CASE STUDY: Reproductive Performance of Beef Cows Fed Whole Soybeans Before the Breeding Interval

INTRODUCTION

Supplying energy and protein to cows after calving to meet nutritional requirements has a great influence on reproductive performance of the cow herd (Wiltbank et al., 1964; Spitzer et al., 1995; Wetteman et al., 2003). Postpartum reproduction is of great economic importance to the cow- calf producer because of the invest- ment made in producing replacement females and maintaining mature cows (Bellows et al., 2002). This is especial- ly true for heifers entering their first breeding season and for primiparous and multiparous cows that may have difficulty in rebreeding during a de- fined breeding season (Ciccioli et al., 2003). Fat supplementation after calv- ing increased first service conception rate (Bader et al., 2000), pregnancy rates (De Fries et al., 1998), and the number of cows exhibiting estrus within 60 d after calving (Webb et al., 2001). Feeding heifers corn gluten feed with a rumen-protected fat source for 60 d before the breeding interval resulted in a trend for improved preg- nancy rates (Long et al., 2007). Lipids and oilseeds have been used to influ- ence postpartum reproduction. The use of safflower seed fed during the postpartum period improved subse- quent conception rates in primiparous cows (Lammoglia et al., 1997), and soybean oil improved ovarian follicu- lar growth compared with either tal- low or fish oil (Thomas et al., 1997). Fall-calving cow and calf gain performance were improved when cows were fed drought-stressed soybeans with hay, but pregnancy rates were not improved by supplementation (Steele et al., 2007). Whole raw soybeans were fed to beef heifers (Howlett et al., 2003; Harris et al., 2008), resulting in no differences in AI conception rates compared with heifers fed diets with other protein and energy supplements. Banta et al. (2008) fed whole soybean or soybean meal-hull diets to beef cows and reported reduced BW gain during mid to late gestation and inconsistent effects of diets on reproductive performance. The present research was designed to investigate the effects of feeding whole raw soybeans to grazing beef cows before spring breeding on cow reproductive and cow and calf gain performance.

MATERIALS AND METHODS

Animals and Experimental Procedures

Reproductive and gain performance resulting from feeding raw soybeans to cows on pasture were determined using primiparous (PC; 508.8 ± 35.8 kg BW) and multiparous (MC; 581.4 ± 66.6 kg BW) beef cows in a single year. The cows were purebred Angus (PC, n = 17; MC, n = 36) and Polled Hereford (PC, n = 12; MC, n = 11). These cows calved from December 30 to February 24. On d 1 (February 11) all cows that had calved to this point were stratified by parity, breed, cow and calf initial BW, and calf age before cow-calf pairs were randomly assigned to either control or soybean supplement treatments, with one treatment group for each of 4 pas- tures. On d 10 (February 24) addi- tional cows (MC = 12; PC = 7) that had calved between d 1 and 10 were added to the study. They were ranked in the same manner as the original cows and randomly assigned to the treatment groups. The control cows (PC = 14, MC = 25) were assigned to pastures 1 and 3, respectively, with no supplemental feeding, except hay. The soybean supplemented cows (PC = 14, MC = 24) were assigned to pastures 2 and 4, respectively, and they were fed cracked whole soybeans as a supplement. Treatments were nonreplicated and intake of cracked whole soybean was assumed to be equal for all animals on the soybean treatment. The soybeans, containing 23.1% crude fat (Table 1), were fed at 2.12 kg DM/cow daily for 70 d, immediately before and at the start of breeding season. Both groups were as- signed to dormant Coastal bermuda- grass (cv. Coastal, Cynodon dactylon) pastures that were sod-seeded with annual ryegrass (cv. Passerei, Lolium multiflorum Lam.; 26.7 kg/ha; Pen- nington Seed Co., Madison, GA). Cows and calves on pastures 1 and 2 and pastures 3 and 4 were rotated twice to minimize any pasture effects. Bermudagrass hay (Table 1) was provided free-choice in hay feeders to each group, and hay disappearance was calculated. All cattle were managed under procedures approved by the University of Georgia Animal Care and Use Committee Guidelines (IACUC AUP No. A2005-10179-C1; Univ. of Georgia Animal Welfare Assurance No. A3437-01).

