| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Animal Breeding and Genetics Group and
and
Animal Nutrition Group, Wageningen Institute of Animal Sciences, Wageningen University, 6700 AH Wageningen, The Netherlands;
and
Institute for Pig Genetics BV, Beuningen, The Netherlands; and
and
DLO-Institute for Animal Science and Health, ID-Lelystad, The Netherlands
3 Correspondence:
P.O. Box 338 (phone: +31(0)317-482335; fax: + 31(0)317-483929; E-mail:
Egbert.Kanis{at}wur.nl).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Key Words: Backfat • Feed Intake • Litter Size • Reproductive Performance • Sows • Weight Losses
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Continued selection for reproduction and leanness, without increasing lactation feed intake, will likely result in a larger proportion of young sows consuming an insufficient amount of feed to adequately support lactation (Cameron et al., 2002). Inadequate intake will lead to increased losses of backfat and BW, probably resulting in more culling of sows due to reproductive failures and a reduced lifetime performance (Eissen et al., 2000). Inadequate feed intake during lactation is particularly evident in primiparous sows because, relative to multiparous sows, they have smaller body reserves, a lower daily feed intake, and need extra energy for BW gain (NRC, 1987).
An experiment was conducted with primiparous sows of three sow lines to evaluate effects of high litter sizes and feed intake on changes in BW and backfat thickness and on growth of the first litter. Ad libitum feed intake during first lactation was recorded to study whether a larger feed intake can prevent possible negative effects of large litters on subsequent weaning-to-estrus interval and litter size, and whether selection for a higher lactation feed intake should be recommended.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
shows the distribution of these 268 sows over genotypes and litter sizes. Note that the number of Genotype 2 sows was about half the number of sows in the other two genotypes.
|
During gestation, three diets were used containing 12.58 to 12.84 MJ of DE, 140 to 162 g of CP, and 5.2 to 7.8 g of digestible lysine per kilogram, respectively. Sows were fed according to a commercial Dutch feeding regimen to attain a target of 16 to 20 mm of backfat thickness (assessed subjectively) at farrowing. The amount of feed offered daily gradually increased from about 1.8 kg at insemination to 3.2 kg during late gestation. Depending on a visual assessment of fatness, individual sows received extra feed per day varying from 0 to 20%. During lactation, three diets also were used containing 13.34 to 14.09 MJ of DE, 163 to 180 g of CP, and 6.8 to 8.8 g of digestible lysine per kilogram, respectively. Feed intake per sow was linearly transformed to feed intake of a diet containing 13.72 MJ of DE/kg, which was the energy content of the intermediate diet. The level of feeding was progressively increased from parturition onward to reach ad libitum from about d 4 after farrowing. Feed was given at 0700 and 1500, in approximately equal amounts, with more or less added to the feeders to ensure that feed was always available and fresh. The water nipple was above the trough, and water was available ad libitum to the sows.
Minimal room temperature in the farrowing crates was maintained between 17 and 21°C, depending on stage of lactation. Maximal room temperature varied with the outside temperature. Farrowing crates had floor heating for piglets. During the first 3 d after farrowing, an infrared heater also provided supplemental heat in winter. Artificial lighting was present from 0700 to 1530. The amount of natural lighting was limited. Routine management procedures were followed in caring for the sow and litter during parturition and lactation.
Piglets were weaned at about 28 d of lactation. From 3 or 4 d after weaning onward, sows were checked each morning for the onset of standing estrus using direct exposure to a boar. Sows were also checked for other signs of estrus, such as vulval swelling and reddening and reaction to back pressure. About two-thirds of the sows were inseminated on the day(s) of standing estrus. After weaning, sows were fed the lactation diet (close to) ad libitum until reaching first estrus. When sows did not show estrus within the first week, daily feed allowance was set at a maximum of 2.5 kg.
Litter size at birth of the second litter was available for 169 sows (45, 40, and 84 sows of Genotypes 1, 2, and 3, respectively). This information should be used and interpreted with caution, however, since these sows formed a selected subset of the total population.
Traits
Sows were assumed to have reached their maximal feed intake capacity by d 10 of lactation (Revell et al., 1998). Total ad libitum feed intake by sows was recorded from d 10 (mean 10.4; SD 1.8) of lactation until weaning at d 28 (mean 27.8; SD 1.9). Sows and litters were weighed when feed intake recording started and ended. At these times, sows’ backfat depth was also measured by ultrasound at three positions along the back at both sides. Piglets that died between d 10 of lactation and weaning were recorded and weighed. Weaning-to-estrus interval was defined as the interval from weaning to standing estrus. A weaning-to-estrus interval up to 7 d was considered normal (Ten Napel et al., 1995), whereas longer intervals were considered prolonged.
