Search
Author
Title
Vol.
Issue
Year
1st Page

Journal of Animal Science - Animal Production

Effects of residual feed intake classification and method of alfalfa processing on ewe intake and growth

 

This article in JAS

  1. Vol. 92 No. 2, p. 830-835
     
    Received: May 31, 2013
    Accepted: Dec 07, 2013
    Published: November 24, 2014


    1 Corresponding author(s): reid.redden@ndsu.edu
 View
 Download
 Share

doi:10.2527/jas.2013-6768
  1. R. R. Redden 1,
  2. L. M. M. Surber,
  3. A. V. Grove and
  4. R. W. Kott
  1. Department of Animal Science, North Dakota State University, Fargo 58108
    Department of Animal and Range Sciences, Montana State University, Bozeman 59717
    AG Research LLC, White Sulphur Springs, MT 59645

Abstract

The objective of this research was to evaluate the effects of residual feed intake (RFI) determined under ad libitum feeding conditions on DMI and performance of yearling ewes fed either chopped or pelleted alfalfa hay. In Exp. 1, 45 ewe-lambs had ad libitum access to a pelleted grower diet for 63 d and individual DMI was determined using an electronic feed delivery system. Residual feed intake values were assigned to each ewe-lamb as a measure of feed efficiency. Sixteen ewe-lambs with the most positive RFI values were classified as high RFI (inefficient) and 16 ewe-lambs with the most negative RFI values were classified as low RFI (efficient). In Exp. 2, half of the ewes from each efficiency group were placed into 1 of 2 pens and provided ad libitum access to either pelleted or chopped alfalfa hay. Individual DMI was again determined using an electronic feed delivery system. Body weight, LM area (LMA), and 12th-rib back fat thickness (BF) were measured at the beginning and end of both experiments. In Exp. 1, DMI by ewe-lambs in the low RFI group was 9% less (P = 0.01) than by ewe-lambs in the high RFI group (2.21 vs. 2.43 kg/d); however, ADG and initial and final BW, LMA, and BF did not differ (P > 0.27) among RFI groups. In Exp. 2, there were no feed processing × RFI group interactions (P > 0.14) for any trait. By design, RFI values were lower (P < 0.01) by yearling ewes in the low than high RFI group (–0.27 vs. 0.27); however, RFI values did not differ (P = 1.0) between yearling ewes fed chopped versus pelleted alfalfa. Dry matter intake was 22% less (P < 0.01) by yearling ewes in the low than high RFI group (2.5 vs. 3.2 kg/d) and 59% less (P < 0.01) by yearling ewes fed chopped versus pelleted alfalfa (2.2 vs. 3.5 kg/d). Initial and final BW, ADG, and G:F did not differ (P > 0.45) between RFI groups but were greater (P < 0.01) by yearling ewes fed pelleted alfalfa compared to chopped alfalfa. Final LMA did not differ (P = 0.77) between RFI groups, but final BF tended to be greater (P = 0.06) for high than low RFI yearling ewes (0.63 vs. 0.57 cm). Final LMA and BF were greater (P < 0.01) by yearling ewes fed pelleted than chopped alfalfa. Low RFI ewes had lower DMI and BF compared to high RFI ewes. Ewe efficiency as determined by RFI was repeatable on subsequent forage based diets; however, differences in intake and efficiency were more apparent when alfalfa was pelleted.



