Abstract
The effect of increasing the feed level (1.8, 2.5, and 3.2 kg/d) during early gestation in parity 1 (PO1) and parity 2 (PO2) sows on maternal growth and litter size was evaluated. A total of 361 sows were group-housed and fed a diet based on a corn–soybean meal (3.15 Mcal Metabolizable Energy (ME) per kg and 0.68% standardized ileal digestible lysine) from day 6 of gestation until day 30. Sows were weighed at weaning and on day 30 of gestation. Farrowing rate, number of total piglets born, piglets born alive, stillborn piglets, and mummified fetuses were recorded at farrowing. The effect of feed level on the total number of piglets born was also evaluated according to classes of body weight (BW), body condition score (BCS), backfat (BF), and caliper unit at weaning for each parity order. There was no evidence for significant effect of the interaction between feed level × parity on the variables related to maternal growth and reproductive performance (P ≥ 0.128). Greater feed levels linearly increased the gains in body condition (i.e., BW, BCS, BF, and caliper unit) between weaning and day 30 of gestation (P < 0.001). Farrowing rate was not influenced by the feed level (P ≥ 0.200) and parity (P ≥ 0.209). The number of total piglets born decreased linearly as the feed level increased (P = 0.041), whereas no evidences for differences were observed on piglets born alive among treatments (P ≥ 0.317), neither between parities (P ≥ 0.904). For PO1 sows, the total piglets born on BW classes (≤183 vs. >183 kg) and on classes of BF (≤11.5 vs. > 11.5 mm) were quadratically affected by the feed levels (P ≤ 0.041). In contrast, the number of total piglets born was marginally affected (linear; P ≤ 0.094) by the feed level in the different classes of BW, BCS, and caliper unit in PO2 sows. There was no evidence for differences for the interaction of feed level and classes of body condition (P ≥ 0.199) for PO2 sows. Similarly, no interactions between feed level and classes of BW, BCS, and BF at weaning were observed (P ≥ 0.233) for PO1 sows; however, the total piglets born were affected by an interaction between feed level and caliper unit class (P = 0.042). In conclusion, increased feed intake from day 6 of gestation until day 30 resulted in increased maternal BW gain but reduced the number of total piglets born. Furthermore, lighter and in a poor BCS PO1 sows at weaning produced fewer total born piglets with no benefits from greater amounts of feed.
Keywords: earlier gestation, feeding, reproduction, total born, young sows
Introduction
Young parity sows are sensitive to body weight (BW) changes during lactation and more prone to subsequent reproductive impairments. The period of early and mid-gestation is commonly used to recover body reserves, especially for young parity sows (NRC, 2012). Lactation performance, and mostly the voluntary feed intake, is affected by body condition at farrowing (Kim et al., 2013; Mallmann et al., 2019). The subsequent reproductive performance may be affected by body losses during lactation and body condition at weaning (Schenkel et al., 2010).
Feeding strategies post-insemination have been discussed over the years, mostly due to their potential impact on reproductive performance (Leal et al., 2019). Early gestation is a critical period for embryo development (Geisert and Schmitt, 2002) because nutritional management and metabolic changes can affect embryo survival and litter size (Langendijk et al., 2017). In a study performed by Jindal et al. (1996), embryo survival in gilts was negatively affected (84.7% vs. 64.5%) by increasing the feed allowance (1.9 vs. 2.6 kg/d) soon after insemination and up to day 15 of gestation. Conversely, Quesnel et al. (2010) reported that increasing feed allowance (2.0 vs. 4.0 kg/d) for gilts, between insemination and day 7 of gestation, did not affect embryo survival (87% vs. 84%). Consistent with these latter results, the litter size in parity 1 (PO1) and 2 (PO2) sows increased by two embryos when the feed amount was increased from 2.5 to 3.25 kg/d between days 3 and 32 of gestation (Hoving et al., 2011).
It has been reported that feed restriction after insemination should not be recommended in gilts, whereas higher feed amounts appear to benefit young parity sows of contemporary genotypes (Leal et al., 2019). However, most of the studies were performed with gilts under experimental conditions that used a limited number of females per group. Young weaned sows experience lactation catabolism, and thus different approaches are necessary to recover their body condition and maintain the pregnancy. However, how the interaction between feeding strategy used after breeding and the recovery of body reserves affects the subsequent performance of swine females remains unclear. Therefore, understanding the female response according to its body condition at weaning will be necessary to extrapolate the results for practical use in swine systems.
The hypothesis of the present study is that greater amounts of feed earlier in gestation may benefit young parity sows or sows with a poor body condition at weaning. This study aimed to evaluate the effect of three different feed levels (1.8, 2.5, and 3.2 kg/d), offered between days 6 and 30 of gestation, on maternal growth, farrowing rate, and litter size of PO1 and PO2 sows.
Materials and Methods
The protocol used in the present study was approved by the Ethics Committee of Animal Utilization (CEUA) of the Federal University of Rio Grande do Sul (UFRGS), under Process no. 32657.
Location
The study was conducted in a herd with 5,500 swine females, located in the Midwest of Paraná State (24°55′04″S, 50°05′50″W), Brazil, between January and April (average, minimum, and maximum temperature were 23.5, 16.1, and 34.0 °C, respectively, with 85.3% relative humidity). The two periods correspond to summer and early autumn, respectively, in the southern hemisphere.
Housing and feeding
The sows were individually housed in stalls (2.2 × 0.7 m) during the lactation period and weaning-to-estrus interval (WEI). During both phases, the sows were fed a corn–soybean meal diet with 3.45 Mcal Metabolizable Energy (ME)/kg, 20.0% crude protein (CP), and 1.10% standardized ileal digestible lysine (SID Lys). Feed was provided ad libitum during lactation; during the WEI, a total amount of 3.5 kg/d was provided in four meals. The sows were moved to gestation pens at 0.8 ± 0.1 d (0 to 3 d) after the last insemination and housed in static groups during gestation (around 70 females per pen). Pens with 140 m2 provided 2.0 m2 per sow and were equipped with one electronic feeding station (ESF; SowComp, WEDA Dammann & Westerkamp GmbH, Germany). Feed intake during the treatment period was recorded daily by the ESF system. Ad libitum access to water was provided throughout the experiment.
