Animals selected for postweaning weight gain rate have similar maintenance energy requirements regardless of their residual feed intake classification

Abstract

Data of comparative slaughter were used to determine Nellore bulls’ net energy requirements classified as efficient or inefficient according to residual feed intake (RFI) and selection lines (SL). Sixty-seven Nellore bulls from the selected (SE) and control (CO) lines of the selection program for postweaning weight gain were used. The animals underwent digestibility trials before being submitted to the finishing trial. Sixteen bulls were slaughtered at the beginning of the finishing trial, and their body composition was used as the baseline for the remaining animals. For body composition determinations, whole empty body components were weighed, ground, and subsampled for chemical analyses. Initial body composition was determined with equations developed from the baseline group using shrunk body weight, fat, and protein. The low RFI (LRFI) and CO animals had a lower dry matter (DMI) and nutrient intake (P < 0.05) than high RFI (HRFI) and SE animals, without alterations in digestibility coefficients (P > 0.05). During the finishing trial, DMI remained lower for LRFI and CO animals. Growth performance was similar between RFI classes, except for empty body weight gain that tended to be higher for LRFI than HRFI (P = 0.091). The SE animals had less fat content on the empty body (P = 0.005) than CO. Carcasses tended to be leaner for LRFI than HRFI (P = 0.080) and for SE than CO (P = 0.066) animals. LRFI animals retained more energy (P = 0.049) and had lower heat production (HP; P = 0.033) than the HRFI ones. Retained energy was not influenced by SL (P = 0.165), but HP tended to be higher for SE when compared to CO (P = 0.075) animals. Net energy requirement for maintenance (NE_m) was lower for LRFI than HRFI (P = 0.009), and higher for SE than CO (P = 0.046) animals. There was an interaction tendency between RFI and SL (P = 0.063), suggesting that NE_m was lower for LRFI+CO than HRFI+CO (P = 0.006), with no differences for SE (P = 0.527) animals. The efficiency of ME utilization for maintenance (k_m) of LRFI and HRFI animals were 62.6% and 58.4%, respectively, and for SE and CO were 59.0% and 62.1%, respectively. The breeding program for postweaning weight has not improved feed efficiency over the years, with RFI classification not being a promising selection tool for SE animals. Classification based on RFI seems to be useful in animals that have not undergone the breeding program, with LRFI animals having lower energy requirements than the HRFI ones.

Introduction

The use of tools to increase livestock feed efficiency without losses in profitability has become more critical given the increasing pressure to reduce land use by livestock, environmental impacts, and food allocation for human consumption to animal feeding. Residual feed intake (RFI) is defined as the difference between observed and predicted intake based on energy requirements (Koch et al., 1963). Unlike growth rate, dry matter intake (DMI), and feed conversion rate, RFI is independent of body size, weight, production, gender, and age (Kenny et al., 2018). According to Elolimy et al. (2018), low RFI (LRFI; more efficient) is associated with a reduction of feed intake and enteric methane emission, with no effect on body size or growth rate.

Asher et al. (2018) suggested that the primary mechanism separating LRFI from high RFI (HRFI, less efficient) calves is the protein-to-fat ratio in the deposited tissues. Carcasses of LRFI animals are leaner than the ones from their contemporaneous HRFI animals (Tedeschi et al., 2006; Lancaster et al., 2009; Asher et al., 2018; Cantalapiedra-Hijar et al., 2018). Consequently, more efficient animals have considerably greater net energy requirements for maintenance (NE_m) than less efficient animals since the energy cost to maintain lean tissue is higher than that for adipose tissue (Asher et al., 2018). However, metabolic processes such as mitochondrial activity (Benedeti et al., 2018) and protein degradation (Elolimy et al., 2019) require less energy in LRFI animals, resulting in similar or lower NE_m. The energy requirements of divergent RFI animals are mostly obtained through equations because empirical studies such as comparative slaughter, heart rate, and indirect calorimetry chambers are scarce in the literature due to a high cost and labor intensity and mainly involve Bos taurus breeds (Basarab et al., 2003; Nkrumah et al., 2006; Hafla et al., 2013; Asher et al., 2018). A comparative slaughter that directly measures the metabolizable energy intake (MEI) retained energy (RE) in Nellore cattle diverging in RFI was performed by Menezes et al. (2020).

The National Academies of Sciences, Engineering, and Medicine (NASEM, 2016) recommend 77 kcal per metabolizable shrunk body weight (SBW) per day as NE_m and a 10% decrease for some Zebu breeds, excluding Nellore cattle. Valadares et al. (2016) evaluated Zebu and crossbred cattle databases to develop nutrient requirement recommendations for Brazilian cattle. The recommended NE_m was 75 kcal per metabolizable empty body weight (EBW) per day, with no difference between the feedlot and pasture-raised animals. Assuming that EBW is 89.1% of SBW, the 77 kcal/kg^0.75 SBW/d would be equivalent to 84 kcal/kg^0.75 EBW/d, which is about 10% less for Nellore cattle (i.e., 75/84), but still within the confidence interval of the measurement (Tedeschi and Fox, 2020). The accumulated productivity improvements of Nellore cattle over the years (Chud et al., 2014) combined with the use of efficient animals (LRFI) could potentially result in lower NE_m, closer to that recommended by NASEM (2016) for Zebu cattle (i.e., 77–69.3 kcal/kg^0.75 SBW/d). Brazil has the largest commercial cattle herd globally, being the second largest producer and the largest exporter of beef in the world (USDA, 2020). The Nellore breed accounts for 80% of the Brazilian herd (Oliveira and Millen, 2014); this improvement in efficiency could significantly contribute to increasing global beef production.

This study aimed to quantify the energy requirements of Nellore cattle selected for postweaning weight gain rate and classified by RFI. We hypothesized that 1) LRFI animals have lower NE_m and greater energy efficiency than HRFI animals, 2) Nellore cattle selected for postweaning weight gain rate have greater NE_m than the control line, and 3) Nellore cattle selected for postweaning weight gain rate and classified as LRFI have the same NE_m as control animals.

