Factors for high reproductive performance of sows in commercial herds

Introduction

With information technology, commercial herds have collected and stored many data. New technologies are expanding the possibilities for data collection, information-exchange, collaboration and data analysis. However, the use of these data have been limited. Farm data analysis can help veterinarians and producers to identify a production problem that they did not recognize and make a better decision about solutions. Furthermore, data analysis could increase the dissemination of useful information to improve herd productivity and stable outputs in breeding herds. It could identify important factors associated with reproductive performance in order to maximize sows’ reproductive potential. In this paper, we review factors associated with sow performance and herd productivity in commercial herds. The factors include both sow level factors and herd level factors. With regard to sow level factors, there are risk factors and other factors. Reproductive performance is not a disease and there is a case that the risk factor is not an appropriate term. For example, PBA is not a risk factor, but actually a predicting factor for lifetime prolificacy of sows (Iida et al., 2015). Factors at the herd level include high-performing herds, herd management factors and herd size. Additionally, boar factors should also be considered in order to improve reproductive performance of sows.

A. Forty pigs weaned per sow per year

The number of pigs weaned per sow per year (PWSY) is commonly used as a benchmarking measurement to compare the productivity of breeding herds between herds or countries. The target values for PWSY have been increased from 20 to 30 pigs over the last three decades, and it is likely that 40 PWSY will be the next target in the swine industry. To achieve 40 PWSY, it is necessary to obtain both 17.3 pigs weaned per sow and 2.3 litters per sow per year by assuming 28 days of lactation, 115 days of gestation and 36 days of non-productive sow days (Dial et al., 1992; Almond et al., 2006). It is likely that genetics and sow management can increase PWSY up to 40 pigs in the near future. However, even though PWSY is a good measurement for herd productivity in the short term, it is not the best measurement for the longevity nor the welfare of piglets or sows. There is a concern that herds with 40 PWSY may produce many runts or small piglets. So welfare in piglets may be compromised when we genetically increase sow prolificacy to such a high level, unless genetic improvement is directed to increasing the uterine capacity, number of functional teats and milk production in sows.

B. Reproductive performance in commercial herds

Sow reproductive performance

In the productivity tree of breeding herds (Dial et al., 1992), there are two branches: one is the number of pigs weaned per sow, and the other is the number of litters per sow per year. The number of pigs weaned depends on the number of pigs born alive and preweaning mortality; the number of litters per sow per year depends on non-productive days, lactation length and gestation length.

Sow reproductive performance includes both fertility (e.g. farrowing rate: FR and weaning-to-mating interval: WMI) and prolificacy (e.g. pigs born alive: PBA). In terms of fertility, the number of litters per sow per year is affected by FR and WMI as well as reservice interval and culling interval, via their effects on non-productive days. Meanwhile, prolificacy is determined by the effects of PBA and preweaning mortality on the number of pigs weaned.

Sow mortality is related to fertility because increased mortality increases death intervals and non-productive days which decreases lifetime fertility in sows. Also, abortion occurrences in commercial herds increase non-productive days of gilts and sows (Iida et al., 2016).

Lifetime performance

Mean sow life days is approximately 1,000 days in southern European countries and Japan. It is important for producers to maximize reproductive potential during sows’ lifetime in order to decrease production costs and economic inefficiency in commercial breeding herds (Stalder et al., 2012).

Lifetime performance includes longevity, that is measured as the number of parity or age at culling or removal, and also lifetime PBA, lifetime number of pigs weaned and lifetime non-productive sow days (Sasaki et al., 2011). Annualized lifetime PBA is an integrated prolificacy measurement for sows that combines lifetime PBA with lifetime non-productive sow days. In contrast, annualized lifetime pigs weaned can be considered as an integrated measurement of sows’ lifetime reproductive productivity that combines sow performance (i.e. PBA and preweaning mortality) with lactation management including nursing and fostering techniques. For example, annualized lifetime pigs born alive per sow is calculated as the number of lifetime pigs born alive divided by the sow’s reproductive herd life days x 365 days. The sow’s reproductive herd life days is the number of days from the date that the sow was first-mated to its removal. For replacement gilts, the date of first-mating appears to be more appropriate than herd entry date, because the herd entry date varies between herds.

