Corn processing, flake density, and starch retrogradation influence ruminal solubility of starch, fiber, protein, and minerals

Abstract

Five ruminally cannulated steers (body weight = 390 ± 7.86 kg) were used in three experiments to evaluate effects of corn processing, flake density, and starch retrogradation on in situ ruminal degradation. In experiment 1, corn was left whole or processed with no screen, ground through a 6-mm screen, or ground through a 1-mm screen. In experiment 2, we produced steam-flaked corn at four densities: 309, 335, 360, and 386 g/L. These four flake densities were sifted for 20 s through a 4-mm screen to produce two particle sizes within each flake density: sifted flakes (>4 mm) and sifted fines (<4 mm). In experiment 3, sifted flakes (335 g/L) were stored for 3-d at either 23 °C (starch availability = 55%) or 55 °C to induce starch retrogradation (starch availability = 41%). All samples for each of the three experiments were weighed into nylon bags and ruminally incubated for 0-h to estimate the soluble fraction. The residue remaining was analyzed for nutrient composition. In experiment 1, whole shelled corn had lesser (P < 0.01) ruminal solubility of all nutrients measured compared with ground corn. Corn ground with a screen (6 and 1 mm) had greater (P < 0.01) ruminal solubility of all nutrients measured compared with corn ground with no screen. Corn ground through a 1-mm screen had greater (P < 0.03) ruminal solubility of DM, total starch, CP, ADF, AHF, P, Mg, K, S, Zn, Fe, and Mn compared with corn ground through a 6-mm screen. In experiment 2, increasing flake density linearly decreased (P < 0.02) the soluble fraction of DM, total starch, CP, ADF, AHF, P, K, S, and Zn of sifted flakes. The soluble DM fraction of sifted fines tended to decrease (P = 0.06) linearly with increasing flake density. Total starch, CP, NDF, and Zn soluble fractions of sifted fines were not influenced by flake density. In experiment 3, storage of sifted flakes in heat-sealed foil bags at 55 °C for 3-d decreased (P < 0.04) the soluble fractions of DM, total starch, CP, NDF, P, Mg, K, S, and Fe. With each increase in the degree of corn processing, there was an increase in the solubility of nutrients. Increasing flake density can decrease ruminal solubility of flakes; however, the soluble fraction of sifted fines is not influenced as much by changes in flake density. Inducing starch retrogradation decreases ruminal solubility of starch, nonstarch OM, and minerals.

Grain processing methods that result in changes in particle size, flake density, or starch retrogradation can affect the solubility of nutrients in the rumen differently.

Introduction

In feedlot operations, corn is the primary grain source used for finishing diets for beef cattle, and steam-flaking is the primary grain processing method (Samuelson et al., 2016). Steam-flaking has become a common practice because it improves starch availability, nutrient utilization, and the overall feeding value (Zinn et al., 2002). Additionally, grain processing method, flake density, particle size, and the degree of starch retrogradation influence the soluble fraction of steam-flaked corn.

We previously demonstrated that increasing flake density from 257 to 412 g/L decreased the soluble dry matter (DM) fraction of sifted flakes from 44.2% to 25.9% and increased the potentially degradable DM fraction from 51.6% to 68.9% (Trotta et al., 2021b). The fines produced as a by-product of the flaking procedure can make up to 15% of the total steam-flaked corn output on a wet basis, depending on flake density, roll use, and roll corrugations. The fines have a greater soluble DM fraction and lesser potentially degradable fraction compared to flakes (Trotta et al., 2021b). Starch retrogradation decreases the soluble fraction and increases the potentially degradable fraction of both flakes and fines (Trotta et al., 2021b).

Modeling digestion in ruminants is a function of the rate and extent of ruminal degradation and the rate of passage. Heterogeneity of diets, feedstuffs, and nutrients can lead to different rates of degradation and passage within the rumen, which must be accounted for when modeling. Because of the unique kinetic properties of different feed particles, digestion models in ruminants typically use a 3-fraction model containing the soluble fraction, potentially degradable fraction, and the undegraded fraction. The soluble fraction is assumed to either ferment immediately or exit the rumen at a rate equivalent to liquid outflow.