INTRODUCTION

Feed costs account for the largest input cost in livestock production systems and identification of animals that require less feed for normal production would clearly increase productivity. Residual feed intake (RFI) is the concept of developing an alternative feed efficiency measurement that is independent of growth traits and was first proposed by Koch et al. (1963). Residual feed intake is the difference between actual feed intake and predicted feed intake based on maintenance of BW and ADG by linear regression. Numerous research efforts have shown that there is considerable individual animal variation in RFI in cattle (reviewed by Archer et al., 1999; Herd et al., 2003) and sheep (Snowder and Van Vleck, 2003; Cammack et al., 2005). Most RFI testing has been evaluated postweaning on animals fed medium-to-high energy diets ad libitum, but few researchers have evaluated the effects of RFI classification on intake and performance of ruminants fed a different diet than that offered during RFI testing, particularly in sheep. Kelly et al. (2010b) reported that RFI values were repeatable in cattle fed grower and finisher diets; however, others have reported that >50% of steers and replacement heifers tested for RFI in 1 period were categorized into a different RFI group in a second feeding period when fed the same or a different diet (Durunna et al., 2011a, 2012). Similarly, Redden et al. (2011) reported that there was no correlation between RFI of ewe-lambs fed an ad libitum pelleted diet and the same animals fed grass hay ad libitum as yearling ewes. Forage processing can affect intake and digestibility of rations (Heaney et al., 1968) and contribute to variation in efficiency (Herd et al., 2004); therefore, the objective of the current study was to measure DMI, ADG, and carcass characteristics of RFI tested ewe-lambs fed a chopped or pelleted alfalfa diet as yearling ewes.


MATERIALS AND METHODS

Use of animals was approved by the Montana State University Animal Care and Use Committee.

Experiment 1: Determination of Residual Feed Intake

A 63-d trial was initiated in October 2010 using 45 Targhee ewe-lambs (average initial BW = 43 ± 4 kg and average initial age = 6 mo of age) to determine RFI using the GrowSafe feed intake system (GrowSafe Systems, Ltd., Airdrie, AB, Canada). Ewe-lambs were selected randomly from the Montana State University Red Bluff Research Ranch 2010 spring-born lamb crop. Ewe-lambs were from a single source flock and had been managed as a contiguous flock from birth. Ewe-lambs selected for the study were housed together in a 10 by 10 m pen with 4 GrowSafe bunks at the Montana State University Nutrition Center. Elevated platforms were constructed to modify GrowSafe beef cattle stanchions and feed bunks for sheep. Ewe-lambs were allowed a 2-wk adaptation period during which all feed bars were removed so that multiple ewes could eat at the same time. After 1 wk, feed bars were placed in the feed trough so that only 1 animal could have access to the feed bunk at any given time. Ewe-lambs were allowed ad libitum access to water and a pelleted grower diet (Table 1). Feed allowances were checked and more feed offered daily at 1500 h. Recording of feed disappearance data was initiated after the 2-wk adaptation period and daily feed intake was computed using the Process Intakes and Export Behavior Data routine of the GrowSafe Data Acquisition software. Average assigned feed disappearance was 99.1% and only 1 daily assigned feed disappearance value was less than 95%. Ewe-lambs were weighed at 7 d intervals and 2 consecutive day BW were collected at the beginning and end of the test. Growth rates of individual animals were modeled by linear regression of lamb BW by day using the PROC GLM procedure of SAS (SAS Inst. Inc., Cary, NC), and regression coefficients were used to compute modeled ADG, initial and final BW, and metabolic BW (MBW; midtest BW0.75) as described by Lancaster et al. (2009). Expected feed intake was modeled (PROC GLM of SAS) by linear regression of DMI (using chemical analysis of feed samples) against MBW and ADG during the trial (Koch et al., 1963). Residual feed intake was determined by subtraction of expected feed intake minus actual feed intake. To further characterize RFI, the 16 ewe-lambs with the most negative RFI values were classified as low RFI (efficient) and the 16 ewes with the highest, most positive RFI vales were classified as high RFI (inefficient).


View Full Table | Close Full ViewTable 1.

Ingredient and chemical composition of diets used in Exp. 1 and 2

 
Item Diet or feedstuff
Grower pellet Chopped alfalfa Pelleted alfalfa
Ingredient, % as fed basis
    Alfalfa sun-cured 43.60 100 100
    Corn 23.93
    Wheat middlings 15.00
    Soybean hulls 10.00
    Molasses cane 4.00
    Calcium carbonate 2.02
    Ammonium chloride 0.60
    Pelletizing agent 0.50
    Premixes1 0.35
Chemical composition, DM basis
    DM, % 88.5 91.1 86.8
    CP, % DM 15.8 15.5 15.8
    TDN, % DM 75.0 57.2 62.4
1Contained 2,000 mg/kg Mo, 44,000 IU/kg vitamin E, 9% Fe, 10% Zn, 6% Mn, 3.4% I, 0.03% Co, 0.02 IU/mg Se, 23,700 IU/mg vitamin A, and 2,250 IU/mg vitamin D.

