Performance and ruminal fermentation of second-trimester pregnant beef cows fed short-season high-moisture corn stover or barley greenfeed during winter in western Canada

Abstract This study evaluated performance and ruminal fermentation for cows fed short-season high-moisture shelled corn stover with corn dried distillers’ grains with solubles (COR) or barley greenfeed (BAR) that was either swathed (S-COR; S-BAR; Experiment 1) or baled (B-COR; B-BAR; Experiment 2) as winter-feeding systems. In Experiment 1, cows were randomly assigned to S-COR or S-BAR and fed for 52 days in the fields where crops were grown. Body weight (BW), subcutaneous rib and rump fat, and body condition score (BCS) were measured, with no differences detected between treatments (P > 0.05). However, S-COR had a lower estimated dry matter intake (DMI) than S-BAR (P ≤ 0.03). In Experiment 2, cows were assigned to B-COR or B-BAR for 42 days and fed in field paddocks. Cows fed B-COR had less (P ≤ 0.01) DMI, final BW, rib fat, rump fat, and BCS than B-BAR, with no differences (P > 0.05) for ruminal pH. Total SCFA concentration was greater (P ≤ 0.05) on day 21 for B-BAR than B-COR, but not on day 42. Under western Canadian conditions, COR may reduce DMI and the performance of pregnant cows, suggesting that additional preservation and supplementation strategies should be investigated.


Introduction
Extensive winter grazing systems have gained popularity in western Canada, with many producers utilizing bale grazing, crop residue grazing, or swath grazing (McGeough et al. 2018;Western Beef Development Centre 2018).Additionally, the development of low-heat unit short-season corn (Zea mays L.) varieties have contributed to the growing trend of grazing standing corn as a winter-feeding strategy (Lardner et al. 2017;Jose et al. 2020;Anderson et al. 2023).Generally, wholeplant corn has greater yield and energy concentration than whole-plant barley (Hordeum vulgare L.), the latter being a popular option for swath grazing, but corn has less crude protein (CP) (Lardner et al. 2017;Jose et al. 2020).Cows grazing standing short-season whole-plant corn have been reported to eat less dry matter (DM) without differences in body weight (BW) or body condition as compared to those grazing swathed whole-plant barley (Jose et al. 2020).This supports the concept that whole-plant short-season corn has greater energy density than whole-plant barley.
In the United States, corn stover has been widely utilized, and there is extensive knowledge on harvesting and grazing systems (Updike et al. 2015;King et al. 2017;Conway et al. 2020), time of use (Gutierrez-Ornelas and Klopfenstein 1991), and its feed value (Redfearn et al. 2019).In fact, it has been estimated that over 4 million ha/year of corn stover are grazed by livestock, and an additional 0.81 million ha/year are mechanically harvested in the United States (Schmer et al. 2017).With increasing production of short-season corn and further development of short-season corn varieties, corn grain production and the use of corn stover in western Canada are anticipated to increase.However, exposure to early frost or drought may limit short-season corn productivity, alter corn stover composition, and impact the management of the stover.That said, growing seasons with adequate temperatures and moisture may warrant the harvest of corn grain (Guyader et al. 2018;Ferreira et al. 2021;Major et al. 2021).
To help minimize environmental risk, strategies for early corn grain harvest may utilize high-moisture grain.Harvesting short-season corn grain at approximately 30% kernel moisture shortens the growing season for grain production and increases neutral detergent fiber (NDF) digestibility and the CP concentration of stover (Masoero et al. 2006;Owens and Basalan 2012).The resulting stover could be utilized as a winter-feeding strategy for cattle in western Canada, but the CP and energy density may necessitate supplementation, such as with the use of corn-dried distillers' grains with Fig. 1.Field sizes and arrangements of corn (COR) and barley (BAR) in Experiments 1 and 2. Snaplage and silages were not used in this study.solubles (DDGS; NASEM 2016).Despite the growing potential for corn harvest and production of corn stover in western Canada, to the authors' knowledge, there are no published research results evaluating the performance of cows fed shortseason corn stover.
Therefore, we hypothesized that utilizing stover remaining after harvest of short-season corn, whether swathed (S-COR; Experiment 1) or baled (B-COR; Experiment 2), along with supplemental corn DDGS, would decrease DM intake (DMI) but not impact the body condition of second-trimester multiparous cows when compared to utilizing barley greenfeed that was swathed (S-BAR; Experiment 1) or baled (B-BAR; Experiment 2).Additionally, feeding short-season corn stover bales supplemented with corn DDGS would not affect ruminal total short-chain fatty acid (SCFA) concentration, SCFA proportions, ammonia concentrations, or pH.

