Effects of feeding hybrid rye silage as a replacement for barley silage on feed intake, ruminal fermentation, and the site and extent of nutrient digestion in growing beef heifers

TL;DR

Replacing barley silage with hybrid rye silage in beef heifer diets reduces feed intake but improves fiber digestion without affecting energy availability.

Contribution

This study demonstrates the tradeoff between reduced dry matter intake and improved NDF digestibility when using hybrid rye silage as a barley silage replacement.

Findings

Increasing hybrid rye silage inclusion linearly decreased dry matter intake but increased ruminal degradability of DM, OM, and aNDFom.

Hybrid rye silage increased acetate molar proportion while decreasing propionate and butyrate in ruminal fermentation.

Total tract digestibility of DM, ADF, and aNDFom improved with hybrid rye silage inclusion, but gross energy digestibility remained unaffected.

Abstract

The objective of this study was to evaluate the effects of feeding hybrid rye silage (HRS) as a replacement for barley silage (BARS) on dry matter intake (DMI), ruminal fermentation, and the site and extent of digestion when fed to growing beef heifers. Eight ruminally cannulated Hereford × Simmental heifers (519 ± 25.8 kg BW) were used in a replicated 4 × 4 Latin square design with 28-d periods, including 21 d for dietary adaptation and 7 d for data and sample collection. Treatments included a control diet (BCON) that contained 59.62% BARS (harvested at the soft dough stage of maturity), 38.61% barley grain, and 1.78% of a vitamin and mineral supplement on a dry matter (DM) basis. The HRS (harvested at the boot stage of maturity) replaced 33% (RLOW), 67% (RMED), or 100% (RHIGH) of the BARS on a DM basis. Increasing HRS inclusion linearly (P = 0.04) decreased DMI, linearly increased water intake (P = 0.02) and mean ruminal pH (P < 0.01) and linearly decreased the duration that ruminal pH was < 5.8 (P = 0.04). There was no effect of HRS inclusion on total short-chain fatty acid (SCFA) concentration (P = 0.46) in ruminal fluid, but the molar proportion of acetate linearly (P < 0.01) increased while propionate and butyrate linearly decreased (P < 0.01). The NH3-N concentration linearly increased (P = 0.03) with increasing HRS inclusion. Increasing HRS inclusion increased ruminal degradability of DM, OM, and aNDFom, while there was no effect on intestinal digestibility of nutrients, either expressed as a percent of intake or percent of digesta flow. Total tract digestibility of DM (P ≤ 0.01), ADF (P ≤ 0.01), and aNDFom (P ≤ 0.01) increased with increasing HRS inclusion, but inclusion did not affect gross energy (GE) digestibility (P = 0.15) and the dietary digestible energy (DE) concentration (P = 0.17). In conclusion, the use of HRS, harvested at the boot stage, as a direct replacement for BARS, harvested at the soft dough stage, is likely to reduce DMI but improve NDF digestibility without affecting dietary DE concentration. As there was no difference in GE digestibility or dietary DE digestibility, these results demonstrate a tradeoff between greater NDF digestibility of HRS compared to BARS but reduced DMI with increasing inclusion.

Replacing barley silage with hybrid rye silage in diets for growing beef cattle decreased DMI but increased ruminal OM and NDF digestion, leading to no difference in the dietary digestible energy concentration.

Tables

Table 1.. Chemical composition, fermentation products of barley silage and hybrid rye silage1

Item	Ingredient2
	Barley silage	Hybrid rye silage
Chemical composition, % DM3
n	4	4
DM, %	29.62 ± 1.066	33.42 ± 1.013
OM	91.10 ± 0.650	90.04 ± 0.736
CP	14.58 ± 0.126	15.15 ± 0.173
ADF	27.38 ± 1.406	30.10 ± 1.490
NDF	47.50 ± 1.160	52.48 ± 0.818
aNDFom	45.90 ± 1.344	49.90 ± 0.812
uNDF240-h	15.64 ± 0.219	14.18 ± 0.980
Lignin	4.63 ± 0.296	4.58 ± 0.277
Starch	5.33 ± 0.568	0.15 ± 0.129
Ether extract	4.09 ± 0.106	3.92 ± 0.217
Ca	0.36 ± 0.067	0.44 ± 0.050
P	0.26 ± 0.008	0.36 ± 0.010
Particle size distribution, %
>19.0 mm	23.31 ± 4.345	26.69 ± 8.846
19.0 to 8.0 mm	65.78 ± 2.626	67.94 ± 6.507
8.0 to 4.0 mm	7.40 ± 0.864	3.15 ± 1.274
<4.0 mm	3.50 ± 1.500	2.22 ± 1.140
Silage fermentation
n	2	2
pH	4.02 ± 0.02	4.33 ± 0.04
Total acid, % of DM	10.09 ± 1.35	8.50 ± 1.03
Lactic acid, % of DM	8.70 ± 1.27	6.15 ± 0.07
Acetic acid, % of DM	1.39 ± 0.08	1.22 ± 0.76
NH3, % of DM	1.53 ± 0.03	1.98 ± 0.33

Table 2.. Ingredients, chemical composition, and physical analysis of the experimental diets1

