on SCFA absorption may be due to reduced feed intake, increased ruminal pH, and reduced total SCFA concentration. Dijkstra et al. (1993) reported that acetate absorption across the reticulorumen epithelium in dairy cattle was lower when the incubation buffer contained 20 mM compared to 50 mM of acetate. They also observed the negative impact of increased ruminal pH on ruminal absorption rate of acetate, propionate, and butyrate. However, it should be acknowledged that the results obtained by Dijkstra et al. (1993) are acute effects of the buffer and do not represent effects caused by dietary treatment. Thus, future research is required to determine whether the dietary S concentration influences SCFA absorption or whether subtle changes in the ruminal SCFA concentration may induce such a response.

In conclusion, dietary F:C ratio does not seem to influence S metabolism in beef cattle, and the risk of S-induced PEM therefore may not be different among cattle fed diets with different F:C. Dietary S inhibits ruminal SCFA absorption.




Be the first to comment.

Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.