Feed content was calculated monthly based on analyses of ingredients. Dietary samples were collected every 2 wk for 4 mo and analyzed in triplicate for CP, total Aminoacids, and dry matter (AOAC International, 2012). Samples were also analyzed for crude fiber, ash, ether extract, calcium, and phosphorus (AOAC International, 2012; CBAA, 2017).
Experimental and treatment design
A total of 361 sows (PO1 and PO2; PIC Camborough, Hendersonville, TN, Landrace × Large White crossbred) were selected at weaning according to general health status and body condition score (BCS) between 2.0 and 4.5 (1 to 5 scale; Young et al., 2004). After weaning, the sows were checked for estrus once a day (0900 hours) by the back-pressure test in the presence of a mature boar. The first insemination was performed at estrous onset and repeated at 24-h intervals during estrus. Each female received 2.2 ± 0.1 semen doses (1.5 × 109 sperm cells; total volume 50 mL).
The day of the first insemination was considered day 0 of gestation. From day 0 until day 5, and from day 31 to farrowing, all sows were fed 1.8 kg/d of a corn–soybean meal diet with 3.15 Mcal ME/kg, 15.0% CP, and 0.68% SID Lys (Table 1). The analyzed diet was considered consistent with formulated values based on analytic variability (Table 2). The selected sows were uniformly distributed, according to BW at weaning, number of piglets born in previous farrowing, weaned piglets, and days of WEI, into the following treatments: 1) 1.8 kg/d (5.7 Mcal ME and 12.2 g/d SID Lys; 2) 2.5 kg/d (7.9 Mcal ME/d and 17.0 g/d SID Lys); and 3) 3.2 kg/d (10.1 Mcal/kg and 21.8 g/d SID Lys). The different feed levels were provided between days 6 and 30 of gestation. The sows were moved to farrowing rooms on day 110 of gestation.
Table 1.
Composition of the experimental diets (as-fed basis)
| Ingredient | Gestation1 |
|---|---|
| Corn | 49.6 |
| Oats | 14.0 |
| Soybean meal, 46 % CP | 11.1 |
| Wheat, 14.5% CP | 10.0 |
| Rice bran defatted, 16% CP | 5.0 |
| Corn dried distillers grains with solubles, 6% fat | 5.0 |
| Molasses | 1.2 |
| Vitamin and mineral premix2 | 0.2 |
| Monocalcium phosphate | 0.35 |
| Limestone | 1.02 |
| Salt | 0.25 |
| Sodium bicarbonate | 0.30 |
| l-Lys | 0.29 |
| dl-Met | 0.02 |
| l-Thr | 0.09 |
| Soybean oil | 1.2 |
| Phytase3 | 0.02 |
| Others | 0.36 |
| Total | 100.00 |
| Calculated analysis | |
| SID Lys, % | 0.68 |
| SID Met: Lys, % | 37 |
| SID Met and Cys: Lys, % | 78 |
| SID Thr: Lys, % | 81 |
| SID Trp: Lys, % | 23 |
| SID Val: Lys, % | 110 |
| CP, % | 15.03 |
| Ca, % | 0.73 |
| STTD P4, % | 0.40 |
| Na, % | 0.25 |
| Cl, % | 0.20 |
| ME, Mcal/kg | 3,151 |
1Diet was fed from day 6 to 30 of gestation.
2Vitamin composition per kg of diet: vitamin A: 12,500 IU; vitamin D3: 2,500 IU; vitamin E: 125.0 IU; vitamin K3: 4.5 mg; vitamin B1: 2.5 mg; riboflavin (B2): 7.5 mg; pyridoxine (B6): 3.5 mg; vitamin B12: 33.8 µg; niacin: 50.0 mg; pantothenic acid: 25.0 mg; folic acid: 2.4 mg; biotin: 0.26 mg; choline: 1.25 g. Mineral composition: selenium: 0.64 mg; iron: 75.0 mg; copper: 21.7 mg; manganese: 61.4 mg; zinc: 183.4 mg; iodine: 1.5 mg.
3Aela (Auster Animal Nutrition, São Paulo, Brazil) provided 1,000 phytase units per kg of diet with release of 0.19% STTD P.
4STTD, standardized total tract digestible.
Table 2.
Chemical analysis of the diets (as-fed basis)
| Ingredient | Gestation1 |
|---|---|
| Proximate analysis, % | |
| DM | 88.55 (87.78) |
| CP | 14.89 (15.03) |
| Crude Fiber | 5.23 (3.85) |
| Fat | 3.38 (4.38) |
| Ash | 5.87 (4.87) |
| Ca | 0.99 (0.73) |
| P | 0.64 (0.40) |
| Total AA, % | |
| Lys | 0.77 (0.78) |
| Ile | 0.71 (0.58) |
| Leu | 1.29 (1.37) |
| Met | 0.26 (0.28) |
| Met and Cys | 0.56 (0.57) |
| Thr | 0.64 (0.63) |
| Trp | 0.15 (0.16) |
| Val | 0.67 (0.75) |
| His | 0.29 (0.40) |
| Phe | 0.82 (0.64) |
1Values in parentheses indicate those calculated from diet formulation and are based on the values from NRC (2012).
Measurements
All sows were weighed and evaluated for BCS, backfat (BF), and caliper unit on the first day after weaning and at day 30 of gestation. The weight was measured with a 500 g precision scale (EW6, Tru Test, Auckland, New Zealand). The BF measurement was performed on the P2 point (6.5 cm away from the midline of the vertebral column at the last rib level) with an A-mode ultrasound (Renco Lean Meter—Renco Corporation, Minneapolis, MN) within a range of 2 mm. Caliper unit was measured on the same BF point with the caliper equipment in a unit range from 1 to 25 (Knauer and Baitinger, 2015). Reproductive performance criteria were recorded using Agriness S2 software (Agriness, Santa Catarina, Brazil). The following reproductive responses were collected: farrowing rate, number of total piglets born, piglets born alive, stillborn piglets, and mummified fetuses.
Statistical analysis
Statistical analysis system software, version 9.3 (SAS Institute Inc., Cary, NC), was used to perform the statistical analysis. All models included feed level as a fixed effect. The week of feed treatment onset was included as a random effect. Polynomial contrasts were used to evaluate the linear and quadratic effects of the dose–response (different feed levels offered daily), parity, and their interactions.