Item	Hybrid rye silage inclusion rate
	BCON	RLOW	RMED	RHIGH
Ingredient, % DM
Barley silage	59.62	39.94	19.67	-
Hybrid rye silage	-	19.67	39.94	59.62
Barley grain	38.61	38.61	38.61	38.61
Vitamin/mineral premix2	1.78	1.78	1.78	1.78
Chemical composition3, % DM
DM, %	40.55 ± 0.012	41.83 ± 0.011	43.24 ± 0.011	44.71 ± 0.012
OM	92.98 ± 0.350	92.76 ± 0.302	92.54 ± 0.419	92.34 ± 0.610
CP	14.29 ± 0.952	14.40 ± 0.934	14.51 ± 0.917	14.63 ± 0.902
ADF	19.73 ± 0.849	20.27 ± 0.709	20.82 ± 0.703	21.36 ± 0.835
NDF	39.35 ± 2.063	40.32 ± 1.997	41.33 ± 1.939	42.31 ± 1.894
aNDFom	38.14 ± 2.016	38.92 ± 1.826	39.73 ± 1.681	40.53 ± 1.605
uNDF	12.23 ± 0.410	11.95 ± 0.400	11.65 ± 0.487	11.36 ± 0.635
Starch	24.10 ± 0.962	23.08 ± 0.983	22.03 ± 1.012	21.01 ± 1.049
Ether extract	3.33 ± 0.960	3.30 ± 0.104	3.26 ± 0.130	3.23 ± 0.165
Ca	0.54 ± 0.086	0.55 ± 0.088	0.57 ± 0.091	0.58 ± 0.094
P	0.31 ± 0.008	0.33 ± 0.005	0.35 ± 0.003	0.37 ± 0.003
Particle size distribution4, %
> 19.0 mm	13.90 ± 2.590	14.56 ± 3.230	15.25 ± 4.196	15.91 ± 5.274
< 19.0 > 8.0 mm	39.48 ± 1.368	39.90 ± 1.814	40.34 ± 2.757	40.77 ± 3.815
< 8.0 > 4.0 mm	26.37 ± 4.742	25.53 ± 4.741	24.67 ± 4.748	23.83 ± 4.761
< 4.0 mm	20.26 ± 6.167	20.01 ± 6.055	19.75 ± 5.942	19.50 ± 5.833

Table 3.. Effect of feeding hybrid rye silage as a replacement for barley silage on DMI, water intake, ruminal fermentation, and particle sorting of growing beef heifers

Item	Hybrid rye silage inclusion rate1				SEM2	P-value
	BCON	RLOW	RMED	RHIGH		Treatment	Linear	Quadratic
Initial BW, kg	564	560	563	563	16.5	0.99	0.86	0.79
Final BW, kg	598	597	591	586	17.1	0.94	0.55	0.89
DMI, kg/d	10.14	10.22	9.70	9.22	0.420	0.15	0.04	0.40
DMI, % of BW	1.75	1.77	1.69	1.60	0.065	0.07	0.02	0.24
Total water intake3	33.2	35.9	34.7	33.8	2.04	0.50	0.89	0.18
Water intake (liquid), kg/d	18.6	21.7	21.9	22.5	1.69	0.06	0.02	0.25
Water intake (feed), kg/d	14.6	14.1	12.8	11.3	0.52	<0.01	<0.01	0.24
Particle sorting4, %
> 19.0 mm	96.44	101.20	100.88	99.66	2.391	0.21	0.23	0.09
< 19.0 > 8.0 mm	95.82a	97.64	96.68a	95.21	1.650	0.47	0.59	0.16
< 8.0 > 4.0 mm	105.01a	102.88a	104.17a	105.89a	1.454	0.24	0.41	0.08
< 4.0 mm	103.50	99.44	100.74	102.69	2.750	0.19	0.86	0.04

Table 4.. Effect of feeding hybrid rye silage as a replacement for barley silage on ruminal fermentation of growing beef heifers1

Item	Hybrid rye silage inclusion rate1				SEM2	P-value
	BCON	RLOW	RMED	RHIGH		Treatment	Linear	Quadratic
Ruminal pH
Minimum pH	5.77	5.81	5.97	6.18	0.097	<0.01	<0.01	0.26
Mean pH	6.36	6.44	6.53	6.54	0.051	0.02	<0.01	0.42
Maximum pH	6.88	6.91	6.98	6.90	0.069	0.72	0.63	0.36
Duration < 5.8, min/d	94.5	59.1	8.4	4.2	34.66	0.19	0.04	0.64
Area < 5.8, (pH × min)/d	23.7	7.1	0.8	1.3	10.63	0.36	0.11	0.41
Rumen fermentation
Total SCFA3, mM	105.6	106.5	105.6	104.2	2.84	0.78	0.46	0.47
SCFA proportions, mol/100 mol
Acetate	62.11	63.57	64.07	64.93	0.526	<0.01	<0.01	0.52
Propionate	18.74	17.84	17.68	17.41	0.371	0.04	<0.01	0.47
Butyrate	13.94	13.46	12.76	12.34	0.409	0.03	<0.01	0.94
Isobutyrate	1.08	1.06	1.10	1.12	0.026	0.24	0.10	0.39
Isolvalerate	1.71	1.51	1.66	1.43	0.101	0.19	0.13	0.89
Valerate	1.72	1.73	1.88	1.95	0.053	<0.01	<0.01	0.57
NH3-N, mg/dL	8.75	9.03	9.61	10.24	0.581	0.16	0.03	0.72

Table 5.. Effect of feeding hybrid rye silage as a replacement of barley silage on nutrient intake, digesta flow to omasum, and ruminal apparent degradability of nutrients of growing beef heifers1