Despite large changes in the soluble and potentially degradable fractions with flake density, particle size, and retrogradation, the potential extent of ruminal DM degradation was not influenced by any of the factors mentioned (Trotta et al., 2021b). With all of this information considered, it remains unclear how flaking procedures that modify the soluble DM fraction influence solubility of individual nutrients. The objective of this experiment was to characterize how corn processing, flake density, particle size, and starch retrogradation influence the soluble fraction of starch, protein, fiber, and minerals.

Materials and Methods

Animal care and management protocols followed the recommendations of the Guide for the Care and Use of Agricultural Animals in Research and Teaching, 4th Edition.

Experiment 1

Five ruminally cannulated Holstein × Angus crossbred steers (initial body weight = 390 ± 7.86 kg) were used to determine the effects of corn processing on nutrient disappearance of the soluble fraction using the in situ nylon bag technique. The experimental design was a randomized complete block design with rep as a blocking factor. Steers were pen-fed a starter diet [53.3% DM, 35.8% total starch, 21.9% acid detergent fiber (ADF), 13.3% crude protein (CP)] ad libitum (Trotta et al., 2021b). Bulk samples of whole-shelled corn (~5 kg) were collected from a storage bin and either left whole or processed by grinding with no screen, grinding through a 6-mm screen, or grinding through a 1-mm screen (SM-100 Cutting Mill; Retsch GmbH, Haan, Germany).

Samples of each corn treatment were weighed (92.7 ± 6.48 g) into nylon bags (10 cm × 20 cm; 50 μm pore size; R1020 Forage Bag; ANKOM Technology, Macedon, NY). One nylon bag for each treatment was placed into a wash bag (25 cm × 31 cm; k2107; HomeAide Delicate Wash Bag) for each steer. The wash bag was immersed in the ventral rumen and swirled for approximately 10 seconds to ensure complete exposure to ruminal fluid. This process corresponds to the 0-h measurement of an in situ digestibility experiment which is used to calculate the soluble fraction. Nylon bags were quickly immersed in cold water and placed into an ice-water bath to stop fermentation. Nylon bags were rinsed 5 times in a washing machine with 1-min rinse and 2-min spin cycles (Coblentz et al., 1997). The bags were then placed in a 55 ºC forced-air oven for 72-h to determine in situ nutrient disappearance. The experiment was replicated twice across days as recommended by Vanzant et al. (1998).

Experiment 2

The same 5 steers were used to evaluate the effects of flake density on soluble fraction disappearance of sifted flakes and sifted fines. The experimental design was a randomized complete block design with rep as a blocking factor. The same source of whole-shelled corn used in experiment 1 was used to produce steam-flaked corn treatments in experiment 2. The rolls of a steam flaker (61 cm × 142 cm roll size) were adjusted to produce steam-flaked corn corresponding to 309 g/L (24 lb/bu), 335 g/L (26 lb/bu), 360 g/L (28 lb/bu), and 386 g/L (30 lb/bu) flake densities. Bulk samples of steam-flaked corn were collected using a metal bucket (15.2 cm diameter × 6.35 cm height) on a handle that was extended under the rolls to collect fresh flakes. The collected bulk samples were transferred to standard testing sieves (30.5 cm diameter, 4-mm screen; Advantech Manufacturing, Inc., New Berlin, WI) fitted above a collection pan. Steam-flaked corn was sieved gently for 20 s. The large particles retained in the top portion (>4-mm; sifted flakes) and the small particles that passed-through the sifting sieve to the bottom collection pan (<4-mm; sifted fines) were air equilibrated for 24-h and then stored in plastic bags. Each treatment (79.6 ± 6.74 g) was weighed into nylon bags and one bag for each treatment was placed into a wash bag for each steer, as described in experiment 1. Ruminal incubation, rinsing, and drying procedures were conducted as previously described in experiment 1. The experiment was replicated twice across days (Vanzant et al., 1998).

