Exploring the cause of reduced production responses to feeding corn dried distillers’ grains in lactating dairy cows

An experiment was conducted to identify the factors that cause reduced production of cows fed a diet with high corn distiller’s grains with solubles (DDGS). We hypothesized that the factors could be high S content in DDGS which may directly (S toxicity) or indirectly [dietary cation-anion difference (DCAD)] cause reduced production. We also hypothesized that high polyunsaturated fatty acids (PUFA) in DDGS could be another major factor. In a randomized complete block design, 60 lactating cows (15 primiparous and 45 multiparious; average ± SD at the beginning of the trial: milk yield, 44.0 ± 6.9 kg/d; DIM, 123 ± 50; BW, 672 ± 82 kg) were blocked and cows in each block were randomly assigned to one of the following treatments: SBM [4.7% fatty acids (FA), 0.22% S, and 178 mEq/ kg DM of DCAD], a diet containing soybean meal as the main protein source; DG, SBM replacing mainly soybean byproducts and supplemental fat with DG at 30% dietary DM (4.7% FA, 0.44% S, and 42 mEq/kg DM of DCAD); SBM+S, SBM with sodium bisulfate for additional dietary S (4.8% FA, 0.37% S, and 198 mEq/kg DM of DCAD); SBM+CO, SBM with corn oil (4.7% FA, 0.23%, and 165 mEq/kg DM of DCAD); and DG+DCAD, DG with increased DCAD (4.7% FA, 0.40% S, and 330 mEq/kg DM of DCAD). Due to the limited tie stalls, the blocks of 1 to 6 started the experiment first as phase 1 and the rest of the blocks as phase 2 started the experiment after phase 1. All cows were fed the SBM diet for 10 d as a covariate period followed by the experimental period for 35 d. Data were analyzed using the PROC MIXED of SAS, block and phase were random effects and treatments, repeated wk, and interaction were fixed effects. There was an interaction of wk by treatment for DMI. While milk yield did not change, milk fat concentration tended to decrease (2.78 vs. 3.34%) for DG compared with SBM. Dry matter, OM, NDF, and CP digestibilities were lower when cows were fed the DG diet compared with SBM. Additionally, cows fed DG had lower blood concentrations of HCO 3-, base excess, and tCO 2 compared with SBM. The SBM+S diet did not affect production, nutrient digestibility, or blood parameters when compared with SBM. The SBM+CO diet decreased milk fat concentration and yield compared with SBM. The DG+DCAD diet tended to increase milk fat yield and concentration (1.24 vs. 1.47 kg/d; 2.78 vs. 3.37%) and increased ECM (40.9 vs. 45.1 kg/d) compared with DG but did not improve nutrient digestibility. However, blood HCO 3-, base excess, and tCO 2 were greater for DG+DCAD compared with DG. In conclusion, the indirect role of S - , altering DCAD, along with the high PUFA content in DDGS appears to be the factors causing reduced production responses to a high DDGS diet. Increasing DCAD to 300 mEq/kg DM in a high DDGS diet can be a feeding strategy to alleviate the reduced production responses.


INTRODUCTION
With feed costs remaining relatively high and milk prices slowly declining (USDA, 2023a), dairy producers may be looking for alternative feed sources to reduce feed expenses.The most common by-product derived from ethanol production, corn dried distillers' grains with solubles (DDGS), can be an economical substitution for soybean meal in lactating diets.However, reduced production responses are somewhat consistent when cows were fed a diet with high DDGS (>20% in dietary DM; Hippen et al., 2010;Benchaar et al., 2013;Ramirez-Ramirez et al., 2016;Zynda et al., 2022).
The high PUFA content in DDGS has commonly been associated with the reduced production responses (Hippen et al., 2010;Ramirex-Ramirez et al., 2016).When a diet containing a high PUFA concentration is present in the rumen, trans-10 intermediates increase and suppress milk fat synthesis (Bauman and Griinari, 2003;Bauman et al., 2011).Reduced-fat DDGS (RFD-DGS) are available and are about 40 to 60% lower in crude fat compared with traditional DDGS (DuFour, 2017).Reduced-fat DDGS have been indicated to mitigate milk fat depression (MFD; Ramirez-Ramirez et al., 2016).While results have been variable among studies, Morris et al. (2018a) reported reduced production responses, including MFD, when feeding RFDDGS (30% of dietary DM).This suggests that other dietary factors within DDGS causing the reduced production responses may exist.
Sulfur concentration in DDGS is relatively high compared with soybean meal (0.71 ± 0.144 vs. 0.41 ± 0.033%; NASEM 2021); therefore, replacing mainly soybean meal with DDGS in a diet increases dietary S concentration.We previously observed (Morris et al., 2018a) a 0.2%-unit increase in dietary S (0.21 vs 0.41% S, DM basis) when DDGS (29% of dietary DM) fully replaced soybean meal and soyhulls.The recommendation for dietary S for lactating cows is 0.15 to 0.25% to reach maximum fiber digestibility (NASEM, 2021).Excess S can cause direct and indirect negative effects on dairy cows.As an oversupply of S can directly reduce rumen pH and fiber digestibility (Drewnoski et al., 2014) and act as a trace mineral antagonist (van Ryssen et al., 1998;Ivancic and Weiss, 2001;Ritcher et al., 2012, Pogge et al., 2014).Indirectly, S decreases dietary cation-anion difference (DCAD; mEq/kg DM = Na + K -Cl -S); therefore, adding DDGS to a diet decreases DCAD due to the high S concentration.Typically, DCAD is about 200 mEq/kg DM for lactating diets that meet the requirements of minerals without adding cation supplements for the purpose of increasing DCAD; however, diets containing 20 to 30% (DM basis) of DDGS reduces DCAD to around 0 to 100 mEq/kg DM due to the high S content.In our previous studies (Morris et al., 2018a;Zynda et al., 2022), when DDGS was added at 20 and 30% of dietary DM, DCAD decreased from 185 to 62 and 195 to 65 mEq/kg DM, respectively.A meta-analysis (Iwanuik and Erdman, 2015) showed that DMI, milk and fat yields, and DM and NDF digestibilities were positively associated with DCAD (0 to 500 mEq/kg DM).By manipulating DCAD in a DDGS diet via cation supplementation, Zynda et al. (2022) observed a numerical increase in milk fat yield (1.24 vs. 1.43 kg/d) as DCAD was increased in a DDGS diet (20% DM basis) from 65 to 187 mEq/kg DM.However, nutrient digestibility was not improved, and more studies are needed to determine if increasing DCAD >200 mEq/kg DM can further alleviate reduced production responses and nutrient digestibility.
