Effect of live weight development and reproduction in first parity on reproductive performance of second parity sows - Part 1
L.L. Hovinga,b,∗, N.M. Soedea, E.A.M. Graatc, H. Feitsmad, B. Kempa
a Adaptation Physiology Group, Wageningen Institute of Animal Sciences (WIAS), Wageningen University, The Netherlands
b 荷兰Varkens KI Nederland公司
b Varkens KI Nederland, The Netherlands
c Quantitative Veterinary Epidemiology Group, WIAS, Wageningen University, The Netherlands
d Institute for Pig Genetics (IPG), The Netherlands
窝产仔数 Litter size
受胎率 Pregnancy rate
繁殖性能 Reproductive performance
二胎母猪 Second parity
与第一胎母猪相比，第二胎母猪的繁殖性能若受损，则会降低繁殖效率，可能还会影响母猪的使用寿命。本研究旨在量化一胎母猪的增重及繁殖情况对其二胎繁殖性能（受胎率和窝产仔数）的影响。本研究对2个试验场的母猪的体重变化(初次授精、分娩和断奶时的体重)和繁殖情况(窝产总仔数、断奶到配种间隔、哺乳期长、断奶仔猪数)进行了测定和记录。采用逻辑回归分析对“第一次断奶后第一次授精但受孕失败”的二元结果进行了分析。采用一般线性回归分析方法对二胎母猪第一次授精后的窝产仔数进行分析。返情复配的母猪未计入二胎母猪窝产仔数的分析，因为断奶到配种间隔的延长有助于提高窝产仔数。不同猪场的母猪增重情况差异显著，因此每个猪场的数据都进行了分析。与A场的后备母猪相比，B场的后备母猪在第一次授精(B场：275±0.9天，145±0.8 kg；A场：230±0.6 天，124±0.5 kg)）、第一次分娩(B场：189±1.1kg；A场：181±0.9kg)和首次断奶(B场：165±1.1kg；A场：156±0.9kg)时的日龄更大，体重更高。两个猪场在妊娠期间的体重损失情况相似(分别为24.9±0.7和23.7±1.0 kg)。然而，与B场相比，A场的后备母猪在第一次授精至第一次断奶期间体重增加更多(A场：36.1±0.8 kg；B场：20.9±1.3 kg)。二胎时A场受孕失败的母猪比例为11%，B场为15%。A场第一胎和第二胎的窝产仔数分别为10.7±0.1和11.6±0.2，B场为11.8±0.1和11.6±0.1。造成受孕失败和二胎产仔数不同的相关变量在两场之间也存在差异。在A场，受孕失败和二胎产仔数主要与母猪体重有关，而在B场，受孕失败和二胎产仔数与第一胎的总产仔数和母猪品系等变量有关。在两个猪场都发现，从第一次授精到第一次断奶期间母猪增重越多，其受孕失败的比例就越低(A场：母猪每增重10公斤，其受孕失败的比例就降低0.7；B场：母猪每增重10公斤，其受孕失败的比例就降低0.8)，而A场还发现，在此情况下二胎的窝产仔数也会更高(每增重10公斤，产仔数高0.42头)。研究结果表明，母猪的增重情况会影响第二胎的繁殖性能，特别是在A场，后备母猪在第一次授精时体重相对较轻，日龄较小。对这些动物的管理应以优化第一次配种时的日龄和体重的发育和增加第一次配种到第一次断奶之间的体重为目标，以优化第二胎的繁殖性能。
An impaired reproductive performance in second parity compared to first parity sows, decreases reproductive efficiency and, perhaps, longevity of sows. This study aims to quantify the effect of live weight development and reproduction in first parity on reproductive performance of second parity sows, i.e. pregnancy rate as well as litter size. Measures of sow development (live weight at first insemination, farrowing and weaning) and reproduction (total number of piglets born, weaning to insemination interval, lactation period, number piglets weaned) were recorded on two experimental farms. Logistic regression analysis was done for the binary outcome ‘non-pregnancy from first insemination after first weaning’(yes/no). General linear regression analysis was used for litter size from 1st insemination in second parity. Repeat breeders were omitted from the analysis on litter size in second parity, since a prolonged period between weaning and conception can positively influence litter size. Farms differed significantly in measures of sow live weight development and therefore data were analyzed per farm. Compared with gilts from farm A, gilts from farm B were older and heavier at: first insemination (275±0.9 days and 145±0.8 kg for farm B vs. 230±0.6 days and 124±0.5 kg for farm A), first farrowing (resp. 189±1.1 vs. 181±0.9 kg) and first weaning (resp. 165±1.1 vs. 156±0.9 kg). Weight loss during pregnancy was similar for both farms (resp. 24.9±0.7 and 23.7±1.0 kg). Gilts from farm A, however, gained more weight in the period between first insemination and first weaning compared with gilts from farm B (resp. 36.1±0.8 and 20.9±1.3 kg). Non-pregnancy in second parity was 11% for farm A and 15% for farm B. Litter sizes in first and second parity were, respectively, 10.7±0.1 and 11.6±0.2 for farm A and 11.8±0.1 and 11.6±0.1 for farm B. Variables associated with non-pregnancy and litter size in second parity differed between farms. On farm A, mainly sow live weight development was associated