Energy and nitrogen utilization of lactating dairy cattle fed increasing inclusion of a high-protein processed corn coproduct*

Advancing technologies of the corn dry-milling ethanol production process includes the mechanical separation of fiber-containing particles from a portion of plant-and yeast-based nitrogenous particles. The resulting high-protein processed corn coproduct (HPCoP) contains approximately 52% crude protein (CP), 36% neutral detergent fiber (NDF), 6.4% total fatty acids (TFA). The objective of this experiment was to examine the effects of replacing nonenzymatically browned soybean meal with the HPCoP on dry matter intake (DMI), energy and N utilization, and milk production of lactating Jersey cows. Twelve multiparous Jersey cows were used in a triplicated 4 × 4 Latin square de-sign consisting of four 28-d periods. Cows were blocked by milk yield and assigned randomly to 1 of 4 treat-ment diets that contained HPCoP (dry matter [DM] basis) at (1) 0%; (2) 2.6%; (3) 5.4%; and (4) 8.0%. Diets were formulated to be isonitrogenous and thus replace nonenzymatically browned soybean meal with HPCoP in the concentrate mix, while forage inclusion remained the same across diets. Increasing the concentration of HPCoP had no effect on DMI (mean ± SE; 19.9 ± 0.62 kg/d), but tended to linearly increase milk yield (27.8, 28.5, 29.8, and 29.0 ± 1.00 kg/d). Although no difference was observed in the concentration of milk protein with increasing inclusion of HPCoP (3.40% ± 0.057%), the concentration of fat linearly increased with the inclusion of HPCoP (5.05%, 5.19%, 5.15%, 5.47% ± 0.18%). No differences were observed in the digestibility of DM, NDF, CP, TFA, and gross energy averaging 66.6% ± 0.68%, 49.0% ± 1.03%, 66.1% ±


INTRODUCTION
In 2020, the United States supplied 53% of the total global production of grain-based fuel ethanol (McCaherty et al., 2021).This production resulted in 33.1 million metric tons of distiller grains, gluten feed, and gluten meal; together these coproducts contributed approximately $34.7 billion dollars to the nation's Gross Domestic Product (McCaherty et al., 2021).Development of new or specialized coproducts provides diversity in feed offerings originating from grain-ethanol production.High-protein coproduct (HPCoP) is created to be more concentrated in protein through modifications in the production process via sieving and elutriation of coproduct streams (Srinivasan et al., 2005).For example, during corn milling, after fermentation the remaining fiber is mechanically separated from the kernel protein through sieving.The remaining yeast-based nitrogenous particles traveling with the kernel protein fraction to become the HPCoP (Srinivasan et al., 2005).This process differs from historically produced high-protein distiller grains (HPDDGS), as protein is fractionated postfermentation instead of prior.Based on an average of 10 samples collected over a months' time, the resulting HPCoP contains approximately 54% CP and 7.2% total fatty acids (TFA) on a DM basis (Carroll et al., 2022).The CP of the HPCoP is greater than the HPDDGS, containing 39% CP listed within NASEM (2021).Although several high-protein cornmilling coproducts have historically been available, production of the HPCoP are different than previous HPDDGS which were produced through the removal of bran and germ before fermentation.This is in contrast to the current postfermentation process, as the metal fermenting vessel is evacuated, and protein is concentrated via multiple phases of sieving.
Whole-animal energy and N balance experiments have examined both wet distiller grains with solubles (Birkelo et al., 2004) and reduced-fat distiller grains (DDGS; Foth et al., 2015).The replacement of nonenzymatically browned soybean meal (NEBSBM) with HPDDGS has been previously examined in dairy cattle (Brown and Bradford, 2020) but because the chemical composition of the HPCoP differs from that of HPD-DGS traditionally produced, controlled feeding experiments are needed to examine the effects of HPCoP on milk production as well as whole-animal energy utilization.The objective of this experiment was to examine the effects of replacing NEBSBM with the HPCoP on feed intake, energy and N utilization, and milk production in lactating Jersey cows.

