Evaluation of kernel processing and processor type in whole-plant sorghum silage: effects on nutrient digestibility and animal performance in backgrounding beef heifers

Abstract

Two experiments were conducted to assess the effects of feeding whole-plant sorghum silage (WPSS) with different kernel processing techniques (KP). Experiment 1 contrasted KP for WPSS on intake and apparent total tract digestibility (ATTD) in beef heifers (n = 24, 13 ± 1 mo, 267 ± 10.9 kg of initial body weight [BW]) housed in individual pens (36 m²). Grain sorghum was harvested at hard dough, switching the kernel processor to obtain the WPSS treatments: A) unprocessed (UNP), B) conventionally processed (CONV), and C) shredlage processed (SHRD). Heifers (8/treatment) received ad libitum WPSS from their respective treatment, plus soybean meal top-dressed at 0.5% BW/d (DM basis). Feed, and feces were collected for 5 d; feed was offered once daily, and orts were collected the following day. Fecal samples were collected twice daily, and ATTD was determined using indigestible neutral detergent fiber (NDF) as a marker. Data were analyzed as a completely randomized design, with heifer as the experimental unit, and the following contrasts were performed 1) Processing: UNP vs. (CONV + SHRD) and 2) Processor: CONV vs. SHRD. Processing WPSS increased the ATTD of starch by 4.5% (P = 0.01), reduced fecal starch by 27.5% (P = 0.01), and reduced the change of NDF from feed to orts by 39% (P < 0.01). Heifers fed SHRD had 6.6% greater ATTD of NDF than CONV-fed heifers (P = 0.04). Experiment 2 evaluated the effects of feeding either SHRD or CONV-processed WPSS on growth performance of beef heifers. Whole-plant grain sorghum was harvested at the hard-dough stage, switching the KP as in experiment 1. Angus heifers (n = 96, 15 ± 1 mo, 249.6 ± 28.6 kg of BW) were blocked by initial BW, and randomly assigned to pens (8 heifers/pen, 6 pens/treatment). Diets consisted, all on a DM basis, of WPSS, either SHRD or CONV, at 90.5%, expeller soybean meal at 7.0%, and a vitamin–mineral–protein concentrate at 2.5%. After 14-d of adaptation, growth was measured for 56 d, and feed was offered once daily. Data were analyzed using a randomized complete block design with the pen as the experimental unit. Heifers fed CONV had a 9.6% greater gain-to-feed ratio (P = 0.05) and a 7.4% greater Kleiber ratio (P = 0.05) than SHRD-fed heifers. Apparent net energy of gain tended to be 7.1% greater in CONV-fed heifers (P = 0.06). In conclusion, kernel processing WPSS increased starch digestibility and reduced fecal starch concentration. Using SHRD increased NDF digestibility and feeding CONV-processed WPSS resulted in enhanced growth performance.

Using either a shredlage or a conventional kernel processor increases the apparent total tract starch digestibility of sorghum silage fed to beef heifers. The use of a shredlage processor increased fiber digestibility, while the use of a conventional kernel processor enhanced growth performance of beef heifers.

Introduction

The use of whole-plant sorghum silage (WPSS) is a viable alternative for cattle production systems due to its reduced production costs and greater production potential in limited moisture regimes, as compared with corn (Abdelhadi and Santini, 2006; Getachew et al., 2016; Venkatesh Bhat, 2019). In the last 3 yr (2020 to 2022), approximately 147,710 ha of whole-plant sorghum were harvested every year in the United States, with an average yield of 29.4 as-fed metric tons per ha (USDA-NASS, 2018). Recent studies conducted by our group at the University of Florida showed that the use of WPSS-based diets for backgrounding beef cattle is a feasible alternative to byproducts and whole-plant corn silage-based diets (Podversich et al., 2023; Tarnonsky et al., 2023), with adequate body weight (BW) gain for this stage of growth. The ideal weight gain for replacement beef heifers is about 600 to 700 g/d to prevent future issues with reproduction and milk production (Funston et al., 2011; Patterson et al., 2016; D'Occhio et al., 2018).

Within the sorghum plant, starch is a significant source of energy, and most of it is contained within the endosperm of the kernels. However, the robust pericarp surrounding the endosperm impairs the utilization of starch by the ruminal microbes and intestinal enzymes. Among cereal grains, sorghum generally has the lowest starch digestibility (Rooney and Pflugfelder, 1986). Therefore, breakage of the kernels through on-board kernel processors during the harvesting of sorghum silage enhances starch utilization (Johnson et al., 2017; McCary et al., 2020) and has the potential to improve feed efficiency (Abdelhadi et al., 2022). Differences inherent to the kernel processor settings may result in different overall nutrient digestibility (Johnson et al., 2017; McCary et al., 2020). In 2008, the multicrop shredlage processor was developed with the intention of improving grain processing while maintaining greater proportions of physically effective fiber. The distinct characteristic of the shredlage processor is that its rolls have a counter-rotating spiral groove shape. Because of that, as the crop passes through the rollers, it gets pulled apart by the sideways movement of the teeth, which helps to crush the forage material (Ferraretto and Shaver, 2012; Vanderwerff et al., 2015). Feeding whole-plant corn silage processed with the shredlage kernel processor technology enhanced nutrient digestibility and improved the performance of dairy cows as compared to feeding whole-plant corn silage processed with a conventional kernel processor (Ferraretto and Shaver, 2012; Vanderwerff et al., 2015). However, scarce information is available regarding the effects of the shredlage processor on WPSS. For the case of sorghum, we consider that if the shredlage processor results in an effective processing method for sorghum silage harvesting, an advantage is that the on-board processor does not have to be replaced when alternating crops. The innovative aspect of this study is that, to our knowledge, it is the first to compare the effectiveness of conventional and shredlage-processed WPSS in beef cattle in vivo trials. Our objectives for these 2 experiments were to contrast the effects of the shredlage processor against a conventional sorghum processor on intake, nutrient digestibility, and growth performance of beef heifers fed a WPSS-based diet. Our hypotheses were that for experiment 1: A) the use of a kernel processor will increase nutrient digestibility of WPSS compared to unprocessed WPSS, B) there will be minimal differences in terms of intake and nutrient digestibility between shredlage or conventionally processed WPSS; for experiment 2: there will be minimal differences in intake and growth performance between shredlage or conventionally processed WPSS.

Material and Methods

All procedures were carried out according to the Guide for the Care and Use of Agricultural Animals in Agricultural Research and Teaching (FASS, 2011) and with the approval of the Institutional Ethics and Security Committee (Protocol No. CAES-292/16, Facultad de Ciencias Veterinarias, Universidad Nacional del Litoral, Santa Fe, Argentina). These experiments were conducted at “El Encuentro” Experimental Farm (35°23’ S; 58°25’ W), Ranchos, Buenos Aires, Argentina. This experimental farm provided the plant material, the animals, and the facilities for the experiment. The technical support representative of Claas Argentina conducted all the adjustments to the silage harvesters.