Item	Hybrid rye silage inclusion rate1				SEM2	P-value
	BCON	RLOW	RMED	RHIGH		Treatment	Linear	Quadratic
Nutrient intake3, kg/d
OM	9.45	9.51	9.00	8.53	0.387	0.11	0.02	0.39
CP	1.45	1.47	1.41	1.35	0.059	0.26	0.08	0.37
ADF	1.98	2.07	2.02	1.94	0.089	0.69	0.65	0.29
aNDFom	3.88	4.04	3.90	3.74	0.192	0.73	0.51	0.40
aNDFom, % BW	0.67	0.68	0.68	0.65	0.027	0.49	0.50	0.17
uNDF	1.22	1.24	1.13	1.03	0.053	<0.01	<0.01	0.15
Starch	2.47	2.34	2.12	1.99	0.123	<0.01	<0.01	0.98
Ether extract	0.34	0.33	0.31	0.29	0.014	0.05	<0.01	0.43
Digesta flow to omasum, kg/d
DM	7.88	7.12	6.77	6.56	0.363	0.03	<0.01	0.38
OM	5.98	5.23	4.97	4.78	0.291	<0.01	<0.01	0.25
aNDFom	2.47	2.10	1.94	1.94	0.175	0.05	0.01	0.23
Starch	0.08	0.10	0.06	0.10	0.019	0.29	0.88	0.53
Ruminal degradability, %
DM	22.34	30.39	30.02	28.85	1.926	0.02	0.03	0.02
OM	36.72	45.24	44.79	44.08	1.930	<0.01	0.01	0.01
aNDFom	35.96	48.31	50.18	47.94	4.313	0.04	0.03	0.06
Starch	95.39	96.99	95.73	96.39	0.787	0.43	0.58	0.52

Table 6.. Effect of feeding hybrid rye silage as a replacement for barley silage on apparent intestinal digestibility of nutrients of growing beef heifers

Item	Hybrid rye silage inclusion rate1				SEM2	P-value
	BCON	RLOW	RMED	RHIGH		Treatment	Linear	Quadratic
Intestinal digestibility3, % of intake
DM	41.53	35.88	35.97	38.92	1.872	0.11	0.34	0.03
OM	30.82	23.98	23.88	26.21	1.928	0.04	0.09	0.02
aNDFom	14.07	7.42	5.99	12.28	4.306	0.40	0.69	0.11
Starch	3.15	1.68	2.75	2.94	0.885	0.60	0.91	0.31
Intestinal digestibility, % of digesta flow to omasum
DM	53.34	51.47	51.47	54.35	1.496	0.39	0.63	0.11
OM	48.37	43.79	43.20	46.30	2.135	0.26	0.45	0.07
aNDFom	18.30	13.17	10.73	21.03	5.961	0.52	0.81	0.17
Starch	54.20	43.98	55.39	61.49	10.828	0.69	0.46	0.42

Table 7.. Effect of feeding hybrid rye silage as a replacement of barley silage on total tract apparent digestibility of nutrients, DE, and fecal output of growing beef heifers

Item	Hybrid rye silage inclusion rate1				SEM2	P-value
	BCON	RLOW	RMED	RHIGH		Treatment	Linear	Quadratic
Digestibility3, % of intake
DM	63.87	66.24	65.99	67.77	1.030	0.04	<0.01	0.74
OM	67.54	69.19	68.67	70.28	1.058	0.28	0.09	0.99
CP	64.50	66.17	65.04	65.45	1.255	0.72	0.71	0.54
ADF	34.49	42.87	44.48	49.13	2.005	<0.01	<0.01	0.33
aNDFom	50.03	55.57	56.17	60.22	1.502	<0.01	<0.01	0.61
Starch	98.18	98.28	98.03	98.37	0.273	0.77	0.77	0.63
Ether extract	64.17	66.53	67.06	68.02	2.262	0.58	0.19	0.74
GE	64.57	66.45	65.55	67.15	1.127	0.34	0.15	0.89
DE, Mcal/kg	2.82	2.90	2.87	2.93	0.053	0.40	0.17	0.86
Fecal output, kg/day	3.65	3.46	3.28	2.96	0.149	<0.01	<0.01	0.62

Full text

Introduction

Whole-crop barley (Hordeum vulgare L.) silage (BARS) is commonly used for feedlot cattle in Western Canada (Nair et al., 2016; Johnson et al., 2020). While barley has been a staple forage source, the development of new varieties of hybrid fall rye (Secale cereale L.) allow for fall seeding and earlier silage harvest than barley (KWS Production Guide Hybrid Rye, 2023), thereby spreading production risk and capitalizing on fall and early spring moisture. Growing fall rye for silage has been reported to produce similar yield when compared to spring-seeded barley (Zhang et al., 2025), or even greater yield for hybrid rye relative to other small cereal grain forages seeded in fall and early spring (Geiger and Miedaner, 2009; KWS Production Guide Hybrid Rye, 2023). Equal or greater yield for hybrid rye relative to spring-seeded barley, along with the earlier harvest date, may allow for a second crop to be seeded in the same growing season (Darambazar et al., 2022). While production of rye silage has been adopted, few studies have evaluated the use of rye harvested as silage (HRS) for feedlot cattle (Stefanyshyn-Cote, 1993; Zhang et al., 2025).

Hybrid rye silage harvested at the late milk stage of maturity was reported to have greater concentrations of acid detergent fiber (ADF) and neutral detergent fibre (NDF), but less starch than BARS harvested at the soft dough stage of maturity (Zhang et al., 2025). Previous studies have characterized how the physiological growth stage for small grain species (barley, oats, rye, wheat, and triticale) harvested as forage affects dry matter (DM) yield, morphological and anatomical characteristics, and forage quality (Edmisten et al., 1998a, 1998b; Sartin et al., 2022). Results from those studies indicate that when harvested at the same maturity, the in vitro dry matter disappearance (IVDMD) of cereal rye silage was generally less than that of barley, oat, and wheat throughout the milk to hard dough stages (Edmisten et al., 1998b). Moreover, fall rye silage harvested either at flowering or at mid-milk resulted in lesser DMI for growing steers, and lesser in vivo digestibility of most nutrients at the mid-milk stage when compared to barley silage harvested at mid-dough (Stefanyshyn-Cote, 1993). However, we are unaware of in vivo studies that evaluated the site and extent of nutrient digestibility for hybrid rye silage.