Real-time ultrasound measurements of LM area (LMA) and 12th-rib back fat thickness (BF) were determined at the beginning, middle, and end of the experiment by a National Sheep Improvement Program certified technician using an Aloka SSD-500V instrument with a 3.5-MHz, 12.5 cm linear array transducer and stand-off guide (Hitachi Aloka Medical, Ltd., Wallingford, CT). Longissimus muscle area and BF were modeled by linear regression to compute muscle and fat deposition rates to estimate the effects of lean and fat deposition on RFI (Knott et al., 2008).

Diet samples were collected weekly, composited, and stored for chemical analysis. Dry matter content was determined by drying samples in a 60°C forced-air oven for 48. Diet samples were analyzed for CP and TDN by an independent laboratory (Midwest Laboratories Inc., Omaha, NE).

Data were analyzed as a completely random design with RFI group (low and high) as a fixed effect using the PROC GLM procedure of SAS. Means were determined and separated using the LSMEANS and PDIFF statements of SAS, respectively, and were considered different if P < 0.05. Because only 32 ewe-lambs from this experiment (16 low RFI and 16 high RFI) were used in Exp. 2, only data from these 32 ewe-lambs were analyzed and reported for Exp. 1.

Experiment 2: Yearling Ewe Intake and Performance

After the conclusion of Exp. 1, ewe-lambs were maintained together at the Red Bluff Research Ranch and allowed to graze native range. On March 8, 2011, the 32 ewes classified as low or high RFI in Exp. 1 were returned to the GrowSafe facility and half the ewes from each RFI group were placed in 1 of 2 pens that contained 2 GrowSafe feed bunks and allowed ad libitum access to either chopped or pelleted alfalfa hay. Alfalfa hay was purchased from a single source and every other bale was sent to a custom pellet mill for pelleting. The remaining bales were chopped to a 1-cm length at the Montana State University Feed Mill. Yearling ewes (11 mo of age; 56 ± 4 kg BW) were allowed a 1-wk period to become reacclimated to the facility. Daily feed intake data was collected as previously described for 57 d. Average assigned feed disappearance was 98.9%. Feed sampling, BW, LMA, and BF were collected as described in Exp. 1. Residual feed intake was also calculated as previously described; however, we did not attempt to include rate of LMA and BF deposition in the RFI model in Exp. 2. Data were analyzed as a completely randomized design using PROC GLM with RFI group, alfalfa processing (chopped or pelleted), and their interaction as fixed effects. Initial BW was included as a covariate for growth, intake, and carcass traits to account for differences in initial BW between treatment groups. Means were determined and separated using the LSMEANS and PDIFF statements of SAS, respectively, and were considered different if P < 0.05.


RESULTS

Experiment 1

Ewe-lambs categorized in the low RFI group consumed 9% less feed (P < 0.01) than ewe-lambs in the high RFI group (2.21 vs. 2.43 kg/d; Table 2). Ewe-lamb ADG and initial and final BW, LMA, and BF did not differ (P > 0.27) among RFI groups in Exp. 1.


View Full Table | Close Full ViewTable 2.

Performance and intake by ewe-lambs according to residual feed intake (RFI) classification in Exp. 1

 
Item1 Low RFI High RFI SEM P-value
No. of ewe-lambs 16 16
Growth traits
    Initial BW, kg 42.5 43.2 1.46 0.75
    Final BW, kg 61.1 61.6 1.90 0.84
    ADG, g/d 294 293 12.7 0.95
    DMI, kg/d 2.21 2.43 0.061 0.01
Feed efficiency trait
    RFI, kg/d –0.10 0.11 0.013 0.0001
Carcass traits
    Initial LMA, cm2 11.6 12.0 0.44 0.44
    Final LMA, cm2 17.6 17.1 0.38 0.41
    Initial BF, cm 0.22 0.25 0.014 0.28
    Final BF, cm 0.63 0.64 0.033 0.89
1Modeled values are reported for BW and ADG. LMA = LM area; BF = 12th-rib back fat thickness.