Whole-plant sorghum silage harvesting

For experiment 1, the whole-plant silage material was obtained from 12 ha of grain sorghum hybrid ADV1250IG (Advanta Seeds, Irving, TX), planted in dryland at a rate of 250,000 seeds/ha, and 35-cm row spacing. The sorghum was harvested at the hard-dough stage using 2 self-propelled harvesters from a private contractor (Silajes La Promesa S.A., Cañuelas, Argentina). The harvesters used were 1) A Claas Jaguar 960 (Claas of America, Columbus, IN)—Type 498—with an Orbis 750 header, equipped with a multicrop shredlage processor (Claas of America; SHRD) of 250-mm diameter rolls, and 110 and 145 teeth in the front and rear rolls, respectively, configured to operate at a speed differential of 60%; and 2) a Claas Jaguar 930—Type 494—with an Orbis 600 header, equipped with a multicrop processor (Claas of America; CONV) of 196-mm diameter rolls, with 125 teeth in both rolls, configured to operate at a speed differential of 50%. The Claas multicrop processor with 196-mm diameter rolls and 125 teeth is the specific processor designed for sorghum harvesting. The silage material for the unprocessed (UNP) treatment was generated by removing the processor in the Jaguar 960 harvester. The theoretical length of cut (TLC) was set at 15 mm for the 3 treatments, and the roll gap for both treatments using kernel processors (SHRD and CONV) was 1 mm. The 3 silages were treated with a homofermentative silage inoculant (Ecosyl, Volac International Limited, Byron, IL) to provide a minimum of 100,000 live Lactobacillus plantarum bacteria per gram of wet forage; and were stored in separate side-by-side 2.74-m-diameter silo bags and allowed to ferment for 30 d prior to the feeding trial.

For experiment 2, the whole-plant silage material was obtained from 25 ha of grain sorghum hybrid ADV2450IG (Advanta Seeds, Irving, TX), planted in dryland at a rate of 220,000 seeds per hectare, and 35-cm row spacing. The sorghum was harvested at the hard-dough stage using a self-propelled harvester from a private contractor (Lopez Seco S.A., Coronel Brandsen, Buenos Aires, Argentina). The harvester was a Claas Jaguar 950 (Claas of America, Columbus, IN)—with a Domai MPD-612 header. The material for the 2 treatments was obtained by switching the processors: multicrop shredlage processor (Claas of America; SHRD) with 250-mm diameter rolls and 110 and 145 teeth in the front and rear rolls, respectively, operating at a speed differential of 60%; and a multicrop processor (Claas of America; CONV) of 196-mm diameter rolls, with 125 teeth in both rolls, operating at a speed differential of 50%. In both cases, the TLC was set at 15 mm and the roll gap at 1 mm. Both silages were treated with a heterofermentative silage inoculant (Lacto Germs AA, Beneficial Germs S.A., Moreno, Buenos Aires, Argentina) to provide 2 × 10¹⁰ CFU of a mix of L. plantarum, Pediococcus acidilactici, Lactococcus lactis sub Lactis, and Lactobacillus buchneri per gram of wet forage; and were stored in separate side-by-side 2.74-m-diameter silo bags and allowed to ferment for 90 d prior to the feeding trial.

Experimental designs, animals, and diets

Experiment 1 was conducted as a completely randomized design (CRD), where 24 black Angus heifers (13 ± 1 mo of age, 267 ± 10.9 kg of BW were stratified by BW and randomly assigned to treatments (8 heifers per treatment). Heifers were housed in 24 pens of 36 m² each, with individual access to feed and water. The treatment diets consisted of ad libitum WPSS of each respective processing type (SHRD, CONV, or UNP), plus expeller soybean meal top-dressed at 0.5% (on a DM basis) of BW daily to meet or exceed the rumen degradable protein and metabolizable protein requirements (NASEM, 2016). The analyzed chemical composition of these ingredients can be found in Table 1. Along with the soybean meal, a vitamin–mineral concentrate (vitamin A: 1,000,000 IU/kg, vitamin D3: 200,000 IU/kg, vitamin E: 6,500 IU/kg, vitamin B1: 650 ppm, manganese: 12,000 ppm, zinc: 12,000 ppm, copper: 6,000, cobalt: 40 ppm, selenium: 60 ppm, iodine: 200 ppm; Vetifarma, Buenos Aires, Argentina) was provided at a rate of 20 g/d to meet or exceed daily requirements of microminerals and vitamins.

Table 1.

Chemical composition and particle size of the different ingredients fed to beef heifers (experiment 1)

Item	Whole plant sorghum silage			Expeller soybean meal
	Unprocessed	Shredlage processed	Conventionally processed
DM, % as fed	34.2	34.1	34.9	94.0
OM, % of DM	93.7	94.0	93.6	95.4
NDF, % of DM	38.9	42.2	42.8	13.0
ADF, % of DM	20.3	21.4	22.3	8.0
CP, % of DM	6.9	6.5	6.2	44.1
Starch, % of DM	30.0	29.4	32.8	5.5
Particle size1, % of as fed retained
19.0 mm	33.4	35.3	5.8	-
8.0 mm	43.4	42.0	64.7	-
1.18 mm	22.5	21.7	28.2	-
Pan	0.8	1.0	1.4	-

¹Particle size was measured using the Penn State Particle Size Separator (Nasco, Fort Atkinson, WI) as described by Kononoff et al. (2003).

The experimental period for the digestibility trial (experiment 1), consisted of a 15-d adaptation period plus 5 d of feed and fecal sample collection. Day 0 was considered the beginning of the treatment diet feeding period (i.e., the beginning of the adaptation). Feed was offered once daily during the collection period, and orts were collected the following day. To measure intake, feed offered, and orts were weighed using a portable scale (Mini Crane Scale OCS-L, IVS-engineering, Bulgary). Feed samples were collected from days 15 to 18, and orts were collected from days 16 to 19 to determine DM and nutrient composition. Fecal samples were collected twice daily (at 0800 and 1900 hours) for 4 consecutive days (from 16 to 19) directly from the pen surface, avoiding contamination with dirt from the pen.

Experiment 2 was conducted using a randomized complete block design (RCBD), using 96 black Angus heifers (15 ± 1 mo of age, 249.6 ± 28.6 kg of BW). On day 14 (beginning of the adaptation period), heifers were blocked by initial BW (6 blocks) and assigned randomly to pens (8 heifers/pen, 6 pens/treatment). Pens had a surface of 96 m², and a linear bunk space of 1 m per heifer. The experimental diets consisted of (on a DM basis) WPSS of each respective type (SHRD or CONV) at 90.5%, expeller soybean meal at 7.0%, and a vitamin–mineral–protein concentrate at 2.5% (“Premix MEAT with Urea”, Vetifarma S.A., Abasto, BA, Argentina). The analyzed chemical composition of the diets can be found in Table 2.

Table 2.