Material and Methods

The experiment was conducted at the Centro Avançado de Pesquisa de Bovinos de Corte of the Instituto de Zootecnia, Sertãozinho, São Paulo, Brazil. The experimental procedures followed the guidelines for animal welfare and humane slaughter (São Paulo State Law #11.977).

Animals

A genetic selection program of the Nellore breed began in 1976 in the Instituto de Zootecnia research unit based on a higher selection differential for growth. Control (CO) and selection (SE) lines were established, using bulls with a high selection differential for the postweaning weight for SE, and bulls with a selection differential for postweaning weight around zero for CO. Additional detailed description of the Breeding Program is available in Mercadante et al. (2003).

The program is in progress, and the animals (tested for feed efficiency in 2007 and slaughtered in 2008—Y08; and tested for feed efficiency in 2008 and slaughtered in 2009—Y09) evaluated in this study belong to the 26th and 27th calf crops of the program. For the feed efficiency trials, 121 animals (258 ± 22.3 days of initial age and 199 ± 33.7 kg of initial BW) were used (60 from Y08, being 41 SE and 19 CO; and 61 from Y09, being 35 SE and 26 CO). After the feed efficiency test, 67 animals (33 from Y08 and 34 from Y09) were randomly sampled for the subsequent trials, being 39 SE and 28 CO; and 34 classified as LRFI and 33 classified as HRFI. From the 67 randomly selected animals, 16 (5 LRFI+SE, 3 HRFI+SE, 3 LRFI+CO, and 5 HRFI+CO) were slaughtered at the beginning of the finishing trial, and their body composition was used as the baseline for the remaining animals. The other 51 animals were used for the digestibility and finishing trials (446 ± 20.2 days of initial age and 283 ± 34.7 kg of initial BW).

The 51 animals were assigned to two feeding groups: restricted feeding group receiving 65 g DM/kg^0.75 BW, and ad libitum feeding group. In summary, 33 animals were assigned to ad libitum feeding group (11 HRFI+SE, 9 LRFI+SE, 5 HRFI+CO, and 8 LRFI+CO) and 18 were assigned to restricted feeding group (5 HRFI+SE, 6 LRFI+SE, 4 HRFI+CO, and 3 LRFI+CO).

Feed efficiency tests

The RFI was evaluated using the same method for Y08 and Y09. The animals were housed in individual pens for 102 d, including 28 d for adaptation, followed by 74 d of feed intake and performance evaluation. The 74 d is within the recommended period to determine animal performance without losing accuracy (Archer and Bergh, 2000). The Y08 and Y09 feed efficiency tests diets were listed in Table 1. The RFI was calculated as the residual of the regression equation of observed DMI on average daily gain (ADG) and mid-test metabolic body weight (BW), as shown in Eq. (1). The animals were classified as LRFI (<0.5 standard deviations [SD] from the mean), medium RFI (± 0.5 SD from the mean), and high RFI (>0.5 SD from the mean).

Table 1.

Nutritional composition of the feed efficiency test diet (1), and digestibility and finishing trials diet (2)

	Year of evaluation1
Composition, g/kg of DM	Y08-1	Y08-2	Y09-1	Y09-2
Brachiaria brizantha hay	449	186	445	186
Corn	319	394	322	394
Cottonseed	-	123	-	123
Cottonseed meal	215	77	214	77
Citrus pulp	-	182	-	182
Urea	-	12	4.80	12
Mineral mixture2	17.0	26	14.5	26
Chemical composition, g/kg as fed
Dry matter	903	853	888	881
Crude protein	112	141	131	151
Ether extract	21.8	50.3	21.9	52.3
Ash	40.2	56.0	39.9	40.5
Neutral detergent fiber	516	300	527	324
Roughage:concentrate	45:55	19:81	45:55	19:81
Dietary energy3, Mcal/kg
Metabolizable energy	2.40	2.69	2.44	2.73
Net energy for maintenance	1.52	1.79	1.54	1.81
Net energy for gain	0.93	1.17	0.97	1.20

¹First (Y08-1) and second (Y09-1) year of feed efficiency test; first (Y08-2) and second (Y09-2) year of digestibility and finishing trials.

²Composition of the mineral mixture (per kg of product): 180 g calcium; 90 g phosphorus; 10 g magnesium; 13 g sulfur; 93 g sodium; 145 g chlorine; 17 mg selenium; 1,000 mg copper; 826 mg iron; 4,000 mg zinc; 1,500 mg manganese; 150 mg iodine; 80 mg cobalt; 900 mg fluoride.

³The dietary metabolizable energy, net energy for maintenance and gain were estimated with the equations proposed by the NASEM (2016; empirical solution).

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{D}\mathrm{M}\mathrm{I}={\beta }_0+{\beta }_{\mathrm{W}}\times \mathrm{B}{\mathrm{W}}^{0.75}+{\beta }_{\mathrm{G}}\times \mathrm{A}\mathrm{D}\mathrm{G}+\epsilon ,}\end{array}}$$ (1)

where β₀ = intercept of the regression; β_W = linear regression coefficient of metabolic BW; β_G = linear regression coefficient of ADG, and ε = residual of the equation, that is, the RFI.