Given that previous studies have suggested rye silage reduces DMI (Fisher and Lessard, 1987; Stefanyshyn-Cote, 1993; Terler et al., 2025) including a recent study where HRS replaced barley silage in diets for growing beef cattle (Zhang et al., 2025), it was hypothesized that increasing dietary HRS as a replacement for BARS would decrease DMI, increase ruminal pH, reduce the total ruminal short-chain fatty acid (SCFA) concentration, and alter the site of digestion due to increased concentrations of NDF and decreased starch in HRS relative to BARS. The objective of this study was to evaluate the effects of feeding HRS as a replacement for BARS on DMI, ruminal fermentation, and the site and extent of nutrient digestion when fed to growing beef heifers.

Materials and Methods

The use of animals and experimental procedures was approved by the University of Saskatchewan Animal Research Ethics Board (protocol 20210069) according to the guidelines of the Canadian Council of Animal Care (Ottawa, ON, Canada).

Animal management, experimental design, and dietary treatments

Eight Hereford × Simmental-cross heifers that were previously fit with a 10-cm cannula (model 9C; Bar Diamond Inc., Parma, ID, USA) and had an initial body weight (BW) of 519 ± 25.8 kg were housed at the University of Saskatchewan Livestock Research Barn (Saskatoon, SK, Canada). Heifers were ovariectomized at the time of ruminal cannulation and were approximately 8 mo old at the time of surgery. Heifers were housed in individual pens (9 m^2^) with ad libitum access to water and rubber mats on the floor throughout the study. Pens were scraped and washed daily at 0730 hours to remove manure. Heifers were permitted 2 h/d of exercise in an outdoor pen during pen cleaning and diet preparation, except during sample collections.

Heifers were stratified by BW into 1 of the 2 squares, and within the square, randomly assigned to 1 of the 4 treatment sequences in a replicated 4 × 4 Latin square design balanced for carry-over effects. Each experimental period lasted 28 d, including 21 d for adaptation and 7 d for sampling and data collection. Dietary exposure occurred abruptly at the start of each period. The control diet (BCON) contained (DM basis) 59.62% BARS in combination with 38.61% dry-rolled barley grain, and 1.78% of a vitamin and mineral premix. The HRS silage was included (DM basis) at 19.67% (RLOW), 39.94% (RMED), and 59.62% (RHIGH) by replacing 33%, 67%, or 100% of the BARS. The BARS in this study were harvested at the soft dough stage of maturity and the HRS was harvested at the boot stage. Diets were formulated to meet or exceed nutrient requirements of the heifers with a target weight gain of 1.3 kg/d according to the NASEM (2016) based on initial analysis of the feed ingredients collected prior to the start of the study. All diets included monensin (Rumensin premix; Elanco Animal Health, Ontario, Canada) at a targeted concentration of 33 mg/kg DM, which was consistent with Zhang et al. (2025). Dry rolled barley grain was purchased commercially from a single supplier and rolled to a processing index of 70% (Yang et al., 2000). Silages were sampled twice weekly, and the remaining ingredients were sampled once weekly to determine the DM concentrations. The DM coefficients for individual ingredients were adjusted on a weekly basis when the 3-wk running average differed by more than 2 percentage units.

Heifers were weighed on 2 consecutive days at 0900 hours (before feeding) at the start and end of each period. Total mixed rations (TMR) were prepared by hand mixing and the TMR was offered once daily at 0930 hours to ensure at least 5% refusals. The amount of feed offered and refused were recorded daily, and data from days 25 to 28 of each period were used to determine DMI based on the difference between the quantity of DM offered and refused. Water intake was measured using inline water meters (Model DLJSJ75, Daniel L. Herman Co., Hackensack, NJ, USA) for each heifer with meter readings recorded daily at 0700 hours from day 25 of each period to day 1 of the following period. Water flow through the meters was assumed to be equal to water intake (Penner et al., 2020). Ingredient samples (500 g) were collected daily from days 25 to 28, and refusal samples were collected from days 26 to day 1 of the following period and composited for each heifer proportionally to the amount refused. The ingredient and refusal samples were dried in a forced-air oven at 55 °C until achieving a constant weight to determine DM concentration. Subsequently, samples were ground to pass through a 1-mm sieve using a hammer mill (Christie-Norris Laboratory Mill, Christie-Norris Ltd, Chelmsford, UK) and sent to Cumberland Valley Analytical Service (Waynesboro, PA, USA) for chemical analysis including: analytical DM; organic matter (OM); crude protein (CP); NDF; NDF determined using sodium sulfate and α-amylase and then corrected for ash content (aNDFom); ADF; starch; ether extract; calcium; and phosphorus as described by Zhang et al. (2024).

Particle distribution and sorting index

The particle size distribution of feed ingredients and refusals collected from days 25 to 28 of each period was determined using the Penn State Particle Separator (Nasco, Fort Atkinson, Wisconsin, USA) with aperture sizes of 19, 8, and 4 mm, and a bottom pan according to Heinrichs (2013). The sorting index was calculated according to Leonardi and Armentano (2003). Briefly, the actual intake of each Penn State Particle Separator fraction was expressed as a percentage of the predicted intake of that fraction, as determined if no sorting occurred. A lack of sorting is indicated by a sorting index value of 100%, sorting against is indicated by a value of <100%, and sorting for particles is indicated by a value of >100%.