The following variables were analyzed using the GLIMMIX procedure, fitted assuming a normal distribution: BW, BF, caliper unit at different periods and the respective gains and losses, and the number of total piglets born and piglets born alive. BCS and respective gains or losses were evaluated as a multinomial distribution using the GLIMMIX procedure. The farrowing rate was analyzed as a binary distribution using the GLIMMIX procedure. The total number of piglets born alive, percentage of stillborn, mummified fetuses, and the sum of piglets born alive and stillborn piglets were analyzed as binomial distribution using the GLIMMIX procedure.
The effects of the different feed levels on the number of total piglets born and farrowing rate were also evaluated considering different classes of BW, BCS, BF, and caliper unit at weaning, based on the median of these variables. For PO1 sows, classes of body measures at weaning were as follows: BW (≤183 and >183 kg), BCS (≤3 and >3), BF (≤11.5 and >11.5 mm), and caliper unit (≤11 and >11). For PO2 sows, classes of body measures at weaning were as follows: BW (≤208 and >208 kg), BCS (≤3 and >3), BF (≤11 and >11 mm), and caliper unit (≤10 and >10). These analyses were performed mainly to investigate whether feed levels used after breeding and body reserves at weaning would interact to affect the subsequent performance of swine females. Polynomial contrasts were used to evaluate the linear and quadratic effects of the dose–response (different amounts of feed offered daily), the different classes of body condition, and their interactions.
Each female was considered to be an experimental unit in all the analyses. The results were considered significant at P ≤ 0.05; a tendency was considered when 0.05 < P ≤ 0.10. The means were compared using the Tukey–Kramer test.
Results
There was no evidence for differences among treatments at the beginning of the experiment (P ≥ 0.570) for the total number of piglets born from previous farrowing (13.4 ± 0.2), number of weaned piglets (12.3 ± 0.1), and WEI (4.7 ± 0.1 d), respectively.
Influence of feed levels and parity on maternal growth
BW, BCS, BF, and caliper unit at weaning were not different among feed levels (Table 3; P ≥ 0.488), and no evidence for significance was observed for the interaction feed level × parity (P ≥ 0.161). However, there was an expected effect of parity (P < 0.001) on BW at weaning, where PO2 sows were 24.9 kg heavier than PO1. Parity was not evidenced to influence (P ≥ 0.137) the other variables of body condition at weaning (BCS, BF, and Caliper).
Table 3.
Least square means estimates and probability values of the effects of feed levels from day 6 to 30 of gestation on maternal BW and reproductive performance of PO1 and PO2 sows under commercial conditions1
| Item | Feed level, kg/d2 | SEM | PO1 | PO2 | Probability, P-value 3 | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1.8 | 2.5 | 3.2 | L | Q | PO | L × PO | Q × PO | ||||
| n = 122 | n = 122 | n = 117 | |||||||||
| BW, kg | |||||||||||
| Weaning | 197.2 | 197.4 | 197.1 | 2.2 | 184.8 | 209.7 | 0.983 | 0.902 | <0.001 | 0.713 | 0.715 |
| Day 30 | 192.8 | 200.4 | 206.0 | 2.0 | 189.8 | 209.7 | <0.001 | 0.584 | <0.001 | 0.583 | 0.563 |
| Weight change, kg | −4.3 | 2.8 | 9.4 | 1.6 | 5.1 | 0.2 | <0.001 | 0.775 | <0.001 | 0.818 | 0.753 |
| BCS | |||||||||||
| Weaning | 3.1 | 3.1 | 3.1 | 0.03 | 3.2 | 3.1 | 0.665 | 0.854 | 0.228 | 0.859 | 0.161 |
| Day 30 | 3.3 | 3.3 | 3.5 | 0.04 | 3.4 | 3.3 | <0.001 | 0.381 | 0.644 | 0.385 | 0.542 |
| BCS change | 0.1 | 0.2 | 0.3 | 0.03 | 0.2 | 0.2 | <0.001 | 0.354 | 0.589 | 0.502 | 0.603 |
| BF, mm | |||||||||||
| Weaning | 11.4 | 11.3 | 11.2 | 0.3 | 11.6 | 11.2 | 0.652 | 0.510 | 0.137 | 0.739 | 0.439 |
| Day 30 | 13.0 | 13.1 | 14.3 | 0.5 | 13.8 | 13.1 | 0.003 | 0.109 | 0.058 | 0.739 | 0.507 |
| BF change, mm | 1.2 | 1.4 | 2.3 | 0.4 | 1.8 | 1.5 | <0.001 | 0.099 | 0.223 | 0.668 | 0.962 |
| Caliper unit | |||||||||||
| Weaning | 11.0 | 10.8 | 10.8 | 0.3 | 10.9 | 10.9 | 0.488 | 0.918 | 0.992 | 0.471 | 0.472 |
| Day 30 | 12.4 | 13.0 | 13.6 | 0.2 | 13.1 | 12.9 | <0.001 | 0.977 | 0.158 | 0.747 | 0.790 |
| Caliper unit change | 1.3 | 2.1 | 2.8 | 0.3 | 2.1 | 2.0 | <0.001 | 0.967 | 0.389 | 0.143 | 0.502 |
| Reproductive performance4 | |||||||||||
| Farrowing rate, % | 87.6 | 88.4 | 81.4 | 3.0 | 83.5 | 88.3 | 0.200 | 0.389 | 0.209 | 0.365 | 0.368 |
| Total piglets born, n | 14.6 | 14.6 | 13.5 | 0.5 | 13.4 | 15.1 | 0.041 | 0.215 | 0.001 | 0.780 | 0.128 |
| Total piglets born alive, % (n) | 92.0 (13.4) | 91.5 (13.4) | 92.8 (12.5) | 1.0 | 92.0 (12.3) | 92.1 (13.9) | 0.467 | 0.317 | 0.904 | 0.864 | 0.640 |
| Born alive + stillborn, % (n) | 97.6 (14.2) | 97.1 (14.2) | 97.0 (13.1) | 0.5 | 97.0 (13.0) | 97.4 (14.7) | 0.349 | 0.730 | 0.404 | 0.512 | 0.634 |
| Stillborn, % | 5.6 | 5.6 | 4.3 | 0.8 | 5.0 | 5.2 | 0.106 | 0.343 | 0.681 | 0.776 | 0.302 |
| Mummified fetuses, % | 2.4 | 3.0 | 3.0 | 0.5 | 3.0 | 2.6 | 0.349 | 0.730 | 0.404 | 0.511 | 0.634 |
1A total of 361 females (Landrace × Large White) were used, with 122, 122, and 117 females for the treatments 1.8, 2.5, and 3.2 kg/d, respectively.