C. Sow level information- factors for sow performance

Low or high parity

Low parity females, especially pregnant gilts and parity 1 sows, have lower reproductive performance than parities 2-5 sows, such as lower FR, higher returns, lower PBA and higher WMI. As the number of parity increases, reproductive performance also increases, reaching a peak between parities 2-5 before it then declines. For example, PBA is highest between parities 3 and 5, whereas FR is highest between parities 2 and 4. Parity 1 sows also have a prolonged WMI which can be explained by the immature endocrine system in these growing young animals, and their low feed intake during lactation which decreases LH secretion (Koketsu et al., 1996a) leading to restricted follicle growth in their ovaries. Parity 1 sows in commercial breeding herds may not consume sufficient nutrients and energy to grow and reach their mature reproductive performance level.

Aged sows also have lower reproductive performance than parities 2-5 sows. There are various reasons for the lower performance. For example, ovulation and fertilization rates decrease in aged sows. Also, their embryonic mortality or pregnancy loss and the number of stillborn piglets increase due to slow responses to the space demands by growing fetuses and to the stimuli from parturition processes (Almond et al., 2006). Additionally, aged sows (parity 5 or higher sows), and also gilts, are at higher risk of having abortion than parities 3-5 sows (Iida et al., 2016).

Season or climate factors

Fertility and prolificacy measurements decrease during summer months. For example, FR is lowest in summer, and also PBA in summer-mated sows is lower than for winter or spring mated-sows. It has been hypothesized that the summer reduction in reproductive performance occurs through a combination of high summer temperatures reducing GnRH secretion, and impaired ovarian follicle development that leads to compromised corpus lutea functions secreting low progesterone concentrations (Bertoldo et al., 2012).

Various studies have highlighted important climatic factors related to the seasonal effects, including daily maximum and minimum temperature, humidity and photoperiod. Climate data in Meteorological stations near studied herds have been used to quantify the association between high temperatures and sow performance (Tummaruk, 2012; Bloemhof et al., 2013; Iida and Koketsu, 2013; 2014b). For example, increased outside temperature decreases FR and total number of pigs born, while it increases returns, WMI and mortality. Also, the impact of the summer effect or outdoor temperature on reproductive performance varies depending on parity number. A previous study showed that as outside temperature increased from 25 to 30 C0, the total number of pigs born to parity 1 decreases by 0.6 pigs at their subsequent parity, whereas for parity 0 females the decreases at subsequent parity was only 0.2 pigs (Iida and Koketsu, 2014b). Another example is that WMI in parity 1 sows increased by 0.8 days as daily maximum temperature rose from 25 to 35 °C, whereas in parities 2 or higher sows the increase in WMI was only 0.3 days (Iida and Koketsu, 2013). These results indicate that that parity 1 sows are more sensitive to such summer changes in climate than gilts or sows at parity 2 or higher. This type of sensitivity appears to be related to the immature endocrine system in parity 1 sows and the low feed intake of parity 1 sows during lactation.

Lactation feed intake and its patterns

Lower lactation feed intake is associated with lower average weaning weight, prolonged WMI, low FR, more returns or more culled sows due to reproductive failure, and fewer PBA at subsequent parity (Koketsu et al., 1996b). This is particularly the case with parity 1 sows where low feed intake during lactation is a detrimental factor related to post weaning reproductive performance such as WMI and FR. Some lactational feed intake patterns (e.g., major dip) are related to prolonged WMI and more culled sows due to reproductive failure. However, increased lactation length and advanced automatic feeders for lactating sows may change these risks to reproductive performance.