Digestibility trials

The digestibility trials were performed using the same facility and methodology in Y08 and Y09. Animals were housed in individual pens with free access to water, and each trial was composed of 20 d of adaptation to the diet (Table 1), followed by 5 days of sample collection. Animals were fed twice a day (0700 and 1600 hours), and feed intake was obtained as the difference between the amount of feed offered and the orts collected after 24 h. Approximately 50 g of fecal samples were collected from each animal immediately after defecation 2, 4, and 6 h after both feedings for 5 consecutive days. Diet ingredients and individual ort samples were collected concurrently and composited by day. The ingredients and composited fecal and ort samples were dried in a forced-air oven at 55°C for 72 h. The samples were ground in a Wiley mill (Tecnal 680, Piracicaba, Brazil) to pass a 1-mm screen for ingredients and orts, and a 2-mm screen for feces. All samples were analyzed for dry matter (DM), crude protein (CP), ether extract (EE), neutral detergent fiber (NDF), ash, and indigestible neutral detergent fiber (iNDF). The contents of DM (AOAC#934.01), ash (AOAC#942.05), and EE (AOAC#954.02) were analyzed according to the AOAC (1990). Nitrogen was quantified using the Dumas method (Etheridge et al., 1998) in a LECO^® nitrogen analyzer (FP-258). The NDF was determined as recommended by Mertens (2002). The iNDF analysis was performed as described by Casali et al. (2008) by the ruminal incubation of samples for 264 h. The total amount of feces was estimated by the ratio of iNDF intake to iNDF content in the feces. Total apparent digestibility was calculated using Eq. (2) and apparent total digestible nutrients (TDN) using Eq. (3).

$${\displaystyle \begin{array}{ll}& {\displaystyle \mathrm{N}\mathrm{u}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{d}\mathrm{i}\mathrm{g}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}=\mathrm{n}\mathrm{u}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{a}\mathrm{k}\mathrm{e}}\\ & -\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{n}\mathrm{u}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{u}\mathrm{t}\mathrm{p}\mathrm{u}\mathrm{t}/\mathrm{n}\mathrm{u}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{a}\mathrm{k}\mathrm{e}\times 100,\end{array}}$$ (2)

$${\displaystyle \begin{array}{ll}{\displaystyle \mathrm{T}\mathrm{D}\mathrm{N}=}& \mathrm{d}\mathrm{i}\mathrm{g}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{b}\mathrm{l}\mathrm{e}\mathrm{C}\mathrm{P}+\mathrm{d}\mathrm{i}\mathrm{g}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{b}\mathrm{l}\mathrm{e}\mathrm{N}\mathrm{D}\mathrm{F}\\ & +\mathrm{d}\mathrm{i}\mathrm{g}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{b}\mathrm{l}\mathrm{e}\mathrm{N}\mathrm{F}\mathrm{C}+(2.25\times \mathrm{d}\mathrm{i}\mathrm{g}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{b}\mathrm{l}\mathrm{e}\mathrm{E}\mathrm{E}),\end{array}}$$ (3)

where nutrient intake and fecal nutrient output are in kilograms; TDN is total digestible nutrients (g/kg DM), CP is crude protein (g/kg DM), NDF is neutral detergent fiber (g/kg DM), NFC is non-fiber carbohydrate (g/kg DM), and EE is ether extract (g/kg DM).

Then, the digestible (DE) and metabolizable (ME) energies were calculated using Eqs. (4) (NRC, 2001) and (5) (Galyean et al., 2016), respectively.

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{D}\mathrm{E}=4.409\times \mathrm{T}\mathrm{D}\mathrm{N},}\end{array}}$$ (4)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{M}\mathrm{E}=0.9611\times \mathrm{D}\mathrm{E}-0.2999,}\end{array}}$$ (5)

where DE is digestible energy (Mcal/kg DM), TDN is total digestible nutrients (g/kg DM), and ME is metabolizable energy (Mcal/kg DM).

Finishing trials

The finishing trials were conducted after the digestibility trials, within studies, in the same facilities using the same diets (Table 1). The experimental period comprised the time required for the ad libitum animals to reach 4 mm of ultrasound subcutaneous backfat thickness measured in the longissimus muscle between the 12th and 13th ribs. Ultrasound measurements were obtained with a veterinary Pie Medical Aquila apparatus (Esaote Europe BV, Maastricht, The Netherlands) equipped with an 18-cm probe and a 3.5-MHz transducer. The images were analyzed using the Echo Image Viewer 1.0 program (Pie Medical Equipment BV, Maastricht, The Netherlands). The measurements were performed at the beginning of the finishing trial and at intervals of 28 d. Vegetable oil was used as an acoustic coupling agent for image acquisition. The initial and final BW recordings were preceded by 16 h of fasting from total solids and were used to calculate ADG.

Slaughter procedures

Baseline animals were slaughtered at the beginning of the finishing trial to develop predictive equations to estimate the initial body composition of the remaining animals in the finishing trial. When two ad libitum bulls reached 4 mm of subcutaneous backfat thickness determined by ultrasound, one restricted feeding bull was randomly chosen and slaughtered on the same day. Before slaughter, all animals had free access to water but were fasted for 16 h to determine the SBW. After slaughter, the gastrointestinal tract was emptied, washed, and weighed to determine EBW. The tissues (head, feet, viscera, soft tissue, and bones) were frozen after slaughter and processed to determine the chemical composition of the empty body and carcass by direct assessment after grinding, sampling, chemical analysis, and combining all tissues (blood, hide, head + feet, viscera, and carcass) of the animal. Tissue samples were analyzed for DM, ash, CP, and EE. Further details of slaughter and chemical analyses were described by Bonilha et al. (2014).