The objective of this study was to compare production responses of cows fed a soybean meal-based diet with a diet containing high DDGS (30% on a DM basis).Additionally, we investigated whether dietary S was another factor causing reduced production responses in addition to PUFA.Lastly, we determined whether increasing DCAD (300 mEq/kg DM) in a DDGS-based diet will alleviate the reduced production responses.We hypothesized the DDGS diet will cause reduced production responses, and both S or PUFA in DDGS are the factors for the reduced production.Furthermore, we hypothesized that reduced production responses to feeding a DDGS diet will be alleviated by increasing DCAD.

Animals and Treatments
All procedures in this study involving animals were approved by The Ohio State University Institutional Animal Care and Use Committee (IACUC: 2022A00000069).The experiment was conducted at the Krauss Dairy Research Center (Wooster, OH).
In a randomized complete block design, 60 lactating Holstein cows (15 primiparous and 45 multiparous; average ± SD at the beginning of the trial: milk yield, 44.0 ± 6.9 kg/d; DIM, 123 ± 50; BW, 672 ± 82 kg) were blocked by parity, DIM, and milk yield into 12 blocks (i.e., 5 cows each).The experiment was organized in 2 phases due to limited tie stalls with 6 blocks in each phase.Cows in each block were randomly assigned to one of the 5 following experimental diets: a soybean meal-based diet (SBM), the SBM diet with DG at 30% (DM basis) by replacing mainly soybean meal, soyhulls, and supplemental fat (DG), the SBM diet with additional dietary S by supplementing sodium bisulfate (SBM+S), the SBM diet with corn oil (SBM+CO), and the DG diet with elevated DCAD (300 mEq/kg) by supplementing sodium bicarbonate and potassium carbonate (DG+DCAD).All diets were formulated to meet or exceed all nutrient requirements for lactating cows according to NASEM (2021).All diets were formulated to have similar fat concentration, using either a commercial saturated fat supplement (Energy Booster 100, Milk Specialties) or corn oil.Replacing mainly soybean meal and soyhulls for DG increased the PUFA and S concentration and decreased DCAD.To examine the direct effect of high S, dietary S was elevated in SBM+S but DCAD was not altered by using sodium bisulfate.Similarly, dietary PUFA was increased in the SBM+CO diet to resemble the FA composition of the DG diet.To determine the effect of indirect S (i.e., DCAD), DCAD was elevated to 300 mEq/kg DM in the DG+DCAD diet.The diet ingredients and chemical composition are shown in Table 1 and 2. The chemical composition of DDGS used in this study is shown in Table 3.
The experiment started with a 10-d covariate where all cows were fed the SBM diet, followed by the experi-mental diets for 5 wk.All diets were prepared daily as a TMR for ad libitum intake with a 5% target refusal rate.Cows were housed in individual tie stalls with free access to water and milked twice daily at 0400 and 1600 h.Body weights were measured on 2 consecutive days during the covariate, wk 3, and wk 5 of the experiment.

Sample Collection and Analysis
Dry matter intake of individual cows was monitored daily for the entirety of the study.Forages were sampled weekly, and DM was determined at 100°C to adjust the as-fed forage inclusion rate in the rations.Refusals were weighed daily, and subsamples were collected on 2 consecutive days in each wk and composited by cow and wk.The composited samples were used to determine DM at 55°C.Individual forages and concentrates were collected on 2 consecutive days weekly, composited by every 3 wk (i.e., 2 composites of ingredient/ phase).Forages were dried at 55°C for 72 h to determine DM.Dried composite forages and concentrates were ground (Ultra Centrifugal Mill, Retsch) to pass through a 1-mm sieve.The ground samples were dried at 100°C overnight to determine DM (method 934.01;AOAC, 2000).Ground feeds were further analyzed for ash (method 942.05;AOAC International, 2000), NDF with heat-stable α-amylase and sodium sulfite (Ankom 200 Fiber Analyzer; Ankom Technology Corp.), starch (Weiss and Wyatt, 2000), FA (Jenkins, 2010), total N using an elemental analyzer (Morris et al., 2019;Flash 2000, Thermo Fisher Scientific, Waltham, MA), and minerals (Rock River Laboratory; https: / / rockriverlab .com/ ) using inductively coupled plasma emission spectroscopy.All ground feeds were weighed into filter bags with a pore size of 25 μm (Ankom Technology Corp.) and were incubated for 288 h in the rumen of 2 cannulated dry cows (Huhtanen et al., 1994).The cannulated cows were fed a dry cow diet (13.7% CP, 54.2% aND-Fom; 6.6% starch, and 86:14 forage: concentrate).After rumen incubation, the bags were rinsed and then analyzed for NDF as previously described.The remaining residuals were considered indigestible NDF (iNDF).Weekly samples of DDGS were collected throughout the experiment and composited (i.e., n = 1).The composite sample was analyzed at Rock River Laboratory (Core nutrients analysis; https: / / rockriverlab .com/pages/ Nutrition -Services .php).