Lactation length

There has been a concern about early weaning systems in the U.S.A. being associated with suboptimal reproductive performance, such as low FR, prolonged WMI and fewer PBA at subsequent parity (Koketsu et al., 1998). Also, short lactation length decreases average feed intake during lactation. However, since 2000, the U.S.A. swine industry has been moving from early weaning to increased lactation length (Knauer and Hostetler, 2013) to increase improve growth performance in nursery and grower pigs. Also, in the European Union the weaning of piglets from the sow at less than 28 days of age has been prohibited since 2013 (European commission: Animal welfare practices, 2015). Meanwhile, there is another concern that some nurse sows with increased lactation length lose their body reserve much due to high milk yields, so they may have increased prolonged WMI and lower FR.

Number of inseminations or matings

Deciding the ideal number of inseminations is related to optimizing reproductive performance and minimizing costs. Single insemination, due in part to late timing, is related to low FR (Kaneko et al., 2013). Inseminating two times with accurate heat detection is more cost-effective than three times inseminations in terms of the costs for labor, semen and a catheter (Takai et al., 2010). However, a GnRH antagonist given intravaginally in gel form has been shown to be effective at advancing and synchronizing ovulation (Knox et al., 2014). Using this GnRH technology, a single insemination has been practiced in the U.S.A. industry enabling reduced costs while still having reproductive performance similar with multiple inseminations.

Peri-partum period or farrowing event

Sow mortality is an indicator of maternal health and animal welfare. Farrowing is a major risk factor for sows in all parities and seasons. A previous study showed approximately 68% sow deaths occurred in the period from 4 weeks before farrowing to 4 weeks after farrowing (Iida and Koketsu, 2014a). As the number of parity increases, the mortality risk for sows also increases. So aged sows in high parity (e.g., parity 6 or higher) in the peri-partum period are at the highest risk of dying (Sasaki and Koketsu, 2008).

It has also been shown that in subtropical climate zones mortality increases in low parity sows increases during summer, whereas in aged sows it increases during winter (Iida and Koketsu, 2014a). It appears that lower parity females that have immature bodies are more sensitive to high ambient temperature than multiparous sows. Pigs are particularly susceptible to heat stress because they have a weak cardiovascular system and limited sweat glands (Frazer, 1970). Distortions in abdominal organs and heart failure are major causes for death in female pigs (Stalder et al., 2012). Similar cardiovascular problems can occur in (human) women; for example, peripartum cardiomyopathy is reported as a disorder in which initial left ventricular systolic dysfunction and heart failure occur (Silwa et al., 2006).

Aged sows are more sensitive to low temperature than low parity females in subtropical climate zones (Iida and Koketsu, 2014a). In humans also, the prevalence of gestational hypertension, pre-eclampsia and eclampsia are highest with delivery in the winter months (TePoel et al., 2011). Such diseases may be related to aged sow responses to cold or to large variation in daily temperature during winter. Additionally, in subtropical climate zones, the facilities for herds such as heating equipment and building insulation do not appear to be sufficient.

Weaning-to-mating interval (WMI)

The WMI is a reproductive performance measurement associated with PBA, FR and returns. Sows with a short WMI that are bred between 3 and 6 days after weaning have higher FR and PBA than those bred between 7 and 20 days post weaning (Hoshino and Koketsu, 2008; Tummaruk et al., 2010). The WMI tends to be increased by short lactation length and low feed intake during lactation (Koketsu et al., 1996b). In addition, prolonged WMI is related to a short duration of estrus and a shorter interval between onset of estrus and ovulation (Weitze et al., 1994; Kemp and Soede, 1996). A consequence of this is an increased risk of inseminating at a suboptimal period, which can be a major cause of low FR and low PBA. As previously mentioned, the use of a GnRH antagonist given to sows intravaginally facilitates a single dose fixed-time insemination in weaned sows. If this practice becomes common, WMI may become a less important factor for reproductive performance.