Calculation of energy requirements

Due to health issues, one Y08 bull (HRFI+CO) from the restricted feeding group was excluded from the data analysis. The procedures used to compute RE and maintenance energy requirements were similar to those described by Tedeschi et al. (2002). The initial EBW and initial body composition were predicted using regressions that relate the body component of interest to SBW obtained with the baseline animals of each class (HRFI or LRFI). The equation of each class on Y08 (Eqs. (6) and (7)) and Y09 (Eqs. (8) and (9)) were presented below:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{I}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{E}\mathrm{B}{\mathrm{W}}_{\mathrm{H}\mathrm{R}\mathrm{F}\mathrm{I}}=0.833\times \mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{S}\mathrm{B}{\mathrm{W}}_{\mathrm{H}\mathrm{R}\mathrm{F}\mathrm{I}}+32.247;R^2=99.0,}\end{array}}$$ (6)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{I}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{E}\mathrm{B}{\mathrm{W}}_{\mathrm{L}\mathrm{R}\mathrm{F}\mathrm{I}}=0.8758\times \mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{S}\mathrm{B}{\mathrm{W}}_{\mathrm{L}\mathrm{R}\mathrm{F}\mathrm{I}}+11.653;R^2=99.9,}\end{array}}$$ (7)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{I}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{E}\mathrm{B}{\mathrm{W}}_{\mathrm{H}\mathrm{R}\mathrm{F}\mathrm{I}}=0.9628\times \mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{S}\mathrm{B}{\mathrm{W}}_{\mathrm{H}\mathrm{R}\mathrm{F}\mathrm{I}}-18.692;R^2=98.5,}\end{array}}$$ (8)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{I}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{E}\mathrm{B}{\mathrm{W}}_{\mathrm{L}\mathrm{R}\mathrm{F}\mathrm{I}}=1.0887\times \mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{S}\mathrm{B}{\mathrm{W}}_{\mathrm{L}\mathrm{R}\mathrm{F}\mathrm{I}}-68.354;R^2=99.4.}\end{array}}$$ (9)

The body tissue energy was obtained from body protein and fat content and their respective caloric equivalents according to ARC (1980): tissue energy (Mcal) = 5.6405 × protein (kg) + 9.3929 × fat (kg). Retained energy was calculated from the increase in body fat and protein composition and their respective caloric values, or similarly, as the difference between initial and final tissue energy. Heat production was calculated as the difference between MEI and RE. As proposed by Lofgreen and Garrett (1968), linear regressions between MEI and log heat production (HP) was used to calculate NE_m. The metabolizable energy requirement for maintenance (ME_m) was estimated using an iterative method, as the point where MEI and HP are equal (Lofgreen and Garrett, 1968). The partial efficiency of ME utilization for maintenance (k_m) was obtained as the ratio between the NE_m and ME_m; the slope of the regression of RE on MEI above maintenance was assumed to be the partial efficiency of metabolizable energy utilization for gain (k_g) (Garrett, 1980; Tedeschi et al., 2002).

Statistical analyses

Data of digestibility, performance, and chemical composition of the final empty body and carcass were analyzed using PROC MIXED procedure of SAS (version 9.4, SAS Institute, Inc., Cary, NC). The effect of RFI class (LRFI and HRFI), SL (SE and CO), and feeding groups (ad libitum and restricted feeding) and their interactions were tested as fixed effects, whereas year was used as a random effect. Feeding groups were not adopted during the feed efficiency tests. For digestibility and performance data, there were no significant interactions, so they were removed from the statistical model. For chemical composition data, a significant interaction was observed between the RFI class and SL; the other nonsignificant interactions were removed from the statistical model. Least-square means were used for means comparison. Significance was declared for P < 0.05 and trends were discussed when 0.05 < P ≤ 0.10. The requirement measures were determined with the PROC MIXED and SOLUTION statements. The effects of RFI, SL, and MEI and their interactions were investigated, and only interaction between RFI and SL was significant. Similarly, year was used as a random effect.

Results

Feed intake and nutrient digestibility

Feed intake and nutrient digestibility were shown in Table 2. During the digestibility trial, LRFI animals consumed 10% less DM/d (P = 0.019) than HRFI animals, and SE animals consumed 17% more DM/d than the CO ones (P < 0.01). In agreement with the lower DMI, CP (0.969 vs. 1.07 kg/d; P = 0.017), NDF (2.13 vs. 2.32 kg/d; P = 0.029), NFC (2.90 vs. 3.21 kg/d; P = 0.017), and EE (0.331 vs. 0.367 kg/d; P = 0.016) intakes were lower for LRFI when compared with HRFI animals. Animals from SE line had higher CP (1.11 vs. 0.927 kg/d; P < 0.01), NDF (2.40 vs. 2.05 kg/d; P < 0.01), NFC (3.36 vs. 2.75 kg/d; P < 0.01), and EE (0.382 vs. 0.316 kg/d; P < 0.01) intakes than the ones from CO line. The difference in DMI detected among the traits in the present study did not influence nutrient digestibility. Average DM digestibility rate was 63.1% (P = 0.459 for RFI classes and P = 0.602 for SL).

Table 2.

Effect of RFI and SL on feed intake and digestibility coefficients of Nellore bulls (n = 51)

	RFI1			SL2			FG3			P-value
	Low	High	SEM	SE	CO	SEM	Ad libitum	Restricted	SEM	RFI	SL	FG
Intake, kg DM/d
Dry matter	6.56	7.22	0.515	7.53	6.25	0.512	7.23	6.56	0.493	0.019	<0.01	0.017
Crude protein	0.969	1.07	0.032	1.11	0.927	0.032	1.07	0.963	0.028	0.017	<0.01	0.009
NDF4	2.13	2.32	0.059	2.40	2.05	0.059	2.33	2.11	0.051	0.029	<0.01	0.011
NFC5	2.90	3.21	0.400	3.36	2.75	0.401	3.20	2.91	0.398	0.017	<0.01	0.031
Ether extract	0.331	0.367	0.017	0.382	0.316	0.016	0.368	0.330	0.016	0.016	<0.01	0.015
Digestibility, %
Dry matter	63.5	62.6	0.943	62.7	63.4	0.940	62.1	63.9	1.07	0.459	0.602	0.180
Crude protein	65.7	64.9	3.98	64.4	66.1	3.98	65.1	65.5	3.91	0.701	0.414	0.829
NDF4	47.2	45.8	6.58	45.4	47.5	6.54	46.6	46.4	6.58	0.327	0.170	0.905
NFC5	77.6	76.6	4.32	77.6	76.9	4.30	74.9	79.5	4.37	0.411	0.671	0.087
Ether extract	84.0	85.7	7.40	84.4	85.3	7.38	84.9	84.7	7.36	0.232	0.556	0.897

¹Treatments were established based on postweaning RFI.