Experiment 3

The same 5 ruminally cannulated steers were used to determine the effects of starch retrogradation on nutrient disappearance of the soluble fraction using the in situ nylon bag technique. The experimental design was a randomized complete block design with rep as a blocking factor. The same source of whole-shelled corn used in experiment 1 was used to produce steam-flaked corn treatments in experiment 3. The rolls of a steam flaker (61 cm × 142 cm roll size) were adjusted to produce steam-flaked corn corresponding to 335 g/L (26 lb/bu). Bulk samples were not sifted (flakes + fines; >4-mm and <4-mm) and then placed into foil bags (Mylar Bags; IMPAK Corporation, Los Angeles, CA) and heat-sealed to prevent moisture loss. Foil bags were stored at either 23 or 55 ºC in a forced-air drying oven for 3-d. The oven temperature setting (55 °C) was chosen because it has been reported that the temperature of the core of piled steam-flaked corn can remain > 55 °C for more than 17-h (Sindt, 2004; Drouillard and Reinhardt, 2006). We previously demonstrated that exposure of steam-flaked corn to 55 °C for 3-d decreased starch availability from 69.4% to 44.6% (Trotta et al., 2021a) and in situ ruminal DM degradability (Trotta et al., 2021b) due to starch retrogradation. Flakes + fines (85.6 ± 6.32 g) were weighed into nylon bags and one bag for each treatment was placed into a wash bag for each steer, as described in experiment 1. Ruminal incubation, rinsing, and drying procedures were conducted as previously described in experiment 1. The experiment was replicated twice across days (Vanzant et al., 1998).

Nutrient analysis

Dried feed samples and feed residues from the ruminal nylon bag incubations were ground to pass through a 1-mm screen using a cyclone sample mill (Model 3010-014; UDY Corporation, Fort Collins, CO). Samples were analyzed for nutrient composition including acid-hydrolyzed fat (AHF; AOAC, 2012; 922.06), available starch and total starch (Trotta et al., 2021a), CP (AOAC, 2012; 990.03), and NDF and ADF (Van Soest et al., 1991). Starch availability was calculated by dividing available starch by total starch and multiplying by 100. Samples were prepared for mineral analyses by pre-digesting samples in HNO₃ followed by H₂O₂ digestion with HCl (Huang and Schulte, 1985; Mills and Jones, 1997) and analyzed using inductively coupled plasma emission spectroscopy (AOAC, 2012; methods 990.08 and 968.08). Nutrient concentrations for each corn treatment are presented in Table 1. The nutrient concentration in each feed sample was used to calculate the initial amount weighed into each bag. The nutrient concentration of each feed residue was used to calculate the final amount of each nutrient remaining. Nutrient disappearance (%) was calculated as the difference between initial nutrient amount (g) and final nutrient amount (g) divided by the initial nutrient amount and multiplied by 100. Nutrient disappearance during the 0-h incubation represents the soluble fraction of feeds that is estimated using first-order digestion kinetic models.

Table 1.

Nutrient composition of corn treatments used in experiments 1, 2, and 3

Item	Nutrient1
	DM %	Total starch % DM	SA %	CP % DM	NDF % DM	ADF % DM	AHF % DM	P % DM	Mg % DM	K % DM	S % DM	Zn ppm	Fe ppm	Mn ppm
Experiment 1 2
WSC	94.1	74.4	9	8.6	10.9	3.7	4.5	0.27	0.10	0.33	0.10	17	20	6
NS	92.2	73.9	9	8.2	10.9	2.4	4.8	0.26	0.11	0.33	0.10	16	20	6
6S	91.8	73.2	9	8.2	10.9	3.6	4.6	0.25	0.10	0.32	0.10	17	22	6
1S	92.7	73.3	8	8.6	10.0	3.8	4.9	0.25	0.10	0.32	0.10	16	20	5
Experiment 2
Flakes
309 g/L	87.1	78.8	77	8.1	6.7	3.4	3.0	0.14	0.05	0.21	0.09	9	16	3
335 g/L	86.4	79.1	68	8.2	7.8	3.3	3.7	0.19	0.06	0.25	0.10	12	19	4
360 g/L	85.9	80.0	60	8.3	8.2	4.1	2.9	0.18	0.06	0.25	0.09	11	20	4
386 g/L	85.7	79.4	51	7.9	8.1	4.6	3.2	0.20	0.08	0.26	0.10	13	20	5
Fines
309 g/L	84.1	39.8	47	13.0	24.3	11.5	15.5	1.18	0.50	1.26	0.14	69	71	23
335 g/L	86.7	37.7	48	13.9	23.3	12.5	18.3	1.26	0.53	1.20	0.15	78	103	23
360 g/L	84.0	40.5	49	14.1	24.2	13.3	18.2	1.20	0.50	1.19	0.14	72	180	23
386 g/L	83.1	41.3	48	13.0	22.4	12.9	17.2	1.32	0.54	1.33	0.16	76	86	24
Experiment 3 3
23 °C	90.0	73.6	55	10.1	10.3	4.8	7.6	0.40	0.17	0.49	0.1	23	33	8
55 °C	89.8	72.3	41	9.9	10.5	5.6	6.9	0.35	0.14	0.43	0.1	23	31	8