Materials and methods
All procedures followed the Canadian Council of Animal Care guidelines (CCAC 2009) and were approved by the Uni-versity of Saskatchewan Animal Research Ethics Board (protocol No. 20200085).This study consisted of two experiments conducted at the Livestock and Forage Centre of Excellence (Clavet, SK, Canada).
fertilizer and herbicide applications for each crop are shown in Table 1.
In the first year, whole-crop barley was swathed at the hard-dough stage, as described by Rosser et al. (2013).Corn was harvested for high-moisture shelled corn when kernel moisture reached 30%, and the remaining stover was windrowed.Environmental losses (high winds and wildlife scavenging) between swathing and grazing time resulted in a substantial loss of stover.As such, in the second year, barley greenfeed was swathed, wilted, and baled (605 ± 31.3 kg DM/bale; 1.52 m × 1.83 m round bales) at the hard-dough stage.Corn was harvested for high-moisture shelled corn when kernel moisture reached 30%, and the remaining stover was baled into large square bales (323 ± 38.6 kg DM/bale; 0.91 m × 0.81 m × 2.29 m square bales).Propionic acid was applied as a mold inhibitor (New Holland Agriculture Crop-Saver; Harvest Tec, Hudson, WI, USA) to the corn stover at the time of baling at a rate of 6.58 kg/ton of forage.Bales were removed from the field, weighed, transported to the feeding site, and stored on dry ground with no cover from the environment.
Experiment 1: experimental design, cow management, diets, and dry matter intake Using a completely randomized design, 60 pregnant, multiparous beef cows (663 ± 51.6 kg, 133 ± 9.5 days pregnant; 10/cows/paddock/treatment) were randomly allocated by BW to one of two treatments, with field grazing initiated on 25 November 2020.Replicate paddocks (n = 3) per forage treatment were constructed using one-strand of electrified fencing in each of the S-BAR and S-COR fields (Table 2).Cows had access to water in insulated portable troughs that were filled and checked daily for ice formation at the trough openings.A wooden temporary windbreak (20% porosity) was placed in each field to provide shelter from the prevailing northwest winds, and straw bedding was provided as necessary on the southeast side of the windbreak.All cattle were supplemented weekly with a commercially available freechoice mineral and vitamin supplement (100 g/cow/day; CO-OP 2:1 Beef Cattle Range Mineral, Federated Co-operatives Ltd., Saskatoon, SK, Canada) and salt (50 g/cow/day) in rubber tubs placed near the windbreaks.The supplement ‡ Calculations for TDN are described by Weiss et al. (1992), and calculations for NE m and NE g are described by NASEM (2016).
provided 1.6 g Ca, 0.8 g P, 0.4 g Na, 0.5 g Mg, 10 100 mg Zn, 150 mg I, 3900 mg Fe, 3500 mg Mn, 3050 mg Cu, 50 mg Co, 3000 mg F, 30 mg, Se, 600 000 IU Vitamin A, 60 000 IU Vitamin D, and 3000 IU Vitamin E per kg of DM.Cows grazing S-COR were provided a corn DDGS (0.50 kg DM/cow/day) supplement in troughs near the windbreak daily to meet the nutrient requirements of second trimester beef cows (668 kg) to gain one body condition score (BCS) in 45 days when NDF intakes were limited to 1.2% of BW (NASEM 2016).When the combined ambient and wind chill temperatures were equal to or less than −30 • C, DDGS supplementation was increased to 1.75 kg DM/cow/day for cows grazing S-COR treatment.A similar supplement strategy was used for those cows grazing S-BAR with DDGS provided at 0.50 kg/cow/day using the same model parameters as above to meet increased energy demands (NASEM 2016).Access to swaths was controlled by moving the portable electric fencing, allocating approximately 3 days of new forage to graze with each movement.
Cows were able to backgraze to access shelter, water, and supplements.Pre-grazing forage availability and post-grazing residue were estimated by collecting and weighing the forage in a set area (S-BAR, 3 m of swath; S-COR, 7.62 m wide, including swath × 4.57 m long) for every 0.81 ha of the paddock, resulting in at least 10 samples/paddock.Pre-grazing measurements were taken on 27 October 2020, before heavy snow covered the swaths.Post-grazing measurements were taken on 14 April 2021, after snow and ice had melted in the fields.Pre-grazing measurements were used to estimate yield, and both pre-and post-grazing measurements were initially used to estimate DMI, as described below.Samples for laboratory analyses were collected and composited by collection date and paddock.The amount of corn DDGS, salt, and min-eral and vitamin supplements delivered was recorded.There were no refusals of corn DDGS, and minimal refusals of the salt and mineral and vitamin supplements.Refusal amounts of the supplement were weighed back, and a subsample was taken to determine DM.Forage and DDGS samples were dried in a forced-air oven at 55 • C for 3 days to determine DM and ground to pass through a 1 mm screen using a Christy & Norris hammer mill (Christy Turner Ltd., Ipswich, UK).Samples were analyzed for CP, NDF, acid detergent fiber (ADF), starch, ether extract, Ca, and P by a commercial laboratory (CVAS; Cumberland Valley Analytical Services, Waynesboro, PA, USA) as described by Rosser et al. (2016), except NDF was expressed on an ash-free basis after discounting the ash remaining in NDF residue after 2 h at 535 • C (aNDFom).Acid detergent insoluble nitrogen was analyzed by CVAS by pressing and wrapping the total residue from the ADF procedure in foil and analyzing it for nitrogen (Leco FP-528 Nitrogen Combution Analyzer; Leco, St. Joseph, MI, USA), and then multiplying by 6.25 to estimate acid detergent insoluble CP (ADICP).Available CP was calculated as CP% × [100 -(ADICP% − 10%)] when ADICP was greater than 10% of CP.Total digestible nutrients (TDNs) were calculated using the equations of Weiss et al. (1992).Net energy for maintenance (NE m ) and gain (NE g ) were calculated according to NASEM (2016).Salt and mineral supplement DM was estimated at 99%, and refusals were dried to determine DM as described above.DMI in each paddock was initially calculated as (DM pre-grazing--DM post-grazing)/(cows in paddock × days on trial) with adjustment for cows removed from trial (one on day 7 and another on day 25 in paddock 2 and 1, respectively) on the S-COR treatment (Jasmer and Holechek 1984).This resulted in DMI values that were unrealistically large for the S-COR treatment, most likely from environmen- tal losses in the fields that could not be measured.Losses due to high winds, migratory birds, and scavenging by local deer were observed between pre-grazing measurements and the start of the experiment.Heavy snows on 7th and 8th November 2022 did not allow for further measurements before the experiment started.This resulted in a snowpack that was approximately 19 cm at the beginning of the experiment, and further accumulation of snow through the experiment resulted in a 26 cm snowpack at the end of the experiment (Environment and Climate Change Canada 2022).Therefore, forage DMIs for both S-BAR and S-COR were estimated using average cow BW in the paddock and the NE m requirements and dietary concentrations, as described in NASEM (2016) and Kelln et al. (2011).Nutrient composition of treatments based on estimated intakes of forage and actual intakes of DDGS, salt, and supplement were calculated (Table 3).A period of above freezing temperatures (Fig. 2) and rain followed by high winds and below freezing temperatures on 16 January 2021 made swaths inaccessible to cows, resulting in an early end to the experiment.