²Selected Nellore (SE) and Control Nellore (CO) from Instituto de Zootecnia, São Paulo, Brazil Nellore breeding program.

³FG, feeding groups: restricted and ad libitum.

⁴NDF, neutral detergent fiber.

⁵NFC, non-fiber carbohydrate = 100 − (% crude protein + % ether extract + % NDF+ % ash).

Performance and chemical composition

Animal performance and chemical composition of the empty body, carcass, and offal were presented in Tables 3 and 4, respectively. The difference in RFI detected between the most and least efficient groups was 0.695 kg/d for postweaning RFI and 0.416 kg/d for finishing RFI. During the feed efficiency test, DMI was 17% lower (P < 0.01) for LRFI when compared with HRFI animals, and during the finishing trial, a 9% lower DMI trend (P = 0.059) was detected for LRFI animals. SE animals had 16% and 18% higher DMI than CO animals (P < 0.01) during the feed efficiency test and finishing trial, respectively. During the finishing trial, the performance was not influenced by RFI class, except for EBW gain, which tended to be higher for LRFI than HRFI (0.627 vs. 0.520 kg/d; P = 0.091) animals. Although similar ADG (0.654 vs. 0.567 kg/d; P = 0.265) was observed, SE animals had higher final BW (433 vs. 373 kg; P < 0.01) and final EBW (396 vs. 344 kg; P < 0.01) than the CO ones. The most efficient animals (i.e., LRFI) had more protein (18.9% vs. 18.0%; P = 0.037) and tended to have less EE (15.5% vs. 17.0%; P = 0.080) on carcass than the least efficient ones (i.e., HRFI), with no difference in empty body chemical composition (P > 0.05). On the other hand, the empty body differed between SL. The SE animals had more water (62.6% vs. 61.7%; P = 0.044), ash (4.93% vs. 4.61%; P = 0.037), and less EE (14.8% vs. 16.2%; P = 0.005) on empty body than CO animals. Also, the carcass of SE animals tended to have less EE (15.4% vs. 17.0%; P =0.066) than the ones from CO animals. A significant interaction between RFI and SL on empty body composition (CP, P = 0.033; EE, P = 0.029) showed that LRFI+SE animals had higher protein (17.8% vs. 17.0%; P = 0.024) and lesser EE (14.5% vs. 17.1%; P < 0.01) than LRFI+CO; no difference was found on HRFI animals. Animals from SE line had heavier offal than the ones from CO line (51.4 vs. 45.9 kg; P = 0.003; respectively) with higher water content (58.5% vs. 56.0%; P = 0.016; respectively). No differences were detected for offal weight and chemical composition between RFI classes.

Table 3.

Effect of RFI and SL on performance of Nellore bulls (n = 51) during the feed efficiency test and finishing trial

	RFI1			SL2			FG3			P-value
	Low	High	SEM	SE	CO	SEM	Ad Libitum	Restricted	SEM	RFI	SL	FG
Feed efficiency test
Postweaning RFI, kg DM/d	−0.335	0.360	0.048	−0.004	0.029	0.047	-	-	-	<0.01	0.585	-
DMI4, kg/d	5.48	6.40	0.141	6.38	5.49	0.152	-	-	-	<0.01	<0.01	-
DMI, % BW5	2.37	2.69	0.047	2.56	2.49	0.048	-	-	-	<0.01	0.060	-
Initial BW, kg	189	194	8.45	202	182	8.86	-	-	-	0.568	0.029	-
Final BW, kg	274	282	7.37	297	259	7.93	-	-	-	0.038	<0.01	-
Mid-test BW, kg	232	238	7.59	249	220	8.32	-	-	-	0.464	0.003	-
ADG6, kg/d	0.766	0.803	0.021	0.866	0.703	0.022	-	-	-	0.168	<0.01	-
Finishing trial
Finishing RFI, kg DM/d	−0.352	0.064	0.292	−0.206	−0.083	0.287	0.361	−0.649	0.283	0.040	0.555	<0.01
DMI, kg/d	5.84	6.35	0.193	6.59	5.60	0.190	7.22	4.96	0.164	0.059	<0.01	<0.01
DMI, % BW	1.52	1.61	0.061	1.58	1.55	0.062	1.80	1.35	0.061	0.026	0.531	<0.01
Initial BW, kg	363	373	11.8	394	342	12.3	377	359	11.0	0.339	<0.01	0.114
Final BW, kg	400	407	13.5	433	373	13.6	424	383	12.7	0.547	<0.01	0.002
Mid-test BW, kg	381	390	12.6	414	358	12.4	400	371	11.9	0.429	<0.01	0.015
ADG, kg/d	0.637	0.583	0.055	0.654	0.567	0.057	0.795	0.426	0.049	0.469	0.265	<0.01
Initial EBW7, kg	328	342	10.4	358	312	10.5	345	325	9.63	0.209	<0.01	0.084
Final EBW, kg	366	374	13.2	396	344	13.4	390	349	12.6	0.450	<0.01	0.001
EBW gain, kg/d	0.627	0.520	0.065	0.595	0.551	0.065	0.750	0.397	0.061	0.091	0.507	<0.01
Gain:feed, g/kg DM	105	89.8	7.84	97.7	97.6	6.93	109	85.9	8.95	0.120	0.998	0.030

¹Treatments were established based on postweaning RFI.

²Selected Nellore (SE) and Control Nellore (CO) from Instituto de Zootecnia, São Paulo, Brazil Nellore breeding program.

³FG, feeding groups: restricted and ad libitum.

⁴DMI, dry matter intake.

⁵BW, body weight.

⁶ADG, average daily gain.

⁷EBW, empty body weight.

Table 4.

