Corn grain is a major source of starch in feedlot diets (Samuelson et al., 2016; Pinto and Millen, 2019). In a survey of Brazilian nutritionists, corn was the grain most cited (i.e., 97%) as an energy source in finishing diets, with all corn grain being considered flint (Silvestre and Millen, 2021). However, the endosperm protein matrix that surrounds the starch granules represents a barrier to the digestion of starch (Owens et al., 1986). Ensiling high-moisture (HMC) or rehydrated corn gain (RCGS) is an efficient strategy to improve starch digestibility and feed efficiency (FE) in feedlot cattle (Owens and Basalan, 2013), especially for flint corn grain (Jacovaci et al., 2021).

The optimal efficiency of feedlot diets containing corn grain silage is attained when starch digestion is maximized without causing ruminal acidosis (NASEM, 2016; Gouvêa et al., 2019). Digestibility of RCGS is primarily affected by moisture content, storage period, and particle size, which taken together have additive effects (Gomes et al., 2020). A moisture content of 350 g/kg and a storage period of at least two months have been recommended to improve the preservation and digestion of flint RCGS (da Silva et al., 2019; Fernandes et al., 2021; Gomes et al., 2020); however, the ideal particle size for RCGS from flint corn grain remains debatable (Gouvêa et al., 2019). In finishing high-concentrate diets, large grain particles might decrease starch digestion and FE, whereas fine particles may depress dry matter intake (DMI) and reduce average daily gain (ADG), due to the high load of rumen degraded starch. Summarizing 15 finishing cattle trials, Owens and Thornton (1976) concluded that DMI decreased linearly by about 0.4% for each 10 g/kg increase in moisture content of HMC from 230 to 300 g/kg moisture. Based on DMI depression, the authors stated that acidosis should be of concern with finely ground HMC.

Currently in Brazil, many beef farmers are ensiling rehydrated corn grain after rolling, due to the higher milling yield and reduced energy costs of rolling instead of grinding, and the availability of bagging machines mounted with roller mills. Meanwhile, the geometric mean particle size (GMPS) of rolled RCGS is often larger than those ground RCGS. Therefore, we hypothesized that processing RCGS more intensively would improve the total-tract digestibility and, in turn, the performance of feedlot cattle. The objective of this work was to evaluate the effect of processing method of flint RCGS on the total-tract digestibility and performance of finishing beef cattle.

All animal care and handling procedures were approved by the Ethics Committee on Animal Use (protocol number 5843150421). The experiment was carried out in Maringá, PR, Brazil (23^o 25' S; 51^o 57' W, 550 m elevation).

Nutrient concentration was similar between ground and rolled RCGS, but GMPS was 43% greater for the RCGS processed in the roller mill (Table 1). Experimental TMR had similar composition of nutrients, whereas particle size distribution on the PSPS indicated a slightly coarser ration for RRC (Table 2).

Dietary treatments had no effect (P>0.10) on initial SBW, final SBW, DMI, ADG, FE, HCW, dressing, ribeye area, marbling score, back fat thickness, biceps femoris fat thickness, and liver abscess score (Table 3). However, the GRC induced greater DMI variation than the RRC (P≤0.05). The GRC also resulted in lower meal length and first meal duration than the RRC (P≤0.05; Table 4). Chewing (min/d) tended to be lower for GRC than RRC (P<0.10). Eating, ruminating, and chewing time, meal frequency, meal size, meal interval, intake rate, particle sorting index, and diet NDF sorting were similar between treatments (P>0.10).

Table 3

Item	Treatment1		SEM	P-value
	GRC	RRC
Performance
Initial shrunk body weight (kg)	412	413	6.7	0.848
Final shrunk body weight (kg)	543	544	10.5	0.918
Dry matter intake (kg/d)	11.2	11.2	0.33	0.869
Dry matter intake variation (g/kg DM)	87.2	68.9	6.26	0.046
Average daily gain (kg/d)	1.92	1.92	0.086	0.966
Feed efficiency	0.170	0.171	0.005	0.792
Carcass traits
Hot carcass weight (kg)	301	306	6.1	0.584
Carcass dressing	0.555	0.562	0.0049	0.331
Ribeye area (cm2)	84.1	85.2	1.93	0.699
Marbling score (0–10)	3.60	3.46	0.105	0.348
Back fat thickness (mm)	5.99	6.12	0.304	0.763
Biceps femoris fat thickness (mm)	7.96	8.20	0.353	0.639
Liver abscess frequency (n/19)2	0/19	1/19	-	-

DM - dry matter; NDF - neutral detergent fiber; SEM - standard error of the mean.