Abbreviations: DM = dry matter; SA = starch availability; CP = crude protein; NDF = neutral detergent fiber; ADF = acid detergent fiber; AHF = acid-hydrolyzed fat.

Abbreviations: WSC = whole shelled corn; NS = ground corn (no screen); 6S = ground corn (6-mm screen); 1S = ground corn (1-mm screen).

Samples were stored for 3-d in heat-sealed foil bags at either 23 or 55 °C to induce starch retrogradation.

Statistical analysis

All variables were checked for normality using the UNIVARIATE procedure of SAS (version 9.4; SAS Institute Inc., Cary, NC). For experiment 1, the data were analyzed using the GLM procedure of SAS as a randomized complete block design for effects of corn processing and rep (blocking factor). To determine the effect of grinding, a contrast statement was analyzed for whole-shelled corn vs. no screen, 6-mm screen, and 1-mm screen. To determine the effect of grinding with a screen, a contrast statement was analyzed for no screen vs. 6-mm screen and 1-mm screen. To determine the effect of screen size, a contrast statement was analyzed for 6-mm screen vs. 1-mm screen.

For experiment 2, sifted flakes and sifted fines data were analyzed separately using the GLM procedure of SAS as a randomized complete block design for effects of flake density and rep. Linear and quadratic contrast statements were generated.

For experiment 3, data were analyzed as a randomized complete block design using the GLM procedure of SAS for fixed effects of rep and starch retrogradation. Least squares means and their standard errors were computed for each fixed effect included in the models. Pairwise differences of least squares means were separated using the Tukey–Kramer adjustment, protected by a significant F-test. Results were considered significant if P ≤ 0.05. Tendencies were declared when 0.05 < P ≤ 0.10.

Results

Experiment 1

Whole shelled corn had lesser (P < 0.01) ruminal solubility of all nutrients measured compared with ground corn (Table 2). Corn ground with a screen (6 and 1 mm) had greater (P < 0.01) ruminal solubility of all nutrients measured compared with corn ground with no screen. Corn ground through a 1-mm screen had greater (P ≤ 0.05) ruminal solubility of DM, total starch, CP, ADF, AHF, P, Mg, K, S, Zn, Fe, and Mn compared with corn ground through a 6-mm screen. The size of the screen used to grind corn (6-mm vs. 1-mm) did not influence (P = 0.38) the soluble fraction of NDF.

Table 2.

experiment 1: effects of corn processing on the soluble fraction (%) of dry matter, starch, crude protein, fat, fiber, and minerals

Nutrient disappearance, %	Whole shelled corn	Ground corn			SEM1	Contrast P-value2
		No screen	6-mm screen	1-mm screen		Grinding	Screen	Screen size
Dry matter	0.00	7.75	22.0	38.1	0.559	<0.01	<0.01	<0.01
Total starch	1.04	9.20	22.5	38.3	0.813	<0.01	<0.01	<0.01
Crude protein	-0.384	-1.12	12.1	30.6	0.765	<0.01	<0.01	<0.01
Neutral detergent fiber	3.60	12.4	34.0	36.9	2.26	<0.01	<0.01	0.38
Acid detergent fiber	3.17	5.63	28.9	41.2	2.85	<0.01	<0.01	<0.01
Acid-hydrolyzed fat	12.9	22.5	36.1	53.7	6.2	<0.01	<0.01	0.05
P	-0.792	2.86	27.3	63.4	1.39	<0.01	<0.01	<0.01
Mg	-6.00	7.85	21.2	56.7	1.62	<0.01	<0.01	<0.01
K	6.63	24.3	54.4	85.5	0.904	<0.01	<0.01	<0.01
S	-0.0322	5.95	18.2	30.7	1.14	<0.01	<0.01	<0.01
Zn	4.10	2.07	30.7	60.2	1.63	<0.01	<0.01	<0.01
Fe	7.99	8.40	43.6	63.8	6.38	<0.01	<0.01	0.03
Mn	8.31	14.0	35.0	56.8	2.09	<0.01	<0.01	<0.01