Total feed intake was adjusted linearly to total feed intake from d 10 to 28 of lactation (weaning) using a within-genotype regression of feed intake on day numbers of lactation. Average daily feed intake was calculated as the total feed intake from d 10 to 28 divided by 18. Sow and litter weights and average backfat thickness of sows not measured exactly at d 10 or d 28 of lactation were adjusted linearly to d 10 and 28, respectively, using within-genotype regression. The daily loss of backfat thickness and BW of sows during lactation was defined as the total decrease from d 10 to 28 divided by 18. The BW of piglets that died between d 10 and 28 was added to the litter weight of the sow at d 28. Litter weight gain was defined as the total increase in litter weight from d 10 to 28 divided by 18. Piglet weight gain was defined as litter weight gain divided by litter size during lactation. Litter size during lactation (from d 10 to 28) of a sow was corrected for piglets that died between d 10 and weaning. Each dead piglet was counted as: (day number of lactation of dying — 10)/(28 — 10) piglet. For example, litter size during lactation was 12.33 when litter sizes at d 10 and weaning were 13 and 12, respectively, and one piglet died at d 16 of lactation.
Statistical Analyses
Analysis was performed in three steps. In the first step, effects of litter size and genotype on lactation feed intake and sow performance during lactation and the subsequent reproductive phase were studied. Effect of lactation feed intake on sow performance was studied in the second step, and the effect of body condition on reproductive performance after weaning of the first litter was examined in the third step. Litter size was used as a continuous and not as a class variable in the analyses because a substantial number of sows had a litter size unequal to 8, 11, or 14 piglets at d 10 of lactation and/or at weaning (Table 1
); this was due to incomplete standardization after farrowing and piglet mortality. Furthermore, treating litter size as a continuous variable enabled correction of litter size during lactation for piglet mortality, as illustrated above.
In Step 1, each trait measured during lactation was analyzed using the GLM procedure of SAS (SAS Inst., Inc., Cary, NC). The initial model was:
 | [1] |
where Y = dependent variable, G = genotype of the sow, LS = litter size during lactation (from d 10 until weaning), AF = age at farrowing, and YS = year and season of farrowing (clusters of 3 mo; nine classes). Genotype, the linear effect of litter size, and the effect of year and season of farrowing were included in the model for each trait measured during lactation, independent of significance. Backward elimination was used for modeling of other terms (P < 0.10).
To determine risk factors affecting occurrence of prolonged weaning-to-first-estrus intervals, logistic regression analyses were performed using the LOGISTIC procedure of SAS. Genotype, litter size during lactation, and year and season of farrowing were included in the model, independent of significance (Model [1a]). Other effects included in Model [1a] (only if P < 0.10) were age at farrowing and lactation length. Odds ratios, defined as the relative increase (odds ratio > 1) or decrease (< 0 odds ratio < 1) in probability of having a prolonged interval (e.g., for Genotype 1 compared to Genotype 2) were estimated. To estimate odds ratios for litter size per genotype, Model [1a] without the genotype effect was applied for the three genotypes separately.
Because of the smaller data set, effects of genotype and first litter size on total number of piglets born in the second litter were analyzed with a model similar to Model [1], but excluding all interactions, whereas age at farrowing was replaced by interval from weaning to estrus (Model [1b]).
For Step 2 (effects of lactation-feed intake), Model 1 was extended as follows:
 | [2] |
To study the effect of feed intake on prolonged weaning-to-estrus interval, logistic regression analysis was applied with ADFI during lactation included in Model [1a] (Model [2a]). Similarly, effect of feed intake on size of the second litter was analyzed by including ADFI during lactation in Model [1b] (Model [2b]).
In Step 3 (effects of body condition and condition loss during first lactation on the probability of a prolonged interval from weaning to estrus and on the total number of piglets born in the second litter), the statistical model was:
 | [3] |
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
shows the effects of genotype and litter size on sow performance. Figures 1 to 3 present the effects of litter size on feed intake and body tissue losses of the sow and litter and piglet weight gain. Interactions were found between the effects of genotype and the linear and quadratic effects of litter size on ADFI and on backfat loss (P = 0.07 and P < 0.01, respectively; see Table 2). Therefore, results are plotted per genotype for feed intake and backfat loss (Figures 1 and 2), and across genotypes for litter weight gain, sow weight loss, and piglet weight gain (Figure 3). Genotype 3 sows had a higher ADFI than Genotype 1 sows (Table 2). Based on Model [1], ADFI was quadratically (P < 0.05) affected by litter size only in Genotype 2 sows (Figure 1), although the course of feed intake was similar for Genotype 3 sows. Genotype 2 sows reached a maximal feed intake at 10.8 piglets a litter. Feed intake of Genotype 1 and 3 sows was not affected by litter size (Model [1]; P > 0.05).