2Feed levels: 1.8, 2.5, and 3.2 kg/d from day 6 to day 30 of gestation.
3 P-values for linear (L) and quadratic (Q) effect of feed level, parity (PO), and their interactions.
4Reproductive performance is related to the farrowing of the same cycle that the treatments were performed.
On day 30 of gestation, BW, BCS, BF, and caliper unit increased linearly as the feed level increased (Table 3; P ≤ 0.003). No evidence for differences in the BCS and caliper unit (P ≥ 0.158) between parities was observed; however, the PO2 sows were 19.9 kg heavier (P < 0.001) and showed a slightly lower BF (P = 0.058) at day 30 of gestation, compared with PO1 sows. There was no evidence for the interaction feed level × parity and for BW, BCS, BF, and caliper unit at day 30 (P ≥ 0.385). Greater feed levels linearly increased body gains (i.e., BW, BCS, BF, and caliper unit gains) between weaning and day 30 of gestation (P < 0.001). PO1 sows gained 4.9 kg more weight than PO2 sows (P < 0.001) but no evidence for differences was observed (P ≥ 0.223) in changes of BCS, BF, and caliper unit between weaning and day 30 of gestation. The body gains were not affected by the interaction feed level × parity (P ≥ 0.143).
Influence of feed levels and parity on farrowing rate and litter size
Farrowing rate did not differ with the feed level (Table 3; P ≥ 0.200) and between parities (P = 0.209). The number of total piglets born decreased linearly as the feed level increased (P = 0.041). PO2 sows had 1.7 more total piglets born (P = 0.001) compared with PO1 sows (Table 3). Piglets born alive, sum of piglets born alive and stillborn piglets, stillborn piglets, and mummified fetuses were not affected by the feed levels (P ≥ 0.106), neither by the parity order (P ≥ 0.404). No evidences for the significant effect of interaction feed level × parity were observed for all reproductive responses (P ≥ 0.128).
Influence of body condition at weaning and their interaction with feed levels on total piglets born
For PO1 sows, there was a significant feed level × Caliper class interaction (P = 0.042; Table 4) for the number of total piglets born. For females with ≤11 caliper units, the total piglets born were quadratically affected, with the highest value observed in 2.5 kg/d level; while for females with >11 caliper units, no evidence for difference was observed. Regarding the other body condition classes, the two-way interactions were not evidenced to be significant (P ≥ 0.233). However, feed level had quadratic effect in the models considering BW class (P = 0.033) and BF class (P = 0.041), whereas in the model considering BCS class, only a marginally effect of feed level was observed (P = 0.062; Table 4). The total number of piglets born was affected by the classes of BW and BCS at weaning in PO1 sows (P ≤ 0.023). Lighter PO1 sows (≤183 kg) and in a poor BCS (≤3 kg) at weaning have produced 1.6 and 1.5 fewer piglets than heavier sows (>183 kg) or in a better BCS (>3), respectively. A similar response was observed for females with >11.5 mm of BF, which tended to produce 0.8 more piglets than those with ≤11.5 mm of BF (P = 0.075).
Table 4.
Total piglets born for PO1 sows submitted to different feed levels from day 6 to 30 of gestation according to BC classes at weaning1
| Class variables at weaning | n | Feed level, kg/d2 | SEM | BC class | Probability, P-value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1.8 | 2.5 | 3.2 | L | Q | BC Class3 | L × BC Class | Q × BC Class | ||||
| BC classes | |||||||||||
| BW, kg | 0.190 | 0.033 | 0.017 | 0.774 | 0.779 | ||||||
| ≤183 | 84 | 12.7 | 13.4 | 11.5 | 0.9 | 12.5 ± 0.6 | |||||
| >183 | 91 | 14.0 | 15.2 | 13.2 | 0.9 | 14.1 ± 0.6 | |||||
| BCS | 0.224 | 0.062 | 0.023 | 0.963 | 0.325 | ||||||
| ≤3 | 85 | 12.4 | 14.0 | 11.5 | 0.9 | 12.6 ± 0.6 | |||||
| >3 | 90 | 14.4 | 14.5 | 13.4 | 0.9 | 14.1 ± 0.6 | |||||
| BF, mm | 0.150 | 0.041 | 0.075 | 0.233 | 0.236 | ||||||
| ≤11.5 | 87 | 13.0 | 14.2 | 10.9 | 0.9 | 12.7 ± 0.6 | |||||
| >11.5 | 88 | 13.8 | 14.3 | 13.6 | 0.9 | 13.9 ± 0.5 | |||||
| Caliper | 0.244 | 0.070 | 0.053 | 0.578 | 0.042 | ||||||
| ≤11 | 93 | 12.2 | 14.5 | 11.7 | 0.8 | 12.8 ± 0.5 | |||||
| >11 | 82 | 14.8 | 13.9 | 13.4 | 0.9 | 14.1 ± 0.6 |
1A total of 175 PO1 females (Landrace × Large White) were used, divided into different body classes at weaning to be fed on 1.8, 2.5, and 3.2 kg/d from day 6 to 30 of gestation.
2Feed level: 1.8, 2.5, and 3.2 kg/d from day 6 to day 30 of gestation.
3Different classes for body condition (BC; BW, BCS, BF, and caliper unit).
In PO2 sows, there was no evidence for significant interaction of feed level × classes of body condition (P ≥ 0.199; Table 5), neither for the isolated effect of classes of body condition (P ≥ 0.417) on the number of total piglets born. However, the number of total piglets born was marginally affected (linear; P ≤ 0.094) by the feed level in the different models considering the classes of BW, BCS, and caliper unit (Table 5). No evidences for differences of the feed levels were observed in models with BF class (P = 0.108).