Ruminal fermentation

An indwelling ruminal pH measurement system (Dascor, Escondido, CA, USA) was placed in the ventral sac of the rumen in each heifer at 1900 hours on day 21 and removed after 0100 hours on day 1 of the following period to ensure completion of sample collection (Penner et al., 2006). Prior to insertion and after removal from the rumen, the pH systems were standardized in pH buffers 7 and 4 at 39 °C to characterize the linear relationship between mV and pH. The pH meters were programmed to measure and record every 5 min over the collection duration. The raw data obtained were converted from mV to pH using the beginning and ending standardizations derived from the linear regression between the pH buffer and the observed mV value, along with the assumption of linear drift between the starting and ending standardizations (Penner et al., 2006). Data were then restricted to 96 h of continuous pH measurement starting at 0700 hours on day 22 and ending at 0655 hours on day 26. Ruminal pH data were used to determine the daily average minimum, mean, and maximum pH values, and the duration and area that ruminal pH was <5.8.

Ruminal digesta was manually collected every 12 h with a 15-h offset between samples collected on consecutive days. Samples were collected at 0200 and 1400 hours on day 25, 0500 and 1700 hours on day 26, 0800 and 2000 hours on day 27, and 1100 and 2300 hours on day 28. At each timepoint, three spot samples (250 mL/region) of mixed ruminal digesta were collected from the cranial central, central, and caudal central regions, respectively. Digesta from the three regions were pooled and strained through 2 layers of cheesecloth. Subsequently, 10 mL of strained ruminal fluid was transferred into two 15-mL tubes, with one containing 2 mL of metaphosphoric acid (25% w/v) and the other containing 2 mL of sulfuric acid (1%). After gently and repeatedly inverting and mixing, an equal volume from each sample was used to create a composite sample for each heifer. Samples were stored frozen at −20 °C until analysis for SCFA (Khorasani et al., 1996) and ammonia-N concentrations (Broderick and Kang, 1980).

Diet digestibility, site, and extent of digestion

Heifers were tethered during the collection period to facilitate ruminal infusion starting at 0900 hours on day 22 and the collection of omasal digesta and feces starting at 0200 hours on day 25. To measure omasal flow and calculate the site and extent of nutrient disappearance, Cr-ethylenediaminetetraacetic acid (Cr-EDTA; Udén et al., 1980) and ytterbium chloride (YbCl_3_; Siddons et al., 1985) were used as digesta markers. Markers were infused into the rumen using a peristaltic pump (Model 205U; Watson and Marlow, Cornwall, UK) to target at a constant infusion rate of 1 L/heifer/day, which provided 2.77 g of Cr and 2.77 g of Yb per day, respectively. Infusions were initiated on day 22 at 0900 hours through day 28 until finishing the collection. A 500-mL priming dose of each marker solution was infused on the first day of marker infusion. The marker solutions were weighed daily at 0900 hours to determine the actual infusion rate and a sample of each marker solution was collected prior to the initiation of ruminal infusions during each period for determination of Cr, and Yb concentrations.

Omasal digesta samples and representative feces samples were collected at the same time points as for ruminal digesta collection. At each sampling time point, 500 mL of omasal digesta was collected as described by Chibisa et al. (2012). Digesta samples collected within a period were pooled for each heifer into a common pail and stored frozen at −20 °C for further analysis. The composited omasal digesta samples (4 L for each heifer per period) were thawed at room temperature and then separated into the large particle, small particle, and fluid phases according to Brito et al. (2009). Each phase was then oven-dried at 55 °C until reaching a constant weight and ground to pass through a 1-mm screen (Retsch ZM 200; Haan, Germany). Additionally, 200 g of feces was collected from the rectum of each heifer at each collection time point. Samples were placed in a common pail for each heifer and stored at −20 °C until the end of the sampling period. Samples of feces were thawed and dried at 55 °C in a forced-air oven until reaching a constant weight. Feces were routinely mixed during drying to facilitate even drying and prevent mold growth. Once dried, all fecal samples were ground to pass through a 1-mm screen using a hammer mill (Christie-Norris Laboratory Mill; Christie-Norris Ltd, Chelmsford, UK) and sent to Cumberland Valley Analytical Service (Waynesboro, PA, USA) for chemical analysis as described above. The gross energy (GE) of all feed ingredients, refusals, and fecal samples were measured using bomb calorimetry (6,400 Automatic Isoperibol Calorimeter, Parr Instruments Company, Moline, IL, USA) at the University of Saskatchewan (Saskatoon, SK, Canada). Then GE digestibility and diet digestible energy (DE) concentration were determined as (GE intake—GE excreted in feces)/GE intake × 100, and (GE intake—GE excreted in feces)/DMI, respectively (NASEM, 2016).