Ingredients and analyzed¹ chemical composition of the silage-based diets fed to beef heifers (experiment 2)

Item	Diets2
	CONV3	SHRD4
Ingredients, % DM basis
Sorghum silage conventionally processed	90.5	—
Sorghum silage shredlage processed	—	90.5
Expeller soybean meal	7.0	7.0
Vitamin–mineral–NPN concentrate	2.5	2.5
Nutrient composition5
DM, % as fed	31.3	29.7
OM, % of DM	90.6	90.4
NDF, % of DM	46.2	45.5
ADF, % of DM	30.2	31.0
CP, % of DM	11.9	12.0
Starch, % of DM	13.8	10.4
EE, % of DM	3.2	2.7
Lignin, % of DM	2.8	2.4
Calcium, % of DM	0.59	0.66
Phosphorus, % of DM	0.18	0.16
Magnesium, % of DM	0.18	0.17
Potassium, % of DM	1.43	1.51
Sulfur, % of DM	0.12	0.13
In situ ruminal starch digestibility
At 0 h, % of starch	41.7	37.6
At 7 h, % of starch	65.3	59.6
Energy6
TDN, %	65.1	65.6
ME, Mcal/kg of diet DM	2.40	2.39
Calculated from ME7
NEm, Mcal/kg of diet DM	1.43	1.42
NEg, Mcal/kg of diet DM	0.85	0.84
Calculated from tabular values8
NEm, Mcal/kg of diet DM	1.34	1.34
NEg, Mcal/kg of diet DM	0.67	0.67

¹Rock River Laboratory Argentina, Santa Fe, SF Argentina.

²DIETS: Both diets were comprised, on a DM basis, of WPSS 90.5%, expeller soybean meal 7.0%, micro concentrate (Premezcla meat con urea, Vetifarma S.A, Abasto, BA, Argentina) 2.5%.

³CONV: Conventionally processed sorghum silage-based diet.

⁴SHRD: Shredlage processed sorghum silage-based diet.

⁵Nutrients: DM, dry matter; OM, organic matter, NDF, neutral detergent fiber; ADF, acid detergent fiber; CP, crude protein; EE, ether extract.

⁶TDN, total digestible nutrients. TDN = 4.4 Mcal Digestible energy (DE). ME, metabolizable energy based on Galyean et al. (2016), with adjustments for nutrients (ME = 0.929 × DE − 0.0056 × CP + 0.0343 × EE + 0.0042 × starch − 0.3612). NEm, net energy of maintenance; NEg, net energy of gain.

⁷Based on NASEM (2016): NEm = 1.37 ME − 0.138 ME² + 0.0105 ME³ − 1.12; NEg = 1.42 ME − 0.174 ME² + 0.0122 ME³ − 1.65.

⁸Based on Tabular values by Preston (2017).

The experimental period for the performance trial (experiment 2) consisted of a 14-d adaptation period (days −14 to −1) followed by a 56-d growing period (days 0 to 56). Shrunk BW was obtained on days −14, 0, and 56; additionally, interim unshrunk BW was obtained on day 28. Shrunk BW consists of a 16-h period of water and feed withdrawal (NASEM, 2016). Experimental diets were delivered once daily at 0800 hours using a tractor-pulled mixer equipped with ± 5kg precision scale (Mixer SENOR 228-10, Roldan, SF, Argentina). Throughout the trial, to adjust feed delivery, bunk scores were assessed every morning at 0800 hours for each pen based on a visual appraisal of feed remaining using the following scoring system: 0 = empty bunk; 0.5 = thin layer of feed; 1 = thick layer of feed; 2 = 25% to 50% of feed offered; 3 = >50% of feed offered, with a target score of 0.5. Adjustments on daily feed delivery were made as follows: for score 0, increased 4 kg/pen as fed, and for score 1, reduced 4 kg/pen. Feed was delivered, orts were removed, and bunk scores were manually recorded, and biweekly feed samples were collected to further determine DM and nutrient composition. At the final BW measurement (day 56), blood samples were collected via coccygeal venipuncture in glass tubes (Vacutainer, Becton Dickinson, Franklin Lakes, NJ) from a subsample of 4 heifers randomly selected from each pen for further analysis of blood metabolites. Tubes were centrifuged for 10 min at 2,000 × g to separate the serum and then stored at −20 °C until submitted to the Applied Cellular and Molecular Biology Lab, Litoral Institute of Veterinary Sciences (ICIVET Litoral), University of Litoral/National Council for Scientific and Technical Research (UNL-CONICET), Esperanza, Argentina.

Nutrient digestibility determination (experiment 1)

The apparent total tract digestibility (ATTD) of nutrients was determined using indigestible NDF (iNDF) as an internal marker. Immediately after being collected, feed, orts, and fecal samples from each heifer at each time point were individually stored and frozen at −20 °C; feed and orts samples were weighed before being stored in the freezer. At the end of the experiment, all samples were thawed and dried at 55 °C for at least 48 h in a forced-air oven in the Universidad Nacional de La Plata, BA, Argentina. Once dried, feed and orts samples were re-weighed to determine the moisture content, and all samples were ground in a Wiley mill (Arthur H. Thomas Co., Philadelphia, PA) to pass a 2-mm screen. Feed, orts, and fecal samples were composited per heifer on an equal weight basis to determine nutrient and marker concentration.

Laboratory analysis

The DM content of feed and orts was determined by weighing the entire sample prior to and after oven drying. To determine sample DM and organic matter (OM) concentration, approximately 0.5 g of the sample was weighed in duplicate, dried in a forced-air oven at 100 °C for 24 h, and ashed at 550 °C for 6 h. For determination of the fibrous component, samples were weighed in duplicate into F57 bags (Ankom Technology Corp., Macedon, NY) and analyzed for neutral detergent fiber (NDF), using heat-stable α-amylase and sodium sulfite, and subsequently for acid detergent fiber (ADF) as described by Van Soest et al. (1991) in an Ankom 200 Fiber Analyzer (Ankom Technology Corp.). The concentration of crude protein (CP) was determined by rapid combustion using a macro elemental N analyzer (Vario Max CN, Elementar Americas Inc., MT. Laurel, NJ) following official method 992.15 (AOAC, 1995). Starch concentration was measured by an enzymatic-colorimetric method as Hall et al. (2015) described. To analyze concentrations of indigestible neutral detergent fiber (iNDF) in feed and feces, the method described by Cole et al. (2016) with the modification proposed by Krizsan and Huhtanen et al. (2013) was used. Briefly, 0.5 g of sample were weighed into F57 filter bags (Ankom Technology Corp.), and then incubated into mesh laundry bags within the rumen of a cannulated steer for 288 h (12 d). After incubation, samples were rinsed and analyzed for NDF concentration in an Ankom 200 Fiber Analyzer (Ankom Technology Corp.) using sodium sulfate and heat-stable α-amylase.