Table 5.
Total piglets born for PO2 sows submitted to different feed levels from day 6 to 30 of gestation according to BC classes at weaning1
| Class variables at weaning | n | Feed level, kg/d2 | SEM | BC Class of | Probability, P-value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1.8 | 2.5 | 3.2 | L | Q | ClassBC3 | L × BC class | Q × BC class | ||||
| BC classes | |||||||||||
| BW, kg | 0.087 | 0.955 | 0.922 | 0.469 | 0.498 | ||||||
| ≤208 | 64 | 15.4 | 15.4 | 14.7 | 0.7 | 15.2 ± 0.4 | |||||
| >208 | 68 | 16.1 | 14.8 | 14.4 | 0.7 | 15.1 ± 0.4 | |||||
| BCS | 0.094 | 0.935 | 0.530 | 0.929 | 0.990 | ||||||
| ≤3 | 56 | 15.6 | 15.0 | 14.5 | 0.7 | 15.0 ± 0.4 | |||||
| >3 | 76 | 16.0 | 15.3 | 14.8 | 0.8 | 15.4 ± 0.4 | |||||
| BF, mm | 0.108 | 0.831 | 0.860 | 0.659 | 0.199 | ||||||
| ≤11 | 59 | 16.0 | 14.7 | 15.1 | 0.8 | 15.2 ± 0.4 | |||||
| >11 | 73 | 15.7 | 15.6 | 14.2 | 0.7 | 15.1 ± 0.4 | |||||
| Caliper unit | 0.081 | 0.867 | 0.417 | 0.757 | 0.986 | ||||||
| ≤10 | 54 | 15.8 | 16.0 | 14.6 | 0.9 | 15.5 ± 0.4 | |||||
| >10 | 78 | 15.8 | 14.5 | 14.5 | 0.7 | 14.9 ± 0.4 |
1A total of 150 PO2 females (Landrace × Large White) were used, divided into different BC classes at weaning to be fed on 1.8, 2.5, and 3.2 kg/d from day 6 to 30 of gestation.
2Feed levels: 1.8, 2.5, and 3.2 kg/d from day 6 to day 30 of gestation.
3Different classes for body condition (BC; BW, BCS, BF, and caliper unit).
Discussion
The main objective of this study was to evaluate the effects of different feed levels provided during early gestation on reproductive performance and the capacity to replenish the body reserves in PO1 and PO2 sows. Furthermore, the intention was to evaluate whether the effects of feed levels were dependent on the body condition of sows at weaning. This subject has been a focus of discussion in production systems because feeding restriction of gilts and sows immediately after insemination has been recommended. However, as recently reviewed, 75% of the experiments indicated that there is no detrimental effect of high energy levels post-breeding (above maintenance) on embryo survival (Leal et al., 2019). Contrarily, in the remaining experiments, a lower embryo survival was reported in swine females that received a higher feed amount during gestation (Leal et al., 2019). However, those trials were usually performed under experimental conditions that used only two feed levels for comparison and mostly with gilts. The focus of the present study was on young weaned sows, which are more sensitive to body changes during lactation. Young parity sows (PO1 and PO2) were subjected to a dose–response study with three different feed levels (1.8, 2.5, and 3.2 kg/d).
Effects of feed levels and parity on maternal growth
The first month of gestation aims to maintain the pregnancy and ensure embryo survival. For early parity, however, these goals become a challenge because sows do not reach their mature body until the fourth or fifth parity (Thomas et al., 2018). Additionally, PO1 and PO2 sows are more prone to have body lactation losses, being necessary to increase the feed intake to recover their body reserves. As expected, using high feed levels earlier in gestation increased the body reserves, regardless of the parity. However, even if the feed levels provided approximately 1, 1.5, and 2 times the estimated energy requirements for maintenance, based on their average metabolic BW (100 kcal × BW^0.75; NRC, 2012), sows gained less weight than expected. Sows were weighed at weaning, but treatments commenced on gestational day 6. Thus, it is necessary to consider the body losses during WEI (Koutsotheodoros et al., 1998; Werlang et al., 2011). In the present study, BW loss between weaning and gestation housing (~7 d later) was around 4.5% (approximately 9 kg; data not shown), comparable to BW losses between 3.8% and 4.2% previously reported by Koutsotheodoros et al. (1998) and Werlang et al. (2011), respectively. It means that at the beginning of the experiment (day 6 of gestation), females were at least 9 kg lighter than at weaning. So, the real BW gains during the treatment period were, therefore, close to 5, 12, and 18 kg for sows fed 1.8, 2.5, and 3.2 kg/d, respectively. Hoving et al. (2011) reported BW gains of 15.5 and 24.4 kg for females fed 2.5 and 3.25 kg/d between days 3 and 32 of gestation; however, those females were genetically different and lighter than those of the present study. Despite being lighter, their females had more BF at weaning and gained less BF than those in the present study. Similarly, another study reported greater BW gains (15.2 and 20.2 kg) and lower BF gains (1.2 and 1.6 mm) for PO1 sows fed 2.5 and 3.25 kg/d respectively, between days 3 and 35 of gestation (Hoving et al., 2012a).
Regarding parity, PO1 sows gained more weight than PO2 sows during the treatment period, as also reported by Thomas et al. (2018), who attributed it to the greater requirements for maintenance in PO2 sows until day 74 of gestation. Considering the BW at weaning and the feed levels performed, PO1 sows had more energy available to grow because more energy above maintenance was provided (NRC, 2012). It is important to mention that PO1 sows were 24.9 kg lighter than PO2 sows at the beginning of the experiment; so, fewer nutrients and energy for maintenance were needed. Hoving et al. (2011) reported similar BW gain between PO1 and PO2 sows; however, the difference in BW between parities was lower at the beginning of the experiment, compared with our study.