Experiment 2

There were no feed processing × RFI group interactions (P > 0.14) for any intake, growth, efficiency, or carcass traits (Table 3).


View Full Table | Close Full ViewTable 3.

Effects of residual feed intake (RFI) classification and method of feed processing on yearling ewe growth, intake, and carcass traits in Exp. 2

 
Item1 Chopped alfalfa
Pelleted alfalfa
SEM P-value
Low RFI High RFI Low RFI High RFI Alfalfa RFI Alfalfa × RFI
No. of yearling ewes 8 8 8 8
Growth traits and intake
    Initial BW, kg 53.4 53.1 58.6 60.5 4.27 <0.01 0.72 0.63
    Final BW, kg 60.4 61.4 73.2 73.2 2.53 <0.01 0.68 0.73
    ADG, g/d 71 89 300 301 45.13 <0.01 0.68 0.73
    DMI, g/d 2.0 2.4 3.0 3.9 0.38 <0.01 <0.01 0.26
Feed efficiency traits
    RFI, kg/d –0.15 0.15 –0.38 0.38 0.31 1.00 <0.01 0.15
    G:F, g/kg 26.1 32.3 99.1 79.2 18.32 <0.01 0.46 0.17
Carcass traits
    Initial LMA, cm2 17.5 17.5 17.6 18.1 1.03 0.56 0.61 0.71
    Final LMA, cm2 17.1 17.6 20.5 20.4 1.19 <0.01 0.77 0.63
    Initial BF, cm 0.48 0.51 0.40 0.45 0.058 0.06 0.23 0.77
    Final BF, cm 0.42 0.52 0.72 0.71 0.062 <0.01 0.06 0.18
1Modeled values are reported for BW and ADG. LMA = LM area; BF = 12th-rib back fat thickness.

Initial and final BW, ADG, and G:F by yearling ewes did not differ (P > 0.45) between RFI groups. Dry matter intake was 22% less (P < 0.01) by yearling ewes in the low than high RFI group (2.5 vs. 3.2 kg/d). Initial and final LMA did not differ (P > 0.60) between RFI groups. Initial BF did not differ (P = 0.23) between RFI groups, but final BF tended to be greater (P = 0.06) by high than low RFI yearling ewes (0.63 vs. 0.57 cm).

Despite the fact that ewe-lambs were managed together between the end of Exp. 1 and beginning of Exp. 2, initial BW was greater (P < 0.01) for yearling ewes assigned to the chopped alfalfa versus pelleted alfalfa diets (53 vs. 60 kg). Final BW, ADG, DMI, and G:F were greater (P < 0.01) by yearling ewes fed pelleted alfalfa compared to chopped alfalfa. Residual feed intake values did not differ (P = 1.0) between yearling ewes fed chopped versus pelleted alfalfa. Initial LMA did not differ (P = 0.56) between alfalfa processing treatments, but final LMA was greater (P < 0.01) by yearling ewes fed pelleted than chopped alfalfa. Initial BF tended to be greater (P = 0.06) for yearling ewes assigned to the chopped versus pelleted alfalfa treatment; however, final BF was greater (P < 0.01) by yearling ewes fed pelleted versus chopped alfalfa.


DISCUSSION

In the current study, MBW and ADG accounted for most of the variation in DMI (P < 0.0001) and the R2 for this model was 0.84. Rate of LMA and BF deposition were not significant sources of variation in the RFI model (P = 0.72 and P = 0.24, respectively) and their inclusion did little to improve the R2 (0.85); therefore, these variables were not included in the final model used to predict DMI and calculate RFI. The R2 values we observed for RFI models were greater than those reported in sheep by other researchers (R2 = 0.33 to 0.73; Knott et al., 2008; Redden et al., 2011, 2013). Knott et al. (2008) observed a 10% increase in the R2 value when fat and muscle depth were used to predict DMI; however, in agreement with our results, most researchers reported only slight improvements in the R2 value (0.3 to 3%) when LMA and BF were included in RFI models to predict DMI of sheep (Redden et al., 2011) and beef cattle (Lancaster et al., 2009; Kelly et al., 2010a; Hafla et al., 2012).