Concentrations of albumin (ALB), blood urea nitrogen (BUN), creatinine (CRE), total protein (TP), triglycerides (TG), cholesterol (CHOL), gamma-glutamyl transpeptidase (GGT), and total bilirubin (TB) were determined enzymatically by using an automatic biochemistry analyzer (CM 250, Wiener Lab Group, Rosario, Argentina). The concentration of non-esterified fatty acids (NEFA) and total antioxidant status (TAS) was assessed spectrophotometrically by using commercial kits (Randox Laboratories Ltd, United Kingdom) with validated micro-methods for small volumes. The reactions were read on a microplate reader (SPECTROstar Nano, BGM LABTECH, Ortemberg, Germany). The concentration of beta-hydroxy butyrate (BHB) was assessed by using reactive strips (FreeStyle Optium Xceed, Abbott Diabetes Care Ltd, Oxon, United Kingdom). Similar determinations were conducted by Gareis et al. (2018) and Angeli et al. (2023).

Calculations

For both trials, dry matter intake (DMI) was determined by the difference between feed offered and orts on a DM basis. For the digestibility trial (experiment 1), the ATTD of DM, OM, NDF, ADF, CP, and starch were calculated using the following formula:

$${\displaystyle \begin{array}{l}{\displaystyle 100-100\times [\left(\frac{\mathrm{m}\mathrm{a}\mathrm{r}\mathrm{k}\mathrm{e}\mathrm{r}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{e}\mathrm{d}}{\mathrm{m}\mathrm{a}\mathrm{r}\mathrm{k}\mathrm{e}\mathrm{r}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}}\right)\times \left(\frac{\mathrm{n}\mathrm{u}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{s}}{\mathrm{n}\mathrm{u}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{f}\mathrm{e}\mathrm{e}\mathrm{d}}\right)]}\end{array}}$$

In the same study, the percentage difference in NDF concentration between diet and refusals was calculated using the following formula:

$${\displaystyle \begin{array}{l}{\displaystyle 100\times \left(\frac{\mathrm{N}\mathrm{D}\mathrm{F}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{r}\mathrm{e}\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{a}\mathrm{l}\mathrm{s}-\mathrm{N}\mathrm{D}\mathrm{F}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{t}}{\mathrm{N}\mathrm{D}\mathrm{F}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{e}\mathrm{t}}\right)}\end{array}}$$

For the performance trial (experiment 2), initial BW and final BW were obtained after a shrink period of 16-h on days 0 and 56, respectively. Mid-test BW was calculated as (initial BW + final BW)/2. Metabolic BW (MBW) was determined as mid-test BW^0.75. Total BW gain was calculated as the difference between the final BW and the initial BW, and average daily gain (ADG) was calculated as total BW gain divided by the number of days on trial (56 d). The gain-to-feed ratio (G:F; kg/kg) was computed as the ratio of ADG to daily DMI. Residual feed intake (RFI; kg) was calculated as actual DMI minus expected DMI (Koch et al., 1963), with expected DMI derived from the regression of actual DMI on ADG and mid-test MBW using the regression equations provided by the GLM procedure of SAS 9.4 (SAS Inst. Inc., Cary, NC). Kleiber ratio (KR; g/kg) was calculated as the ratio of ADG to MBW (Kleiber, 1936).

The net energy (NE) for each diet was calculated from performance data using the equation proposed by Zinn and Shen (1998) based on pen intake and ADG. Energy gain (EG) was calculated as EG = (0.0557 × EqSBW^0.75) × EBG^1.097, where EG is daily energy deposited (Mcal/d), EqSBW is equivalent shrunk BW, calculated as mid-test shrunk BW × (reference BW / final shrunk BW). The reference BW is 478, from the reference NRC 1984 choice medium frame steer calf (NASEM, 2016). For the final shrunk BW, an adjusted final body weight (AFBW) to 28% empty body fat (EBF) was obtained using carcass data from a set of contemporary heifers that were finished and harvested, using the following equation: AFBW = (Empty body weight + [(28—EBF) × 14.26])/0.891. Where Empty body weight = (1.316 × Hot carcass weight) + 32.29, and EBF = 17.76207 + (4.68142 × rib fat) + (0.01945 × Hot carcass weight) + (0.81855 × quality grade) − (0.06754 × longissimus area), based on Guiroy et al. (2002). EBG is empty ADG calculated as shrunk ADG × 0.956 (NASEM, 2016). Expended maintenance energy (EM; Mcal/d) was calculated as EM = 0.077 × BW^0.75 (Lofgreen and Garret, 1968). Using the calculated amounts of energy required for maintenance and gain, apparent Net Energy of maintenance (NEm) of each diet was obtained by the quadratic equation:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{N}\mathrm{E}\mathrm{m}=\frac{[-b\pm {(b^2-4ac)}^{0.5}]}{2a}}\end{array}}$$

where a = −0.41 × EM, b = 0.877 × EM + 0.41 × DMI + EG, and c = −0.877 × DMI (Zinn and Shen, 1998). Apparent metabolizable energy (ME) of each diet was obtained as ME = 0.896706 + (0.847878 × NEm_diet) + 0.100045 × NEm_diet^2-0.003842 × NEm_diet³. Apparent Net Energy of gain (NEg) of each diet was obtained by the equation Zinn and Shen (1998):

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{N}\mathrm{E}\mathrm{g}=0.877\times \mathrm{N}\mathrm{E}\mathrm{m}-0.41}\end{array}}$$

Statistical Analysis

Both experiments were analyzed using the MIXED Procedure of SAS 9.4 (SAS Institute Inc., Cary, NC). For experiment 1, data were analyzed as a CRD, heifer was considered the experimental unit, and the model included the fixed effect of treatment. Initial BW was tested as a covariate, remaining in the model when significant. Contrasts were used to determine the effects of A) Processing: UNP vs. (SHRD + CONV); and B) Processor type: SHRD vs. CONV. For experiment 2, data were analyzed as a RCBD, pen was considered the experimental unit, and the model included the fixed effect of treatment and the random effect of block. In both experiments, significance was declared at P ≤ 0.05, and tendencies were considered when 0.10 > P > 0.05. Finally, correlation coefficients with animals as the experimental unit were determined for the correlations among serum metabolites and growth performance parameters by using the CORR procedure of SAS, either as Pearson or Spearman ranks.

Results

Digestibility trial (experiment 1)

For this experiment, all the results will be presented in the form of the contrasts A) Processing: Unprocessed vs. Processed [Shredlage + Conventional], and B) Processor type: Shredlage vs. Conventional).