Milk samples were collected on 2 consecutive days weekly to determine milk fat, true protein, lactose, and MUN (B2000 infrared analyzer, Bentley Instruments, Chaska, MN) by DHIA (Columbus, OH).Moreover, AM and PM milk samples were collected on 1 d during the last wk of the experiment and proportionally composited by milk weights for individual cows.An aliquot of composited milk was centrifuged (17,200 × g at 4°C for 30 min) to separate the milk fat cake and was frozen (−20°C) until milk FA analysis.Milk FA composition was quantified using a 2-step methylation process (Jenkins, 2000) with separation by gas chromatography (HP 5890 series, Agilent Technologies) with Spot fecal and urine samples were collected for 3 consecutive days during the last wk.Six spot samples were taken in 12 h increments (i.e., 1000 and 2200 on d 1, 0600 and 1600 h on d 2, and 0200 and 1400 h on d 3) to represent every 4-h sampling in a 24-h period (Morris et al., 2018b;Lee et al., 2019).Spot urine samples were collected by massaging the vulva.Samples collected were composited by cow (i.e., 6 time points/cow), acidified using 2M H 2 SO 4 , and frozen until further analysis.Spot fecal samples were obtained from the rectum and composited per cow.A subsample of composited feces was dried at 55°C for 72 h to determine DM content.Dried composite fecal samples were ground (Ultra Centrifugal Mill; Retsch Mill, Haan, Germany) to pass through a 1-mm sieve.Ground samples were dried at 100°C overnight to determine DM and further analyzed for OM, N, NDF, iNDF, and FA as previously mentioned.Using iNDF as an internal marker, fecal DM outputs of individual cows were estimated (Morris et al., 2018b).Fecal DM output estimates were used to determine DM, OM, CP, NDF, and FA digestibility.
Blood samples were collected via coccygeal venipuncture into evacuated EDTA tubes before feeding and 6 h after feeding in the last wk .Samples were immediately centrifuged (2,500 × g 4°C for 20 min) and the plasma was stored at −20°C until further analysis.All the plasma samples were assayed for AA composition using high performance liquid chromatography (Perkin Elmer Series 200 HPLC system, Perkin-Elmer Inc.) with a fluorescent detector after deproteinization with 5M perchloric acid.The column (WAT052885), amino acids standards (186009300), and derivatization kit (WAT052880) were obtained from Waters Corp.The assay procedure followed the manufacturer's instruction except that the mobile phase constituents and their gradient mode were prepared according to Jaworska et al. (2012).Additionally, another blood sample was collected into lithium-heparin vacutainers 6 h after feeding.An aliquot of whole blood at 6 h after feeding were determined for blood parameters using an    ), and base excess (BE).

Statistical Analysis
All data were analyzed using the Mixed Procedure of SAS (Version 9.4, SAS Institute).Dry matter intake and milk yield data were averaged weekly and weekly DMI, milk yield, and milk composition were analyzed using the following model: where Y ijkl = dependent variable; μ = overall mean; Cov = measures during the covariate period; b i = random effect of block (b = 1 to 12); p j = random effect of phase (P = 1 to 2); T j = fixed effect of dietary treatment (T = 1 to 5); W k = fixed effect of wk (W = 1 to 5); the interaction of dietary treatment and wk (TW) kl ; and e ijkl = random residual error.Week was the repeated measure and an auto-regressive first order covariance structure was used (smallest Akaike Information Criterion).The Kenward-Rogers method was applied to calculate the denominator degrees of freedom.Preplanned contrasts were applied to determine differences between SBM vs. DG, SBM vs. SBM+S, SBM vs. SBM+CO, and DG vs. DG+DCAD.When analyzing blood and nutrient digestibility data, the covariate period and fixed effect of wk and its interactions were removed from the model.The model for the plasma AA composition was the same as that for the blood data except that fixed effects of sampling time points (0, 6 h after feeding) and their interactions with treatments were included with time points as repeated measures (auto-regressive first order covariance structure).All data were presented as least squares means and statistical differences were determined at P ≤ 0.05, while trends were considered at 0.05 < P ≤ 0.10.

RESULTS
There was no difference in initial BW (Table 4) at the start of the experiment (i.e., covariate period); however, BW tended to be greater (P = 0.09) in the last wk of the experiment for DG compared with SBM.The BW change, however, did not differ among treatments.A wk effect for most production variables (milk yield, milk fat concentration and yield, milk protein concentration and yield, milk lactose concentration and yield, MUN, feed efficiency, and NE L ) was observed (P < 0.01).An interaction between treatment and wk was observed only for DMI (P = 0.05; Figure 1).The interaction occurred due to differences among treatments in wk 3, 4, and 5 which did not occur in earlier wk.In the later wk, DMI was greater (P < 0.05) or tended to be greater (P < 0.10) for DG or DG+DCAD compared with SBM or SBM+CO.Milk fat concentration tended to be lower (2.78 vs. 3.34%; P = 0.06) for DG and consequently ECM per unit of DMI was lower (1.48 vs. 1.62 kg/kg; P = 0.03) compared with SBM.However, milk fat yield was not different (P = 0.13) between DG and SBM.Lactose concentration differed between DG and SBM (4.92 vs. 4.85%, P < 0.01) without a different in lactose yield (P = 0.14).Milk protein concentration and yield were not affected by DG versus SBM.There were no differences in production variables for SBM+S compared with SBM.Milk yield tended to be greater (46.7 vs. 43.8kg/d; P = 0.07) for SBM+CO compared with SBM, resulting in greater milk yield per unit of DM (1.74 vs. 1.61 kg/kg; P < 0.01).Milk protein concentration was not affected, but milk protein yield was greater (1.49 vs. 1.41 kg/d; P = 0.05) for SBM+CO compared with SBM.Milk fat concentration (2.27 vs 3.34%; P < 0.01) and yield (1.06 vs 1.43 kg/d; P < 0.01) were lower for SBM+CO compared with SBM.As a result, ECM per unit of DMI tended to be lower (1.50 vs. 1.62 kg/kg; P = 0.06) for SBM+CO compared with SBM.While there were no differences in milk yield, milk fat concentration and yield tended to be greater (P = 0.06; P = 0.07) for DG+DCAD compared with DG.The increase in milk fat yield resulted in a 3.8 kg/d increase (P = 0.03) in ECM for DG+DCAD compared with DG.