We observed a 9% decrease in DMI by low versus high RFI ewe-lambs in Exp. 1 and a 22% decrease in DMI by low versus high RFI yearling ewes in Exp. 2. Other researchers have observed similar reductions in DMI between low and high RFI groups during RFI testing ranging from 15 to 29% in sheep (Muro-Reyes et al., 2011; Redden et al., 2011, 2013) and 10 to 17% in beef cattle (Lawrence et al., 2011; Durunna et al., 2011b, 2012). In agreement with our results, Redden et al. (2011) reported that reductions in DMI between low and high RFI groups were of a lower magnitude in ewe-lambs than yearling ewes (19 and 29%, respectively). Differences in DMI between low and high RFI groups during RFI testing, however, were not observed when sheep (Redden et al., 2011) and cattle (Herd et al., 1998; Meyer et al., 2008; Lawrence et al., 2012) were evaluated under production settings in subsequent studies. This was true when DMI was estimated for groups of sheep (Redden et al., 2011) or cattle (Meyer et al., 2008) or for individual grazing animals (Herd et al., 1998; Lawrence et al., 2012).

Residual feed intake is independent of production traits (Herd and Arthur, 2009); therefore, it is not surprising that we and others observed no differences in BW or ADG between low and high RFI groups during initial RFI testing in sheep (Muro-Reyes et al., 2011; Redden et al., 2011, 2013) and cattle (Basarab et al., 2003; Lancaster et al., 2009; Kelly et al., 2010a), during subsequent RFI testing periods (Redden et al., 2011; Durunna et al., 2011b, 2012), or when previously RFI tested animals were evaluated in production settings (Meyer et al., 2008; Lawrence et al., 2012; Redden et al., 2013). In contrast to these results, Herd et al. (1998) reported that cattle with predetermined low RFI had greater BW than high RFI cattle when grazing pasture.

We observed no difference in feed conversion (G:F) between RFI groups of yearling ewes. Similar to our results, Lawrence et al. (2011, 2012) observed no difference in feed conversion by forage-fed cattle during RFI testing; however, feed:gain was greater for low than high RFI cattle fed diets based on grain and silage (Basarab et al., 2003; Lancaster et al., 2009; Kelly et al., 2010a). Redden et al. (2013) reported greater feed:gain by high versus low RFI ewe-lambs fed a pelleted grower diet, but no differences in feed:gain were observed between RFI groups when these same sheep were fed chopped grass hay as yearling ewes. It is intuitive that lower DMI but similar ADG by low versus high RFI animals would result in improved feed conversion for the low RFI group; however, results from these studies suggest differences in feed conversion between RFI groups when animals are fed medium- to high-energy diets but not forages.

In agreement with our results, most researchers also observed no differences in LMA between high and low RFI groups during RFI testing (Lancaster et al., 2009; Kelly et al., 2010a; Hafla et al., 2012) or subsequent feeding trials (Durunna et al., 2012; Redden et al., 2011, 2013; Lawrence et al., 2012). In contrast, results for BF between RFI groups have been more variable. Ultrasound BF differed between RFI groups during RFI testing in heifers (Kelly et al., 2010a; Durunna et al., 2012) and bulls (Lancaster et al., 2009; Hafla et al., 2012) but not in steers (Durunna et al., 2011b) or sheep (Redden et al., 2013) and not when previously RFI tested sheep (Redden et al., 2013) and cattle (Herd et al., 1998; Durunna et al., 2011b, 2012) were evaluated in subsequent feeding periods. Other researchers have reported increased BF by high compared to low RFI cattle that was of a similar magnitude (0.06 to 0.08 cm) to what we observed in the current study (Lancaster et al., 2009; Kelly et al., 2010a; Hafla et al., 2012). Similarly, Redden et al. (2013) reported that change in BF was numerically greater for high versus low RFI sheep that were limit fed after RFI testing.