Number of pigs born alive (PBA)

The PBA in parity 1 is a factor that can help producers to identify high prolific sows at an early stage (Iida et al, 2015). A sow’s PBA is determined by environmental or management factors and genetic potential (Hoving et al., 2011). Sows that have a high PBA in parity 1 typically produce high PBA throughout all the subsequent parities, and have high FR up to parity 3. These high prolific sows also have high lifetime reproductive performance.

Birth weight and preweaning growth rate

Litter-of-origin in sows includes their birth weights and preweaning growth rate. Gilts with lower age at puberty have heavier birth weights and higher preweaning growth (Vallet et al., 2016). These characteristics may affect subsequent reproductive performance of sows. Preweaning growth is affected by sow milk production, whereas heavier birthweights are associated with fewer pigs born in the litter.

Number of pigs weaned

An increased number of pigs weaned or heavier litter weight at weaning could impair a sow’s post weaning reproductive performance due to increased loss of body reserves in the lactating sow. Therefore, the use of fostering and nurse-sow techniques impairs the metabolic state of sows and decreases post weaning reproductive performance (Quesnel et al., 2007). Also, sows that fostered 3 or more piglets have been found to have prolonged WMI (Usui and Koketsu, 2013).

Age of gilts at first-mating (AFS)

Gilt development and management is a critical to optimize the reproductive performance of sows. It is useful to record age of gilts at first estrus and dates of heat-no-serve in gilt development and management. However, the age of gilts at first estrus and dates of heat-no-serve are hardly recorded in commercial swine herds in North America, whereas AFS is commonly recorded (Patterson et al., 2010). Therefore, AFS in herd data analysis is still a factor for PBA and lifetime performance in commercial herds.

Increased AFS is associated with increased PBA in parity 1 (Iida et al., 2015). In the U.S.A., southern E.U. and Japan a typical AFS of approximately 240 days has been practiced to increase body weights and more body reserves of replacement gilts to be first-mated.

Number of stillborn piglets

By definition, stillborn piglets are those piglets that are alive at the initiation of farrowing but die intrapartum (Dial et al., 1992). In practice, the stillborn piglets in commercial herds are categorized as piglets found dead behind the sow at the first check up after parturition, with no sign of decomposition (Vanderhaeghe et al., 2013). Like AFS or WMI, the number of stillborn piglets is a factor in sow performance and is related to other aspects of reproductive performance. For example, abortion risk for sows has been found to be associated with sows having stillborn piglets (Iida et al., 2016). Such an association between an increase in abortions and the number of stillborn piglets could be explained by having infectious agents, such as porcine parvovirus, and also porcine reproductive and respiratory syndrome virus (Almond et al., 2006).

D. Herd-level information

Herd-level information includes various useful factors that can be used to characterize a production system. Herd characteristics, management practices, production systems and facility types can all be analyzed as herd level information.

Herd size

Herd size is an indicator of how advanced a production system is, including the amount of investment and the quality of the facilities and human resources and the level of genetic improvement. Larger herds were associated with high PWSY (King et al., 1998). The herd size itself does not appear to directly increase PWSY, but large herds tend to be able to hire more specialized workers and use better facilities than small herds. Also, there may be more rapid genetic improvement and a better production system in larger herds.

High-performing herds

The concept of high-performing herds is related to best-practice benchmarking, which has been used to provide values for target performance and efficiency (Koketsu, 2007). Herds can be categorized into two herd categories based on PWSY: high-performing herds and ordinary herds. In southern European countries, high-performing herds have 4-7% higher FR and 4-6% lower return risks across parity than ordinary herds. Consequently, these high-performing herds have fewer non-productive days, such as reservice interval and culling interval. Also, the high-performing herds have 0.6-0.9 pigs more PBA and 0.8-0.9 more pigs weaned across parities than ordinary herds. With regards to culling management, the high-performing herds have lower culling rates from parities 0 to 5 but higher culling rates in parity 6 or higher than ordinary herds.