On d -68 (December 4), d 1 (February 11), d 28 (March 11), and d 60 (April 11), cows and calves were weighed, cow ultrasound subcutaneous fat depth was measured at the 12th rib, and a BCS (scale 1 to 9; 1 = emaciated, 9 = obese) was assigned to each cow. On d 1, 28, and 60, calf BW were recorded, and a blood sample was taken from each cow by tail venipuncture. The same procedures were followed for the cow-calf pairs that calved later than the original group of cows when they entered the experiment on d 10. Cow BW was recorded, BCS were assigned to cows, and ultrasound fat depth measurements were recorded for cows on June 16. Additional cow and calf BW were recorded at 28-d intervals from June 16 to September 19 when the last cow BW were recorded, and calves were weaned on September 8 (d 210). Calf weaning weights were adjusted for age of dam, calf sex, and to 205 d of age using standard Beef Improvement Federation (2006) adjustments.

The breeding interval began on d 63 (April 14), consisting of a 44-d AI interval followed by a 27-d natural service interval using mature fertile beef bulls. AU cows regardless of prebreeding treatment or parity were combined into one group on d 77 and remained in this group throughout the AI interval. Cows were monitored for estrual activity by visual observation for 30 min twice daily at 12-h intervals. A Heatwatch System (DDX Inc., Denver, CO) was used to detect estrus, and cows were bred 12 h after the onset of estrus, defined as 2 mounts in 4 h detected by the Heatwatch system. If cows were visually detected in estrus, and there was no indication of estrus by Heatwatch observations, these cows were bred 12 h after initial visual observation. At the end of the AI interval, cows were separated by breed, and Polled Hereford cows were exposed to a mature Angus bull, and Angus cows were exposed to a mature Polled Hereford bull from May 28 to June 24; bulls had passed a breeding soundness examination before assignment to cow herds. On September 17 pregnancy examinations via rectal palpation were completed for each cow, and the approximate date of conception was determined by aging of the fetus by an experienced technician.

Chemical Analyses of Diets and Blood Components

Samples of all major dietary components were chemically analyzed for DM and CP (AOAC, 1990). The ADF and NDF were determined using the methods outlined in Van Soest et al. (1991). The TDN and NEm were determined using equations in NRC (1996). Three randomly selected round bales of hay were sampled in 3 different locations of each bale using a core sampling tube, and then ground (1-mm screen), composited samples for each bale were chemically analyzed (Table 1). Soybeans were sampled from several sites in the load and analyzed in duplicate (Table 1). Ryegrass forage was sampled by cutting at ground level on d 18 and 48 of the supplementation period at 3 different locations within each pasture. After drying (60°C for 48 h), samples were ground through a 1-mm screen, composited by pasture, and duplicates were chemically analyzed (Table 1).

All blood samples were refrigerated immediately after collection and cooled overnight at 4°C before being centrifuged (2,860 × g for 20 min at 4°C) to separate serum. Serum was collected, frozen, and stored for later analysis. A Boehringer Mannhiem/ Hitachi 912 analyzer was used to analyze serum for cholesterol, highdensity lipoprotein, and triglycerides (Roche Diagnostics, Indianapolis, IN). The low-density lipoprotein was calculated using the formula: total cholesterol - [high-density lipoprotein × (triglycerides/5)] . Serum leptin was determined using an ovine leptin RIA that had been validated for bovine serum (Delavaud et al., 2000). Each sample was analyzed in triplicate and leptin was reported as the average of the triplicate.