Effects of feed levels and parity on litter size and farrowing rate
In a recent systematic review, while gilts had no detrimental effects on embryo survival when the energy level of diet in early pregnancy is provided above the maintenance, a positive effect on embryo survival is observed in PO1 sows (Leal et al., 2019). Nevertheless, the negative effect of the increased feed level on total piglets born contrasts with the results of other studies that reported no evidences of differences (Virolainen et al., 2005a; Quesnel et al., 2010; Hoving et al., 2012a) or reported an increased embryonic survival or piglets born (Hoving et al., 2011; Athorn et al., 2013). The increase in feed amount from 1.8 to 3.6 kg/d did not affect the litter size in gilts (Virolainen et al., 2005a). Embryo survival and development on day 27 of gestation were not affected by a high feed level (2 vs. 4 kg/d) offered during the first 7 d of gestation in gilts (Quesnel et al., 2010). There was greater embryo survival on day 10 after insemination in gilts that received 2.8 compared with 1.5 kg/d (Athorn et al., 2013). Hoving et al. (2011) observed that sows that received extra feed (+0.75 kg) had fewer litters with ≤13 piglets and more litters with ≥17 piglets compared with a control group. In that study, the litter size increased from 13.2 to 15.2 piglets in PO1 and PO2 when the amount of feed was increased from 2.5 to 3.25 kg/d, between days 3 and 32 of gestation. However, in a second study, Hoving et al. (2012a) were unable to increase the litter size in primiparous sows fed 3.25 compared with 2.5 kg/d. In the present study, the finding that 30.9% of sows fed 3.2 kg/d had <11 piglets, whereas these percentages were only 18.7% and 17.9 % in the 1.8 and 2.5 kg/d groups, respectively (P = 0.055; data not shown), indicates a detrimental impact on litter size of overfed young sows. Consistently, Virolainen et al. (2005b) observed that embryo recovery was lower in multiparous sows fed 4 kg/d than sows fed 2 kg/d.
The reduction in piglets born by the higher-fed sows in the present study may be explained by a higher metabolic clearance of progesterone and its impact on embryo survival (Prime and Symons, 1993). Jindal et al. (1996) reported that a high feed level reduced systemic progesterone and embryo survival in gilts. The effect of feed level on progesterone concentration has been controversial because there are differences in the progesterone profile according to the site of blood collection (caudal vena cava vs. systemic circulation). It has been reported that progesterone concentration from the caudal vena cava is higher compared with the jugular vein (Virolainen et al., 2005a; Athorn et al., 2013). The vena cava drains the blood flow from the uterus and ovaries, the sites of local production of progesterone, before hepatic metabolism (Virolainen et al., 2005a). The nutritional regime significantly affected the progesterone concentration in the jugular vein (i.e., systemic measure), with no effect in the caudal vena cava (Virolainen et al., 2005a). These authors also observed a lower postprandial compared with preprandial progesterone concentration in the jugular vein in gilts. This finding provides support for the increased metabolism on portal blood flow. Athorn et al. (2013) observed that greater feed intake increased the number of progesterone pulses on day 9 of gestation. Furthermore, the vena cava progesterone concentration tended to be greater on day 6 (before embryo implantation) in high-fed gilts, but there was no difference in the jugular vein. The embryo survival was 15% higher on day 10 post-insemination in high-fed compared with low-fed gilts. Thus, Athorn et al. (2013) speculated that embryo survival benefits by the local production of progesterone, ensured by the counter-current flow and lymphatic pathways in high-fed gilts. In primiparous sows, Hoving et al. (2012a) reported no difference in progesterone concentration (jugular vein) between the animals offered 2.5 or 3.25 kg/d between days 3 and 35 of gestation. However, the progesterone concentration tended to be lower during the first 15 d of pregnancy in multiparous sows (PO2 to 11) fed 4 compared with 2 kg/d (Virolainen et al., 2005b). Hoving et al. (2011) hypothesized that embryonic survival may be affected, and even cause failed maternal recognition of the pregnancy, if the progesterone concentration is reduced in the pre-attachment period. Although we did not perform hormonal evaluations, the fact that litter size was reduced with high feed levels allows us to speculate that the progesterone concentration was affected. However, this variable should be better explored in further studies.
It is worth mentioning that most of the studies that used multiparous sows were performed with a reduced number of sows. Furthermore, the feed levels did not consider the extremes, that is, below and above the requirements for maintenance in the same experimental design. Even though Leal et al. (2019) reported that embryo survival was compromised only when the energy levels were greater than 12.9 Mcal/d, the reduced litter size observed with the highest feed level of the present study (10.1 Mcal/d; 3.2 kg) indicate that providing 7.9 Mcal/d (2.5 kg/d), between days 6 and 30 of gestation, ensured better results for total piglets born by PO1 and PO2 sows. Even though considering that the number of total piglets born observed for PO2 was higher compared with PO1 sows, no evidence for interaction with feed level was observed on the main reproductive variables. However, it is important to consider this response within parity taking into account the body condition of the females at weaning.
Influence of body condition at weaning and their interaction with feed levels on total piglets born
Better results for litter size were expected when sows with poor body condition at weaning received higher feed levels. However, the lack of interaction between feed level and classes of BW, BCS, and BF demonstrated that sows with lower body reserves did not benefit from a greater feed amount. This result indicates that current strategies for increasing the feed level post-insemination for body recovery should be reviewed because they implicate in losing potential on the performance of young weaned sows. The reductions of 1.6, 1.5, 1.2, and 1.3 piglets born in PO1 sows with lower BW, BCS, BF, and caliper units at weaning, respectively, regardless of the feed level, are consistent with reduced litter sizes in PO1 sows with a BW less than 177.9 kg or with a poor BCS, as reported by Schenkel et al. (2010). Young sows still need to grow to achieve a target weight at first farrowing; if the feed level is not adequate, growth might be prioritized instead of reproduction (Hoving et al., 2010). In contrast to PO1 sows, whose body reserves are more affected by nutritional level, the litter size was not affected in PO2 sows with lower body reserves. Schenkel et al. (2010) attributed the reduction in subsequent litter size of primiparous sows to both body reserves at weaning and body losses during the previous lactation. Prunier et al. (2003) suggested that subsequent poor reproductive performance is more related to lactational events rather than post-weaning events. In fact, reproductive performance (i.e., total embryos, embryo survival, pregnancy rate) was negatively affected in PO1 sows that experienced high weight losses during lactation (Hoving et al., 2012b). Unfortunately, changes in body reserves and the metabolic profile during lactation were not evaluated in the present study. However, lighter sows at weaning are suggestive of poor body condition management during lactation or even during the previous gestation. Hoving et al. (2010) observed that for each 10 kg higher BW gain between the first insemination and first weaning, the total number of piglets born was increased by 0.42 piglets.