Body weight, ultrasound, and BCS measurements
Cows were individually weighed on two successive days, immediately before the start of the experiment and again at the end of the experiment.The averages of the 2-day weights were considered starting and ending BW. End-of-study measurements were conducted 52 days after the start of the experiment.Cows were weighed every 21 days to monitor performance and adjust feed allocation.The date of calving was recorded the following spring, and these data were used to adjust the starting and final BW for the estimated conceptus weight (Kelln et al. 2011;NASEM 2016).
Subcutaneous fat thickness was measured by ultrasound (SSD-500 Echo Camera; Aloka Co., Ltd., Tokyo, Japan) between the 12th and 13th ribs (rib fat) and at the intersection of the deep gluteus medius and superficial gluteus medius muscles (rump fat) at the start and end of the trial.Additionally, BCS was assessed at the start and end of the trial by two researchers blinded to treatments using a 1 (emaciated) to 5 (obese) scale with 0.5-point increments (Morris et al. 2002).The average of the two evaluator scores was considered the starting or ending BCS value for each cow.
Experiment 2: experimental design, cow management, diets, and dry matter intake The second experiment was a completely randomized design with six field paddocks (six pregnant multiparous cows, including one ruminally cannulated cow/paddock; 667 ± 63.2 kg, 126 ± 20.3 days pregnant) per treatment.The cows were assigned to groups to balance for BW, and groups were randomly allocated to one of two treatments in field paddocks.Cows in each paddock had free access to a heated, float-activated water trough.Shelter from prevailing north-west winds was provided by wooden windbreaks (20% porosity).Straw bedding was provided as necessary on the south-east side of the windbreaks.Treatments included diets based on B-BAR or B-COR forage, with the experiment initiated on 30th November 2021 (Table 2).Bales were individually cored, weighed, and placed in hay rings for consumption as needed, with the bale placement being recorded.Hay rings were moved randomly when new bales were added to distribute manure throughout the paddock.Corn DDGS were supplemented in portable troughs to cows assigned to B-COR at 1.96 kg DM/cow/day and increased to 2.80 kg DM/cow/day when the combined ambient temperature and wind chill were at or below −30 • C as in Experiment 1 (NASEM 2016).Corn DDGS supplement was delivered at ap-proximately 0900 h each morning, except during sampling days when cows were supplemented at 1200 h due to staff availability.Cows fed B-BAR forage were not supplemented with DDGS.Bale cores were analyzed for DM, as previously described.Cores were then composited by treatment to generate two-week composites.Cores and DDGS were analyzed for nutrient content, as previously described.All cows received 50 g DM/day of salt and 100 g DM/day of a mineral and vitamin supplement (NLM Re-Charge 2:1 Beef Cow Premix; New-Life Mills, Listowel, ON, Canada) weekly in rubber tubs located near the windbreaks.The supplement provided 1.60 g Ca, 0.75 g P, 0.6 g Na, 0.4 g Mg, 6000 mg Zn, 100 mg I, 4800 mg Mn, 2500 mg Cu, 15 mg Co, 2000 mg F, 30 mg Se, 500 000 IU vitamin A, 50 000 IU vitamin D, and 4500 IU vitamin E per kg of DM.Salt and supplement were estimated to be 99% DM, with no refusals.The end of the experiment was declared when barley greenfeed bales were all fed on 10th January 2022 (42 days of feeding).After spring thaw, all visible forage refusals throughout the paddock were raked by hand and loaded onto a platform scale mounted to a sled, where each load was weighed and sampled.Load samples were combined by paddock and analyzed for DM and nutrient concentrations.DMI was calculated as (DM delivered − DM refused)/(6 cows × 42 days on trial), and the nutrient composition of total intake by treatment was calculated (Table 3).Sorting indices for NDF and ADF were calculated using the methods of Leonardi and Armentano (2003), where values equal to 100% indicated no sorting, less than 100% indicated selective refusals of a nutrient, and more than 100% indicated preferential consumption of a nutrient.