Animals and Treatments
The University of Nebraska-Lincoln Animal Care and Use Committee approved animal care and experimental procedures.Twelve multiparous Jersey cows averaging 95 ± 7.3 DIM in parity 2.5 ± 0.5 at the start of the experiment were housed in individual tiestalls in a climate-controlled environment (20°C) at the University of Nebraska-Lincoln Dairy Metabolism Facility in the Animal Science Complex.Stalls were equipped with rubber mats, and cows were milked at 0700 and 1800 h.All cows were less than 134 d pregnant at the end of the last experimental period.
The experimental design was a triplicated 4 × 4 Latin square balanced for carryover effects.The experiment consisted of 4 periods of 28 d in length including 24 d of diet adaptation followed by 4 d of collection (Kononoff and Hanford, 2006).Before the initiation of the experiment, cows were fed a common diet then grouped by milk yield and randomly assigned 1 of 4 treatment diets within a given treatment sequence that differed in the inclusion of a HPCoP (POET, Fairmont, NE).Nonenzymatically browned soybean meal was used as a comparison to the HPCoP, as the CP content of NEBSBM closely resembles that of the HPCoP.Also, while the rumen undegradable protein content of HPCoP is unknown, DDGS also produced from the dry corn-milling process have traditionally been shown to contain more bypass protein similar to NEBSBM (NASEM, 2021).Treatments were as follows: 0% HP-CoP (0CRTL); 2.6% HPCoP on DM basis (2.6L); 5.4% HPCoP on DM basis (5.4M); or 8% HPCoP on DM basis (8.0H).Two concentrate mixes were used in the study where one concentrate mix provided 0% HPCoP and 8% NEBSBM, and the second provided 8% HPCoP and 0% NEBSBM on a DM basis.These 2 concentrate mixes were added in a ratio of 33% and 67% for the 2.6L and 67% to 33% for the 5.4M, whereas basal forage mixture remained the same between treatments (Table 1).Dietary ingredients (corn silage, alfalfa hay, and concentrate) were placed in a Calan Data Ranger (American Calan Inc. Northwood, NH), mixed and fed at 0930 h.Cows were fed each day to a target daily refusal rate of 5%.During the 4-d collection period individual cattle were fed at 100% of the prior week's average intake.