SEM = standard error of the mean (n = 10).

Grinding = whole shelled corn versus others; screen = no screen versus 6-mm screen and 1-mm screen; Screen size = 6-mm screen versus 1-mm screen.

Experiment 2

The mass of sifted flakes after sieving increased linearly (P < 0.01) with increasing flake density (Table 3). Sifted fines mass responded quadratically (P = 0.01) where increasing flake density from 309 to 360 g/L decreased sifted fines mass and then increased from 360 to 386 g/L. The proportion of sifted fines responded quadratically (P < 0.01) where it decreased from 309 to 335 g/L and then plateaued.

Table 3.

experiment 2: effects of flake density on sifted flakes and sifted fines mass and the proportion of sifted fines¹

Item	Flake density, g/L				SEM2	P-value
	309	335	360	386		Linear	Quadratic
Sifted flakes mass, g	535	561	590	617	19.3	<0.01	0.99
Sifted fines mass, g	67.0	60.9	57.1	64.5	2.48	0.33	0.01
Sifted fines, % of sample	11.2	9.79	8.85	9.48	0.330	<0.01	<0.01

Sifted flakes = particles retained in the top-portion of the sieve (>4-mm) after sifting steam-flaked corn samples for 20 s; sifted fines = particles retained in the bottom-portion of the sieve (<4-mm) after sifting steam-flaked corn samples for 20 s.

SEM = standard error of the mean (n = 10).

With sifted flakes, increasing flake density linearly decreased (P ≤ 0.02) the soluble fraction of DM, total starch, CP, ADF, AHF, P, K, S, and Zn (Table 4). The NDF soluble fraction increased from 309 to 335 g/L, remained similar to 360 g/L, and then decreased from 360 to 386 g/L (P = 0.02). The soluble Mg and Mn fractions decreased (P = 0.04) from 309 to 360 g/L and then increased from 360 to 386 g/L.

Table 4.

experiment 2: effects of flake density on in situ ruminal solubility (%) of dry matter, starch, crude protein, fat, fiber, and minerals of sifted flakes¹

Nutrient disappearance, %	Flake density, g/L				SEM2	P-value
	309	335	360	386		Linear	Quadratic
Dry matter	48.6	44.4	41.9	35.6	1.23	<0.01	0.39
Total starch	53.0	48.5	48.8	42.9	1.34	<0.01	0.61
Crude protein	27.0	21.5	20.0	12.2	1.54	<0.01	0.45
Neutral detergent fiber	-4.91	5.38	6.19	-0.918	3.57	0.43	0.02
Acid detergent fiber	26.7	17.8	31.5	36.9	2.26	<0.01	<0.01
Acid-hydrolyzed fat	46.6	50.1	25.8	26.5	3.87	<0.01	0.72
P	60.4	60.3	49.3	50.6	2.10	<0.01	0.75
Mg	41.5	38.0	23.8	39.7	3.36	0.20	<0.01
K	90.2	88.7	85.7	83.6	0.712	<0.01	0.73
S	25.9	28.4	16.7	19.8	1.68	<0.01	0.88
Zn	35.6	35.3	19.3	29.4	3.13	0.02	0.11
Fe	33.4	23.7	31.3	33.5	5.33	0.74	0.27
Mn	36.7	36.1	27.2	39.4	3.03	0.95	0.04

Particles retained in the top-portion of the sieve (>4-mm) after sifting steam-flaked corn samples for 20 s.

SEM = standard error of the mean (n = 10).