Ruminal fermentation characteristics
The ruminally cannulated cows were utilized to characterize ruminal pH, SCFA, and ammonia concentrations.Ruminal pH was measured by inserting an indwelling ruminal pH meter (Dascor, Escondido, CA, USA) into the ventral sac on day 0 and measuring ruminal pH every 30 min from day 14 through day 20.While in the rumen, debris broke the glass on four sensors, and one probe had a faulty battery.These data were removed from analysis, resulting in four working sensors in cows fed B-BAR forage and three sensors in cows fed B-COR forage.Threshold evaluation at pH 5.8 and 5.5, standardization methods, and calculations are described by Penner et al. (2007), Joy et al. (2021), and Carey et al. (2023), respectively.For ruminal SCFA and ammonia concentrations, 750 mL of digesta samples were obtained from the caudal ventral, central, and cranial sacs at the fluid-rumen mat interface on days 21 and 42 before DDGS supplementation, when all cows were weighed.Digesta was strained through two layers of cheesecloth, and two 10 mL subsamples were mixed with 2 mL of 25% w/v metaphosphoric acid or 1% H 2 SO 4 , respectively.Subsamples were stored at −20 • C until analysis of SCFA concentration using gas chromatography (Agilent 6890, Mississauga, ON, Canada; Khorasani et al. 1996) and ammonia concentrations using a phenol-hypochlorite assay (Broderick and Kang 1980) modified for a plate reader.