Intake data for experiment 1 is shown in Table 3. Initial BW did not differ among treatments and averaged 273.1 kg (P > 0.10). Neither processing nor processor type affected the intake of DM, ADF, and CP (P > 0.10) expressed in absolute terms (kg/d). However, heifers fed kernel-processed silage tended to consume 5% more NDF (P = 0.10; 3.03 vs. 3.20 kg/d) and 3% less starch (P = 0.09; 2.40 kg/d vs. 2.33 kg/d); while no differences were observed between the processors (P > 0.10). In addition, expressed as a percentage of BW, heifers fed processed silage tended to consume 5% less total DM (P = 0.10; 3.31% vs. 3.14% BW), while no differences were observed between processors (P > 0.10). Even though no differences were observed for intake of NDF and ADF expressed as a percentage of BW (P > 0.10). Finally, processing the sorghum silage, reduced the percentage change of NDF concentration between diet offered and refusals by 40% (P < 0.01; 30.8% vs. 18.7% change of NDF) without differences (P = 0.76) between the shredlage or conventional processor (Figure 1).

Table 3.

Effects of feeding WPSS either unprocessed, conventionally processed or shredlage processed, on intake and nutrient digestibility of beef heifers (experiment 1)

Item	Treatment1				P-value2
	UNP3	CONV4	SHRD5	SEM6	Processing7	Processor type8
Heifers, n	8	8	8	—	—	—
BW, kg	273.5	272.0	273.7	1.03	0.63	0.24
Intake, kg/d
DM	9.08	8.32	8.77	0.250	0.74	0.14
OM	8.55	7.82	8.27	0.234	0.30	0.46
NDF	3.03	3.12	3.27	0.129	0.10	0.17
ADF	1.60	1.63	1.67	0.073	0.57	0.73
CP	1.07	1.05	1.08	0.016	0.63	0.34
Starch	2.40	2.41	2.25	0.109	0.09	0.19
Intake, as % of BW
DM	3.31	3.05	3.22	0.083	0.10	0.22
NDF	1.11	1.15	1.20	0.045	0.31	0.43
ADF	0.59	0.60	0.61	0.027	0.53	0.75
Digestibility, % of intake
DM	71.4	67.9	70.7	1.00	0.11	0.07
OM	73.1	70.2	72.3	1.02	0.16	0.17
NDF	53.6	50.0	53.3	1.09	0.15	0.04
ADF	50.7	45.4	48.2	1.84	0.10	0.30
CP	64.9	63.0	66.3	1.51	0.91	0.13
Starch	86.3	90.1	90.2	1.14	0.01	0.92

¹Heifers were fed ad libitum sorghum silage plus expeller soybean meal at 0.5% of the initial BW, on a DM basis.

²Observed significance level (n = 8 heifers/diet) for the contrasts.

³Unprocessed sorghum silage-based diet.

⁴Conventionally processed sorghum silage-based diet.

⁵Shredlage processed sorghum silage-based diet.

⁶Standard error of the mean, n = 8 heifers/diet.

⁷Contrast the nonuse of processor (UNP), vs. using a kernel processor (CONV + SHRD).

⁸Contrast the 2 processors (CONV vs. SHRD).

Figure 1.

Beef heifers were fed ad libitum whole-plant sorghum silage either unprocessed (UNP, conventionally processed (CONV; Multicrop Kernel Processor, Claas) or shredlage processed (SHRD; Shredlage processor, Claas). Kernel processing (UNP vs. [CONV + SHRD]) reduced (P < 0.01, SEM = 3.48) the percentage change of NDF concentrations from diet to refusals; without differences (P = 0.76, SEM = 4.03) between the type of processor utilized (CONV vs. SHRD). Percentage change of NDF concentration = (NDF concentration in orts − NDF concentration diet)/NDF concentration diet × 100.

Digestibility data for experiment 1 is shown in Table 3. As compared with non-processed silage, the use of kernel processor did not affect the digestibility of DM, OM, NDF, and CP (P > 0.10); while ADF digestibility tended to be lower for processed silage-fed heifers (P = 0.10; 50.7% vs. 46.8%). Yet, as expected, processing the sorghum silage increased starch digestibility (P = 0.01; 86.3% vs. 90.2 %). When comparing the 2 processors evaluated, DM digestibility tended to be greater (P = 0.07; 67.9% vs. 70.7%) and NDF digestibility was greater (P = 0.04; 50.0% vs. 53.3%) for SHRD-fed heifers as compared to CONV-fed heifers; without differences for OM, ADF, CP and starch digestibility (P > 0.10). Moreover, the concentration of fecal starch was reduced in the heifers fed processed silage (P < 0.01; 12.0% vs. 8.7%), without differences (P = 0.71) between the processors evaluated (Figure 2).

Figure 2.

Beef heifers were fed ad libitum whole plant sorghum silage either unprocessed (UNP), conventionally processed (CONV; Multicrop Kernel Processor, Claas) or shredlage processed (SHRD; Shredlage processor, Claas). Kernel processing (UNP vs. [CONV + SHRD]) reduced (P < 0.01, SEM = 0.96) fecal starch concentration; processor type (CONV vs. SHRD) did not affect (P = 0.71, SEM = 1.11) this parameter.

Performance trial (experiment 2)

Growth performance data for experiment 2 is shown in Table 4. Initial BW, final BW, and mid-test MBW did not differ among treatments (P > 0.10). Heifers fed the SHRD diet tended to gain less weight than CONV-fed heifers, either expressed as total BW (P = 0.08; 44.0 kg vs. 47.2 kg) or as ADG (P = 0.07; 0.77 kg/d vs. 0.83 kg/d). No differences were observed for DMI between treatments, expressed either in absolute terms or as percentage of BW (P > 0.40; average 10.2 kg/d and 3.59% of BW, respectively). Heifers fed the SHRD diet exhibited a decreased G:F ratio (P = 0.05; 0.075 vs. 0.083 kg/kg) and KR (P = 0.04; 11.2 g/kg vs. 12.1 g/kg) than those fed the CONV diet. However, RFI did not differ between diets (P = 0.64). Based on growth performance, the apparent energy content of the SHRD diet was less than the CONV diet (P = 0.01); with 3.5% less for apparent ME (P = 0.01; 2.02 Mcal/kg vs. 1.95 Mcal/kg), 6% less for apparent NEm (P = 0.01; 1.17 Mcal/kg vs. 1.10 Mcal/kg), and 9.7% less for apparent NEg (P = 0.01; 0.62 Mcal/kg vs. 0.56 Mcal/kg), when compared to the CONV diet. The apparent NEg (performance calculated) approximates the tabular expected NEg based on Preston (2017) values, with an observed to expected NEg ratio of 0.92 and 0.83 for the CONV and SHRD diet (P = 0.01), respectively.

Table 4.