Sample Collection and Analysis
Individual feed ingredients were sampled daily during collection periods and frozen at −20°C.All feed ingredients were composited and dried at 60°C and were ground through a 1-mm screen.(Wiley Mill; Arthur A. Thomas Co.; Philadelphia, PA).A subsample of ground feed was sent to Cumberland Valley Analytical Services Inc. (Waynesboro, PA) for analysis of DM (method 930.15, AOAC International, 2000), N (Leco FP-528 Nitrogen Combustion Analyzer; Leco; St. Joseph, MI), soluble CP (Krishnamoorthy et al., 1982), acid detergent insoluble CP, and neutral detergent insoluble CP (Van Soest et al., 1991;coupled with Leco FP-528 Nitrogen Combustion Analyzer;Leco;St. Joseph, MI), ADF (method 973.18, AOAC International, 2000), NDF with sodium sulfite and α amylase corrected for ash contamination (aND-Fom; Van Soest et al., 1991), lignin (Goering andVan Soest, 1970), sugar (DuBois et al., 1956), starch (Hall, 2009), ash (method 942.05, AOAC International, 2000), minerals (method 985.01, AOAC International, 2000), and TFA (Sukhija and Palmquist, 1988).The HPCoP was also analyzed for AA (method 994.12, AOAC International, 2000).Feed samples were also analyzed for gross energy (GE) content using a bomb calorimeter (Parr 6400 Calorimeter, Moline, IL) at the University of Nebraska-Lincoln.The chemical composition of the diets and feed ingredients are listed in Table 2. Total mixed rations were sampled on d 1 of each collection pe-riod and used to determine particle size using the Penn State Particle Separator (Heinrichs and Kononoff, 2002) and reported in Table 1 on a DM basis (60°C for 48 h).During each collection period refusals were sampled daily and composited on a weight basis.Refusals were analyzed for DM, CP, NDF, aNDFom, starch, ash, fatty acids, and GE according to the same methods as feeds described previously in this section.
Total fecal and urine outputs were collected from each individual cow during the collection period for 4 consecutive days and composited as described by McLain et al. (2021), where total urine output was collected with a bladder catheter.After collections, approximately ~600 g of feces were dried at 60°C for 48 h and ground to pass through a 1-mm screen (Wiley Mill; Aurthur A. Thomas Co.; Philadelphia, Treatments: 0CTRL = 0% high-protein coproduct; 2.6L = 2.64% high-protein coproduct; 5.4M = 5.36% high-protein coproduct; 8.0H = 8% high-protein coproduct.Where L, M, and H translate to low, medium, and high. 2 Alfalfa was ground through a 20.3-cm screen, and corn was ground using a 9.53-mm and 12.7-mm screen.PA).The ground feces were analyzed for DM, CP, NDF, aNDFom, ash, fatty acids, GE, and urine was analyzed for CP using the same methods as described for feeds.Milk production was measured daily while animals were milked in tiestalls, and milk samples were collected during the morning and evening milking of collection periods as described by McLain et al. (2021).Milk samples were then sent to Heart of America DHIA (Kansas City, MO) for near infrared spectroscopy analysis of the percentage of milk fat, protein, and lactose.Each period, milk was composited at the 700h and 1800h milking based upon the weight of milk produced, and samples were analyzed for N as previously described for feeds and fatty acids (Hara and Radin, 1978;Chouinard et al., 1999).To determine BW, cows were weighed before feeding at 0800 h the first day of the collection period.On the last day of each collection period, feed was withheld until BW was determined at 1000 h.

Heat Production and Energy Utilization
Heat production was determined indirectly through 23-h composite gas samples in headbox-type indirect calorimeters as described previously (McLain et al., 2021).However, total volume of gas flow through the headbox was measured using mass flow meters (MCW-1000SLPM-D Whisper, Alicat Scientific) and corrected to standard temperature and pressure (0°C, 101.3 kPa) with adjustment for moisture content of exhaust air (Nienaber and Maddy, 1985).Before experiment initiation, animals were trained and acclimatized to headboxes.To do so, animals were trained with an 8-h period then two 23-h periods within the headboxes.Individual animal water and feed consumptions were monitored with individual water meters, and refusals were weighed after the period within the headboxes.System efficiency (head box and gas analyzer) was determined by burning 100% ethyl alcohol and measuring

Energy Calculations
The respiratory quotient (RQ) was calculated using the ratio of carbon dioxide produced to oxygen consumed (L/L).Methane energy was estimated by multiplying CH 4 production by its enthalpy (9.45 kcal/L).Tissue energy was adjusted according to NRC (1989), where k T is the efficiency of utilizing body-reserve energy for milk production, k G is the efficiency of utilizing ME intake for tissue gain, and k L is the efficiency of utilizing ME for milk synthesis (Moe et al., 1971).Values of 0.66 and 0.74 and 0.89 were used for k L , k G, and k T , respectively (Moraes et al., 2015;Chris Reynolds, personal communication).Energy utilization for pregnancy (fetal_E) was calculated according to NASEM (2021) where: Gravid uterine weight at parturition [GrUter_wt (parturition) ] = calf birth weight (kg) × 1.825.[12]

Statistical Analysis
The UNIVARIATE procedure of SAS (9.4; SAS Institute Inc.) was used to determine outliers from the data set.An outlier was determined if an observation was greater than 2.5 standard deviations from the mean.Data were analyzed with PROC GLIMMIX in SAS (9.4).The model included fixed effect of treatment, period nested in square, and square and the random effect of cow nested in square according to Kononoff and Hanford (2006).A type III ANOVA with Kenward-Rodger denominator degrees of freedom was completed using the PROC GLIMMIX function of SAS.Linear, quadratic, and cubic contrasts were created with the IML procedure in SAS (9.4) due to uneven treatment spacing.All data are presented as least squares means ± largest standard error.Significance was declared with a P-value ≤ 0.05 and trends at a P-value > 0.05 but ≤0.10.