With sifted fines, increasing flake density did not influence the soluble fractions of total starch, CP, NDF, or Zn (Table 5). The soluble DM fraction of sifted fines tended to decrease (P = 0.06) linearly with increasing flake density. Increasing flake density resulted in a linear increase (P < 0.01) in the soluble ADF fraction but the soluble AHF fraction decreased (P < 0.01). Quadratic responses for P, Mg, K, S, and Mn of the sifted fines decreased (P < 0.01) and then increased as flake density increased to 386 g/L. The soluble fraction of Fe increased (P = 0.02) from 309 to 360 g/L and then decreased from 360 to 386 g/L.

Table 5.

experiment 2: effects of flake density on in situ ruminal solubility (%) of dry matter, starch, crude protein, fat, fiber, and minerals of sifted fines¹

Nutrient disappearance, %	Flake density, g/L				SEM2	P-value
	309	335	360	386		Linear	Quadratic
Dry matter	43.6	42.0	40.3	41.6	0.893	0.06	0.13
Total starch	55.9	53.3	57.4	55.0	1.42	0.82	0.93
Crude protein	24.8	26.9	24.5	24.4	1.13	0.50	0.35
Neutral detergent fiber	33.3	29.0	33.6	36.0	3.21	0.39	0.30
Acid detergent fiber	32.8	26.7	40.3	41.3	2.84	<0.01	0.21
Acid-hydrolyzed fat	34.7	42.3	35.2	30.0	1.67	<0.01	<0.01
P	42.8	36.4	33.7	40.1	1.28	0.11	<0.01
Mg	32.5	24.8	20.7	28.4	1.56	0.02	<0.01
K	78.5	72.7	72.8	75.9	0.635	0.01	<0.01
S	28.9	29.3	25.1	30.6	1.04	0.04	<0.01
Zn	18.6	15.6	12.6	17.8	2.49	0.63	0.10
Fe	9.27	22.7	49.0	23.7	5.39	<0.01	0.01
Mn	19.5	10.6	10.7	18.5	3.35	0.84	0.02

Particles retained in the bottom-portion of the sieve (<4-mm) after sifting steam-flaked corn samples for 20 s.

SEM = standard error of the mean (n = 10).

Experiment 3

Inducing starch retrogradation by storing flakes + fines in heat-sealed foil bags at 55 °C for 3-d decreased (P < 0.04) the soluble fractions of DM, total starch, CP, NDF, P, Mg, K, S, and Fe (Table 6). Storage temperature did not influence the soluble fraction of ADF, AHF, Zn, or Mn.

Table 6.

experiment 3: effects of starch retrogradation on in situ ruminal solubility (%) of dry matter, starch, crude protein, fat, fiber, and minerals of flakes + fines (335 g/L)¹

Nutrient disappearance, %	3-d storage temperature		SEM	P-value
	23 °C	55 °C
Dry matter	39.7	28.7	0.663	<0.01
Total starch	40.8	28.7	0.974	<0.01
Crude protein	31.1	23.2	0.74	<0.01
Neutral detergent fiber	24.2	18.1	1.95	0.04
Acid detergent fiber	38.9	43.5	1.88	0.11
Acid-hydrolyzed fat	67.2	64.6	2.48	0.46
P	73.9	64.8	2.04	<0.01
Mg	69.6	56.7	2.51	<0.01
K	92.3	88.4	0.654	<0.01
S	28.4	20.2	0.937	<0.01
Zn	56.1	52.3	2.86	0.36
Fe	49.3	40.9	1.76	<0.01
Mn	54.9	47.4	3.15	0.11

Steam-flaked corn samples obtained from directly under the rolls that were not sifted.

SEM = standard error of the mean (n = 10).

Discussion

Corn processing

The treatments used in experiment 1 were designed to produce corn fractions that represent common feeds fed to finishing cattle. Grinding with no screen produced a fraction that resembled coarsely cracked corn. Grinding with a 6-mm screen produced a fraction that resembled coarsely ground corn and grinding with a 1-mm screen produced a fraction that resembled finely ground corn. A possible limitation with the in situ ruminal digestibility procedure was that mastication was avoided by placing samples directly into the rumen. It is plausible to assume that feeds with greater particle sizes may be masticated to a greater extent which may alter in vivo degradability characteristics compared with the use of the in situ procedure.