Effects of feeding WPSS either conventionally processed or shredlage processed, on intake and growth performance of beef heifers (experiment 2)

Item	Diets1			P-value5
	CONV2	SHRD3	SEM4
Pens, n	6	6	—	—
Heifers, n	48	48	—	—
Initial BW, kg	260.7	264.1	1.76	0.22
Final BW, kg	307.8	308.0	2.27	0.94
Mid-test MBW, kg	69.2	69.5	0.36	0.55
Total BW gain, kg	47.2	44.0	1.01	0.08
Average daily gain, kg	0.83	0.77	0.018	0.07
Intake, kg/d	10.1	10.3	0.17	0.46
Intake, % of BW	3.56	3.61	0.041	0.41
Gain-to-feed, kg/kg	0.083	0.075	0.0020	0.05
Residual Feed Intake, kg	−0.04	0.04	0.127	0.64
Kleiber Ratio, g/kg	12.1	11.2	0.232	0.04
Performance calculated dietary energy
ME, Mcal/kg of diet	2.02	1.95	0.025	0.01
NEm, Mcal/kg of diet DM	1.17	1.10	0.023	0.01
NEg, Mcal/kg of diet DM	0.62	0.56	0.020	0.01
Observed to expected NEg6
Expect. From ME calculated	0.73	0.66	0.024	0.02
Expect. Tabular	0.92	0.83	0.030	0.01

¹Both diets were comprised, on a DM basis, of WPSS 90.5%, expeller soybean meal 7.0 %, micro concentrate (Premezcla meat con urea, Vetifarma S.A, Abasto, Argentina) 2.5%.

²Conventionally processed sorghum silage-based diet.

³Shredlage processed sorghum silage-based diet.

⁴Standard error of the mean (n = 6 pens/diet).

⁵Observed significance level for diet (n = 6 pens/diet).

⁶Observed to performance calculated NEg, expected from ME (NASEM 2016), expected from tabular (Preston 2017).

Serum metabolite data for experiment 2 is shown in Table 5. For nitrogen metabolism end products, no differences were observed for serum concentrations of TP and BUN (P > 0.10); yet serum albumin concentration of SHRD-fed heifers tended to be greater than CONV-fed heifers (P = 0.08; 3.75 g/dL vs. 3.62 g/dL). For blood energy metabolism parameters, no differences were observed for serum BHB and TG (P > 0.10); conversely, SHRD-fed heifers exhibited lower serum CHOL (P = 0.02; 160.5 mg/dL vs. 180.5 mg/dL) and lower NEFA concentrations (P = 0.03; 0.53 mmol/L vs. 0.55 mmol/L). Also, no differences were observed for the homeostatic indicators CRE, GGT, and TAS (P > 0.10); however, SHRD-fed heifers exhibited a decreased serum TB (P = 0.04; 0.090 mg/dL vs. 0.105 mg/dL).

Table 5.

Effects of feeding WPSS either conventionally processed or shredlage processed, on serum metabolites of beef heifers (experiment 2)

Metabolites1	Diets2			P-value6
	CONV3	SHRD4	SEM5
Pens, n	6	6	—	—
Heifers, n	24	24	—	—
Nitrogen metabolism
ALB, g/dL	3.62	3.75	0.040	0.08
TP, g/dL	7.96	8.20	0.111	0.19
BUN, mg/dL	34.7	31.1	2.53	0.36
Energy metabolism
CHOL, mg/dL	180.5	160.5	4.09	0.02
BHB, mmol/L	0.11	0.13	0.012	0.52
NEFA, mmol/L	0.55	0.53	0.005	0.03
TG, mg/dL	40.8	39.2	1.43	0.45
Homeostasis
CRE, mg/dL	1.27	1.18	0.056	0.31
TB, mg/dL	0.105	0.090	0.0040	0.04
GGT, IU/L	11.3	11.2	0.79	0.97
TAS, mmol/L	0.91	0.86	0.066	0.59

¹ALB, albumin; BHB, beta-hydroxy butyrate; BUN, blood urea nitrogen; CRE, creatinine; CHOL, cholesterol; GGT, gamma-glutamyl transferase; NEFA, non-esterified fatty acids; TAS, total antioxidant status; TB, total bilirubin; TG, triglycerides; TP, total proteins.

²Both diets were comprised, on a DM basis, of WPSS 90.5%, expeller soybean meal 7.0%, micro concentrate (Premezcla meat con urea, Vetifarma S.A, Abasto, Argentina) 2.5%.

³Conventionally processed sorghum silage-based diet (196-mm diameter rolls, with 125 teeth rolls and 60% of speed differential).

⁴Shredlage processed sorghum silage-based diet (250-mm diameter rolls, 110 and 138 teeth roll, and 50% of speed differential).

⁵Standard error of the mean (n = 6 pens/diet).

⁶Observed significance level for Diet (n = 6 pens/diet).

Correlations (r) among serum metabolites and growth performance parameters for experiment 2 are shown in Table 6. No correlations were observed between any of the serum metabolites analyzed and MBW (P > 0.10, data not shown) and between the serum concentrations of BUN, CHOL, and TG with any of the growth performance variables evaluated (P > 0.10, data not shown). On the other hand, there was a strong negative correlation between serum concentration of NEFA, and RFI (r = −0.71, P = 0.01). Also, BHB tended to be positively correlated with RFI (r = 0.55, P = 0.06) and negatively correlated with GF (r = −0.52, P = 0.08) and apparent NEg (r = −0.53, P = 0.08). Conversely, serum ALB was negatively correlated with GF (r = −0.59, P = 0.04), and apparent NEg (r = −0.67, P = 0.02), and positively correlated with RFI (r = 0.68, P = 0.02). Additionally, TP concentration was negatively correlated with GF (r = −0.60, P = 0.04) and apparent NEg (r = −0.62, P = 0.03), and tended to be negatively correlated with KR (r = −0.51, P = 0.09). Furthermore, there was a strong positive correlation between serum TB with ADG (r = 0.73, P = 0.01), GF (r = 0.62, P = 0.03), KR (r = 0.65, P = 0.02), and apparent NEg (r = 0.67, P = 0.02). In contrast, serum GGT tended to be negatively correlated with ADG (r = −0.54, P = 0.07) and KR (r = −0.51, P = 0.09), and was negatively correlated with GF (r = −0.62, P = 0.03) and apparent NEg (r = −0.60, P = 0.04). Similarly, serum CRE tended to be negatively correlated with RFI (r = −0.53, P = 0.08).

Table 6.

Correlations between serum metabolites and growth performance parameters of beef heifers fed sorghum silage-based diets (experiment 2)

	Growth performance
Serum metabolites	DMI	ADG	GF	KR	NEg	RFI
Pearson
NEFA
r	−0.48	−0.14	0.20	0.02	0.28	−0.71
P-value	0.11	0.67	0.53	0.94	0.39	0.01
Spearman rank
ALB
r	0.41	−0.48	−0.59	−0.47	−0.67	0.68
P-value	0.18	0.11	0.04	0.13	0.02	0.02
TP
r	0.05	−0.46	−0.60	−0.51	−0.62	0.44
P-value	0.89	0.14	0.04	0.09	0.03	0.15
BHB
r	0.38	−0.40	−0.52	−0.33	−0.53	0.55
P-value	0.22	0.20	0.08	0.29	0.08	0.06
CRE
r	−0.41	−0.24	0.08	−0.03	0.14	−0.53
P-value	0.19	0.45	0.80	0.93	0.66	0.08
TB
r	−0.11	0.73	0.62	0.65	0.67	−0.27
P-value	0.73	0.01	0.03	0.02	0.02	0.39
GGT
r	0.24	−0.54	−0.62	−0.51	−0.60	0.34
P-value	0.46	0.07	0.03	0.09	0.04	0.28

¹DMI: Dry matter intake as percentage of BW, ADG: average daily gain, GF: gain-to-feed, KR: Kleiber ratio, NEg: apparent net energy of gain, RFI: residual feed intake.