Statistical analysis
Data from both experiments were analyzed independently using the MIXED procedure of SAS (version 9.4; SAS Institute, Cary, NC, USA) as completely randomized designs.Paddock was considered an experimental unit.The fixed effects of the models included diet; random effects of paddock were included in the model; and denominator degrees of freedom were calculated using Satterthwaite approximation.The data were checked for normality using the Shapiro-Wilk test and met normality assumptions.Repeated measures were applied to ruminal pH, SCFA, and ammonia concentrations in Experiment 2 with the single ruminally cannulated cow in each paddock as the subject.Ruminal pH data were averaged for each hour in a day.Effects of hour (pH) or day (SCFA and ammonia) and their interaction with diet were added to the fixed effects of the model, and covariance structures used were determined based on the one that yielded the least Akaike and Bayesian information criterion values.The interaction of diet × hour and hour, while significant, did not change the ranking of treatments and only reflected the magnitude of the response.As these changes were considered to have limited novelty when analyzing ruminal pH, only diet effects are presented.When appropriate, mean separation was performed using the LSMeans option of SAS using the Tukey's honestly significant differences test.Means were declared different when P ≤ 0.05.Additionally, sorting indices were also compared to 100 (no sorting) with a two-tailed t-test, and sorting was declared when P ≤ 0.05.

Results
Experiment 1: swath grazing barley greenfeed and high-moisture shelled corn stover Crop yield for S-BAR was greater (P ≤ 0.02) on an as-is and DM basis than stover yield for S-COR.Crop and stover DM at harvest were greater (P < 0.01) for S-BAR than S-COR.Daily DMI (kg/day and %BW) was greater (P ≤ 0.03; Table 4) for cows grazing S-BAR than S-COR, but daily predicted NDF intake as a percentage of BW did not differ (P = 0.20).There were no differences (P ≥ 0.14) in BW, rib fat, rump fat, or BCS measurements between cows grazing either S-BAR or S-COR.
Experiment 2: feeding baled barley greenfeed and high-moisture shelled corn stover Crop yield for B-BAR was less (P = 0.03) than stover yield for B-COR on an as-is basis, but not on a DM basis.Crop and stover DM at harvest were greater (P < 0.01) for B-BAR than B-COR.For cows fed B-COR, daily DMI was less (P < 0.01; Table 5) than those fed B-BAR, with no differences (P ≥ 0.16) in NDF intake, forage NDF sorting index, forage ADF sorting index, or percentage of forage refused.Both B-BAR and B-COR cows sorted against NDF (P ≤ 0.04), but only B-COR cows sorted against ADF (P < 0.01).There were no differences (P ≥ 0.63) in starting BW, rib fat, rump fat, or BCS between treatments.BW on days 21 and 52 and ending rib fat, rump fat, and BCS were lesser (P ≤ 0.04) for cows fed B-COR than B-BAR.There were no differences (P ≥ 0.59; Table 6) in ruminal pH measurements between diets.Diet × day interactions (P ≤ 0.05; Table 7) were found for total SCFA concentrations, the molar proportion of propionate, and the acetateto-propionate ratio.Total SCFA concentration was greater (P ≤ 0.05) on day 21 for B-BAR than B-COR, but not on day 42.Cows fed B-BAR had a greater acetate-to-propionate ratio on day 21 than on day 42, which did not differ (P > 0.05) from B-COR on either day.Cows fed B-COR had a greater (P ≤ 0.05) proportion of propionate on both sampling days than those fed B-BAR, and those fed B-BAR had lesser (P ≤ 0.05) proportions of propionate on day 21 than on day 42.While cows fed B-COR had greater (P < 0.01) acetate proportions than those fed B-BAR, B-BAR-fed cows had greater (P < 0.01) proportions of butyrate, valerate, and caproate and concentrations of ammonia than cows fed B-COR.There were no differences (P ≥ 0.06) between treatments in proportions of isobutyrate or isovalerate.