|
|
|
|
). Increasing litter size during lactation reduced backfat thickness and BW of sows at weaning in a linear (P < 0.001) and quadratic (P < 0.10) manner, respectively. Daily backfat loss was quadratically affected by litter size for Genotype 2 sows only (Figure 2; P < 0.01); backfat losses decreased when litter size increased up to 9.8 piglets, after which losses increased again. When only the linear effect of litter size was included in the analysis, loss of backfat thickness increased at 0.020 (P < 0.001) and 0.0073 (P < 0.05) mm/piglet for Genotype 1 and 3 sows, respectively. The daily loss of BW during lactation was higher for Genotype 2 than Genotype 1 sows (Table 2; P < 0.05) and increased linearly (P < 0.001) with litter size (Figure 3). Litter and piglet weight gains during lactation were larger for Genotype 3 sows than for both other genotypes (P < 0.05). Litter weight gain increased linearly (P < 0.001) with litter size, whereas piglet weight gain decreased quadratically (P < 0.01) with litter size (Table 2 and Figure 3). Of the 279 sows, 60, 20, and 24 sows of Genotype 1, 2, and 3, respectively (57, 37, and 20%), had a prolonged interval from weaning to estrus. Logistic regression analysis (Model [1a]) revealed that genotype and litter size significantly affected the probability of a prolonged interval. Genotype 1 and 2 sows were 5.47 (P < 0.01) and 2.03 (P < 0.10) times as likely to have a prolonged interval than Genotype 3 sows. The odds ratio for litter size was 1.23 (P < 0.01), indicating that a sow was 1.23 times as likely to have a prolonged interval when nursing one extra piglet during lactation. For example, a sow nursing 11 piglets is 1.51 (1.232) times as likely to have a prolonged interval as a sow nursing nine piglets. When logistic regression was performed for each genotype separately, the odds ratio for litter size remained significantly > 1 for Genotype 1 sows (odds ratio = 1.37; (P < 0.01), whereas the odds ratios for Genotype 2 and 3 sows (1.21 and 1.13, respectively) did not differ from unity.
No significant genotype effects on size of the second litter at birth were found, but size of the second litter was quadratically (P < 0.05) associated with litter size during first lactation (Model [1b]). Maximal size of second litter was found with 9.9 piglets nursed during first lactation.
Relationships Between Daily Feed Intake, Body Condition, and Sow Performance
An increase in voluntary ADFI during lactation reduced backfat (P < 0.001) and BW (P < 0.01) losses of sows and increased (P < 0.001) litter weight gain (Table 3). The favorable effect of higher feed intake on losses of backfat and BW was smaller with larger litter sizes. For example, a 1-kg (21%) increase in daily feed intake reduced daily backfat loss by 0.022 mm (15%) at a litter size of 10 piglets (-0.091 + [10 x 0.0069]), whereas backfat loss was not reduced (+0.0056 mm/d) at a litter size of 14 piglets. Furthermore, a 1-kg increase in daily feed intake reduced daily weight loss of sows by 0.13 kg (13%) at a litter size of 10 piglets (-0.43 + [10 x 0.030]), whereas weight loss was only reduced by 0.015 kg/d (1%) at a litter size of 14 piglets. The regression coefficient of litter weight gain on feed intake was larger for Genotype 2 than Genotype 1 sows, but was not affected by litter size (Table 3).
|
11 piglets, respectively, when analyzed as two subsets differing in number of piglets per sow. The latter results suggest that a greater feed intake is less beneficial for the sow at a larger litter size.
Effects of Body Condition on Reproductive Performance
Of the four sow condition traits analyzed (Model [3]), only BW loss during first lactation seemed to have significant effects on subsequent reproduction. The odds ratio for the effect of BW loss on the probability of a prolonged interval from weaning to estrus was 2.59 (P < 0.001), indicating that a 1-kg lower BW loss per day results in a 2.59 times lower risk (63.4% reduction) of a prolonged interval. Furthermore, a 1-kg reduction in daily BW loss of the sow resulted in 1.28 piglets more in her second litter (P < 0.01). Although, this last figure is based on a selected subset of 178 sows, the results emphasize the importance of reducing BW loss during first lactation for the subsequent reproductive performance of sows.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
). Differences in BW and body composition between sows of the three genotypes may have contributed to different responses. Also, effects of genetic heterosis may have contributed as Genotype 1 sows were purebred animals, whereas Genotype 2 and 3 sows were crossbred animals.