Processing corn by grinding, grinding with the use of a screen, and decreasing screen size all influenced the soluble fractions of different nutrients of corn. The soluble DM and starch fractions responded similarly because starch is the principal component of corn (total starch ≥73.2% of DM) and thus, relationships between ruminal DM and starch digestibility are similar for high-starch feeds (Offner et al., 2003). Interestingly, with each increase in the degree of processing, the soluble starch fraction increased by 8.2, 13.3, and 15.8 percentage units relative to whole-shelled corn. However, with each increase in the degree of processing the relative increase was at a decreasing efficiency. With each increase in the degree of processing, the starch soluble fraction increased by 8.9-fold, 2.5-fold, and 1.7-fold. These data suggest that grinding corn without a screen is the most efficient method to increase starch solubility; but, fine grinding can achieve the largest increase in ruminal starch solubility.

Many studies that have evaluated the effects of decreasing corn particle size in finishing beef cattle diets have replaced the entire corn portion of the diet with finely ground corn (Macken et al., 2006; Corona et al., 2006; Loe et al., 2006; Swanson et al., 2014). In general, previous studies did not find improvements in average daily gain or gain:feed when the entire corn portion of the diet was replaced by finely ground corn (Macken et al., 2006; Corona et al., 2006; Loe et al., 2006; Swanson et al., 2014). In the current study, the soluble DM fraction of finely ground corn was similar to the soluble DM fraction of corn steam flaked to a density of 360 g/L (Buttrey et al., 2016; Trotta et al., 2021b). Because the soluble DM fraction of finely ground corn is similar to that of corn steam-flaked to 360 g/L, future research should compare these two feeds directly to evaluate whether or not ruminal degradability characteristics are similar. If so, fine grinding corn could potentially improve nutrient utilization similar to that of steam-flaked corn when included in finishing diets at low inclusion levels and/or when included in diets with other forms of processed grains. Grinding coarseness, dietary inclusion level, interactions with other dietary components, roughage inclusion level, DM intake, and the rate of ruminal passage could all be important factors when deciphering impacts of fine ground corn inclusion in finishing cattle diets.

Flake density

As flake density increased, the proportion of sifted fines decreased, which is in agreement with previous findings by Hales et al. (2010). In the current study, increasing flake density from 309 to 386 g/L decreased the soluble DM fraction of sifted flakes from 48.6% to 35.6%. Similarly, we previously demonstrated that increasing flake density from 257 to 412 g/L decreased the soluble DM fraction of sifted flakes from 44.2% to 25.9% (Trotta et al., 2021b). In general, increasing flake density decreased the soluble fraction of most nutrients measured in the current study. As flake density decreases, the extent of gelatinization, starch availability, and ruminal disappearance increases (Trotta et al., 2021b). However, it is widely accepted that flaking to a lighter density (<300 g/L) does not result in economically favorable outcomes because of limited improvements in total-tract starch digestibility and feed efficiency, decreased mill production rate, increased mill energy consumption, and increased risk of ruminal acidosis and bloat (Reinhardt et al., 1997; Brown et al., 2000; Drouillard and Reinhardt, 2006). In contrast, flaking to a heavier density (>400 g/L) can increase mill production rate and decrease mill energy consumption but also can decrease ruminal and/or total-tract digestibility (Zinn 1990; Theurer et al., 1999). Therefore, consulting feedlot nutritionists recommended flaking corn to an intermediate bulk density of 350 g/L (Samuelson et al., 2016) to improve starch digestibility while maintaining a feasible mill production rate and low to moderate risk of digestive disturbances.

In agreement with above, Zinn et al. (2002) suggested that steam-flaking not only increases starch digestibility compared with other grain processing methods, but also increases digestibility of nonstarch OM. Moreover, Zinn et al. (1995) noted that the increase in nonstarch OM digestibility was proportional to the increase in starch digestibility. These observations were most notable when comparing steam-flaked corn digestibility with other forms of grain processing. In the current study, our data support this concept and expand our current understanding by demonstrating that decreasing flake density increases ruminal solubility of several nonstarch OM components and minerals.