Statistical Analyses

The cow data in the experiment were statistically analyzed using PROC MIXED (SAS Institute, 2002). All analyses began with full models including treatment, parity, cow breed, interaction (treatment × parity), and the appropriate linear covariate, except for BW, BCS, and ultrasound fat depth on d -68 (December 4), which used interaction and cow breed only. The analyses were examined, and terms with an F- value less than 1.0 were removed from the analysis; a final analysis was then conducted for each variable (Carmer et al, 1969). The BW, ADG, BCS, and ultrasound fat depth measurements and reproductive data for the cows (Table 2) were analyzed as a 2 × 2 factorial with cow as the individual experimental unit as the source of error (Ciccioli et al., 2003; Lents et al., 2005; Long et al., 2007). Cow breed was removed as a class effect covariable, and interaction (treatment × parity) was left in the model when determined to be important. The cow data were centered (Table 2; Draper and Smith, 1981) as follows: pre-experiment BW on d -68, centered on 568.68 kg; initial BW on d 1, centered on 554.79 kg; initial BCS on d 1, centered on 5.171052; pre-experiment ultrasound fat thickness on d -68, centered on 0.82; initial ultrasound fat thickness on d 1, centered on 0.63. Calf birth weight, ADG, and weaning weights were analyzed as a 2 × 2 factorial (treatment and breed as factors) using PROC MIXED (SAS Institute, 2002). Least squares means for calf birth weight, calf sex (bulls = 24; heifers = 28; steers = 24), and calf age at weaning were used as covariables when appropriate, using reduced statistical models (Carmer et al., 1969). From the preliminary analyses, it was determined that when age of dam was used as a covariable replacing parity in the model, error MS was reduced for calf data. The calf data were centered (Table 3; Draper and Smith, 1981) as follows: birth weight was centered on 35.17 kg; calf age at weaning was centered on 225.53 d; and age of dam was centered on 3.24 yr. The cow age group distribution was 2 yr = 28, 3 yr = 19, 4 yr = 12, >4 yr = 17.

Additional analyses were conducted for the serum components as a 2 × 2 × 3 factorial (breed, treatment, and date as factors) using PROC MIXED (SAS Institute, 2002), with cow being the experimental unit and cow within treatment and parity as the source of error. Repeated measures effects of date of sampling were examined for each serum component variable. AU analyses began with full models including all 4 interactions (treatment × cow breed, treatment × sampling date, cow breed × sampling date, and treatment × cow breed × sampling date), using cow age as a covariable. Any effect that had an F-value less than 1.0 was removed from the model (Carmer et al., 1969), and new analyses were conducted with reduced models. The age of dam was centered on 3.24 yr, using the same value that was used in calf analyses.

RESULTS AND DISCUSSION

Nutrient Intake Effects on Cow and Calf Performance

The effects of feeding raw soybeans to cows before the breeding interval on cow BW gain, BCS, and pregnancy rates are shown in Table 2. During the supplementation interval before breeding began, cows were fed moderate quality bermudagrass hay while grazing sod-seeded ryegrass with high levels of CP and TDN (Table 1). Although treatment × parity interaction means are displayed for each variable (Table 2), the majority of the variables were unaffected by treatment × parity interactions (P > 0.10), as noted in Tables 2, 4, and 6. Cow breed was removed as a class effect covariable if it significantly affected treatment or parity. On d 1, initial BW was greater (P < 0.05; Table 2) for the cows that were assigned to the soybean treatment compared with the cows assigned to the control treatment, and MC cows were heavier than PC cows. Cow ADG during the supplementation period did not differ (P > 0.10) for treatments and parity, although younger PC cows tended to gain more than MC cows. The bermudagrass hay (Table 1) had greater CP and NDF concentrations than the average hay fed to beef cows in the southeastern United States. It was fed in addition to sod-seeded ryegrass during the supplementation period, which prob- ably reduced the differences between treatment groups that might have been observed if lower quality hay had been fed to cows. This greater quality diet probably assisted control PC cows to increase their ADG to comparable ADG of MC cows dur- ing this interval. Cow breed affected breeding interval ADG, with Angus cows losing BW, and Polled Her- eford cows gaining BW [ADG (kg) for Angus vs. Polled Hereford was -0.38 vs. 0.14, P < 0.01]. During the 73-d breeding interval following the supplementation period, cows on the soybean treatment had lower (P < 0.05) BW losses than control cows, and younger PC cows tended (P > 0.10) to lose more BW than MC cows. Calves were weaned on September 8 (d 210), and 210-d ADG of cows was unaffected (P > 0.10) by prebreeding supplementation treatments and parity. The initial cow BCS (Table 2) did not differ for the supplementation treatments, but greater (P < 0.01) BCS was observed for MC than PC, and a similar pattern of observations continued through d 128. Visual BCS declined for PC cows by d 1 of the supplementation period because many cows had already calved and were nursing calves. On d 1 and on subsequent scoring dates, soybean supplementation did not affect BCS scores (P > 0.10). Cow ultrasound fat depth increased for all cows from d 1 to 60 (Table 2). The ultrasound data supports visual BCS, with similar (P > 0.10) ultrasound fat depth for control and soybean treatments on d 1, 60, and 128. At d 128 MC tended (P < 0.12) to have greater fat depth than PC.