The absence of a significant effect of litter size on daily feed intake for Genotypes 1 and 3 sows is in agreement with the literature, in which litter size during lactation was six or 10 piglets (Yang et al., 1989) or ranged from six to 14 piglets (Auldist et al., 1998). It should be noted that daily feed intake was restricted to a maximum of 7 and 5 kg per sow in the experiments of Yang et al. (1989) and Auldist et al. (1998), respectively. The larger feed intake of Genotype 3 compared with Genotype 1 sows may have caused the larger litter weight gain, as backfat and BW losses during lactation were not different (Table 2; P > 0.05). Moreover, for the heavier (larger) Genotype 3 sows, 1 mm of extra backfat loss meant a greater loss of subcutaneous fat than for both other genotypes. The positive effect of feed intake on litter weight gain was also found across genotypes (Table 3) and agrees with literature (for a review see Eissen et al., 2000).
O’Grady et al. (1985) and Koketsu et al. (1996a) found an increase in feed intake with increasing litter size for small and medium size litters, which was only found for Genotype 2 sows in the present study. Across parities, O’Grady et al. (1985) estimated a maximal feed intake at a litter size of 12.8 to 14 piglets, and Koketsu et al. (1996a) concluded that feed intake did not significantly increase beyond litter sizes of 11 piglets. The increase in daily feed intake with increasing litter size for small and medium size litters was larger in Genotype 2 in the present study than in both other studies. For example, daily feed intake increased by 0.44 (based on the formula presented for Genotype 2 in Figure 1), 0.16, and 0.18 kg when litter size increased from 8 to 11 piglets in the present study, the study of Koketsu et al. (1996a), and the study of O’Grady et al. (1985), respectively. This relatively large increase for Genotype 2 sows was matched by a small decrease in backfat loss (Figure 2).
Breed differences in duration of weaning-to-estrus interval are known to exist (Vesseur et al., 1996) and may have contributed to differences in odds ratios for a prolonged interval between the three genotypes, as well as differences in heterosis. Further, each extra piglet in the litter increased the risk of a prolonged interval by 23%. Also, some other studies indicated an unfavorable association between litter size during lactation and weaning-to-estrus interval (e.g., Sterning et al., 1990; Vesseur et al., 1994; 1996), whereas others did not (e.g. Tubbs et al., 1990; Koketsu and Dial, 1997). An increasing weaning-to-estrus interval with increasing litter size may be attributed to increased losses of BW, in particular of body protein (King et al., 1993; Touchette et al., 1998; Guedes and Nogueira, 2001). This is consistent with our result that sows with a reduced BW loss had a lower probability of a prolonged weaning-to-estrus interval (P < 0.001) and that backfat loss had no effect on this interval (P > 0.05).
Sterning and Lundeheim (1995) found that the probability of having a large second litter was low for sows with a large first litter. This is in line with the present results, which show that an increasing litter size from 9.9 piglets onward during the first lactation was associated with a decreasing size of the second litter. Also Morrow et al. (1992) concluded that a large first litter was the major determinant of fewer piglets being born alive in the second litter. The actual relation between second and first litter size may be stronger than observed here since the sows that produced a second litter were on average in better condition than sows that did not produce a second litter, and carry-over effects can be expected especially in sows with a poor condition.
Feed Intake and Body Condition
Sows with a greater lactation feed intake showed significantly smaller backfat and BW losses, a higher litter weight gain (Table 3), and a reduced probability of a prolonged weaning-to-estrus interval. However, size of second litter was not influenced by feed intake during first lactation (P > 0.05), which may be due to selection for body condition after the first parity. In the literature, relationships between feed intake and sow performance have been mostly studied in experiments in which feed intake was restricted during part of lactation. In a number of these experiments, litter weight gain was not affected by feed intake since sows that are more feed restricted mobilized more body reserves to compensate for the lower feed intake (e.g. Reese et al., 1982; Prunier et al., 1993; Zak et al., 1997). In other experiments, feed-restricted sows showed a lower litter weight gain (e.g., Reese et al., 1982, Eastham et al., 1988; Mullan and Williams, 1989), possibly because the amount of sow body reserves was insufficient or less able to be mobilized, or the level of feed intake restriction was more severe compared to experiments where litter weight gain was not affected. Koketsu et al. (1997b) found in ad libitum fed primiparous sows that a 1-kg higher daily feed intake during wk 2 and 3 of lactation resulted in an increase of litter weight gain by 0.035 kg/d. This is less than in the present experiment, which may be due to different lactation lengths (on average 19 d in Koketsu et al. vs. 28 d in the present experiment). Pluske et al. (1998) superalimented primiparous sows and found that extra feed (energy intake of superalimented sows was 38% higher than that of ad libitum-fed sows) was partitioned to the sow’s body rather than into milk production, since litter growth of superalimented and ad libitum-fed sows was similar.