Table 4

Item	Treatment1		SEM	P-value
	GRC	RRC
Feeding behavior
Eating (min/d)	167	186	9.1	0.138
Ruminating (min/d)	370	394	13.5	0.231
Chewing (min/d)	540	580	14.6	0.063
Chewing (min/kg DM)	49.6	52.2	1.88	0.333
Meal frequency (meals/d)	12.4	12.3	0.54	0.799
First meal duration (min)	29.9	41.4	3.08	0.012
Meal length (min/meal)	13.1	15.8	0.73	0.014
Meal size (g DM/meal)	922	987	57.1	0.428
Meal interval (min)	108	109	13.2	0.934
Intake rate (g DM/min)	68.8	64.2	3.99	0.425
Particle sorting index
8-19 mm	1.04	1.02	0.008	0.314
4-8 mm	1.01	1.01	0.002	0.281
Pan	0.996	0.995	0.001	0.478
Diet NDF sorting index	1.01	1.01	0.002	0.760

DM - dry matter; NDF - neutral detergent fiber; SEM - standard error of the mean.

Total-tract apparent digestibility of DM, OM, CP, EE, and starch, and fecal pH were higher (P≤0.05), while fecal starch was lower (P≤0.05) for GRC than RRC (Table 5). Apparent digestibility of NDF, fecal score, and diet energy content based on animal performance were similar between treatments (P>0.10).

Table 5

Item	Treatment1		SEM	P-value
	GRC	RRC
Apparent total-tract digestibility
Dry matter (DM)	0.746	0.728	0.046	0.009
Organic matter	0.763	0.743	0.045	0.003
Neutral detergent fiber	0.565	0.569	0.011	0.811
Crude protein	0.725	0.708	0.054	0.038
Ether extract	0.801	0.759	0.014	0.043
Starch	0.956	0.936	0.003	<0.001
Fecal traits
Fecal score (1–5)	3.83	3.92	0.076	0.450
Fecal pH	6.20	5.93	0.060	0.003
Fecal starch (g/kg DM)	72.2	93.2	3.36	0.001
Diet energy (MJ/kg DM)2
Net energy for maintenance	8.23	8.23	0.113	0.988
Net energy for gain	5.50	5.50	0.099	1.000
Metabolizable energy	12.3	12.3	0.13	0.990
ME intake (MJ/d)	139	137	4.1	0.723

SEM - standard error of the mean.

Dietary treatments had no effect (P>0.10) on ruminitis score, number of papillae/cm² of rumen wall, papillae surface, rumen surface, epithelium thickness, keratinized layer thickness, and nonkeratinized layer thickness, but the GRC induced lower (P≤0.05) papillae height, and higher (P≤0.05) mean papillae width than the RRC (Table 6). Plasma glucose and serum D-lactate were similar (P>0.10) between treatments, whereas plasma urea concentration tended to be lower for GRC than RRC (P<0.10).

Table 6

Item	Treatment1		SEM	P-value
	GRC	RRC
Rumen morphology
Macroscopy
Ruminitis score (0–10)	0.368	0.500	0.114	0.433 (χ2)
Number of papillae (n/cm2)	38.6	38.3	2.63	0.939
Mean papillae surface area (cm2)	0.402	0.396	0.026	0.865
Mean papillae height (cm)	1.06	1.18	0.420	0.042
Mean papillae width (cm)	0.376	0.331	0.014	0.034
Rumen surface area (cm2/cm2 of wall)	15.9	15.8	1.09	0.985
Microscopy
Epithelium thickness (μm)	104	105	3.2	0.758
Keratinized layer thickness (μm)	20.7	20.2	0.84	0.700
Nonkeratinized layer thickness (μm)	83.1	84.9	3.32	0.691
Blood metabolites
Plasma glucose (mg/dL)	77.9	79.9	3.21	0.668
Plasma urea (mg/dL)	26.9	30.0	1.25	0.086
Serum D-lactate (mmol/L)	1.20	1.30	0.165	0.685

SEM - standard error of the mean.