Differences in plasma AA concentration were obvious between SBM and DG.The DG diet increased (P < 0.05) the concentrations of His, Leu, Met, Phe, Pro, Asp, Glu, and Tyr but decreased (P < 0.05) Lys, Thr, Arg, Gly, Orn, and Tau and tended to decrease (P = 0.08) Ileu compared with SBM (Table 5).However, no difference in plasma AA concentration was observed between DG and DG+DCAD.In addition, elevating S concentration of the SBM diet (i.e., SBM+S) did not alter plasma AA concentration except for Tau which was lower (P = 0.02) for SBM+S versus SBM.Supplementation of SBM with corn oil (SBM+CO) decreased (P < 0.05) plasma concentrations of Ileu, Thr, Val, Gln, Ser, and Tyr compared with SBM.
Milk FA composition results are displayed in Table 6.Greater concentrations of C18:2 cis-9, trans-11 (P = 0.01) and C18:2 trans-10, cis-12 (P = 0.03) were present in the DG diet compared with the SBM.However, the concentrations of C18:1 trans-11 and trans-10 did not differ between DG and SBM.In the DG diet compared with SBM, the concentration of saturated and PUFA were lower and greater (P < 0.01), respectively.Furthermore, OBCFA in milk were lower (3.67 vs. 5.33; P < 0.01) for DG compared with SBM.There were no differences in biohydrogenation intermediates, saturated FA, PUFA, and OBCFA for the SBM+S diet compared with SBM.Various FA concentrations were altered for SBM+CO compared with SBM.The concentrations of trans-10 intermediates from the alternative biohydrogenation pathway were greater (P < 0.01) for SBM+CO compared with SBM, resulting in greater (P < 0.01) PUFA content in milk.Additionally, FA concentration of 16 carbons or less were lower (52.77 vs. 66.54%;P < 0.01) and FA more than 16 carbons were higher (44.37 vs. 31.93%;P < 0.01) in milk for SBM+CO.The DG+DCAD diet had greater concentrations of C18:1 trans-11 (P = 0.03) and C18:2 cis-9, trans-11 (P = 0.05) compared with the DG diet.However, the DG+DCAD did not alter the proportions of FA of, less than, or more than 16 carbons compared with DG.
Apparent total-tract nutrient digestibilities are in Table 7.While DM and CP intakes tended to increase (P ≤ 0.08) and NDF intake increased (P = 0.04) in DG diet compared with SBM during the last wk of the study, total-tract apparent digestibility of DM, OM, NDF, and CP were lower (P < 0.05).Additionally, FA intake was not different between DG and SBM, but DG had lower 16-carbon FA and greater 18-carbon FA intakes compared with SBM.However, compared with the SBM diet, C18 and C16 digestibility tended to be (P = 0.06) and were greater (P < 0.01) for the DG diet, respectively.For SBM+CO compared with SBM, intake of C16 FA was lower (0.17 vs. 0.26 kg/d; P < 0.01) and digestibility was greater (P < 0.01).Digestibility of all nutrients was not altered when additional dietary S was provided (SBM+S vs. SBM) or when DCAD was elevated in the DG diet (DG+DCAD vs. DG).
There were no differences in blood pH, Na + , K + , BUN, and hemoglobin throughout dietary treatments (Table 8).The DG diet decreased (P ≤ 0.05) tCO 2 , BE, and HCO 3 -, and tended to increase (P < 0.10) anion gap and pO 2 compared with SBM.Blood parameters did not differ between SBM+S and SBM.There was also no difference in blood parameters between SBM and SBM+CO, except that K + tended to be lower (P = 0.10) in the SBM+CO diet compared with SBM.Cows

DISCUSSION
Nutrient analysis of DDGS, soybean meal, and soyhulls were conducted before the start of the study, and all diets were formulated successfully to have similar concentrations of CP, FA, NDF, and starch (Table 2).When the SBM diet was formulated to meet the requirements of all nutrients, the level of DCAD was 170 mEq/kg of DM.The SBM+S and SBM+CO diets had DCAD similar to SBM (165 and 198 mEq/kg of DM, respectively).By the experimental design, substituting mainly soybean meal and a fat supplement for DDGS in the DG diet resulted in an increase in dietary unsaturated FA without altering total FA.This rise in unsaturated FA is common and was observed in DDGS-based diets (Ramirez-Ramirez et al., 2013;Morris et al., 2018a;Zynda et al., 2022).To investigate the PUFA effect as a factor of DDGS, total FA content was formulated to be similar between SBM+CO and SBM but resemble the PUFA content and composition between SBM+CO and DG.To investigate the direct S effect as a factor of DDGS, sodium chloride in the SBM diet was replaced with sodium bisulfate.This resulted in an increase in S concentration in SBM+S compared with SBM and resembled the S concentration in the DG diet.However, by the experimental design, the SBM+S had the DCAD similar to that of SBM.By increasing the level of potassium carbonate and sodium bicarbonate in the DG+DCAD diet, K and Na concen-  tration was elevated compared with the DG diet, which resulted in increased DCAD to about 300 mEq/kg DM.