Discussion
While corn stover systems have been described in predominant corn-growing regions of the USA (Larson et al. 2009;Funston et al. 2010;Watson et al. 2015;Meteer et al. 2018;Petzel et al. 2018;Lehman et al. 2021), we are unaware of previous research evaluating the use of corn residues in western Canada.Weighted average harvested estimates for 19 of the major corn-producing states in the United States have been reported to be 3.6 t/ha as-is (Schmer et al. 2017).Our observed harvested yield was 4.20 t/ha, as-is for B-COR; however, the yield estimates presented by Schmer et Table 4. Barley and stover yield (n = 3) and pregnant, multiparous beef cow performance when grazing barley greenfeed swaths (S-BAR; n = 3) or high-moisture corn stover swaths (S-COR; n = 3) over 52 days of winter-feeding in Experiment 1. † Dry matter intake was estimated using equations based on net energy for maintenance of the diet ingredients and average cow BW on a paddock basis (Kelln et al. 2011; NASEM 2016).‡ BW has been adjusted for conceptus weight (NASEM 2016).
§ Subcutaneous fat thickness between the 12th and 13th ribs.
Subcutaneous fat thickness at the intersection of the deep gluteus medius and superficial gluteus medius muscles.¶ Body condition score was judged using a 1 (emaciated) to 5 (obese) scale.Note: DDGS, corn dried distillers' grains with solubles; SEM, standard error of the means; NDF, neutral detergent fiber; ADF, acid detergent fiber; BW, body weight.† BW has been adjusted for conceptus weight (NASEM 2016).‡ Calculations described by Leonardi and Armentano (2003), where values equal to 100% are no sorting, less than 100% are selective refusals, and more than 100% are preferential consumption.No DDGS supplementation was refused, so sorting indices apply only to the forages.§ Subcutaneous fat thickness between the 12th and 13th ribs.
Subcutaneous fat thickness at the intersection of the deep gluteus medius and superficial gluteus medius muscles.¶ Body condition score was judged using a 1 (emaciated) to 5 (obese) scale.a Values with a superscript are different from 100.0% (no sorting; P ≤ 0.05) using a two-tailed t-test.* P-values describe the diet effect, day effect, and diet × day interaction.a,b,c Different superscripts denote significantly different means (P ≤ 0.05) within row when diet × day interaction is significant.Mean separation was performed using the LSMeans option of SAS (version 9.4) with a Tukey adjustment.
al. ( 2017) were influenced by partial harvest, approximately 50% of the residue, to ensure that plant organic matter was returned to the soil.Partial harvest was not a consideration in Experiment 2, as measuring total yield was a primary goal.The observed corn stover yields in the present study, when combined with data from the United States, suggest that if an adequate yield of corn grain is present to warrant corn harvest, the remaining stover may justify utilization by cattle.
Although initial stover yields were reasonable at harvest, we observed substantial field loss due to wind and wildlife scavenging, which likely contributed to the unreasonably high DMI values that were calculated when comparing original residue weights to residue weights following grazing in Experiment 1.The loss of corn stover due to environmental conditions is supported by previous research in Germany, highlighting that leaving corn residue for grazing rather than harvesting can result in losses as great as 5.6 t/ha (Fleschhut et al. 2016).These environmental challenges were the primary reason for baling barley and corn stover in Experiment 2, potentially limiting low replication in Experiment 1.Additionally, whole-crop barley green feed yields were larger than stover yields when severe drought was not a factor, but, in 2021, crops were subject to severe drought, and there were no differences in DM yield when comparing corn stover and barley greenfeed.While we cannot separate drought and postharvest management, these data suggest that during severe drought years, stover yields may be less affected than barley greenfeed yields, particularly when losses are minimized by baling.
The short growing season and limited corn heat units available in some years (Lardner et al. 2017;Ferreira et al. 2021;Major et al. 2021) are a major challenge for corn production in western Canada, and frost routinely prevents corn from reaching full maturity (Daynard 1978;Major et al. 2021).In addition to an immature plant, whole-plant and residue drying are limited by freezing temperatures (Daynard 1978).This is further evidenced by the low DM concentration in B-COR forage.Although direct comparisons are not possible, there were numerical differences between years in the composition of corn stover and barley greenfeed.The most notable numerical difference in barley greenfeed was the lower starch content in Experiment 2 compared to Experiment 1 (8.03 vs. 20.33%DM).This difference in starch most likely reflects the incomplete grain fill in Experiment 2 due to the severe drought in 2021 (Jamieson et al. 1995).Otherwise, nutrient composition was numerically similar in Experiments 1 and 2, suggesting baling the comparatively dry barley greenfeed does not change composition compared to swathing.Conversely, numerical differences in the corn stover between experiments were apparent for DM, ADICP (and subsequently, available CP), NDF, ADF, and starch.DM was greater (77.42 vs. 70.95%) in the swathed stover than the baled stover, most likely due to more extensive drying in the field (Bernardes et al. 2018).Lesser concentrations of NDF (59.93 vs. 68.40%DM) and greater concentrations of starch (8.27 vs. 2.23% DM) in Experiment 1 stover compared to Experiment 2 probably arise due to the drought in 2021 (Ferreira et al. 2021) coupled with higher proportions of less digestible anatomical fractions such as stalks as opposed to leaves and dropped ears picked up by the baler.In Experiment 2, on the day before baling, hand-grab sub-samples (n = 10) from each corn field (composited to n = 1) were analyzed for nutritional composition to predict DDGS supplementation needs (data not shown).Dry matter was measured at 52.7 ± 4.72%, CP and available CP at 10.37 ± 1.305% DM, ADICP at 9.67 ± 1.079% CP, NDF at 64.60 ± 2.12% DM, ADF at 38.07 ± 0.929% DM, and starch at 4.17 ± 1.350% DM.Before baling, the stover in Experiment 2 was more similar to Experiment 1 regarding ADICP, available CP, and ADF.Therefore, the greater ADICP (37.80 vs. 7.47% CP) and ADF (44.37 vs. 36.10%DM) and lesser available CP (6.67 vs. 10.53%DM) observed in Experiment 2 stover compared to Experiment 1 stover were most likely an effect of spontaneous heating of B-COR bales during storage (Coblentz and Hoffman 2009).While not measured, mold, blackened forage, and caramel odor were observed in most corn stover bales, also indicating spontaneous heating (Coblentz and Hoffman 2009).Researchers also observed that the ruminal contents (solids and fluid) of cows consuming B-COR were dark brown.