Early research has examined the effects of particle size of HMC (prepared with US dent corn) on cattle performance (Secrist et al., 1995; Secrist et al., 1996; Macken et al., 2006; Coulson et al., 2021), but this present study is the first to examine the effects of grinding or rolling dried flint corn to prepare RCGS for finishing feedlot cattle. As expected, RCGS that was rolled before ensiling had greater particle size than RCGS that was ground in a hammer mill with a uniform 8-mm screen. The GMPS of the ground RCGS (1.52 mm) was at the same magnitude of the mean value observed for dry-ground corn in US feedlots (1.82 mm), but our rolled RCGS (2.18 mm) was noticeably finer than the mean dry-rolled corn surveyed in US feedlots (4.53 mm; Schwandt et al., 2015). Approximately 713 g/kg of the rolled RCGS particles were retained above the 2-mm screen but below the 8-mm screen (whole kernel), whereas only 354 g/kg of ground RCGS particles were retained above a 2-mm screen. Such differences in GMPS and particle distribution led to dissimilar responses in feeding behavior, DMI variation, total-tract digestibility, rumen papillae morphology, and plasma urea concentration. However, cattle performance remained unchanged by milling method.

In a previous study, we demonstrated that grinding, rather than rolling, flint RCGS increased in situ ruminal DM degradation by approximately 15% (Gomes et al., 2020). In the current trial, in vivo total tract-digestibility of starch and DM were approximately 20 g/kg greater for GRC than RRC. Whether increased surface exposure during fermentation of the more finely processed grain would complement the particle size effect on digestibility is uncertain. In contrast, Owens et al. (2016) reported in a data analysis that the total-tract starch digestion was weakly associated with ruminal starch digestion in HMC diets for feedlot cattle, but their dataset for HMC was small (n = 6). In our study, a greater ruminal degradability of starch in GRC is further supported by alterations in DMI variation, feeding behavior, plasma urea concentration, and rumen papillae morphology.

Total-tract starch digestibility values observed in the current study were lower than that found by Jacovaci et al. (2021), which reported 0.99 of starch digestibility for cattle fed diets containing ensiled corn grain. Differences in total-tract starch digestibility among studies may be due to diet composition, grain vitreousness, moisture, length of storage, grain particle size, and feed intake level, which ultimately affects passage rate. In our trial, DMI was around 23.5 g/kg BW, whereas in the meta-analysis of Jacovaci et al. (2021), DMI was on average 20.4 g/kg BW in animals fed diets containing ensiled corn grain.

Although DMI was similar between treatments, feeding behavior was altered by milling method in our study. Likewise, cattle fed GRC had greater DMI variation. Such responses strongly suggest a greater ruminal starch fermentability for GRC, which might have increased the portal flux of propionate, and probably altered liver signaling to brain feeding centers. In high-grain diets with a high proportion of ensiled corn as in the current trial, propionate is likely the major anaplerotic metabolite of the tricarboxylic acid cycle within meals (Allen, 2020). Propionate may stimulate greater oxidation of the existing pool of acetyl-CoA increasing hepatic energy charge that presumably could generate a satiety signal according to the hepatic oxidation theory (Allen, 2020). In fact, the GRC induced shorter meal length and first meal duration than RRC. Additionally, more available energy in the rumen often increases microbial protein synthesis and decreases ruminal ammonia and blood urea concentrations (Hristov et al., 2019). As DMI and diet CP were similar among treatments, a tendency of lower plasma urea concentration may have resulted from greater ruminal starch digestion for the GRC.

Recently, Pereira et al. (2021) found that reduced daily DMI variation was related to greater ADG and carcass weight, and lower rumenitis scores, indicating that high DMI fluctuation had adverse effects on growth performance potentially associated with an increased incidence of metabolic disorders. In our study, RCGS milling method did not change serum D-lactate, liver abscess, and ruminitis scores. However, rumen papillae of cattle fed GRC were wider and shorter, which may have reflected a greater ruminal volatile fatty acids (VFA) load and its consequences on epithelium proliferation and differentiation, and mucosal blood flow (Tamate et al., 1974; Gabel et al., 2002; Resende-Junior et al., 2006).