The dried omasal fractions and fecal samples were used to determine the concentrations of Cr and Yb according to the methods as described in Zhang et al. (2024). The concentration of 240 h undigested NDF (uNDF) was determined in the large particle phase of omasal digesta, feed ingredients, and refusals samples using the in situ nylon bag technique according to Nair et al. (2017). Briefly, 1.0 g of feed and refusals, 0.5 g of large particle phase, and 1.2 g of small particle phase were weighed in triplicate into individual 5 cm × 10 cm in situ bags (6 μm pore size; Sefar America Inc., Depew, NY, USA). The bags were assigned randomly to 1 of the 4 laundry bags, each with two 1-kg weights. The laundry bags were then randomly assigned to 1 of the 2 ruminally cannulated lactating Holstein cows and were placed in the ventral sac of the rumen. All bags were incubated for 240 h. Upon removal, all bags were washed 10 times in cold water and then dried in an air-forced oven at 55 °C for 48 h. Subsequently, all samples were analyzed for NDF concentration with the use of heat-stable alpha-amylase and sodium sulfite using the Ankom 200 Fiber Analyzer according to the manufacturer’s protocol (Ankom Technology Corp, 2017). The Cr, Yb, and uNDF concentrations were used to reconstitute the omasal fractions into a single representative omasal sample for each heifer in each period using the triple-marker method developed by France and Siddons (1986). The reconstituted omasal digesta samples and fecal samples were sent to Cumberland Valley Analytical Service (Waynesboro, PA, USA) for determination of analytical DM, OM, CP, aNDFom, ADF, and starch as described by Zhang et al. (2024). The flow of omasal digesta and chemical constituents were calculated based on the triple-maker method as described by France and Siddons (1986). Fecal DM output was determined based on the infusion of Yb and fecal Yb concentration according to (Beauchemin et al., 2001).

Statistical analysis

The ruminal cannula for one heifer on the RLOW diet came out during collections in period 1, and these data were removed. As the heifer was considered the experimental unit, the loss of data resulted in an n = 7 for period 1 and n = 8 for each of the remaining 3 periods. Data were analyzed using the GLIMMIX procedure of SAS 9.4 (SAS Institute Inc., Cary, NC, USA) with fixed effects of dietary treatment, and random effects of period and heifer nested within square. All data and their residuals were tested for normality using the Shapiro–Wilk test and visual assessment of the distribution of the data and residuals. The denominator degrees of freedom were calculated using the Kenward–Roger option and compound symmetry was used as the covariance error structure. Least squares means were obtained using the LSMEANS statement of SAS, and preplanned contrasts were used to test whether HRS inclusion responses were linear, quadratic, or cubic effects. Data were presented as least square means and the largest SEM was reported. The difference between the means were declared significant when P < 0.05, and tendencies are discussed when 0.05 < P ≤ 0.10. Due to the absence of significant cubic effects for the variables of primary interest, the cubic contrast was removed. In addition, a two-tailed t-test was used to confirm whether the particle sorting index for each particle size category and each treatment differed from 100% (version 9.4, SAS Institute Inc., Cary, NC, USA). The difference was considered significant when P < 0.05.

Results

Diet composition

While statistical comparisons were not evaluated for the silages or dietary treatments, HRS had numerically greater DM, ADF, and aNDFom concentrations with numerically less starch than BARS (Table 1). The particle size distribution was similar, although HRS had a slightly greater proportion of particles retained on the 19-mm sieve and less retained on the 8-mm sieve. As this was a direct substitution study, replacing BARS with HRS increased dietary DM, ADF, and aNDFom, and decreased starch (Table 2).

DMI, ruminal pH, and ruminal fermentation

Increasing the HRS inclusion rate linearly decreased DMI, whether reported in kg/d or as a percentage of BW (P ≤ 0.04; Table 3). Water intake consumed from drinking linearly increased with increasing HRS inclusion (P = 0.02), while dietary water intake linearly decreased as HRS inclusion increased (P < 0.01). Total water intake (L/d) did not differ with increased HRS inclusion (P = 0.89).

Sorting index values for particles retained on the 19-mm sieve tended to be affected quadratically, with values increasing and then decreasing such that heifers fed CON sorted against long particles while those fed HRS did not sort for or against; however, none of the sorting index values for the 19-mm sieve differed from 100%. Sorting for particles retained on the 8-mm sieve were not affected by treatment. While heifers in all treatments preferentially consumed particles on the 4-mm sieve, the sorting index tended to respond quadratically (P = 0.08), where sorting for 4-mm particles decreased from BCON to RLOW and then increased with increasing HRS inclusion rate. In addition, a quadratic response was observed for the sorting index for particles retained on the pan, such that heifers fed BCON preferentially consumed these particles while those fed RLOW and RMED had lesser sorting index values, indicating no preferential consumption, and those fed RHIGH again preferentially consumed particles retained on the pan. No sorting index values for the pan differed from 100%.

Increasing the HRS inclusion as a replacement of BARS linearly increased the minimum (P < 0.01) and mean ruminal pH (P < 0.01), but did not affect maximum pH (Table 4). There was also a linear decrease for the duration that ruminal pH was <5.8 (P = 0.04). There was no diet effect on the area where ruminal pH was <5.8. Increasing HRS inclusion at the expense of BARS had no effect on total SCFA concentration (P = 0.46). However, the molar proportion of acetate linearly increased (P < 0.01) while propionate and butyrate linearly decreased (P < 0.01) as HRS inclusion increased as a replacement for BARS. Isobutyrate and isovalerate were not affected by dietary treatments, but the molar proportion of valerate linearly increased (P < 0.01) with increasing HRS inclusion. In addition, the ruminal NH_3_-N concentration linearly increased with increasing HRS inclusion in the diet (P = 0.03).