Pregnancy data for the cows indicated that cows on the soybean treatment tended to have lower pregnancy rates than control cows (Table 2; P= 0.19), and PC cows tended to have lower (P = 0.55) pregnancy rates than MC cows. Cow breed did not affect pregnancy rate (P = 0.27; Angus = 74.6%, Polled Hereford = 61.7%), with greater pregnancy rates for Angus than Polled Hereford cows on both treatments and in both parity groups. The pregnancy rates for Angus cows were 67 and 90%, respectively, for control treatment PC and MC cows, and 75 and 67%, respectively, for PC and MC cows fed soybeans. The Polled Hereford cows had lower pregnancy rates, with control PC and MC cows having 67 and 80% pregnancy rates, respectively; but both PC and MC Polled Hereford cows fed soybeans had pregnancy rates of only 50%, which contributed to the lower pregnancy rates for the soybean treatment. Pregnancy rate variation associated with breed of cow increased the difficulty in drawing conclusions for treatment effects. Pregnancy rates for all herds at the same research farm bred under similar conditions and breeding interval were lower than normal (approximately 70% pregnancy rate for 180 cows), and cows on this experiment fit the overall pattern observed for the Angus and Polled Hereford cows in the breeding herd that year. The relatively low numbers of cows in the experiment and of PC and MC cows on each treatment contributed to inconsistencies in cow pregnancy rates, which added to the difficulty in drawing conclusions regarding pregnancy effects of these treatments.

The estimated fetal age determined after weaning (Table 2) indicated that prebreeding treatments did not affect time of conception after the initiation of the breeding interval, but PC cows apparently conceived about 11 d later (P < 0.10) than MC cows. A treatment × parity interaction (P < 0.10) for fetal age indicated that PC and MC cows fed soybeans during the prebreeding interval conceived at about the same time, but for control cows, MC cows apparently conceived about 22 d earlier than PC cows and about 12 d earlier than either group of soybean-supplemented cows. Bottger et al. (2002) reported no improvement in days to concep- tion rates of primiparous beef heifers individually fed safflower seeds as an energy source from d 3 to 90 postpar- tum. Reproductive performance was not greatly improved when soybeans were fed to cows in the present study, and these results were supported by research in which no differences in pregnancy rates were observed in cows fed raw soybeans (Whitney et al., 2000), soybean oil (Howlett et al., 2003), drought-stressed soybeans that were rolled or fed whole (Steele et al., 2007), and whole corn germ containing 35% ether extract (Martin et al., 2005). Harris et al. (2008) fed whole soybeans, wet corn gluten feed, or corn dried distillers grains to beef heifers and reported no differences in overall dietary effects on reproductive performance of heifers. Banta et al. (2008) reported reduced BW gain of beef cows fed whole soybeans compared with soybean meal-hull supplemented diets, but effects on reproduction were inconsistent.