²ALB: albumin, BHB: beta-hydroxybutyrate, CRE: creatinine, GGT: gamma-glutamyl transferase, NEFA: non-esterified fatty acids, TB: total bilirubin, TP: total proteins.

Discussion

The DMI in both experiments averaged 3.2% and 3.6% of BW, for experiments 1 and 2, respectively, which is slightly greater than what has been observed in other studies with growing cattle fed either corn or sorghum silage (Podversich et al., 2023; Tarnonsky et al., 2023). However, these intake levels are in agreement with a previous study conducted at the same experimental farm, with similar heifers and diets, where DMI averaged 3.6% of BW (Abdelhadi et al., 2022). Several factors may be responsible for the greater than 3% of BW consistently recorded in this experimental farm, including genetics (black Angus), biotype (moderate frame with good body capacity), and climate (temperate weather). In comparison, in a previous study by Gutierrez et al. (1982), feeding either processed or unprocessed sorghum silage diets resulted in a DMI of approximately 2.5% of BW. Also, in the previous studies conducted by our group, the DMI was between 2.5% and 2.6% of BW (Podversich et al., 2023; Tarnonsky et al., 2023) for sorghum silage-based diets. Hart (1987) demonstrated that the DMI of sorghum silage-based diets increases quadratically with a greater proportion of grain. Young et al. (1996) showed a linear increase in DMI by sheep fed a sorghum silage-based diet as grain content increased.

Another noteworthy result from the digestibility trial (experiment 1) is that heifers receiving unprocessed silage tended to consume more feed than those receiving kernel-processed diets. Kernel processing affects changes in the particle size distribution, which may prevent the animals from sorting the forage to consume only the finer particles. Therefore, the observed intakes can be related to potentially greater sorting behavior in the unprocessed diet. Even though this experiment was not designed to test sorting behavior, this could be inferred from the differences in the percentage of change in NDF concentration between the diet offered and refusals. The greater change for the unprocessed silage indicates that heifers could easily sort through the coarser material to select finer particles in that treatment. The finer portions of the sorghum silage, in theory, would have less proportion of stalks (typically greater NDF and lignin concentration, less digestible and slower to ferment in the rumen) and a greater proportion of kernels and leaves (with a greater content of soluble carbohydrates, more digestible and rapidly fermentable). In support of this idea, it was also noted that heifers fed processed diets tended to consume 5% more total NDF (average of 0.165 kg/d more). Gutierrez et al. (1982) also observed that beef steers fed unprocessed sorghum silage consumed more feed than a group fed processed sorghum silage. Andrae et al. (2001) reported that beef steers consuming a diet containing 60% corn silage left a greater proportion of cobs and had increased NDF digestibility when the silage was unprocessed. Also, Kononoff et al. (2003) found that feeding dairy cows with corn silage of longer particle size resulted in greater consumption of small particles. Similarly, Leonardi et al. (2005) demonstrated that feeding dairy cows with longer-fiber alfalfa hay in a TMR diet increased the intake of finer particles.

The use of processors reduced the observed ADF digestibility and this can be related to the fact that for processed silage-fed heifers, the material that the animals end up consuming probably contained a greater proportion of stalks (more lignified, harder to digest) and less proportion of leaves (more degradable). The main advantage of using a processor was the increased starch digestibility, compared to the unprocessed treatment. This finding is supported by the reduced fecal starch concentration observed in heifers fed the processed WPSS diets. In agreement with these results, the study by Johnson et al. (2017) showed that processing WPSS positively affected the 7-h in situ starch digestibility, with greater efficacy at reduced roll gap. Furthermore, a recent study by Abdelhadi et al. (2022), found that heifers fed kernel-processed WPSS had greater feed efficiency than unprocessed WPSS fed heifers. Conversely, Gutierrez et al. (1982), failed to prove differences in apparent digestibility between unprocessed and pre-ensiled rolled sorghum silage. In our study, the greater NDF digestibility of the SHRD diet compared to the CONV diet can explain why there was a tendency for greater DM digestibility in the SHRD diet. In agreement with our findings, Ferraretto and Shaver (2012) found that dairy cows fed TMR diets based on shredlage processed corn silage ate more feed and were more efficient than cows fed TMR diets based on conventionally processed corn silage. The authors suggested that the increased intake may be related to increased NDF digestibility (Ferraretto and Shaver, 2012). However, in the mentioned study, improved total tract starch digestibility of corn silage was observed when shredlage processor was used (Ferraretto and Shaver, 2012), whereas in our study with sorghum silage, this difference was not observed.

To the best of our knowledge, experiment 2 is the first study in which the conventional and shredlage processors for sorghum silage harvesting are contrasted in terms of growth performance in beef cattle. The observed ADG fell in the range between 0.66 and 0.96 kg/d reported in previous studies with similar diets with more than 88% dietary inclusion of sorghum silage on a DM basis (Gutierrez et al., 1982; Podversich et al., 2023; Tarnonsky et al., 2023). The variability in growth performance among studies with similar sorghum silage inclusion could be related to different grain content in the WPSS utilized. The tendency for a reduced ADG for the SHRD-fed heifers, paralleled with no differences in DMI, can explain the reduction in calculated G:F ratio, KR, and performance-based apparent ME, NEm, and NEg. Interestingly RFI was similar between diets, highlighting the different nature of all the previously mentioned feed efficiency parameters. With the information obtained from this study, it cannot be determined the reason why SHRD-fed heifers tended to have reduced BW gain and ADG, and decreased feed efficiency as compared to CONV-fed heifers. However, some potential mechanistic insights could be considered. One plausible explanation is that the growth performance of SHRD-fed heifers is affected by the different starch concentrations between CONV and SHRD sorghum silages used in experiment 2; since no differences in nutrient digestibility were observed between these 2 processors in experiment 1. A second reason could be a shift in the site of digestion for starch. In this sense, based on the results from experiment 1, we could determine that there were no differences in ATTD of starch; however, the site of starch degradation was not determined in the present study. Ferrareto and Shaver (2012) determined that shredlage-processed corn silage had greater ruminal degradation than conventional kernel-processed corn silage, both at 12 and 24 h of in situ ruminal incubations. In line with this, changing the site of digestion towards greater ruminal degradation could lead to enhanced nitrogen utilization, microbial CP synthesis, and feed efficiency. In contrast with our results, in a study conducted at the University of Nebraska with finishing diets where corn silage was used as the roughage source, growth performance and feed efficiency were improved if the silage was shredlage processed compared to conventionally processed silage (Conroy et al., 2020). Possibly, in a finishing diet with lower inclusion of silage in the diet (9% to 14%) as it is the case in the study by Conroy et al. (2020), the enhanced performance can be related to greater rumination time and enhanced ruminal health. Whereas the 90% silage inclusion in our growing diet should not present any challenges in terms of rumination time and ruminal health.