In the current study, the total number of piglets born was also reduced when PO1 sows were fed 1.8 kg/d in comparison to those fed 2.5 kg/d. Insulin and insulin-like growth factor 1 (IGF-1) are hormones that respond, to some extent, to nutritional changes (De et al., 2009); they may act directly on the ovary or systematically on the hypothalamus and interfere with luteinizing hormone (LH) release (Quesnel, 2009). Increases in insulin and IGF-1 concentrations were reported in gilts fed 2-times the maintenance diet compared with those fed 1.2- and 0.6-times the maintenance diet (De et al., 2009). Furthermore, LH pulsatility was negatively affected in pregnant gilts that underwent feed restriction (1.8 vs. 3.6 kg/d; Peltoniemi et al., 1997). Notably, gilts do not experience lactational catabolism, and distinct metabolism should be considered for weaned sows. Indeed, no differences between 2.5 and 3.25 kg/d offered feed during early gestation were reported for progesterone, LH, and IGF-1 concentrations in PO1 sows (Hoving et al., 2012a). However, even considering the absence of interaction of feed level with class of BW, BCS, and BF, the fact that litter size was compromised in PO1 sows (quadratic effect) indicates that nutrients slightly above the maintenance, provided by 1.8 kg/d in the present study, were perhaps insufficient for sows that experience lactational catabolism and are still growing. This effect was ensured for PO1 sows with lower caliper unit (≤11), where a quadratic effect of feed level was observed for total piglets born, with no effect in sows with higher caliper unit (>11). The trend for linear reduction in total born piglets observed in PO2 sows (as the feed level increased) indicates that 3.2 kg/d is also an excessive feed amount to offer in early pregnancy for this category of females.
Conclusion
In young parity sows (PO1 and PO2) with suitable body condition, the gain in body measurements (BW, BCS, BF, and caliper units) increased as the feed intake during the first month of gestation increased from 1.8 to 3.2 kg/d. However, increasing feed allowance negatively impacted the total number of piglets born in PO1 and PO2 sows. Moreover, even considering sows with lower body reserves at weaning, the greater feed intake (3.2 kg/d) during early gestation did not increase the number of piglets born, especially for PO1 sows.
Acknowledgments
We are thankful to Frísia Cooperativa Agroindustrial, especially to all staff from UPL (Carambeí, Paraná, Brazil), for providing facilities to perform this study and Agroceres Pig Improvement Company for the financial support to this project. This study was financed in part by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brazil (CAPES) – Finance Code 001.
Glossary
Abbreviations
- AA
amino acids
- BCS
body condition score
- BF
backfat
- BW
body weight
- CP
crude protein
- ESF
electronic feeding station
- IGF-1
insulin-like growth factor 1
- LH
luteinizing hormone
- ME
metabolizable energy
- PO1
parity 1
- PO2
parity 2
- SID Lys
standardized ileal digestible lysine
- WEI
weaning-to-estrus interval
Conflict of interest statement
The authors declare no real or perceived conflicts of interest.
Literature Cited
- AOAC International 2012. Official methods of analysis of AOAC International. 19th ed. Gaithersburg (MD):AOAC International. [Google Scholar]
- Athorn R. Z., Stott P., Bouwman E. G., Chen T. Y., Kennaway D. J., and Langendijk P.. . 2013. Effect of feeding level on luteal function and progesterone concentration in the vena cava during early pregnancy in gilts. Reprod. Fertil. Dev. 25:531–538. doi: 10.1071/RD11295 [DOI] [PubMed] [Google Scholar]
- CBAA 2017. Compêndio Brasileiro de Alimentação Animal. 5th ed. São Paulo, SP (Brazil):Sindicato Nacional da Indústria da Alimentação Animal. [Google Scholar]
- De W., Ai-Rong Z., Yan L., Sheng-Yu X., Hai-Yan G., and Yong Z.. . 2009. Effect of feeding allowance level on embryonic survival, IGF-I, insulin, GH, leptin and progesterone secretion in early pregnancy gilts. J. Anim. Physiol. Anim. Nutr. 93:577–585. doi: 10.1111/j.1439-0396.2008.00844.x [DOI] [PubMed] [Google Scholar]
- Geisert R. D. and Schmitt R. A. M.. . 2002. Early embryonic survival in the pig: can it be improved? J. Anim. Sci. 80:54–65. doi: 10.2527/animalsci2002.0021881200800ES10009x [DOI] [Google Scholar]
- Hoving L. L., Soede N. M., Feitsma H., and Kemp B.. . 2012a. Embryo survival, progesterone profiles and metabolic responses to an increased feeding level during second gestation in sows. Theriogenology 77:1557–1569, doi: 10.1016/j.theriogenology.2011.11.024 [DOI] [PubMed] [Google Scholar]
- Hoving L. L., Soede N. M., Feitsma H., and Kemp B.. . 2012b. Lactation weight loss in primiparous sows: consequences for embryo survival and progesterone and relations with metabolic profiles. Reprod. Domest. Anim. 47:1009–1016. doi: 10.1111/j.1439-0531.2012.02007.x [DOI] [PubMed] [Google Scholar]
- Hoving L. L., Soede N. M., Graat E. A., Feitsma H., and Kemp B.. . 2010. Effect of live weight development and reproduction in first parity on reproductive performance of second parity sows. Anim. Reprod. Sci. 122:82–89. doi: 10.1016/j.anireprosci.2010.07.009 [DOI] [PubMed] [Google Scholar]
- Hoving L. L., Soede N. M., van der Peet-Schwering C. M., Graat E. A., Feitsma H., and Kemp B.. . 2011. An increased feed intake during early pregnancy improves sow body weight recovery and increases litter size in young sows. J. Anim. Sci. 89:3542–3550. doi: 10.2527/jas.2011-3954 [DOI] [PubMed] [Google Scholar]