Calf performance data are presented for prebreeding treatment and calf breed (Table 3). Calf BW on d 1 were similar for control and soybean treatments, and calf breed did not affect d 1 calf BW (P > 0.10; Table 3). Calf weaning weights were greater than the 205-d Beef Improvement Federation (2006) adjusted weights that were adjusted to a constant calf age, but the trends were the same for both weaning weight estimates. In both cases, prebreeding treatment did not affect (P > 0.10) weaning weights, but Angus calves were heavier (P < 0.01) than Polled Hereford calves. Calf ADG were unaffected (P > 0.10) by treatment during the prebreeding interval (d 1 to 60), during the breeding interval (d 60 to 133), or from d 1 to weaning. Calf breed affected prebreed- ing ADG and total ADG from d 1 to 210 (P < 0.01), with Angus calves having consistently greater ADG than Polled Hereford calves (Table 4). Dur- ing the breeding interval, calf gains were similar (P > 0.10) for prebreed- ing control and soybean treatments, and for calf breeds. Feeding soybeans to the cows during the prebreeding interval apparently did not stimulate additional milk production for calves or cause a positive ADG carryover effect during the breeding interval or extending until weaning. This was probably the result of the greater quality pasture being grazed and the hay being fed (Table 1) to cows dur- ing the prebreeding interval, which apparently minimized effects of the additional energy and protein pro- vided in the supplemental soybeans. Alexander et al. (2002) reviewed research from several sources that indicated no effects of prepartum lipid supplementation on subsequent cow milk production. Martin et al. (2005) observed no differences in actual or adjusted weaning weights of calves whose dams were fed whole corn germ for 45 d after calving. Studies in which soybeans were fed as supple- ments to cows included the work of Steele et al. (2007), who fed soybeans with bermudagrass hay for 100 d during winter- to fall-calving cows. They reported improved calf weaning weights and increased early lactation milk production in these cows com- pared with calf and cow performance of nonsupplemented cows. Banta et al. (2008) observed no differences in calf weaning weights when cows were fed whole raw soybean or soybean-hull supplemented diets for 80 d during mid to late gestation, although calves averaged 2 kg heavier at birth for cows fed raw soybeans.

Soybean Supplementation Effects on Serum Components of Cows

To determine effects of prebreeding control and soybean treatments on serum lipid components (Table 4), blood samples were collected from each cow on d 1, 28, and 60. Serum cholesterol was similar for control and soybean supplemented cows on d 1, but at d 28 and 60 cholesterol concentrations were greater (P < 0.01) for cows fed soybeans than for control cows. By d 28, cholesterol was more than 30 mg/dL greater (P < 0.01) for soybean treatment cows than control cows, and approximately the same margin was noted at d 60. Cholesterol at d 60 was lower for PC than MC cows (P < 0.05; 188.8 vs. 166.2 mg/ dL). A treatment × parity interaction (P < 0.05) occurred for cholesterol at d 60, with soybean supplemented PC cows having cholesterol concentrations similar to MC cows, but control PC cows had substantially reduced cholesterol compared with control MC cows.

Serum triglycerides were similar for control and soybean supplemented cows on d 1 and 28, but at d 60 trig- lyceride concentrations were greater for soybean supplemented cows than for control cows (P < 0.05; Table 4), and soybean supplemented cows had greater (P < 0.01) overall triglyceride concentrations. Triglyceride concen- trations were greater (P < 0.01) for PC cows than MC cows at d 1 and at d 28 and 60 (P < 0.05). Cow breed affected triglyceride concentrations, with Angus cows having more (P < 0.01) triglycerides than Polled Hereford cows (P < 0.06; Table 5). The high-density lipoprotein concentrations were greater (P < 0.05) at d 28 and 60 for soybean supplemented cows than for control cows. At d 60 PC cows fed soybeans had elevated high-density lipoproteins compared with either PC or MC cows on the control treatment (treatment × parity, P < 0.10; Table 4). Concentrations of low-density lipoproteins were greater in soybean supplemented cows than controls at d 28 and 60 (P < 0.01; Table 4), and the overall mean low-density lipoprotein concentrations tended to increase (P < 0.10) over time for all cows. As with cholesterol and high-density lipoproteins, at d 60 low-density lipoproteins had a treatment × parity interaction (P < 0.01), with soybean supplemented PC cows having substantially greater low-density lipoproteins than either control PC or MC cows.