Effect of RFI and SL on the final chemical composition of the empty body, carcass, and offal of Nellore bulls (n = 51)

	RFI1			SL2			FG3			P-value
	Low	High	SEM	SE	CO	SEM	Ad libitum	Restricted	SEM	RFI	SL	RFI*SL	FG
Carcass, kg	242	246	6.92	262	225	6.62	257	231	6.84	0.576	<0.01	0.515	<0.01
Offal4, kg	47.8	49.5	3.20	51.4	45.9	3.57	51.2	46.1	3.27	0.314	0.003	0.102	0.005
Carcass, % of EBW	66.1	65.8	0.488	66.3	65.4	0.489	66.2	65.5	0.511	0.511	0.007	0.200	0.018
Non-Carcass, % of EBW	33.9	34.2	0.488	33.7	34.6	0.489	33.8	34.5	0.511	0.511	0.007	0.200	0.018
Chemical composition (%)
Empty body
Water	62.1	62.2	0.321	62.6	61.7	0.326	61.8	62.4	0.370	0.737	0.044	0.181	0.203
Protein	17.4	17.7	1.34	17.7	17.4	1.35	17.5	17.5	1.34	0.280	0.319	0.033	0.944
Ether extract	15.4	15.7	1.39	14.8	16.2	1.39	15.9	15.2	1.41	0.421	0.005	0.029	0.227
Ash	4.77	4.77	0.110	4.93	4.61	0.121	4.76	4.79	0.191	0.968	0.037	0.849	0.827
Carcass
Water	59.2	59.8	1.51	59.9	59.1	1.50	59.0	60.0	1.52	0.333	0.216	0.173	0.074
Protein	18.9	18.0	0.301	18.7	18.1	0.318	18.2	18.6	0.347	0.037	0.181	0.070	0.368
Ether extract	15.5	17.0	1.27	15.4	17.0	1.29	16.8	15.6	1.31	0.080	0.066	0.090	0.137
Ash	5.86	5.91	0.389	6.04	5.73	0.386	5.94	5.82	0.377	0.835	0.188	0.647	0.607
Offal 4
Water	57.2	57.3	2.74	58.5	56.0	2.76	56.7	57.8	276	0.968	0.016	0.061	0.231
Protein	11.5	12.2	2.29	11.4	12.4	2.31	12.2	11.6	2.32	0.430	0.315	0.728	0.501
Ether extract	30.4	29.8	5.16	29.6	30.7	5.17	31.0	29.1	5.17	0.584	0.410	0.200	0.169
Ash	0.817	0.853	0.040	0.833	0.837	0.042	0.771	0.898	0.042	0.350	0.920	0.481	0.002

¹Treatments were established based on postweaning RFI.

²Selected Nellore (SE) and Control Nellore (CO) from Instituto de Zootecnia, São Paulo, Brazil Nellore breeding program.

³FG, feeding groups: restricted and ad libitum.

⁴Offal: Liver, kidney, empty gastrointestinal tract, other internal organs, tail and kidney, pelvic and heart fat.

Energy components

The energy components were presented in Table 5. The RFI classification had no effect on MEI (P = 0.330), however LRFI animals retained 31.1% more energy (32.1 vs. 22.1 kcal/kg^0.75 EBW/d; P = 0.049) and produced 10% less heat than HRFI animals (159 vs. 175 kcal/kg^0.75 EBW/d; P = 0.033). Selected animals tended to produce more heat (174 vs. 159 kcal/kg^0.75 EBW/d; P = 0.075), with the same RE (P = 0.165). The linear regressions between MEI and log(HP) were represented by follow Eqs. (10)–(13):

Table 5.

Effect of RFI and SL on performance and energy components of Nellore bulls (n = 51)

	RFI1			SL2			FG3				P-value
	Low	High	SEM	SE	CO	SEM	Ad libitum	Restricted	SEM	RFI	SL	RFI*SL	FG
MEI4, kcal/kg0.75 EBW/d	191	197	5.01	198	190	4.98	219	169	5.11	0.330	0.275	0.583	<0.01
RE5, kcal/kg0.75 EBW/d	32.1	22.1	8.3	23.4	30.9	9.4	35.0	19.3	8.92	0.049	0.165	0.075	0.005
HP6, kcal/kg0.75 EBW/d	159	175	21.0	174	159	21.1	184	149	21.3	0.033	0.075	0.490	<0.01
NEm7,11, kcal/kg0.75 EBW/d	71.3	76.2	0.075	75.5	72.0	0.077	-	-	-	0.009	0.046	0.063	-
MEm8, kcal/kg0.75 EBW/d	114	131	-	128	116	-	-	-	-	-	-	-	-
km9, %	62.6	58.4	-	59.0	62.1	-	-	-	-	-	-	-	-
kg10, %	21.8	19.5	0.158	21.8	56.5	0.134	-	-	-	0.751	0.013	0.126	-

¹Treatments were established based on postweaning RFI.

²Selected Nellore (SE) and Control Nellore (CO) from Instituto de Zootecnia, São Paulo, Brazil Nellore breeding program.

³FG, feeding groups: restricted and ad libitum.

⁴MEI, metabolizable energy intake.

⁵RE, retained energy.

⁶HP, heat production.

⁷NE_m, net energy requirement for maintenance.

⁸ME_m, metabolizable energy requirement for maintenance.

⁹k_m, efficiency of metabolizable energy utilization for maintenance.

¹⁰k_g, efficiency of metabolizable energy utilization for gain.

¹¹NE_m regression slope: 0.0019 for LRFI, 0.0020 for HRFI, 0.0022 for SE, and 0.0013 for CO.