included.Although the current study used the DDGS and inclusion rate that are similar to those in Morris et al. (2018a), the interaction of treatment by wk indicated that DMI for DG and DG+DCAD was similar during the entire experiment while SBM and SBM+CO decreased DMI as the experiment progressed.The different response of DMI between Morris et al. (2018a) and the current study is difficult to explain.In the current study, the inclusion of DDGS did not affect milk yield.This agrees with our previous study (Morris et al., 2018a), although many studies observed an increase (Kleinschmit et al., 2006;Hubbard et al., 2009;Benchaar et al., 2013;Ramirez-Ramirez et al., 2016) or decrease (Zynda et al., 2022)   crepancy in milk yield responses across studies could be also due to the feed ingredients that DDGS replaced.For example, replacing both energy (i.e., ground corn) and protein (i.e., soybean meal) sources with a high inclusion of DDGS increased milk yield in Benchaar et al. (2013).Furthermore, effects of DDGS on milk yield might be associated with nutrient digestibility.When forages and ground corn were replaced with DDGS, increases in DM, OM, and NDF digestibilities were observed, which likely increased milk yield in a study by Ramirez-Ramirez et al. (2016).The current study, along with our previous studies (Morris et al., 2018a;Zynda et al., 2022), mainly replaced soybean meal and soyhulls which have higher NDF digestibility than DDGS (Firkins, 1997;Mjoun et al., 2010).As expected, we observed decreases in DM, OM, NDF, and CP digestibilities for the DG diet compared with SBM.Studies in the literature also observed decreases in DM, OM and NDF digestibilities for diets at 20 to 30% DDGS (Benchaar et al., 2013;Morris et al., 2018a;Brown and Bradford, 2020;Zynda et al., 2022).A decline in milk yield in a high DDGS diet could be explained by lower digestibility.The current study observed decreases in DM, OM, and NDF digestibilities by 3.5, 3.0, and 8.2%-units, respectively, for DG versus SBM, but the degree of the decreases in digestibility was not likely large enough to impair milk yield.Milk protein concentration and yield were not different between DG and SBM although large differences in plasma AA concentrations were observed.
The differences in plasma AA concentration likely reflect the difference in AA composition between DG and SBM according to greater concentrations of Leu and Met but lower concentration of Lys for DG versus SBM.In addition, the concentrations may have been affected by total-tract CP digestibility or other factors (e.g., intestinal digestibility of RUP) because changes in concentrations of some AA (e.g., Thr and Phe) do not reflect the composition of AA between DG and SBM.No difference in milk protein concentration and yield between DG and SBM despite the large difference in plasma AA profile may support the previously proposed concept that mammary tissues have large flexibility in utilizing AA available in blood for protein synthesis (Yoder et al., 2019).
While milk fat concentration tended to decrease by 17% (tendency; P = 0.06), milk fat yield was only numerically decreased (14%; P = 0.13) for DG compared with SBM in the current study.The numerical difference in milk fat yield between DG and SBM likely occurred due to 2 kg greater milk yield for DG compared with SBM.The decline in milk fat led to a numerical decrease in ECM and significant decrease in ECM per unit of DMI for DG compared with SBM.Diet-induced MFD commonly occurs when a high DDGS-based diet was fed due to the high supply of dietary PUFA (Beauchemin et al., 2007;Ramirez-Ramirez et al., 2016).In the current study, although FA content was similar, PUFA concentration was 32% greater for DG compared with SBM (Table 2).The excess PUFA supply can increase trans-10 intermediates in the rumen and inhibit milk fat synthesis in the mammary gland (Bauman and Griinari, 2003).Indeed, the concentration of these intermediates (C18:1 trans-10 and C18:2 trans-10, cis-12) in milk were greater numerically and significantly, respectively, for DG versus SBM (Table 6).This agrees with our previous studies where a diet with DDGS at a 20 to 30% inclusion was fed to cows (Morris et al., 2018a;Zynda et al., 2022).It is worth noting that milk fat concentration for the SBM diet in the current study was relatively low (3.34%), which might have occurred due to the relatively high starch (28.4%) and low forage NDF concentration (19.0% of DM).A diet of similar levels of starch and forage NDF with 0.7% soybean oil was used as a diet having moderate risk of milk fat depression previously (Baldin et al., 2018).If it was true that the starch and NDF concentration in the SBM diet decreased milk fat concentration moderately in the current study, milk fat depression that occurred due to inclusion of DDGS for DG (i.e., adding PUFA) may have been greater compared with inclusion of DDGS in a diet that has no risk of milk fat depression.
In our previous experiment (Morris et al., 2018a), because we observed reduced responses in milk fat and protein although reduced-fat DDGS was used (30% of dietary DM), we speculated that a high PUFA content in DDGS might not be the only dietary factor contributing to the reduced production responses.Therefore, we hypothesized that the high S content in a DDGS-based diet may be another dietary factor.At an inclusion of DDGS greater than 20%, S concentration can range from 0.40 to 0.50%, depending on S content in DDGS (DM basis; Ramirez-Ramirez et al., 2016;Morris et al., 2018a).However, we did not observe any differences in intake, milk production, or nutrient digestibility between SBM and SBM+S, suggesting that about 0.4% dietary S in DG was not likely a direct factor associated with the reduced production responses to feeding DDGS.We designed the experiment to examine the direct effect of S in DDGS by maintaining DCAD.The DCAD levels of SBM+S and DG were 170 and 46 mEq/ kg DM, respectively.This suggests that the decrease in DCAD by high S in DDGS could be an indirect factor contributing to reduced production when feeding DDGS.We concluded that high S in DDGS played a role causing reduced production as an indirect factor in the current study.Further discussion on the effects of DG+DCAD will follow later.