Estimated dry matter intake and cow performance
Paddock-based estimates for DMI were lower when cows were fed or grazed short-season corn stover and DDGS over barley greenfeed, and this result was consistent among both experiments.A decrease in DM intake has also been reported in studies examining cows grazing standing corn versus barley swaths (Baron et al. 2014;Jose et al. 2020), as well as those evaluating backgrounding calves fed corn silage versus barley silage (Chibisa and Beauchemin 2018;Sutherland et al. 2021).In the previously mentioned studies, the decrease in DMI observed in stover feeding is most likely related to increased ruminal propionate molar proportions, as described in the hepatic oxidation theory (Allen et al. 2009) rather than distension in the reticulo-rumen (Allen 2000).It is unlikely that ruminal propionate concentrations from corn stover alone would reach similar concentrations as those observed on backgrounding rations, but it may be possible that corn DDGS supplementation provided to cows on corn stover had a substitution effect, which is discussed in a later section.
Cows consuming short-season corn stover or barley greenfeed swaths exhibited no difference in BCS, but when the stover and barley greenfeed were baled, the body condition score of cows fed stover decreased, as indicated by lesser ending BW, rib fat, rump fat, and BCS.Adipose tissue was most likely mobilized to combat the undernutrition (Chilliard et al. 2000) of B-COR.The cows fed B-COR were projected to need 21.46 Mcal/day for maintenance and 1.33 Mcal/day for pregnancy (NASEM 2016).Based on the observed DMI (12.0 kg/day) and measured nutrient composition, cows fed B-COR were predicted to have a 4.31 Mcal/day energy deficit, supporting mobilization of adipose tissue.Cows fed B-COR were likely not able to access parts of the stover bales because of freezing, as 18.84% of the forage was refused, a factor that could account for the loss of body condition.Researchers observed that large portions of the stover bales were frozen as a result of their high moisture content.This was not observed when grazing the short-season corn stover, most likely due to the higher DM content than bales and insulation from the snowpack preventing freezing (Zhu et al. 2022).Additional research into appropriate methods of preserving corn stover bales in western Canadian conditions is needed.While direct comparisons of anatomical part selection between cattle grazing corn stalks and cattle fed corn stalk bales are not available in published research, both grazing cattle and those fed bales seemingly prefer the more digestible leaves and husks than stalks and cobs (Fernandez-Rivera and Klopfenstein 1989;Karls et al. 2022).Comparing across studies, it is possible grazing cattle may have easier access to those desirable parts than cattle fed stover bales, but further research is needed to confirm this hypothesis (Fernandez-Rivera and Klopfenstein 1989; Karls et al. 2022).