Nutrient intake and the site and extent of digestion

Increasing HRS inclusion linearly decreased OM, uNDF, starch, and ether extract intake (P ≤ 0.02; Table 5) while ADF and aNDFom intake did not differ (P > 0.51). Increasing HRS inclusion also tended to decrease CP intake (P = 0.08). Consequently, the omasal flow of DM, OM, and aNDFom decreased linearly (P ≤ 0.01) as HRS replaced BARS. The apparent ruminal degradability of DM and OM responded quadratically, such that the values increased from BCON to RLOW and reached a plateau (P ≤ 0.02). The apparent ruminal aNDFom degradability increased linearly (*P *= 0.03) with increasing HRS inclusion. No difference was detected in ruminal starch degradability (P = 0.58).

Increasing HRS inclusion decreased intestinal digestibility (% of intake) of DM and OM in a quadratic manner (P ≤ 0.03; Table 6) such that digestibility decreased from BCON to RLOW and increased from RMED to RHIGH. There were no differences for the intestinal digestibility (% of intake) of aNDFom or starch (P > 0.10). When reported as a percentage of nutrient flow, increasing HRS inclusion resulted in a tendency for a quadratic effect on OM digestibility (P = 0.07), with RLOW and RMED tending to have lesser digestibility than BCON and RHIGH. There was no dietary effect on intestinal digestibility of DM, aNDFom, or starch (P > 0.10).

Increasing HRS inclusion linearly increased apparent total tract DM, ADF, and aNDFom digestibility (P < 0.01; Table 7), and tended to increase OM digestibility (P = 0.09). Total tract digestibility of CP, starch, and ether extract were not affected by dietary treatment (P > 0.10). Apparent total tract GE digestibility and dietary DE concentration did not differ. Fecal DM output linearly decreased with increasing HRS inclusion (P < 0.01).

Discussion

In the present study, a linear reduction in DMI (kg/day and % of BW) was observed as HRS inclusion increased when used as a replacement for BARS in diets for growing heifers. This aligns with our previous study with growing steers in which HRS was harvested at the late milk stage and replaced BARS harvested at the soft dough stage of maturity (Zhang et al., 2025). The reasons for reduced DMI with increasing HRS inclusion remain unclear; although Zhang et al. (2025) suggested that the increase in dietary NDF concentration with increasing HRS inclusion may lead to greater ruminal fill, or that awns present as part of the spikelet may reduce the palatability of rye. In the present study, aNDFom intake was less than 0.7% of BW, suggesting that it is unlikely that aNDFom intake should limit DMI (Mertens, 1987). Moreover, uNDF intake linearly decreased with increasing HRS inclusion and ruminal aNDFom degradability linearly increased, challenging whether increased aNDFom concentration is a plausible explanation driving reduced DMI.

Several authors have suggested that rye forage has low palatability, particularly as maturity advances (Lardy et al., 2022; O’Reilly, 2024), but we were not able to identify specific factors in the literature that may alter palatability. It should also be noted that as HRS inclusion increased, the dietary DM concentration increased from 40.55% for BCON to 44.71% for RHIGH. Most studies evaluating TMR DM concentration within the observed range for dietary DM in the present study have reported reductions in DMI as dietary DM decreases (Lahr et al., 1983; Miller-Cushon and DeVries, 2009; Felton and DeVries, 2010). Thus, the available data also challenge whether dietary DM explains the reduction in DMI as HRS replaces BARS. However, those studies evaluated the use of water inclusion rather than moisture inherent within a feed ingredient and as such, responses may differ.

The stage of maturity at the time of harvest can affect ensiling characteristics (Kim et al., 2001) and potentially palatability. It has been suggested that fall rye should be harvested at the flag leaf to flowering stage for silage (Kim et al., 2001) and hay production because of the potential decrease in palatability and intake (Stefanyshyn-Cote, 1993; Schlegel, 2013; McLelland and Brook, 2016), and reduction in IVDMD (Kim et al., 2001) with advancing maturity. Stefanyshyn-Cote (1993) suggested that fall rye harvested either as hay or as silage at flowering and mid-milk might display inherent factors that reduce intake when compared to barley silage harvested at the mid-dough stage, but did not describe what these potential factors could be. Awns are one possible inherent anti-palatability factor and are present at the boot stage. When comparing awn structure among cereals, rye awns contain a more lignified, thick-walled cell compared to barley and wheat (Evert, 2006; Ntakirutimana and Xie, 2019), which may influence the palatability of rye when harvested as whole-crop silage. However, the HRS in the present study was harvested at the boot stage, chopped, and ensiled, questioning whether awns may be a plausible factor driving the reduction in DMI. Future research is needed to elucidate factors that affect the palatability of HRS and to determine if this changes with the stage of maturity.

The decrease in DMI and reduction in dietary starch with increasing HRS inclusion were likely the primary contributing factors for many of the observed results in the present study. For example, the increased mean ruminal pH and the reduction in the duration that pH was <5.8 with increasing HRS inclusion could be caused by the reduction in DMI and the lesser dietary starch as HRS inclusion increased. Past studies that purposely imposed low feed intake reported that ruminal pH increases in a dose-dependent manner as feed intake decreases (Zhang et al., 2013), and more recently, a negative correlation between DMI and ruminal pH has been reported (Hartinger et al., 2024). The linear increase for the molar proportion of acetate and linear decrease for the molar proportions of propionate and butyrate are characteristic of diets with greater aNDFom, particularly considering the linear increase in ruminal degradability of aNDFom with increasing HRS inclusion, lesser dietary starch (Russell, 2002; NASEM, 2016; Chibisa et al., 2020), and when cattle decrease DMI (Zhang et al., 2013). Moreover, the reduction in fecal DM output was likely driven by reduced DMI.