The apparent NEg obtained was closer to the expected NEg content based on tabular values by Preston (2017) than that of the NEg calculated as a function of TDN and ME, based on NASEM (2016). A few reasons could be responsible for this discrepancy. NASEM (2016) emphasizes that the relationship between digestible energy (DE) to ME may be variable depending on several factors like age of the cattle, intake level, forage to concentrate, etc. For that reason, the equation by Galyean et al. (2016) was chosen to compute ME from DE, since it corrects for CP, EE, and starch content. Furthermore, NASEM (2016) suggests further studies should reevaluate the accuracy of the conversion between DE and ME. The tabular values from Preston (2017) are generated from performance studies, therefore, these values are in closer agreement with our performance-calculated NEg.

In addition to the growth performance data, serum metabolites of heifers fed with silages from different processing types showed some minor differences, where SHRD-fed heifers tended to have greater ALB and lesser NEFA, CHOL, and TB concentrations. Greater growth, determined by increased ADG and KR, was associated with increased TB. Greater feed efficiency, determined by lower RFI values, was associated with increased NEFA and decreased ALB, and tended to be associated with increased CRE and reduced BHB. Similarly, greater energy utilization, determined by increased performance-based apparent NEg, was associated with reduced ALB, TP, and GGT, and increased TB, while tended to be associated with reduced BHB. Relating the differences in growth performance variables and serum metabolites and their correlations, the most striking metabolite is TB and, to a lesser extent, ALB and NEFA. Among all serum metabolites, TB was the most clearly associated with the observed differences in growth performance between treatments. Bilirubin levels were 17% greater for the CONV-fed heifers, and this metabolite was associated with increased ADG, GF, KR, and apparent NEg, which were also greater for CONV-fed heifers. Conversely, the apparent NEg of the CONV diet was greater; also, CONV-fed heifers tended to have lesser serum ALB concentration, and ALB and apparent NEg tended to be negatively correlated. In a backgrounding study by Kelly et al. (2010a), more efficient heifers (lesser RFI) had increased blood NEFA and reduced blood BHB concentrations. Similarly, in another study from the same group using finishing heifers, blood BHB was correlated with reduced feed efficiency, determined by greater RFI values (Kelly et al., 2010b). In a similar attempt to correlate blood parameters with feed efficiency, Foote et al. (2016) reported that blood cortisol and fecal corticosterone concentrations were positively correlated with RFI in finishing steers and heifers. Additionally, Nascimento et al. (2015) reported that in growing Nellore cattle, blood insulin was negatively correlated with RFI, and the ratio of glucose to insulin was positively correlated with RFI. Comparably, in dairy cows, greater feed efficiency was associated with greater plasma concentrations of fatty acids and BHB (Nehme-Marinho and Santos, 2022).

A possible explanation for the observed differences in growth performance between treatments can be related to altered energy availability, and the antagonism between fat and carbohydrate metabolism at the muscular level, a mechanism known as the “Randle cycle” (Hocquett et al., 1998; Hocquett, 2010). In muscle cells, stimulation of carbohydrate catabolism by insulin reduces the lipolytic activity of the cells (Hocquett et al., 1998; Hocquett, 2010). In the context of our study, the greater NEFA observed in the CONV-fed group, may be related to decreased oxidation of this substrate, which can be the result of enhanced volatile fatty acids availability (particularly propionate) by increased ruminal fermentation. Furthermore, the increment in blood concentrations of CHOL can be associated with the increased TB concentrations, in other words, more CHOL can be the result of increased absorption by greater biliary secretion. Fatty acids and proteins can promote biliary secretions by stimulating the hormone cholecystokinin (Klein, 2013). A change in the ruminal fermentation pattern, resulting in increased ruminal microbial growth could lead to increased protein and fatty acids (constituents of the ruminal microbes) flowing postruminally. However, with the data collected in these experiments, these hypotheses cannot be verified, and further research evaluating ruminal fermentation products is needed to fully understand these mechanisms.

Conclusion

In conclusion, when harvesting WPSS, using a kernel processor increased total tract starch digestibility and seemed to reduce diet sorting by cattle. Furthermore, using the shredlage processor resulted in greater total tract NDF digestibility when compared with a conventional kernel processor for sorghum, while reaching a similar starch digestibility. However, a difference in growth performance was observed between the 2 processors in favor of the conventionally processed silage. Differences in blood concentrations of NEFA, CHOL, and TB parallelled the difference in growth performance. Changes in ruminal kinetics and fermentation parameters resulting from different particle sizes, grain processing, and potential sorting may help explain these findings.

Glossary

Abbreviations

ADF

acid detergent fiber

ADG

average daily gain

ALB

albumin

ATTD

apparent total tract digestibility

BHB

beta-hydroxy butyrate

BUN

blood urea nitrogen

BW

body weight

CHOL

cholesterol

CP

crude protein

CRD

completely randomized design

CRE

creatinine

DM

dry matter

DMI

dry matter intake

EBF

empty body fat

EG

energy of gain

EM

energy of maintenance

EU

experimental unit

EqBW

equivalent body weight

G:F

gain-to-feed ratio

GGT

gamma-glutamyl transpeptidase

iNDF

indigestible neutral detergent fiber

KP

kernel processing

KR

Kleiber ratio

MBW

metabolic body weight

NDF

neutral detergent fiber

NE

net energy

NEFA

non-esterified fatty acids

NEg

net energy of gain

NEm

net energy of maintenance

OM

organic matter

RCBD

randomized complete block design

RFI

residual feed intake

TAS

total antioxidant status

TB

total bilirubin

TG

triglycerides

TLC

theoretical length of cut

TP

total proteins

WPSS

whole-plant sorghum silage

Conflict of interest statement

The authors declare no real or perceived conflicts of interest.

Author contributions

Federico Podversich (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review & editing), Leandro Abdelhadi (Conceptualization, Data curation, Formal analysis, Methodology, Resources, Visualization, Writing—original draft, Writing—review & editing), Sergio Roskopf (Methodology, Writing—original draft, Writing—review & editing), Gleise Medeiros da Silva (Formal analysis, Methodology, Writing—review & editing), Emmanuel Angeli (Formal analysis, Methodology, Writing—review & editing), Gustavo J. Hein (Formal analysis, Writing—review & editing), Hugo Ortega (Formal analysis, Resources, Writing—review & editing), Martin Ruiz-Moreno (Data curation, Formal analysis, Methodology, Writing—review & editing), Jose Dubeux (Formal analysis, Writing—review & editing), and Nicolas DiLorenzo (Conceptualization, Methodology, Project administration, Resources, Supervision, Writing—review & editing)