Despite the differences in total-tract digestion of starch, CP, EE, and total OM in favor of GRC, these changes were not translated into greater performance, and cattle fed GRC or RRC had the same FE. Secrist et al. (1995) also observed no effect of grinding or rolling HMC on DMI, ADG, FE, and carcass characteristics of feedlot cattle. A following study by the same group found tendencies of greater DMI and ADG as particle size of HMC increased (GMPS from 1.28 to 2.12 mm), without changes in FE and nutrient digestibility (Secrist et al., 1996).

More recent studies with diets containing corn grain coproducts also showed no or slight advantages of rolled over ground HMC in fishing diets. Comparing ground (GMPS 0.484 mm) or rolled (GMPS 2.90 mm) HMC in diets with 250 g/kg of wet corn gluten feed (DM basis), Macken et al. (2006) found similar DMI, ADG, FE, fecal starch, and carcass traits. On the other hand, Coulson et al. (2021) observed that HMC processed with a roller mill (GMPS 2.87 mm) improved FE in finishing cattle by 4.7% compared with HMC processed in a hammer mill (GMPS 2.25 mm) in diets containing 200 g/kg DM of wet distillers with solubles. Starch digestibility and beef cattle performance have not always been correlated in the literature (Lundy et al., 2015). Moreover, the responses to particle size of ensiled corn might diverge between feedlot cattle and lactating dairy cows (Ferraretto et al., 2013), likely due to a combination of or interactions among various factors, including shorter retention time in the rumen and large intestine with higher DMI and fiber content in dairy diets (Owens et al., 2016). Whether benefits of reducing the particle size of grain in diets for lactating cows reflects greater sensitivity to ruminal acidosis or a greater impact on DMI on growing cattle than lactating cows remains unclear.

In our experimental conditions, the RRC diet likely increased the amount of starch flowing to the intestines. Although the post-ruminal starch digestion may not have completely compensated for lower ruminal starch fermentation (which led to a lower total-tract digestion), shifting site of starch digestion from the rumen to the small intestine might have reduced the risk of subclinical acidosis and increased energetic efficiency (Owens et al., 2016). If cattle fed GRC experienced subclinical acidosis, it may not have been clearly established in this study as compared with diets containing more ensiled corn grain or less roughage. Energetic efficiency with which starch is used by cattle varies with site of starch digestion, being higher for starch digested to glucose absorbed from the small intestine (energetic efficiency ~0.97) than for starch fermented in the rumen (energetic efficiency ~0.80), due to the greater loss of energy as heat and methane in the rumen (Owens et al., 1986; McLeod et al., 2001; Huntington et al., 2006). Although Kreikemeier et al. (1991) observed that 6 to 8% of the starch infused into the abomasum escaped digestion in the small intestine and passed through the ileum as unpolymerized glucose, in our study, dietary level of starch (about 500 g/kg DM) and post-ruminal starch digestibility may have offset differences from ruminal starch degradability. Previous literature suggests that post-ruminal starch digestibility is typically high for corn that has been ensiled (Szasz et al., 2007; Owens et al., 2016). Therefore, the intestinal starch digestion of RRC might have been sufficiently high in our bulls to secure similar ADG and FE despite having a lower total-tract starch digestibility than GRC.

Particle size of processed grains is greatly influenced by the combination of the milling setting (e.g., motor power, rotation per minutes, screen, etc). Hence, corn grain particle size distribution should be monitored to avoid under or overprocessing when producing RCGS. Practical goals for processing HMC in US feedlots target less than 25 g/kg of whole kernels and less than 200 g/kg of particles <1 mm (Hicks and Lake, 2006). In our study, both rolled and ground RCGS had less than 10 g/kg whole kernels while fine particles (<1.18 mm) were 129 and 267 g/kg, which seem particle distributions of bordering under or overprocessing in rolled and ground grain, respectively. Therefore, processing corn grain more finely than the ground RCGS used in this study may increase the risk of digestive disorders in diets with high proportion of RCGS, whereas processing less intensively to obtain coarser grain than our rolled RCGS might decrease starch digestion and make rehydration difficult at ensiling (i.e., increase the chance of water percolation).

Reducing the geometric mean particle size of RCGS from 2.18 mm (rolled) to 1.52 mm (ground) failed to influence performance of finishing cattle. Hence, RCGS processing method might be selected based on other aspects of the production system, such as availability of machinery and equipment, processing yield, and diet composition. Considering that the particle size of processed grains is greatly influenced by the combination of the milling setting, the particle size distribution should be monitored to avoid under or overprocessing the grain when producing RCGS.