with non-pregnancy and litter size in second parity, whilst on farm B variables like total number born in 1st parity and sow line, were associated with non-pregnancy and litter size in second parity. On both farms, higher weight gain from first insemination to first weaning was associated with a decrease in non-pregnancy (odds ratio 0.7 per 10 kg for farm A and 0.8 per 10 kg for farm B) and on farm A with higher litter size in second parity (ˇ = 0.42 per 10 kg weight gain). Results show that sow live weight development affects reproductive performance in second parity, especially on farm A where gilts are relatively light or young at first insemination. Management of these animals should aim to optimize development at first insemination and to increase growth between first insemination and first weaning in order to optimize production in second parity.
繁殖性能，即分娩率和窝产仔数，应该随着胎次的增加而增加，在3-5胎时达到最高水平(Koketsu et al., 1999; Hughes and Varley, 2003)。然而，与一胎母猪相比，二胎母猪在第一次授精后的受胎率和/或产仔数通常更低(Morrowet al., 1989)。这种现象被称为二胎综合征(SLS)。二胎综合征降低了2胎母猪的繁殖效率，可能也会缩短母猪的使用寿命，因为繁殖失败是导致年轻母猪被淘汰的主要原因(Zak et al., 1997b; Lucia et al., 2000)。从遗传学角度看（基因型），1胎和2胎的产仔数之间存在高度相关性(0.88; Holm et al., 2005 和0.83; Hanenberg et al., 2001)，但表现型相关性较低(0.04; Hanenberg et al., 2001)，表明环境对窝产仔数的影响很大。
Reproductive performance, i.e. farrowing rate and litter size, is supposed to increase as parity increases, reaching the highest levels from parity 3 to 5 (Koketsu et al., 1999; Hughes and Varley, 2003). Second parity sows, however, often have lower pregnancy rates and/or smaller litter sizes from first insemination compared with first parity sows (Morrowet al., 1989). Thisphenomenonis called second litter syndrome (SLS). SLS decreases reproductive efficiency of 2nd parity sows and might decrease sow longevity, as reproductive failure is the main reason for culling in young sows (Zak et al., 1997b; Lucia et al., 2000). Genetically there is a high correlation between litter size in parities 1 and 2 (0.88; Holm et al., 2005) and (0.83; Hanenberg et al., 2001), however phenotypic correlation is low (0.04; Hanenberg et al., 2001), indicating a high environmental influence on litter size.
母猪哺乳期间严重的身体储备消耗是造成母猪繁殖失败的一个众所周知的因素(Prunier et al., 2003)。代谢状态对后续繁殖性能的影响已被广泛研究。有文献记载，在哺乳期限制饲料和蛋白质摄入量，因此导致能量负平衡加剧，就会损害卵泡发育(Quesnel et al., 1998; Clowes et al., 2003)，增加断奶至发情间隔(Zak et al., 1997a)，降低排卵率(Zak et al., 1997a; Vinsky et al., 2006)，降低胚胎存活率(Vinsky et al., 2006)和产仔数(Revell et al., 1998; Prunier et al., 2003)。在这些研究中，母猪通常在妊娠早期被屠宰，对猪场变量(如分娩率和窝产仔数)的影响没有得到很好的记录。此外，在这些研究中，都使用了人为限制母猪采食量的做法引起的能量失衡，这可能与“自然形成的”能量失衡不同。
Severe body reserve depletion during lactation is a well known factor associated with reproductive failure in sows (Prunier et al., 2003). Effects of metabolic status on subsequent reproductive functioning have been extensively studied. Feed restriction as well as protein restriction during lactation, and therefore an increased negative energy balance, have been reported to decrease follicular development (Quesnel et al., 1998; Clowes et al., 2003), to increase weaning to estrous interval (Zak et al., 1997a), to decrease ovulation rate (Zak et al., 1997a; Vinsky et al., 2006), and to decrease embryonic survival (Vinsky et al., 2006) and litter size (Revell et al., 1998; Prunier et al., 2003). In these studies sows are usually slaughtered during early pregnancy and effects on farm variables (e.g. pregnancy rate and litter size) are not well documented. More over, in these studies differences in energy balance are induced by restricted feeding of sows, which might be different from ‘natural’ differences in energy balance.