Increasing inclusion of HRS linearly reduced aNDFom intake but linearly increased ruminal degradability and total tract digestibility of aNDFom. The earlier stage of maturity at harvest for HRS than for BARS likely explains the greater aNDFom digestibility. As noted, the HRS used in the present study was harvested at the boot stage of maturity as recommended by Heuzé et al. (2015) and O’Reilly (2024). When compared to BARS harvested at soft dough stage, HRS contained greater aNDFom but lesser uNDF than BARS, suggesting a greater supply of potentially degradable aNDFom supporting the digestibility responses observed. Kim et al. (2001) reported that the NDF concentration increased and IVDMD decreased for rye from the boot to flowering stages. Likewise, for barley forage, increasing maturity from late milk to hard dough decreased NDF digestibility (Rosser et al., 2016). However, few studies have compared digestibility responses between HRS and BARS when harvested at industry-recommended stages of maturity. In addition to differences in maturity, the reduction in DMI with increasing HRS inclusion may have helped facilitate greater aNDFom digestibility as decreased DMI is expected to extend ruminal retention, thereby enhancing ruminal DM and NDF degradability (Allen and Mertens, 1988; Russell et al., 1992; Allen, 2020).

The observed linear increase in ruminal NH_3_-N concentrations was likely a result of reduced fermentable substrate supply limiting energy availability for ruminal microorganisms to synthesize microbial protein (Russell et al., 1992; Russell, 2002; NASEM, 2016) and is consistent with the reduction in dietary starch and OM intake as HRS inclusion increased. Differences in maturity at the time of harvest between the HRS and BARS may help explain this response. Rye (Kim et al., 2001) and barley (Rosser et al., 2013) forage decrease in CP concentration with advancing maturity (Boudon et al., 2002). As such, it is likely that the reduction in dietary starch, coupled with greater CP with increasing HRS inclusion, led to the greater ruminal ammonia concentration observed. As we did not measure true ruminal CP degradability or urinary N output, the lack of response for apparent total tract CP digestibility is difficult to interpret. Future studies should include the evaluation of N-balance and whether HRS alters microbial protein production.

It is surprising that GE digestibility and dietary DE concentrations did not differ with increasing HRS inclusion rate, particularly when considering previous results evaluating the replacement of BARS with HRS for growing steers (Zhang et al., 2025). The lack of response for GE and DE may be associated with the reduced DMI (Allen, 1996) and improved ruminal degradability of DM, OM, and aNDFom, as well as apparent total tract digestibility of DM, ADF, and aNDFom, with increasing HRS inclusion. Thus, in the present study, the increase in DM, OM, and aNDFom digestibility likely compensated for the decreased starch concentration, resulting in similar GE digestibility and DE concentration amongst treatments.

There are some limitations in the present study. Firstly, forages were harvested at industry-recommended maturities for ensiling (Stefanyshyn-Cote, 1993; Nair et al., 2016) rather than at the same stage of maturity. Maturity at harvest impacts quality and the nutritional value of whole-plant forages along with the resulting silage (Borreani et al., 2018; Nair et al., 2018; Moore et al., 2020). For example, the HRS had greater DM content than BARS. The BARS DM concentration was less than ideal, which could impact ensiling characteristics, challenging whether the lesser DM concentration of the BARS could help explain the reduction in DMI observed with increasing HRS inclusion. However, based on silage pH, organic acid concentrations, and ammonia, both silages appeared to be adequately preserved. Additionally, small grain species are known to have differing digestibility at the same stage of maturity (Stefanyshyn-Cote, 1993), and the changes for chemical composition (e.g., CP and soluble CP, aNDFom, and starch) and digestibility differ by plant species with advancing maturity (Helsel and Thomas, 1987; Stefanyshyn-Cote, 1993; Edmisten et al., 1998b). While the in vitro and in situ degradability of NDF decreases as barley (Rosser et al., 2013; Nair et al., 2018) and fall rye (Helsel and Thomas, 1987; Stefanyshyn-Cote, 1993; Kantar et al., 2011) advance in maturity, the digestibility of rye has been reported to be lesser than barley at all stages of maturity evaluated (Helsel and Thomas, 1987).

Another limitation is that we used a simple ingredient substitution approach for dietary formulation, such that HRS replaced BARS. The approach used resulted in greater dietary DM and NDF, and lesser dietary starch as HRS inclusion increased. While some nutritionists may prioritize the proportion of forage in the growing diet, findings from this study and a related study (Zhang et al., 2025) suggest that dietary formulation strategies should also consider the dietary aNDFom and starch concentrations. It is unclear whether responses for DMI, ruminal fermentation, and nutrient digestibility would be similar if diets were formulated for equal starch and aNDFom. However, ingredient substitution studies are a logical first step when evaluating novel ingredients (Swanson et al., 2017).

Conclusion

Replacing BARS with increasing levels of HRS in growing diets decreased DMI and increased ruminal pH without affecting total SCFA concentration in the rumen. The increased extent of ruminal DM, OM, and aNDFom degradability with increasing HRS inclusion can be interpreted to indicate that harvesting HRS at an earlier stage of maturity than BARS may stimulate ruminal NDF degradability, although the reduction in DMI may have also contributed to this response. Incorporating more HRS in diets increased the ruminal NH_3_-N concentration, which might imply lower efficiency of ruminal fermentation and microbial protein synthesis; however, the latter was not assessed. Given the lack of difference for apparent total tract GE digestibility and the resulting diet DE concentration, there is likely a tradeoff between the greater digestibility of aNDFom in HRS and less starch than in BARS. In conclusion, the use of HRS, harvested at the boot stage, as a direct replacement for BARS, harvested at the soft dough stage, is likely to reduce DMI but improve NDF digestibility without affecting dietary DE concentration.