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{l}\mathrm{o}\mathrm{g}(\mathrm{H}{\mathrm{P}}_{\mathrm{L}\mathrm{R}\mathrm{F}\mathrm{I}})=0.0019\times \mathrm{M}\mathrm{E}\mathrm{I}+1.853;R^2=54.3\mathrm{\%},}\end{array}}$$ (10)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{l}\mathrm{o}\mathrm{g}(\mathrm{H}{\mathrm{P}}_{\mathrm{H}\mathrm{R}\mathrm{F}\mathrm{I}})=0.0020\times \mathrm{M}\mathrm{E}\mathrm{I}+1.882;R^2=49.4\mathrm{\%},}\end{array}}$$ (11)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{l}\mathrm{o}\mathrm{g}(\mathrm{H}{\mathrm{P}}_{\mathrm{S}\mathrm{E}})=0.0022\times \mathrm{M}\mathrm{E}\mathrm{I}+1.878;R^2=72.3\mathrm{\%},}\end{array}}$$ (12)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{l}\mathrm{o}\mathrm{g}(\mathrm{H}{\mathrm{P}}_{\mathrm{C}\mathrm{O}})=0.0013\times \mathrm{M}\mathrm{E}\mathrm{I}+1.856;R^2=19.8\mathrm{\%},}\end{array}}$$ (13)

where log(HP) is the base 10 logarithm of heat production (kcal/kg^0.75 EBW/d) and MEI is metabolizable energy intake (kcal/kg^0.75 EBW/d).

The LRFI animals had lower NE_m on EBW basis than HRFI animals (71.3 vs. 76.2 kcal/kg^0.75 EBW/d; P = 0.009); and SE animals had higher NE_m than the CO ones (75.5 vs. 72.0 kcal/kg^0.75 EBW/d; P = 0.046). A tendency for the interaction between RFI class and SL (P = 0.063; Figure 1) suggested that LRFI+CO animals had lower NE_m than HRFI+CO animals (67.9 vs. 76.0 kcal/kg^0.75 EBW/d; P = 0.006); however, for SE animals, there was no statistical difference in NE_m between RFI classes (74.7 vs. 76.4 kcal/kg^0.75 EBW/d; P = 0.527). The ME_m, computed when HP equals MEI, was 114 and 131 kcal/kg^0.75 EBW/d for LRFI and HRFI, respectively, and for SL ME_m was 128 and 116 kcal/kg^0.75 EBW/d for SE and CO, respectively. The k_m was 62.6% and 58.4% for LRFI and HRFI animals, and 59% and 62.1% for SE and CO animals, respectively. The k_g was similar between RFI classes (P = 0.751), and lower for SE when compared to CO animals (21.8% vs. 56.5%; P = 0.013; respectively).

Figure 1.

Effects of postweaning RFI (low = LRFI; high = HRFI) and selection line (CO = Control; SE = Selected) on net energy requirement for maintenance (NE_m; kcal/kg^0.75 empty body weight, EBW). (a–b) Least squares means within treatments (low vs. high RFI and CO vs. SE) with different superscripts differ (P < 0.05).

Discussion

Dry matter intake is influenced by several different factors, including animal BW, stage of production, forage-to-concentrate ratio, diet quality, and environmental conditions (Tedeschi and Fox, 2020). Since RFI is defined as the difference between actual and predicted feed intake based on maintenance and growth (Koch et al., 1963), the mean difference in RFI detected during the postweaning phase between most and least efficient animals can change until slaughter (Batalha et al., 2020). However, in the present study, a minimum difference of 416 g DM/d between the most and least efficient animals remained during the postweaning efficiency test, digestibility, and finishing trials. Asher et al. (2018) suggested that RFI ranking of animals remains independent of physiological age and diet quality. Regarding the SL, Intituto de Zootecnia Breeding Program selects animals for higher growth rates (Bonilha et al., 2014), resulting in higher DMI due to the large body size of selected animals (Sobrinho et al., 2011).

According to Cantalapiedra-Hijar et al. (2018), more efficient animals have a greater digestive capacity, accounting for 10% of the inter-animal variation in RFI (Herd and Arthur, 2009). Although higher DMI can be related to a greater passage rate and reduced digestibility, the higher DMI of LRFI+SE than HRFI+CO, respectively, did not affect digestibility in the present study. Bonilha et al. (2017) and Johnson et al. (2019) found that LRFI bulls and steers, respectively, had greater DM digestibility than HRFI animals. However, the literature regarding the associations between RFI and DM digestibility is inconsistent (Kenny et al., 2018). Batalha et al. (2020) found lower apparent nutrient digestibility of DM, CP, NDF, and NFC for LRFI when compared with HRFI Nellore bulls. In agreement with our findings, Tedeschi et al. (2006) suggested that animals with a higher growth potential have the digestive and metabolic capacity to utilize more nutrients. Even at the same digestibility rates, more efficient animals should be able to better utilize nutrients due to changes in the expression of genes involved in energy metabolism in the ruminal epithelium (Elolimy et al., 2019). Consequently, energy expenditure is reduced in this tissue and the capacity for uptake of volatile fatty acids is increased, resulting in maximum weight gain at levels below the maximum feed intake (Ferrell and Jenkins,1998). In part, however, the restricted feeding required by the comparative slaughter technique to determine energy requirements near the maintenance, may have prevented HRFI and LRFI animals to express their genetic potential regarding feed intake, the impact of visceral organ size on digestibility, and body composition.

The primary mechanism separating LRFI from HRFI animals is the protein-to-fat-ratio in the deposited tissue (Asher et al., 2018). Lancaster et al. (2009) showed that the inclusion of gain in 12th to 13th rib fat thickness and final longissimus muscle area into the base model of RFI increased R² and accounted for 9% of the variation in DMI not explained by mid-test metabolic BW and ADG. The authors suggested that more efficient animals are leaner than less efficient animals and that carcass-adjusted RFI can be used to select more efficient animals without affecting the rate or composition of gain. The results were later confirmed by Asher et al. (2018), who observed more protein and less fat in LRFI animals’ muscle compared with the HRFI ones. In the present study, the most efficient animals had less EE (LRFI and SE) and more protein (LRFI) in their carcasses than the less efficient ones, agreeing with the authors mentioned above. The leaner carcasses may indicate an association among genetic selection, RFI and maturity patterns and is commonly found in larger, faster gaining animals (i.e., SE; Tedeschi et al., 2006).