以上大多数研究都集中在一胎母猪上。一胎母猪对身体储备的消耗特别敏感，因为它们在第一次生产时没有足够的身体储备，它们的采食量不足以满足哺乳期间的能量需求(Everts, 1994)。此外，年轻的母猪还需要进一步发育成长才能达到成熟。生长发育主要是蛋白质和脂肪的增加，目的是达到一定的“生理上的生长标准”(Everts and Dekker, 1995)。身体储备的缺乏和生长欲望，使年轻母猪对能量负平衡对繁殖产生的负面影响更加敏感(Prunier et al., 2003)。
Most of the studies above have focused on first litter (first parity) sows. First parity sows are especially sensitive to body reserve depletion because they do not have enough body reserves at first farrowing and their feed intake capacity is not sufficient to fulfill energy needs during lactation (Everts, 1994). In addition, young sows still need to grow to reach maturity. Growth mainly consists of protein and fat accretion and aims to reach a certain ‘intrinsic growth standard’(Everts and Dekker, 1995). The lack of body reserves and desire to grow, make young sows more sensitive to the negative effects of a negative energy balance on reproduction (Prunier et al., 2003).
This retrospective study aims to quantify the association between reproductive performance of second parity sows, focusing on non-pregnancy and litter size in relation to measures of sow live weight development in first parity.
材料和方法 Materials and methods
2.1. 总体 General
Between August 1999 and June 2005, sow development and sow reproduction data were recorded on a routine basis on two experimental farms of Wageningen University and Research Centre in The Netherlands. Farm A had a sow population of about 400 Great Yorkshire×Dutch Landrace sows (Y×DL). Farm B had a sow population of about 300 sows, consisting of Dutch Landrace (DL) and Y×DL sow lines.
在泌乳期间，母猪被圈养在分娩定位栏里，并用商品泌乳料饲喂。母猪在哺乳期第一周逐渐提高饲喂水平，而在哺乳期接下来的时间里是自由采食的。未记录每头母猪的实际采食量。在断奶至配种期间，母猪被圈养在定位栏里，饲喂商品妊娠料，每天能最多3.5公斤。每天用一只成熟的公猪对这些母猪进行2次查情。在出现静立发情的第一天，用商品公猪精液进行配种。若第二天母猪持续静立，则进行第二次配种。在配种4周后进行了一次B超孕检。母猪在妊娠期间被圈养在大栏里，用个体料槽饲喂（A场和B场都有），或者饲喂站饲喂（只有A场有）。所用饲料是商品妊娠料。A场妊娠期前60天饲喂水平为每天2.5 kg，第61 – 85天每天2.8 kg，从第86天起每天3.4 kg。B场母猪在妊娠期前85天的饲喂水平为每天2.6kg(1胎)或2.8kg(1胎以上)，之后每天3.0kg(1胎)或3.4kg(2胎)。
During lactation sows were housed in individual farrowing crates and were fed a commercial lactation diet. After a gradual increase in feeding level during the first week of lactation, sows were fed ad libitum for the remaining lactation. Actual feed intake per sow was not recorded. During the weaning to insemination interval, sows were individually housed in crates and were fed a commercial gestation diet with a maximum of 3.5 kg of per day. Sows were checked for estrus twice a day using a mature boar. On the first day of standing estrus sows were inseminated using a commercial AI dose. A second insemination took place when the standing estrus extended to the next day. An ultrasound pregnancy check was done 4 weeks after insemination. During pregnancy sows were housed in stable groups with either feeding stalls (farms A and B) or feeding stations (farm A). Sows were fed a commercial gestation diet. Feeding level on farm A was 2.5 kg per day for the first 60 days of gestation, 2.8 kg per day from days 61 to 85, and 3.4 kg per day from day 86 for the remaining gestation. Feeding level for farm B was 2.6 (parity 1) or 2.8 (parity >1) kg per day for the first 85 days of gestation, and 3.0 (parity 1) or 3.4 (parity 2) kg per day for the remaining gestation.