With no difference in nutrient digestibilities except 16-carbon FA, we observed a tendency for an increase in milk yield and increase in milk yield per unit of DMI for SBM+CO compared with SBM.The effect of supplemental PUFA on DMI and milk yield in the literature is variable and seems to be affected by many factors such as dietary composition (e.g., NDF and starch), lactation stage, the forms or level of supplemental PUFA (Moallem, 2018).For example, a study by Leonardi et al. (2005) observed no change in DMI but an increase in milk yield by 2.6 kg/d with the inclusion of corn oil (1.5% DM basis).On the other hand, Boerman et al. (2014) reported decreases in DMI, milk yield, and feed efficiency with increasing corn oil up to 2.4% of dietary DM.As expected, diet-induced MFD was present for SBM+CO in the current study, as both milk fat yield and concentration declined, while milk and protein yields were not negatively affected.Compared with the SBM, de novo (<C16) and mixed (C16) milk FA decreased and preformed (>C16) FA increased for the SBM+CO diet compared with SBM, which is a typical phenomenon for milk FA during diet-induced MFD (Bauman et al., 2011).Furthermore, increases in C18:1 trans-10 and C18:2 trans-10, cis-12 for SBM+CO by 173 and 150%, respectively, compared with SBM, which is also common for cows on diet-inducted MFD (Bauman and Griinari, 2003).Leonardi et al. (2005) reported decreases in de novo and mixed FA and increases in preformed FA and trans-10 milk FA when corn oil (1.5% of dietary DM) was added to a soybean mealbased diet.In that study, however, although milk FA profile was different, there was no difference in milk fat yield or concentration between the diet with and without corn oil.Another study by Boerman et al. (2014) also observed similar changes in the profile of milk FA and decreases in milk fat concentration and yield for diets supplemented with corn oil (2.8% of dietary DM).In the current study, we did not statistically compare variables between SBM+CO and DG.However, milk fat yield was likely even lower (1.06 vs. 1.24 kg/d) for SBM+CO compared with DG.If that was true, perhaps the corn oil added to SBM+CO had an instant exposure to rumen microbes when the diet was consumed, altering rumen biohydrogenation more considerably as C18:1 trans-10 FA in milk was much greater (7.84 vs. 3.47%) for SBM+CO than DG.However, DDGS in the DG diet may have required more time for digestion and released oil in the rumen relatively slowly as digestion progressed.From the effects of SBM+CO, it was clear that high PUFA in DDGS was the major dietary factor associated with MFD.
In previous studies, milk protein yield and concentration tended to increase when corn oil (1.5%; 720 g/d) or corn oil and palm kernel oil (720 g/d at 75:25 ratio) was added (Leonardi et al., 2005;Parales Girón et al., 2016).Parales Girón et al. ( 2016) credited the 2.4 kg/d increase in milk yield to the increase in milk protein.In contrast, milk protein yield did not differ across diets when corn oil was increased from 0 to 2.8% (Boerman et al., 2016).Another fat supplementation with high PUFA content, linseed oil, has had mixed results on milk protein.There have been linear decreases, increases, and no differences in milk protein yield with supplementation of linseed oil (Bu et al., 2007;Flowers et al., 2008;Benchaar et al., 2012).The increased milk protein yield for SBM+CO versus SBM in the current study likely occurred due to the increase in milk yield.Differences in plasma AA concentrations where some of AA concentrations in plasma decreased and, if not significantly affected, most AA decreased numerically for SBM+CO versus SBM, may reflect greater extraction of available AA in blood for milk protein synthesis.
If a lactation diet is formulated to be adequate in all the minerals (NASEM, 2021), DCAD should be around 200 mEq/kg DM.When DDGS is included in a ration at a high inclusion rate without adjusting cation concentrations, which was the case in most DDGS studies (Benchaar et al., 2013;Zanton et al., 2013;Ramirez-Ramirez et al., 2016;Ranathunga et al., 2018), DCAD should decrease due to the high S content of DDGS.In the current study, when DDGS replaced mainly soybean meal and soyhulls at 30% inclusion (DM basis), DCAD decreased from 178 to 42 mEq/kg DM.This agrees with our previous studies that also replaced soy products with DDGS (Morris et al., 2018a;Zynda et al., 2022).A meta-analysis (Iwaniuk and Erdman, 2015) indicated that increasing DCAD ranging from 0 to 500 mEq/kg of dietary DM in lactating cow diets increased DMI, DM and NDF digestibilities, and milk and milk fat yields.Therefore, by manipulating DCAD in a DDGS diet, there is potential to alleviate the reduced production responses.Previously, we observed numerical increases in milk fat composition (3.15 vs. 3.47%) and yield (1.24 vs. 1.43 kg/d) and ECM (38.0 vs. 40.7 kg/d) when DCAD of a DDGS-diet (20% of DDGS in dietary DM) was increased from 62 to 187 mEq/kg DM by adding potassium carbonate and sodium bicarbonate (Zynda et al., 2022).In that study, however, milk fat yield for the DDGS diet with increased DCAD was numerically lower compared with a soybean meal-based diet (control; 1.55 vs. 1.27 kg/d).Therefore, in the current study, we increased DCAD from 42 to 330 mEq/kg DM to determine if production responses, including milk fat, to feeding a high DDGS diet could be further alleviated.We found that although nutrient digestibility and milk yield were not affected by increased DCAD, milk fat concentration and yield tended to increase compared with the DG Clark et al.: FACTORS OF FEEDING CORN BYPRODUCTS CAUSING REDUCED PRODUCTION diet, resulting in increased ECM.The result confirms that increasing DCAD of a DDGS diet (>20% of DDGS in dietary DM) can be an effective strategy to alleviate or eliminate negative effects that a high DDGS diet has.Previously, effects