Variations in feed efficiency are related to cellular energy-consuming processes such as mitochondrial activities and protein turnover (Cantalapiedra-Hijar et al., 2018). Recent studies demonstrated that more efficient animals have lower mitochondrial activity (Benedeti et al., 2018) and lower protein degradation (Elolimy et al., 2019). Asher et al. (2018) reported lower MEI and HP in male Holstein calves classified as LRFI, suggesting a lower rate of protein degradation in this class compared with HRFI calves. Basarab et al. (2003) reported 10.2% less MEI and 9.3% less HP for LRFI compared with HRFI animals, similar to the 10% lower HP found in the present study. Since the k_m may be associated to body composition (Garrett, 1980), animals with higher proportion of protein on carcass (i.e., LRFI) had lower k_m because the efficiency of energy deposited as protein is lesser than the energy deposited as fat (Tedeschi et al., 2004). In the present study, k_m of 62.6% and 58.4% for LRFI and HRFI, respectively, were found, confirming Asher’s et al. (2018) hypothesis that efficient energy utilization partly explains variation in the efficiency to support leaner tissue on LRFI animals. Having lower HP and higher k_m, lower protein turnover is expected for LRFI when compared with HRFI animals, as suggested by Asher et al. (2018) and Elolimy et al. (2019). Regarding SL, SE animals had higher protein and lesser EE on empty bodies and tended to have higher HP and lower k_m than CO animals, suggesting no difference in protein turnover between the SL. These results were later confirmed by the lower k_g detected for SE animals when compared to the CO ones. According to the nonlinear relationship between k_g and protein proportion in RE devised by Tedeschi et al. (2004), protein proportions in RE were 89% for SE and 12% for CO animals, assuming the efficiencies of protein and fat of 20% and 75%, respectively.

Menezes et al. (2020) did not find differences in Nellore bulls gain composition between the RFI classes, and reported similar values of k_m between LRFI and HRFI animals (64.3% vs. 63%); however, the NE_m of LRFI was 23.2% lower than the NE_m of HRFI animals (63.4 vs. 78.1 kcal/kg^0.75 EBW/d). In the present study, the detected NE_m was 71.3 kcal/kg^0.75 EBW/d for LRFI, 6% lower than the pattern suggested by NASEM (2016) of 75.6 kcal/kg^0.75 EBW/d for Zebu cattle and 17.8% lower than the pattern of taurine animals (84 kcal/kg^0.75 EBW/d). This result reveals a margin for energy improvements in Nellore cattle through RFI identification. However, our data do not support the hypothesis that SE animals classified as LRFI have similar NE_m than CO animals. In our study, NE_m was 74.7 kcal/kg^0.75 EBW/d for LRFI+SE and 72.0 kcal/kg^0.75 EBW/d for CO animals (LRFI and HRFI combined). The NE_m for LRFI+CO animals was 67.9 kcal/kg^0.75 EBW/d; however, the lower carcass weight, rib-eye area, and hindquarter proportion (Bonilha et al., 2014) compared with SE animals yielded the advantage of the lower requirement impracticable.

According to Castro Bulle et al. (2007), animals with high carcass weight divert more energy toward production and growth than lighter carcass animals that divert more energy toward maintenance; thus, demonstrating the possibility of animals with different weights having similar NE_m, as found for HRFI+CO and HRFI+SE. For the group of animals composed only by SE, it is not possible to find a difference between the most and least feed efficient. The SE animals seems unresponsive to RFI classification, with no difference in ME_m, NE_m, or HP between LRFI and HRFI animals. Instituto de Zootecnia Breeding Program began in 1976 with the aim to increase postweaning weight gain and meat production. Over the years, selection resulted in a higher relative growth rate and Kleiber ratio (Sobrinho et al., 2011), as well as heavier carcasses and higher tissue deposition rates (Bonilha et al., 2014), in SE animals compared with CO, mostly when conducted for a group of animals (Tedeschi et al., 2006). The results found in the present study indicated that the breeding program had no influence in feed efficiency, with the same NE_m, ME_m, and HP for LRFI+SE and HRFI+SE, which means no difference between LRFI and HRFI in SE animals. Our analyses showed a clear separation between LRFI and HRFI animals within CO animals, and no differences for SE animals in which the maintenance requirement was similar.

Conclusion

The breeding program for postweaning weight may not have improved feed efficiency over the years, with RFI classification not being a promising selection tool for SE animals. Classification based on RFI seems to be useful in animals that have not undergone the breeding program, with LRFI animals having lower energy requirements than the HRFI ones. The energy requirements indicated lower NE_m than predicted by NASEM (2016), which was based on a literature review (Chizzotti et al., 2008), suggesting that opportunities still exist to improve Zebu cattle energy requirements. Future research studies are encouraged to conduct more comprehensive meta-analytical regressions using a larger sample size.

Glossary

Abbreviations

ADG

average daily gain

BW

body weight

CP

crude protein

DE

digestible energy

DM

dry matter

DMI

dry matter intake

EBW

empty body weight

EE

ether extract

HP

heat production

iNDF

indigestible neutral detergent fiber

k_g

partial efficiency of metabolizable energy utilization for gain

k_m

partial efficiency of metabolizable energy utilization for maintenance

MEI

metabolizable energy intake

ME_m

metabolizable energy for maintenance

NDF

neutral detergent fiber

NE_g

net energy for gain

NE_m

net energy requirements for maintenance

ME

metabolizable energy

RE

retained energy

RFI

residual feed intake

SBW

shrunk body weight

TDN

total digestible nutrients

Conflict of interest statement

The authors declare no real or perceived conflicts of interest.