Data Collection and Chemical Composition
One cow consuming the 2.6% DM inclusion of HPCoP contracted mastitis, and data were not used for the first period.The cow responded to treatment and was used to collect subsequent observations.Additionally, one cow refused to drink water from the water bowl situated within the headbox during training before the experiment, and consequently gas data were not collected on this cow.Diet composition of the 4 treatments are listed in Table 1.Dietary CP averaged 16.1% DM and was similar with increasing HPCoP inclusion.Increasing inclusion of HPCoP increased the concentration of TFA from 5.03% ± 0.32% DM in 0CTRL to 5.27% ± 0.46% DM in the 8.0H treatment.Similarly, 18C fatty acids increased from 2.86% ± 0.12% DM in the 0CTRL to 3.07% ± 0.14% DM in 8.0H.Chemical composition of corn silage, alfalfa hay, concentrate mixes, and HP-CoP are listed in Table 2. On a DM basis the HPCoP contained 36.2% ± 1.63% NDF, 52.4% ± 0.35% CP, 6.44% ± 0.099% TFA.

N Utilization
No difference was observed in N intake averaging 513 ± 16.81 g/d with increasing HPCoP inclusion (Table 5).In comparison, fecal and urinary N excretion tended to increase linearly with increasing HPCoP from 167.5 to 177.2 ± 5.92 g/d and 125.8 to 138.5 ± 7.41 g/d, respectively.No difference was observed for milk N excretion.Urinary N as a proportion of total N intake decreased quadratically from CRTL (25.8% ± 1.77%) to 2.6L (23.0%± 1.77%) and then increased to 8.0H (27.1% ± 1.77%).

Milk Yield and Composition
Dry matter intake averaged 19.9 ± 0.64 kg/d (Table 6), and no difference was observed among treatments.Milk yield tended to increase linearly with increasing HPCoP inclusion from 27.8 to 29.0 ± 0.87 kg/d from 0CTRL to 8.0H.No difference was observed in the concentration of milk protein which averaged 3.39% ±   Treatments: 0CTRL = 0% high-protein coproduct; 2.6 = 2.64% high-protein coproduct; 5.4M = 5.36% high-protein coproduct; 8.0H = 8% high-protein coproduct; where L, M, and H are used to denote low, medium, and high.0.057%; however, milk protein yield tended to increase from 0.93 to 0.99 ± 0.033 kg/d with increasing HP-CoP inclusion.Milk fat percentage increased linearly from 5.05% to 5.47% ± 0.288%, while milk fat yield increased linearly from 1.40 to 1.58 ± 0.051 kg/d.The same response was also observed as ECM increased linearly from 34.3 to 37.4 ± 0.057 kg/d.Concentration of C16:0 in the milk tended to decrease linearly with increasing HPCoP inclusion from 38.5 to 37.7 ± 0.582 g/100 g of fat (Table 7).The concentration of C18:2 in milk increased linearly from 0CTRL to 8.0H (1.98 to 2.35 ± 0.100 g/100 g of fat).The concentration of <16 carbon milk fatty acids averaged 25.3 ± 0.36 g/100g of fat, and no difference was observed among treatments.Milk fatty acids >16 carbons tended to increase linearly from 32.0 to 32.9 ± 0.72 g/100 g of fat.Trans-10 cis-12 conjugated linoleic acid was not detected in any of the milk samples.