of DCAD on positive milk fat yield in lactation cows have been explained with better rumen environments (e.g., pH), leading to more complete biohydrogenation through the normal pathway (Apper-Bossard et al., 2010;Harrison et al., 2012;Guiling et al., 2017).However, C18:1 trans-10 and C18:2 trans-10, cis-12 did not differ in milk between DG and DG+DCAD in the current study, although DG+DCAD tended to increase milk fat concentration and yield.In agreement, Zynda et al. (2022) observed no difference in trans-10 intermediates in milk fat when DCAD was elevated from 62 to 187 mEq/kg DM in a 20% DDGS-diet.This suggests that improving biohydrogenation (i.e., less intermediates from the alternate pathway) in the rumen was not likely the mechanism for elevated DCAD to increase milk fat as speculated in the literature (Apper-Bossard et al., 2010;Iwaniuk and Erdman, 2015).Instead, our results suggest that the effect of DCAD on milk fat likely occurred via a postabsorptive rather than ruminal effect (see the discussion later about blood parameters).It is worth noting that trans-11 intermediates (C18:1 trans-11 and C18:2 cis-9, trans-11) increased for DG+DCAD compared with DG even though trans-10 intermediates in milk fat did not decrease.This may indicate that increased DCAD influenced biohydrogenation in the rumen, resulting in the alternative biohydrogenation intermediates to become diluted.However, because abomasal supply of trans-11 intermediates did not affect milk fat (Lock et al., 2007), further studies are needed to investigate if the dilution of trans-10 intermediates by trans-11 intermediates is related to milk fat synthesis.Lamar (2013) observed decreases in C18:1 trans-10 and C18:2 trans-10, cis-12, and increases in C18:1 trans-11 and C18:2 cis-9, trans-11 in milk fat when DCAD was increased (20 to 300 mEq/kg DM) by adding potassium carbonate to a 27% DDGS diet.Increased proportion of C18:0 and decreased C18:2 cis-9, cis-12 was also observed in that study, suggesting that adding K and increasing DCAD resulted in an increase of ruminal biohydrogenation of C18:2 to C18:0.More research is needed to understand the mechanisms of DCAD and milk fat synthesis.
The increase in milk fat for DG+DCAD compared with DG did not occur due to increasing nutrient digestibility.A meta-analysis by Iwaniuk and Erdman (2015) showed improving DM and NDF digestibilities with increasing DCAD.Additionally, with DCAD ranging from 13 to 436 mEq/kg DM, DM and NDF digestibilities increased 0.73 and 1.54% per 100 mEq/ kg DM increase in DCAD, respectively.Another meta-analysis by Martins et al. (2016) also observed a linear increase in NDF digestibility with increasing DCAD from −71 to 290 mEq/kg DM, likely due to the linear increase in rumen pH 3 h after feeding.However, the current study and other studies (Apper-Bossard et al., 2010;Zynda et al., 2022) did not observe a difference in DM or NDF digestibilities between DG+DCAD and DG.No differences in DM and NDF digestibilities in addition to plasma AA and milk FA profile for DG-DCAD versus DG supports our previous speculation that the increase in milk fat with increased DCAD may have been associated with post-absorptive effects rather than ruminal effects.In addition, no difference in plasma AA concentrations except for His between DG and DG+DCAD may also support that increased DCAD did not affect the rumen environment, i.e., ruminal degradation of dietary protein and passage rates of RUP.
Blood metabolic acidosis can occur from feeding anionic diets, but physiological mechanisms can quickly reestablish the acid-base balance throughout the body.To maintain blood equilibrium when blood anion (Cl -, SO 4 −2 ) concentration is high, blood HCO 3 -must decrease to increase H + (Stewart, 1983;Gelfert et al., 2010).Blood pH is quickly returned to homeostasis with alterations in blood HCO 3 -and pCO 2 by adjusting renal regulation of HCO 3 -and respiratory rate (Tucker et al., 1988).When cows have metabolic acidosis, blood HCO 3 -, tCO 2 , and pCO 2 decrease because of increased CO 2 expiration (Block, 1994).The decline in blood Cl -and increase in HCO 3 -for DG+DCAD versus DG agrees with a previous study (Apper-Bossard et al., 2010).It appears that Cl -is a major anion involved in regulating HCO 3 -, as the Cl -concentration was strongly correlated to HCO 3 -concentration in blood ([HCO 3 -] = 124.2-1.00 [Cl -]; r = 0.94; Apper-Bossard et al., 2010).In the current study, a similar correlation was observed for SBM, DG, and DG+DCAD diets ([HCO 3 -] = 150.0-1.21[Cl -]; R 2 = 0.83).In the DG diet, the large absorption of anions likely reduced blood HCO 3 -concentration and activated renal mechanisms to reabsorb HCO 3 -to maintain acid-base homeostasis.These renal mechanisms were likely not activated in the DG+DCAD because acid-base equilibrium was maintained.According to the fact that no differences in milk FA composition and total-tract nutrient digestibility but differences in the blood acid-base balance between DG+DCAD and DG, we concluded that the effect of DG-DCAD versus DG on increased milk fat occurred due to post-absorptive mechanisms.Therefore, low DCAD in addition to PUFA were likely the major factors that caused MFD when the DG diet was fed.
When the SBM diet was elevated for dietary S, i.e., SBM+S, the anion gap and blood electrolytes were not altered compared with SBM.This suggests that the SBM+S diet did not affect the acid-base balance as expected.Although dietary S was elevated, the DCAD in the SBM+S diet was not altered when compared with SBM because the S supplement was sodium bisulfate which contains both cation and anion, resulting in no change in DCAD.Based on the responses of production and blood electrolytes between SBM+S and DG, we concluded that the dietary S of 0.4% (DM basis) does not have direct negative effects on performance and nutrient digestibility of lactating cows.