Serum leptin was affected by a treatment × parity × date interaction (P < 0.01; Table 6). Serum leptin increased (P < 0.05) over time for PC and MC cows fed soybeans and for control MC cows, but it declined for control PC cows (Table 6). Serum leptin was greater for PC than MC on d 1 for control and soybean treatments (P < 0.05). On d 60, leptin was greater for MC than PC cows on the control treatment (P < 0.05; 9.0 vs. 5.4 ng/mL), but within the soybean treatment, PC and MC cows had similar leptin concentrations (P > 0.10; 8.4 vs. 8.6 ng/mL). Additionally, at d 60, leptin was greater for soybean supplemented PC cows than for control PC cows (P < 0.05; 8.4 vs. 5.4 ng/mL). Cow breed affected serum leptin concentrations (Table 5), with greater leptin concentrations (P < 0.01) in Angus than Polled Hereford cows. Therefore, feeding soybeans to cows during the prebreeding interval increased leptin during the 60-d supplementation period, and PC cows had increased leptin, bringing concentrations in serum of these cows to the levels observed in MC cows. Ciccioli et al. (2003) reported an increase in leptin resulting from feeding a high nutrient level during the postpartum period in primiparous cows. Using mature beef cows, Lents et al. (2005) concluded that amount of nutrient intake had a greater effect than body energy reserves on insulin and leptin concentrations in plasma of gestating beef cows. Long et al. (2007) fed a supplement containing rumenprotected fat and corn gluten feed to heifers on pasture before the breeding interval, which improved reproductive performance and increased serum lipids, including leptin.

The serum lipids in cows were affected by feeding soybeans during the prebreeding interval. Cholesterol, triglycerides, and low- and highdensity lipoproteins were all elevated by d 60 for the soybean treatment (Table 4). Serum lipid increased in response to the supplemental fat contained in the raw soybeans. These results are supported by Whitney et al. (2000), who reported an increase in cholesterol concentrations in cows when feeding soybean oil at 3 and 6% of a supplement that was fed at 2.87% of BW. Thomas et al. (1997) reported an increase in total cholesterol, high-density lipoproteins, and triglycerides when soybean oil was fed at 4% of DMI as part of a complete diet. A new observation in our study indicated that cows of different parities responded differently to soybean supplementation with different serum lipid concentrations. The MC cows had greater concentrations of cholesterol and high-density lipoproteins compared with PC cows in the same treatment groups. The opposite was true for triglycerides, with PC cows having greater concentrations than MC cows. Parity had no affect on low-density lipoproteins.

IMPLICATIONS

The feeding of soybeans as a postpartum supplement is a relatively new idea, and there are few published results supporting or dismissing the practice as a feasible option for improving postpartum reproduction. Feeding soybeans increased the concentrations of serum cholesterol, triglycerides, and high- and lowdensity lipoproteins. The increased serum component concentrations occurred without affecting cow BW, BCS, or ultrasound fat measurements. However, pregnancy rates were not consistently improved in cows on the soybean treatment, which is the most important economic factor to cow-calf producers. Feeding whole soybeans as a short-term postpartum supplement effectively delivered high levels of fat and CP to the cows, but additional research is needed to determine if the practice can reliably improve reproductive performance in beef cows.

© 2008 American Registry of Professional Animal Scientists Provided by ProQuest LLC. All Rights Reserved.