Ruminal fermentation parameters
As hypothesized, mean, minimum, and maximum ruminal pH did not differ between B-BAR and B-COR but were lower than in past studies investigating the use of barley forage compared to corn forage in western Canadian for winter grazing (Jose et al. 2020;Anderson et al. 2023).The low pH values and lack of differences are surprising, given the lesser estimated DMI for B-COR-fed cows relative to B-BAR-fed cows.According to the sorting indices calculated in both experiments, all cows selected against NDF and B-COR cows selected against ADF.It is possible cows were searching for more energy-dense or digestible anatomical components (Rosser et al. 2016) of the barley greenfeed or corn stover, but further research into the nutritional composition and fiber digestibility of the anatomical components could illuminate which components those were.Nonetheless, selection for more digestible carbohydrates could decrease ruminal pH (Chibisa et al. 2020), especially if selection varied daily when a new bale was provided (Rosser et al. 2016;Anderson et al. 2023).
In addition to the selection variability described above, it is possible that the diet × day interaction for total ruminal SCFA concentration and molar proportions of ruminal propionate could have been influenced by bale-to-bale variability.For example, changes in mean kernel weight from harvest location within a field (Jamieson et al. 1995) or selection for seedheads within the drought-stressed barley greenfeed bales (Rosser et al. 2016) could yield variable nutrient composition.On day 21, paddocks fed B-BAR had new bales placed approximately 36 h prior to sampling from an area that was relatively flat, while bales placed approximately 60 h prior to sampling on day 42 came from an area of the field with topographical variation.Therefore, barley bales fed before day 42 could have had more and fuller seedheads with more starch than those placed on day 21 but this was not measured (Jamieson et al. 1995).Cows fed B-COR probably had greater molar proportions of propionate over the B-BAR-fed cows because of corn DDGS supplementation.Previous research has demonstrated that as DDGS supplementation increases when cattle are grazing, the molar proportions of ruminal propionate also increase (Martínez-Pérez et al. 2013;Murillo et al. 2016).The increased ruminal acetate to propionate ratio exhibited by cows on day 21 consuming B-BAR over other diet × day combinations is due to decreased propionate proportions.
Previous researchers have found that increasing DDGS supplementation for grazing cattle also increases ruminal ammonia concentration (Van De Kerckhove et al. 2011;Murillo et al. 2016), but cows supplemented with DDGS in this experiment did not exhibit this response.In fact, B-COR-fed cows had lower ruminal ammonia concentrations than B-BAR-fed cows, probably because more ruminally degradable protein was available to B-BAR-fed cows (Hare et al. 2019) due to the large amount of ADICP in the short-season corn stover bales created by heating during storage (Machacek and Kononoff 2009).When sampled from the corn bales, ADICP and ADF were greater at feeding than when bales were sampled on the field, indicative of spontaneous heating during storage (Coblentz and Hoffman 2009).Additionally, low ammonia concentrations in B-COR-fed cows suggest rumen microbes may have been rapidly utilizing ammonia for microbial protein synthesis, further reflecting the low N availability in corn stover (Pritchard and Males 1985).It is important to note that application of propionic acid at 6.58 kg/t of forage did not preserve the corn stover bales as intended, as indicated by increased ADICP, ADF, and the presence of mold, caramel color, and odor.As stated above, corn stover bale preservation methods in western Canadian conditions need to be investigated.

Conclusion
Stover from short-season corn with corn DDGS supplementation can be used as an alternative feed source to bar-ley greenfeed in extensive winter grazing conditions for dry, multiparous, pregnant cows, but additional supplementation may be required to prevent reductions in body energy status, as indicated by the poorer performance of cows fed B-COR in Experiment 2. Special care must be taken to minimize environmental loss and mitigate freezing and spontaneous heating when a short-season corn stover is baled.Additional research into the most appropriate corn stover feed presentation (swaths, bales, etc.) and preservation in western Canadian conditions is warranted.and editorial decisions regarding this manuscript were handled by another Editorial Board Member.

Fig. 2 .
Fig. 2. Daily minimum, maximum, and average temperatures during Experiments 1 and 2. Data were obtained through Environment and Climate Change Canada (2022; Saskatoon RCS Station).

Table 1 .
Agronomics of barley greenfeed and corn stover grown over 2 years.
* Data obtained through Environment and Climate Change Canada (2022; Saskatoon RCS Station).

Table 2 .
Nutrient composition of barley greenfeed, short-season corn stover, and corn DDGS offered to pregnant, multiparous cows.

Table 6 .
Ruminal pH of pregnant, multiparous cows (body weight = 667 ± 63.2 kg) when fed barley greenfeed bales (B-BAR) or high-moisture corn stover bales and corn dried distillers' grains with solubles (B-COR) over winter measured continuously from days 14 to 21 of Experiment 2. The statistical model included the effects of diet, hours, and their interaction.Hour effects and significant diet × hour effects were considered to have limited novelty; therefore, only treatment means and the standard error of the means are reported. *

Table 7 .
Short-chain fatty acid (SCFA) and ammonia concentrations of pregnant, multiparous cows (body weight = 667 ± 63.2 kg) when fed barley greenfeed bales (B-BAR) or high-moisture corn stover bales and corn dried distillers' grains with solubles (B-COR) over winter on days 21 and 42 of Experiment 2.
Note: SEM, standard error of the means.