A greater daily feed intake reduced the weaning-to-service interval in the studies of Koketsu and Dial (1997) and Koketsu et al. (1997b). Koketsu et al. (1997a) estimated odds ratios for the effect of lactation feed intake ranging from 0.82 to 0.89 for regularly and irregularly returning to service, anestrus, and not farrowing. These results support the finding that a larger feed intake during lactation reduces the risk of postweaning reproductive problems.
Litter weight gain increased more per kilogram increase in feed intake for Genotype 2 sows than for Genotype 1 sows (0.19 vs 0.058 kg/d; Table 3), suggesting that Genotype 2 sows partitioned more of extra feed intake to milk production and therefore less to body reserves. It would be expected that feed intake during lactation has less effect on prolonged weaning-to-estrus intervals in Genotype 2 sows than in Genotype 1 sows. Odds ratios for the effect of feed intake on weaning-to-estrus interval being 0.63 for Genotype 2 and 0.44 for Genotype 1 sows, were in line with this expectation.
Our results indicate that reducing BW loss of sows during first lactation is favorable for the size of the subsequent litter, which was also found by King and Williams (1984) and Kirkwood et al. (1987). At higher litter sizes, however, extra feed eaten by primiparous sows was hardly used to reduce their weight and backfat losses, and litter weight gain was not increased either. This finding suggests that the use of feed for reducing sow body tissue loss and increasing litter weight gain is less efficient at higher levels than at lower levels of litter size and may point to some form of stress with sows nursing large litters, which requires further study. The result that sows with <11 piglets had a lower odds ratio for the effect of feed intake on weaning-to-estrus interval than sows with 11 piglets is in line with the finding that at higher litter sizes, any extra feed taken in is used less efficiently to reduce BW loss. This implies that at high litter sizes, means other than just stimulating feed intake during lactation (e.g., genetic selection) should be used to reduce BW losses of sows.
In The Netherlands in 1998, the number of piglets born alive per litter was, on average, 10.9 for sows of mixed parity (Siva, 1999). It is known that first-parity sows have smaller litters than higher parity sows (e.g., Tummaruk et al., 2001). At current litter sizes, a higher feed intake during lactation, either by nutritional or genetic measures, would therefore reduce backfat and weight losses of primiparous sows and decrease the probability of a prolonged weaning-to-estrus interval, as well as the probability of a reduced second litter.
|
Implications |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
|
Footnotes |
|---|
2 Current address: Nutreco, P.O. Box 240, 5830 AE Boxmeer, The Netherlands.
Received for publication May 1, 2002. Accepted for publication October 22, 2002.
|
Literature Cited |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Bergsma, E. Kanis, M. W. A. Verstegen, and E. F. Knol Genetic parameters and predicted selection results for maternal traits related to lactation efficiency in sows J Anim Sci, May 1, 2008; 86(5): 1067 - 1080. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Crenshaw, R. D. Boyd, J. M. Campbell, L. E. Russell, R. L. Moser, and M. E. Wilson Lactation feed disappearance and weaning to estrus interval for sows fed spray-dried plasma J Anim Sci, December 1, 2007; 85(12): 3442 - 3453. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Koketsu Longevity and efficiency associated with age structures of female pigs and herd management in commercial breeding herds J Anim Sci, April 1, 2007; 85(4): 1086 - 1091. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Quesnel, M. Etienne, and M.-C. Pere Influence of litter size on metabolic status and reproductive axis in primiparous sows J Anim Sci, January 1, 2007; 85(1): 118 - 128. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Guillemet, J. Y. Dourmad, and M. C. Meunier-Salaun Feeding behavior in primiparous lactating sows: Impact of a high-fiber diet during pregnancy J Anim Sci, September 1, 2006; 84(9): 2474 - 2481. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Kanis, K. H. De Greef, A. Hiemstra, and J. A. M. van Arendonk Breeding for societally important traits in pigs J Anim Sci, April 1, 2005; 83(4): 948 - 957. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