Measures of sow development
实验期间测定了母猪整个繁殖期的体重和背脂厚。两个猪场都记录了初次配种时的日龄和体重、妊娠第112天的体重和断奶时的体重。A场也测定了第一次配种、妊娠第112天和断奶时的背脂厚(P2)。根据Everts 和 Dekker (1995)的公式，用体重和背脂厚估算A场母猪的蛋白质储备量：蛋白质(kg)=1.67+0.175×体重(kg)−0.38×P2 (mm)。
Measures of sow development, both body weight and backfat, were taken throughout sow reproductive life. Both farms recorded age and body weight at first insemination, body weight at day 112 of pregnancy and body weight at weaning. On farm A also backfat was measured (P2method) at first insemination, at day 112 of pregnancy and at weaning. Based on body weight and backfat, the protein content of sows on farm A was estimated according to the formula of Everts and Dekker (1995): Protein (kg)=1.67+0.175 × weight (kg) − 0.38 × P2 (mm).
分娩后体重(kg)估计为：产后体重(kg) =妊娠第112天体重(kg)−(总产仔数×1.5 kg)。这1.5公斤是估计的仔猪出生重，包括胎盘和体液。
Weight after farrowing (kg) was estimated as: Weight after farrowing (kg) = weight at day 112 of pregnancy (kg) − (total number of piglets born × 1.5 kg). The 1.5 kg accounts for the estimated average piglet weight at birth including placenta and fluids.
Sow reproduction measurements
For each sow, weaning to insemination interval (WII), number of piglets born alive, number of piglets born dead, and number of piglets weaned were recorded per parity. Total number of piglets born (TNB) was calculated from number born alive and number born dead. Only on farm A individual sow records of cross-fostering and piglet birth weight were registered. A sow was considered pregnant after a positive pregnancy diagnosis at 4 weeks from after insemination.
材料和方法Materials and methods
使用一般线性回归检验了不同胎次和不同猪场之间在母猪发育和繁殖变量方面的差异的显著性(proc glm, SAS Inst. Inc., 2004)。通过检验模型残差来检验正态性假设。
Differences between parities and between farms in sow development and reproduction variables, were tested for significance using general linear regression (proc glm, SAS Inst. Inc., 2004). Assumptions on normality were checked by examining model residuals.
利用两个结果变量分析了母猪发育对繁殖性能的影响。采用逻辑回归分析对“第一次断奶后第一次授精但受孕失败”的二元结果进行了分析(proc logistic, SAS Inst. Inc., 2004)。采用一般线性回归分析方法对二胎母猪第一次授精后的窝产仔数进行分析(proc glm, SAS Inst. Inc., 2004)，其中通过检验模型残差检验正态性假设。返情复配的母猪未计入二胎母猪窝产仔数的分析，因为断奶到配种间隔的延长（母猪更好的恢复）有助于提高窝产仔数(Revell et al., 1998; Prunier et al., 2003)。两个猪场都没有很好地记录母猪淘汰的原因，因此无法对淘汰作出统计推断。
Effects of sow development on reproductive performance were analyzed using two outcome variables. A binary outcome variable ‘non-pregnancy’ (yes/no) was analyzed using logistic regression (proc logistic, SAS Inst. Inc., 2004). Total number born in second parity (TNB2) was analyzed using general linear regression (proc glm, SAS Inst. Inc., 2004), in which assumptions on normality were checked by examining model residuals. Repeat breeders were omitted from the analysis on TNB2, since a prolonged (recovery) period between weaning and conception can positively influence litter size and therefore mask possible effects on litter size (Revell et al., 1998; Prunier et al., 2003). Reason of culling was not registered properly on both farms and, therefore, no statistical inferences on culling could be made.