Thirty-four percent of yearling ewes that had a positive RFI value in Exp. 1 also had a positive RFI value in Exp. 2 (5/16 and 6/16 of yearling ewes fed chopped and pelleted alfalfa, respectively) and 44% of yearling ewes that had a negative RFI value in Exp. 1 also had a negative RFI value in Exp. 2 (6/16 and 8/16 of yearling ewes fed chopped and pelleted alfalfa, respectively). Therefore, 78% of yearling ewes maintained a positive or negative RFI value in both experiments (11/16 and 14/16 of yearling ewes fed chopped or pelleted alfalfa, respectively). In contrast to our results, Durunna et al. (2011a, 2012) reported that 51 to 55% of heifers and steers were classified into a different RFI category when fed the same or different diets in 2 periods. Our results may be higher because we simply looked at the positive or negative sign of the RFI value rather than reclassifying animals into RFI groups.

There is some research evaluating performance of animals previously tested for RFI under different feeding conditions in cattle, but little research has been published in sheep. Redden et al. (2013) reported that there was no correlation between RFI values of sheep fed a pelleted grower diet and then a chopped native grass hay diet. In contrast, Kelly et al. (2010b) reported that RFI scores were repeatable (r = 0.62) in finishing heifers fed a grower and then finisher diet and Durunna et al. (2012) reported a significant rank correlation of 0.52 for RFI values between replacement heifers fed the same roughage-based diet in 2 different periods. Durunna et al. (2011a) also reported significant Spearman correlations by steers fed the same diet in 2 different periods (0.42 for finisher–finisher and 0.44 for grower–grower) or different diets in 2 periods (0.33 for grower–finisher). Correlations were slightly greater when cattle were fed the same diet in both periods (r = 0.42 to 0.52) than when steers were fed different diets in periods 1 and period 2 (r = 0.33; Durunna et al., 2011a, 2012). All these studies in cattle reported moderate values for correlations and repeatability of RFI. The lack of correlation of RFI values between periods and diets in the Redden et al. (2011) study could have been due to diets that were more dissimilar in form and energy content (i.e., pelleted grower diet versus forage diet) than diets used in the cattle studies. Data from these studies suggest that RFI traits are expressed in a similar but not identical manner on different diets.

Dry matter intake by yearling ewes fed pelleted alfalfa was 59% greater (P < 0.01) than by yearling ewes fed chopped alfalfa and explains greater ADG, G:F, and final LMA and BF of yearling ewes fed pelleted versus chopped alfalfa hay. Heaney et al. (1963) reported that sheep fed pelleted alfalfa hay had greater intake of digestible energy compared to sheep fed chopped alfalfa and Meyer et al. (1959) attributed increased ADG in sheep fed pelleted versus chopped alfalfa to increased feed intake from a faster rate of passage. In the current study, there were no alfalfa processing × RFI group interactions (P > 0.14); however, it is interesting to note that within each alfalfa treatment, numerical differences in DMI between RFI groups were greater for the pelleted than chopped alfalfa diet (0.9 and 0.4 kg, respectively) and the range of RFI values was also greater for the pelleted than chopped alfalfa diet (0.76 and 0.30 kg/d, respectively). There also seemed to be a greater increase in BF by high versus low RFI yearling ewes fed chopped alfalfa (0.42 versus 0.52 cm) than between high and low RFI yearling ewes fed the pelleted alfalfa (0.72 to 0.74). Lack of statistical differences for these traits could be due to low numbers of animals and perhaps these observations warrant further investigation. Data from the current study suggests that ewe efficiency as determined by RFI is repeatable on subsequent forage based diets; however, differences in intake and efficiency were more apparent when alfalfa was pelleted.

In summary, we observed a 22% decrease in DMI between low and high RFI yearlings that were RFI tested as ewe lambs. Method of hay processing had a large influence on yearling performance and difference in intake between RFI groups was more apparent on the higher DMI feed. Differences were not detected between RFI groups for growth performance or feed conversion; however, high RFI yearlings tended to deposit more fat than low RFI yearlings. We hypothesize that high RFI ewes have physiological differences from low RFI ewes that control intake. In Exp. 1, no measureable differences were detected for lean or fat deposition between RFI groups; however, additional calories consumed by high RFI yearlings were either partitioned towards fat deposition or used to maintain weight instead of using fat resources for energy. This research indicates that RFI testing does predict differences in yearling intake and warrants further investigation into mechanisms that control intake and how these different mechanisms affect overall productivity of sheep.

 

References

Footnotes


Comments
Be the first to comment.



Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.