Chemical Composition
The objective of this experiment was to examine the effects of NEBSBM with a new HPCoP and to examine the effects on whole-animal nutrient utilization.The production of HPCoP is unique because a portion of fiber is removed through sieving and this concentrates the protein stream resulting in a feed with more CP (Birkelo et al., 2004;Srinivasan et al., 2005).Additionally, nutrients are also decanted to the fermenting vessel leaving behind corn protein which also contains spent yeast cells (Shurson, 2018).This is different from traditional HPDDGS which is usually produced when the hull and germ are removed before fermentation.Although the exact content of yeast cells within the new HPCoP is unknown, but is estimated at 22% based on other HPCoP produced through similar methods (Nazeer et al., 2023).Based upon data published in NASEM (2021) high-protein DDGS typically contain 38% NDF, 6.6% TFA, and 39% CP of which approximately 2.8% is Lys.The test high-protein coproduct in the current study contained similar concentrations of NDF (36.2% ± 1.63%) and TFA (6.44% ± 0.099%), but more CP (52.4% ± 0.35% CP) and this protein contained more lysine (3.39% ± 0.342% CP).The NDF content (36.2% ± 1.63%) of the HPCoP is increased relative to the 30.8%NDF for reduced-fat DDGS and 11.1% for NEBSBM (NASEM, 2021).Relative to the production of DDGS fiber is removed during the sieving process of the HPCoP, but surprisingly this is not reflected by a decreased NDF content.However, NDF is not a direct measure of fiber and other constituents may still be present within the residue but in the current study this was not explored further.

Energy Utilization
In evaluating the HPCoP as a new feedstuff we chose to employ methods of indirect calorimetry.Increasing the dietary inclusion of HPCoP increased the GE content of the diet from 4.25 Mcal/kg in the zero control to 4.31 Mcal/kg in the diet containing 8% HPCoP.This increasing response was largely driven by the  (Tyrrell and Reid, 1965).
in tissue energy which increased in the 2 intermediate treatments but then decreased when cows consumed the highest concentration of HPCoP.This quadratic response was also observed in RQ.The paired observation of the highest RQ and tissue energy on intermediate treatments could be interpreted to mean that more energy was being diverted to adipose tissue, similar to a response observed by (Morris et al., 2020).Energetic conversion of ME to NE L is dependent on several factors including diet chemical composition, nutrient flux, and metabolic status of the animal.However, fixed conversion efficiencies, such as 0.66 in the dairy model of NASEM ( 2021), are used in nutrition models.Given the factors affecting this conversion we know such fixed assumptions may not accurately represent the mechanisms and interactions associated with whole-animal energetics.This is supported by observations in the current experiment in which the conversion of ME to NE L was affected by diet and increased with the inclusion of the HPCoP from 0.689 in the control to 0.714 in the 8.0H treatment.We suggest that the response may be due to one or both of the following reasons: 1) increasing TFA content in the diets; this is because the incorporation of preformed fatty acids into milk is more efficient than de novo lipogenesis (Rico et al., 2014;Boerman et al., 2015;Morris et al., 2020), 2) to an increasing milk fat to protein ratio from 1.51 to 1.60 for 0CTRL and 8.0H, respectively.This shift in production from protein to milk fat may have resulted in reduced heat load, as one kilogram of milk protein synthesis produced 6.2 Mcal of heat relative to 1.8 Mcal from one kilogram of milk fat (Morris et al., 2021).