Particle size: flakes versus fines

During the flaking process, starch is gelatinized and other nutrients are typically segregated into the fines. Improper sampling of steam-flaked corn can lead to under- or over-representation of specific nutrients (Corona et al., 2006; Hales et al., 2010). The flakes and fines have drastically different nutrient profiles (Hales et al., 2010; Trotta et al., 2021a) where the fines contain greater proportions of fiber, protein, fat, and minerals. This is most likely because different nutrients are stored in different anatomical locations of the corn kernel, which can be segregated during the flaking procedure. The endosperm contains large proportions of starch which is gelatinized to produce flakes and the minerals from the germ segregate into the fines. Differences in nutrient profiles and particle size between flakes and fines influence their degradability characteristics in the rumen (Trotta et al., 2021b). Although not directly compared in the current study, sifted fines had a greater soluble fraction for most macronutrients while the sifted flakes had a greater soluble fraction for most minerals. However, because sifted fines contained a greater proportion of most minerals, the amount that disappeared (g) was greater compared with sifted flakes. Ruminal solubility of sifted flakes was dramatically affected by changes in flake density, whereas changes in solubility were much smaller with sifted fines. These data show that it is critical to obtain similar fractions when sampling steam-flaked corn for laboratory analysis. If incorrect sampling of proportions occur, the wrong interpretation/conclusion will be drawn relative to flaker quality control and/or laboratory analytical error.

Starch retrogradation

In cattle feeding operations, starch retrogradation of steam-flaked corn could occur through prolonged heat exposure in the core of corn piled for storage (McAllister et al., 2006), storage in grain bins (Ward and Galyean, 1999), or by method of moving steam-flaked corn from under the rolls into storage bins (McMeniman and Galyean, 2007). Storage of steam-flaked corn at 55 °C for 3-d in heat-sealed foil bags decreased starch availability from 69.4% to 44.6% due to starch retrogradation (Trotta et al., 2021a). Furthermore, inducing starch retrogradation of flakes + fines decreased in situ ruminal DM degradability from 65.7% to 52.0% (Trotta et al., 2021b). However, starch retrogradation decreases the soluble fraction and rate of ruminal degradation without influencing the potential extent of degradation (Trotta et al., 2021b). If starch retrogradation decreases the soluble fraction or the rate of ruminal degradation while maintaining total-tract digestibility, it is plausible that the site of digestion could be shifted from the rumen to the small intestine to improve the efficiency of total-tract starch digestion in feedlot cattle.

Proximate analyses from our previous study (Trotta et al., 2021b) and the current study show that the nutrient composition of steam-flaked corn stored at 23 or 55 °C is similar, with the exception of starch availability. In the current study, inducing starch retrogradation decreased the soluble fraction of DM, starch, CP, ADF, and minerals. These data suggest that starch retrogradation could decrease ruminal digestibility of nonstarch components. It is important to note that the total starch content of flakes + fines was ≥72.3% and thus, reducing the digestibility of nonstarch components in retrograded steam-flaked corn may not be of practical concern because these nutrients do not contribute to a substantial portion of the animal’s intake of CP, fiber, or mineral. More information is needed to understand how starch retrogradation of steam-flaked corn influences the site and extent of starch and nonstarch OM digestibility, growth, and feed efficiency.

Conclusions

Corn processing, flake density, and starch retrogradation influenced the solubility of nutrients in the rumen. Processing corn by grinding, grinding with a screen, and decreasing screen size can increase the soluble fraction of starch, nonstarch OM, and minerals. Increasing flake density linearly decreases the solubility of DM and starch of sifted flakes; however, the soluble DM and starch fractions of sifted fines were not influenced by flake density. The soluble fraction of macronutrients was greater for sifted fines compared with sifted flakes. Storage of steam-flaked corn in heat-sealed foil bags for 3-d at 55 °C to induce starch retrogradation decreased the soluble fraction of most nutrients. These data will be useful for understanding the various factors influencing the soluble fraction of processed corn when modeling.

Glossary

Abbreviations

ADF

acid detergent fiber

AHF

acid-hydrolyzed fat

CP

crude protein

DM

dry matter

NDF

neutral detergent fiber

lb/bu

pounds per bushel

OM

organic matter

Conflict of interest

The authors declare no conflict of interest.