检验的解释变量为：初配时的日龄和体重；第一次分娩时的体重；第一个妊娠期的体重增加，哺乳期体重的减少，第一次断奶时的体重，第一胎的总产仔数，第一胎的断奶仔猪数，配种和分娩的季节，母猪品系（只有B场记录），以及妊娠期的圈舍条件（只有A场记录）。如果以连续量表测量的解释变量与因变量线性相关，则将其作为连续变量进行分析。如果自变量与因变量不线性相关，则将自变量分类分析为类变量(断奶至配种间隔(≤4,5,6 - 20，≥21天)、哺乳期相对体重损失(≤15%和>15%)。
Explanatory variables tested were: age and weight at first insemination; weight at first farrowing; weight gain during first pregnancy, weight loss during lactation, weight at first weaning, total number of piglets born in first parity, number of piglets weaned in first parity, season of insemination and farrowing, sow line (only farm B) and housing during gestation (only farm A). If explanatory variables, measured on a continuous scale, were linearly related to the dependent variable they were analyzed as continuous variables. If independent variables were not linearly related to the dependent variable, they were categorized and analyzed as class variables (weaning to insemination interval (≤4, 5, 6–20, ≥21 days)), relative weight loss during lactation (≤15% and >15%).
在多变量模型中包含高度相关的变量会导致共线性问题。因此，从更大的一组解释变量中，从统计学上选择出最“重要”的预测因子可能是非常困难的。如果分析的目的是预测，这个问题就不是那么严重了，但当分析的目的是解释因果效应时，这就成了一个比较棘手的问题(Hosmer and Lemeshow, 1989)。为了避免多变量模型中的共线性问题，在多变量模型中只包含皮尔逊相关系数为0.5的变量(表1)。如果变量高度相关，则选择与其他变量数量最多的相关变量纳入多变量模型。把变量的数量最小化可以得到一个数值上稳定的模型，并减少标准误差(Neter et al., 1985)。由于我们的目标不是预测，而且由于解释变量与妊娠失败的高度相关性，本研究也给出了单变量分析的结果。所有可能影响二胎总产仔数的变量，以及双向交互作用，都包含在一个多变量模型中。在逆向剔除过程中，将最不显著的交互作用或变量从模型中剔除，直到最终模型只包含显著变量(P≤0.05)。
Including highly correlated variables in a multivariable model results in co-linearity problems. As a consequence of that, it might be difficult to statistically select the most “important” predictors from a larger group of explanatory variables. This is less serious if the purpose of analysis is prediction, but it is a problemwheninterpretation of causal effects is the objective (Hosmer and Lemeshow, 1989). To avoid co-linearity problems in multivariable models, only variables with a Pearson correlation coefficient of <0.5 (Table 1) were included in the multivariable model. If variables were highly correlated, the variable that correlated with the highest number of other variables was chosen to be included in the multivariable model. Minimizing the number of variables results in a model that is numerically stable, and reduces the standard errors (Neter et al., 1985). As our goal was not prediction, and due to high correlations between explanatory variables associated with non-pregnancy, also results of univariable analyses are presented. All variables possibly affecting TNB2, as well as two-way interactions, were included in a multivariable model. In a backward elimination procedure, the least significant interaction or variable was eliminated from the model until the final model only contained significant (P≤0.05) variables.
Table 1: Pearsons correlation coefficients between variables of sow development and reproductive performance in first and second parity sows.
* Indicates significant correlation coefficients.
妊娠失败的结果以妊娠失败率、比值比(OR)和95%置信区间(CI)表示。比值比是相对风险的估计值，即“暴露”组的发生率或累积发生率除以参照组的累积发生率。如果比值比等于1，则该变量与结果变量之间不存在关联；如果比值比小于1，则该变量风险较小；如果或比值比大于1，则该变量风险增加(Frankena and Thrusfield, 2001)。在本文的结果部分，只列出了与妊娠失败和2胎窝产总数有显著相关性的变量(P<0.05)。
Non-pregnancy results are presented as percentage of non-pregnancy, odds ratio (OR) and 95% confidence interval (CI). An OR is an estimate of the relative risk, which is the prevalence or cumulative incidence in the ‘exposed’ group divided by the prevalence of cumulative incidence in the reference group. If an OR equals 1, then there is no association between the variable and the outcome variable; if OR is smaller than 1, the variable imposes a decreased risk; if OR is larger than 1 the variable imposes an increased risk (Frankena and Thrusfield, 2001). In the results section of this paper, only variables significantly (P<0.05) associated with non-pregnancy andTNB2are presented.
To be continued…