Herd management factors

Information relating to herd management factors can be collected by the means of a questionnaire survey. The survey can collect information about gilt development programs, insemination timings, farrowing and lactation management, farrowing spaces and culling guidelines and so on. Analysis of such herd management information shows that herds performing first insemination immediately after first detection for gilts, or within 6-12 hours for sows, had higher FR than those with later times for insemination (Kaneko et al., 2013). Furthermore, the analysis showed that AFM of gilts in the herds using direct boar contact was 14 days less than that in the herds using indirect contact (Kaneko and Koketsu, 2012). Another finding from herd management analysis is that actual culling intervals for mated gilts and sows were at least 30 days longer than the guideline culling interval (Sasaki and Koketsu, 2012).

Within-herd variability for number of mated females or age structure

A consistent flow of pigs through a production facility becomes more important as production systems become more standardized. Using a group measurement of females in breeding herds, within-herd variability in the flow of pigs in a breeding operation can be measured as the number of females mated per week, over a 52-week period. Large within-herd variability in the number of females mated 16-19 weeks previously is associated with lower annual FR, increased non-productive days and decreased herd reproductive productivity (Koketsu et al., 1999) and farrowing space utilization (Koketsu et al., 2015).

Herds that have a stable age structure over 2 years have higher FR than those in herds having an unstable age structure (Koketsu, 2005a). This is because the herds with a stable age structure had higher proportions of parities 3-5 sows and a lower proportion of gilts than the herds with an unstable age structure. Therefore, within-herd variability affects efficiency in breeding herds.

Number of farrowing spaces

A limited number of farrowing spaces is a bottleneck for pig production in most breeding herds. A recent study showed that sows in high-performing herds, based on the number of pigs weaned per farrowing space per year, produced 130 (+ 3.5) pigs and 836 (+ 2.3) kg (Koketsu et al., 2015). A higher farrowing space utilization efficiency is associated with lower within-herd variability measured as the coefficient of variation (%) for the number of females mated 16 weeks previously.

Boar and semen factors

Semen characteristics including motility parameters may affect reproductive performance of sows. However, data about semen characteristics in boar studs have not been combined well with sow performance data in commercial herds (personal communication from Dr. Neil DeBuse, U.S.A.). There needs to be more research integrating field data on boar semen quality with reproductive performance of sows in order to identify the causes of poor performance at boar, sow and herd levels, and also to determine the motility parameters and the optimum number of motile cells in a dose (Broekhuijse et al., 2011).

E. Limitations of data analysis using commercial herd data

There are certain production research areas which can be investigated using epidemiological studies or which are suitable for situations that would require excessive funding or time to conduct in controlled experiments. As information technology advances, production research can evolve by using commercial herd data to disseminate useful information for producers and veterinarians.

However, there are several limitations with non-controlled observational studies that would not occur in controlled experiments led by university researchers. For example, some commercial herd data that are recorded incorrectly, and this means that some exclusion criteria are essential. Also, herd health, nutrition, management practices and genotype may not be well controlled in observation studies. Additionally, sows are not randomly selected and multiple observations per sow are not independent units of observation. One other limitation is that herds’ data are also in a two level structure because management practices, production systems, facilities and herd health programs vary between herds, i.e., sows are not independent of the herd. Even with such limitations, herd data analysis using appropriate exclusion criteria and multi-level statistical models can disseminate practical and readily applicable information to swine veterinarians and producers about production issues that are difficult to be investigated by controlled experiments.

Conclusions

It is critical for veterinarians to know about the factors affecting reproductive performance in order to optimize their clients’ breeding herd productivity. Improving the herd management that controls these factors, together with genetic improvement, will enable us to reach 40 PWMSY. Finally, in order to empower data analysis it is necessary to ensure correct data recording, data collection and data integrity checks.

Presented at the 24th International Pig Veterinary Society Congress. For information on the next edition, click here.