According to the diets and production results, we calculated the income over feed costs (IOFC) for SBM, DG, and DG+DCAD.Milk price ($/cwt) was determined on a component basis (USDA, 2023b): $5.94/ kg of butterfat, $5.63/kg of protein, and $0.55/kg of other solids.Means of DMI, milk yield, milk fat, protein, and solid yields from each treatment were used to estimate income and feed costs.Based on current feed costs, the SBM diet was the most expensive followed by DG+DCAD and DG (9.16,7.46,and $7.00/cwt,respectively).The difference in price between SBM and DG diets was because of the current price of soybean meal and DDGS, $443/ton of soybean meal and $240/ ton of DDGS (5/15/2023; USDA, 2023c).Market prices of soybean meal and DDGS are highly variable, and the diet prices that we obtain can change depending on the day.Increasing cation supplementation in the DG+DCAD diet only increased feed costs by $0.46/ cwt.When calculating milk income, DG+DCAD and SBM were similar ($18.81,$18.82/cwt, respectively) and DG was at $17.18/cwt.When calculating IOFC, the DG+DCAD diet was the highest, followed by DG and SBM ($11.36, $10.18, and $9.65/cwt, respectively).This indicates that replacing mainly soybean meal and soyhulls for high-DDGS is economically beneficial for producers.Although, producers should be cautious as reduced production responses for DG in this study were not as severe as in previous studies (Morris et al. 2018a;Zynda et al., 2022).When deciding on whether to feed DDGS at a high inclusion, it is important to not only consider the price differential of DDGS and soybean meal, but also how milk production responds.For example, Morris et al., (2018a) observed decreases in milk fat and protein concentrations in a diet containing 30% RFDDGS.When calculating IOFC using those production responses along with milk yield, the SBM diet has a better IOFC compared with the DG diet ($10.62 vs. $10.27/cwt).In this case, feeding DDGS to reduce the feed cost was not beneficial as the reduced production responses were more severe.However, feeding high DDGS (replacing soy products) with increased DCAD could be an economically beneficial strategy in practice.

CONCLUSION
This experiment confirmed that MFD occurs when high DDGS are included in a lactating cow diet (30% of dietary DM), and that the high PUFA concentration is a dietary factor associated with the low milk fat.The high S content of DDGS did not appear to have a direct effect on the reduced production responses observed with the DG diet.However, it did have an indirect effect on contributing to MFD.Adding DDGS to a diet decreased DCAD, impairing the acid-base balance of cows, which was likely associated with MFD observed in cows fed the DG diet.The differences in the acid-base balance were obvious when DG+DCAD was compared with DG.Although nutrient digestibility was not improved, increasing DCAD to 300 mEq/kg DM for DG+DCAD eliminated MFD that resulted from the DG diet.We suggest that increasing DCAD (~300 mEq/kg DM) in a high DDGS (>20%) diet can be a useful strategy to lower feed costs and increase incomeover-feed-cost without reduced production responses, such as milk fat depression.
Clark et al.: FACTORS OF FEEDING CORN BYPRODUCTS CAUSING REDUCED PRODUCTION Clark et al.: FACTORS OF FEEDING CORN BYPRODUCTS CAUSING REDUCED PRODUCTION Clark et al.: FACTORS OF FEEDING CORN BYPRODUCTS CAUSING REDUCED PRODUCTION receiving the DG+DCAD diet tended to have greater (P ≤ 0.10) pCO 2 and have increased (P ≤ 0.03) tCO 2 , HCO 3 -, ED, and BE than cows receiving the DG diet.Decreased (P = 0.02) Cl -concentration (P = 0.02) was observed for DG+DCAD compared with DG.
. The variability in DMI could be due to ingredients replaced with DDGS and chemical composition altered by inclusion of DDGS.Benchaar et al. (2013) fed a traditional DDGS and Ramirez-Ramirez et al. (2016) used a traditional DDGS and reduced-fat DDGS in their studies.Both studies mainly replaced soybean meal and corn grain with DDGS and as a result, the DDGS diets increased NDF content at least 3% units compared with a diet with no DDGS.Whereas Morris et al. (2018a) fed a reduced-fat DDGS by replacing soybean meal and soyhulls and had similar NDF contents between the diet with and without reduced-fat DDGS.That study observed a tendency (P = 0.08) for lower DMI when reduced-fat DDGS was Clark et al.: FACTORS OF FEEDING CORN BYPRODUCTS CAUSING REDUCED PRODUCTION
Clark et al.: FACTORS OF FEEDING CORN BYPRODUCTS CAUSING REDUCED PRODUCTION

Table 1 .
Clark et al.: FACTORS OF FEEDING CORN BYPRODUCTS CAUSING REDUCED PRODUCTION Ingredient composition of dietary treatments (% of DM)

Table 2 .
Chemical composition of dietary treatments (DM basis)

Table 4 .
Effects of polyunsaturated fatty acids, sulfur, and corn dried distillers' grains with and without dietary cation-anion difference manipulation on performance of lactating dairy cows

Table 6 .
Effects of polyunsaturated fatty acids, sulfur, and corn dried distillers' grains with and without dietary cation-anion difference manipulation on milk fatty acid profile 1 SBM = soybean meal-based diet; DG = diet containing 29.6% DDGS (DM basis); SBM+S = SBM with increased dietary S; SBM+CO = SBM with corn oil; DG+DCAD = DG with elevated DCAD.

Table 7 .
Effects of polyunsaturated fatty acids, sulfur, and corn dried distillers' grains with and without dietary cation-anion difference manipulation on apparent total-tract digestibility in milk yield.The dis-Clark et al.: FACTORS OF FEEDING CORN BYPRODUCTS CAUSING REDUCED PRODUCTION

Table 8 .
Effects of polyunsaturated fatty acids, sulfur, and corn dried distillers' grains with and without dietary cation-anion difference manipulation on blood urea nitrogen, hemoglobin, electrolytes and acid-base balance 2Least squares means; largest standard error of treatment is shown.