Milk Production
Although no difference was observed on DMI, inclusion of HPCoP tended to increase milk yield and significantly affected milk fat percent and yield and in turn ECM.As noted in the "Energy Utilization" section, it is likely that the response in ME: NE L was supported by the partition of feed energy to milk fat synthesis (Boerman et al., 2015).We did not anticipate the increase in milk fat within the current experiment.This is because factors such as effective fiber (Zebeli et al., 2008) were not manipulated and fiber digestibility was not affected (Oba and Allen, 1999).We suggest a portion of the increase in milk fat could be supported by the increase in intake of C18 fatty acids with increasing inclusion of the HPCoP.Compared with the zero control, intake of C16 and C18 carbon fatty acids increased by 18 and 62 g/d respectively in cows consuming the 8.0H diet.However, the observed digested C16 did not differ across treatments, but digested C18 fatty acids increased 60 g with increasing inclusion of the HPCoP.If all C18 preformed fatty acids were used within the mammary gland this would only account for approximately a third of the observed increase in milk fat and would not explain the 46 g/d increase in < 16C fatty acids and 62 g/d increase in 16C fatty acids observed.The 16C fatty acids can originate from both dietary energy and de novo synthesis (Loften et al., 2014).Because digestible C16 did not differ, dietary energy may have driven the increase in both < 16C and 16C fatty acids.Although the molecular source of this energy is unclear and needs further elucidation, we also speculate the yeast within the HPCoP could have stimulated milk fat production as seen in other experiments (Desnoyers et al., 2009).The increase in the concentration of fat from HPCoP resulted in an increase in energy and preformed fatty acids likely contributing to the overall increase in milk fat.
Increasing the HPCoP in the diet did not affect the concentration of milk protein but did tend to increase the yield of milk protein.Similar to milk fat, this response was not expected as a reduction in milk protein yield in diets containing DDGS is often attributed to a decrease in the supply of lysine relative to other protein sources (Nichols et al., 1998).Using dietary chemical composition and animal responses observed in the current study we evaluated the predictions of the intake and supply of AA with the NASEM (2021) model.We compared the models "entered efficiency" to the "predicted efficiency."NASEM (2021) suggests that if the observed entered efficiency was greater than the target efficiency, that the AA may be in short supply.The model suggested that the 8.0H diet was more limiting than the control for His, Lys, Met, and Trp which does not explain why we observed a trend for an increase in milk protein yield.It should also be noted that differences in ruminal CP digestibility of HPCoP and NEBSBM may have been present but we speculate that the increase in milk protein yield was at least in part due to the increase in energy supply.The fine particle size of HPCoP that makes it difficult to measure the rumen degradability of the feed and compare it to NEBSBM (NASEM, 2021).Therefore, further characterization of the HPCoP is needed to effectively compare the effects of HPCoP relative to other feed ingredients.
Carroll et al.: CORN COPRODUCT AFFECTS MILK ENERGYincreased the conversion of energy from ME to NE L and milk fat production.This response may have been supported by increasing dietary fat provided by the HP-CoP and the subsequent energetic efficiency.Results indicate that the high-protein corn-milling coproduct can be effectively used in diet formulation similar to NEBSBM.

Table 1 .
Ingredient inclusion and chemical composition of experimental diets of lactating Jersey cattle fed increasing inclusion of a high-protein corn-milling coproduct (HPCoP; % of diet DM)

Table 3 .
Increasing inclusion of HPCoP linearly increased dietary GE from 4.25 to 4.31 ± 0.007 Mcal/kg, however no difference was observed in the concentration of either DE or ME.A linear trend was observed for dietary NE L concentration increasing from 1.71 to 1.81 ± 0.055 Mcal/ kg.These same effects were also reflected in measures of DE, ME, and NE L supply.The ratio of NE L to ME increased linearly with increasing HPCoP inclusion from 0.689 to 0.714 ± 0.0172.Milk energy increased linearly with increasing HPCoP inclusion from 23.7 to 26.1 ± 0.764 Mcal/d.No difference was observed in O 2 consumption, CO 2 production, and CH 4 production.However, a quadratic response in RQ was observed with an increase from 0CTRL (1.019 ± 0.011) to 5.4M Carroll et al.: CORN COPRODUCT AFFECTS MILK ENERGY

Table 4 .
Carroll et al.:CORN COPRODUCT AFFECTS MILK ENERGY Apparent total-tract digestibility of nutrients of lactating Jersey cattle fed increasing inclusion of a high-protein corn-milling coproduct 2Least squares means; largest SEM is listed.

Table 5 .
Fecal output, urine output and N utilization of lactating Jersey cows fed increasing inclusion of a high-protein corn-milling coproduct

Table 6 .
Carroll et al.:CORN COPRODUCT AFFECTS MILK ENERGY DMI, milk production and milk composition, water intake, BW and BCS of lactating Jersey cattle fed increasing inclusion of a highprotein corn-milling coproduct

Table 7 .
Milk fatty acid composition of lactating dairy cows fed increasing inclusion of a high-protein corn-milling coproduct