- Jindal R., Cosgrove J. R., Aherne F. X., and Foxcroft G. R.. . 1996. Effect of nutrition on embryonal mortality in gilts: association with progesterone. J. Anim. Sci. 74:620–624. doi: 10.2527/1996.743620x [DOI] [PubMed] [Google Scholar]
- Kim S. W., Weaver A. C., Shen Y. B., and Zhao Y.. . 2013. Improving efficiency of sow productivity: nutrition and health. J. Anim. Sci. Biotechnol. 4:26. doi: 10.1186/2049-1891-4-26 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Knauer M. T., and Baitinger D. J.. . 2015. The sow body condition caliper. Appl. Eng. Agr. 31:175–178. doi: 10.13031/aea.31.10632 [DOI] [Google Scholar]
- Koutsotheodoros F., Hughes P. E., Parr R. A., Dunshea F. R., Fry R. C., and Tilton J. E.. . 1998. The effects of post-weaning progestagen treatment (Regumate) of early-weaned primiparous sows on subsequent reproductive performance. Anim. Reprod. Sci. 52:71–79. doi: 10.1016/s0378-4320(98)00088-8 [DOI] [PubMed] [Google Scholar]
- Langendijk P., Bouwman E. G., Chen T. Y., Koopmanschap R. E., and Soede N. M.. . 2017. Temporary undernutrition during early gestation, corpora lutea morphometrics, ovarian progesterone secretion and embryo survival in gilts. Reprod. Fertil. Dev. 29:1349–1355. doi: 10.1071/RD15520 [DOI] [PubMed] [Google Scholar]
- Leal D. F., Muro B. B. D., Nichi M., Almond G. W., Viana C. H. C., Vioti G., Carnevale R. F., and Garbossa C. A. P.. . 2019. Effects of post-insemination energy content of feed on embryonic survival in pigs: a systematic review. Anim. Reprod. Sci. 205:70–77. doi: 10.1016/j.anireprosci.2019.04.005 [DOI] [PubMed] [Google Scholar]
- Mallmann A. L., Camilotti E., Fagundes D. P., Vier C. E., Mellagi A. P. G., Ulguim R. R., Bernardi M. L., Orlando U. A. D., Gonçalves M. A. D., Kummer R.. et al. 2019. Impact of feed intake during late gestation on piglet birth weight and reproductive performance: a dose-response study performed in gilts. J. Anim. Sci. 97:1262–1272. doi: 10.1093/jas/skz017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- NRC 2012. Nutrient requirements of swine. 11th rev. ed. Washington (DC):The National Academies Press. [Google Scholar]
- Peltoniemi O. A. T., Love R. J., Klupiec C., and Evans G.. . 1997. Effect of feed restriction and season on LH and prolactin secretion, adrenal response, insulin and FFA in group housed pregnant gilts. Anim. Reprod. Sci. 49:179–190. doi: 10.1016/S0378-4320(97)00062-6 [DOI] [PubMed] [Google Scholar]
- Prime G. R., and Symonds H. W.. . 1993. Influence of plane of nutrition on portal blood flow and the metabolic clearance rate of progesterone in ovariectomized gilts. J. Agric. Sci. 121:389–397. doi: 10.1017/S0021859600085580 [DOI] [Google Scholar]
- Prunier A., Soede N. M., Quesnel H., and Kemp B.. . 2003. Productivity and longevity of weaned sows. In: Pluske J. R., Le Dividich J., and Verstegen M. W. A., editors. Weaning the pig. Wageningen (The Netherlands):Wageningen Acad. Publishers; p. 385–419. [Google Scholar]
- Quesnel H. 2009. Nutritional and lactational effects on follicular development in the pig. In: Rodriguez-Martinez H., Vallet J. L., and Ziecik A. J., editors. Control of pig reproduction VIII. Nottingham (UK):Nottingham University; Press; p. 121–134. [PubMed] [Google Scholar]
- Quesnel H., Boulot S., Serriere S., Venturi E., and Martinat-Botté F.. . 2010. Post-insemination level of feeding does not influence embryonic survival and growth in highly prolific gilts. Anim. Reprod. Sci. 120:120–124. doi: 10.1016/j.anireprosci.2010.04.006 [DOI] [PubMed] [Google Scholar]
- Schenkel A. C., Bernardi M. L., Bortolozzo F. P., and Wentz I.. . 2010. Body reserve mobilization during lactation in first parity sows and its effect on second litter size. Livest. Sci. 132:165–172. doi: 10.1016/j.livsci.2010.06.002 [DOI] [Google Scholar]
- Thomas L. L., Goodband R. D., Tokach M. D., Woodworth J. C., DeRouchey J. M., and Dritz S. S.. . 2018. Effect of parity and stage of gestation on growth and feed efficiency of gestating sows. J. Anim. Sci. 96:4327–4338. doi: 10.1093/jas/sky279 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Virolainen J. V., Love R. J., Tast A., and Peltoniemi O. A. T.. . 2005a. Plasma progesterone concentration depends on sampling site in pigs. Anim. Reprod. Sci. 86:305–316. doi: 10.1016/j.anireprosci.2004.07.004 [DOI] [PubMed] [Google Scholar]
- Virolainen J. V., Peltoniemi O. A., Munsterhjelm C., Tast A., and Einarsson S.. . 2005b. Effect of feeding level on progesterone concentration in early pregnant multiparous sows. Anim. Reprod. Sci. 90:117–126. doi: 10.1016/j.anireprosci.2005.01.012 [DOI] [PubMed] [Google Scholar]
- Werlang R. F., Argenti L. E., Fries H. C., Bernardi M. L., Wentz I., and Bortolozzo F. P.. . 2011. Effects of breeding at the second oestrus or after post-weaning hormonal treatment with altrenogest on subsequent reproductive performance of primiparous sows. Reprod. Domest. Anim. 46:818–823. doi: 10.1111/j.1439-0531.2010.01747.x [DOI] [PubMed] [Google Scholar]
- Young M. G., Tokach M. D., Aherne F. X., Main R. G., Dritz S. S., Goodband R. D., and Nelssen J. L.. . 2004. Comparison of three methods of feeding sows in gestation and the subsequent effects on lactation performance. J. Anim. Sci. 82:3058–3070. doi: 10.2527/2004.82103058x [DOI] [PubMed] [Google Scholar]