Data from 22 farms in the Swine Management Services (SMS) database was used for this article. We used two specialized reports developed by SMS called the In-Depth Breeding Analysis Report and Breeding Technician Report to analyze the farms.

These reports give us detailed information about AI (artificial insemination) technicians, semen batches, sow parity, number of matings and services per female bred, time of day females are breed, day of week, etc. We use these reports with several of our customers to give them detailed feedback on AI technicians. It breaks down the current breeding protocols to review results, and then we make suggestions for making minor or major changes to the current protocols to improve farrowing rate.

In this article, we will focus on the effect the number of services a female is given before culling has on farrowing rate. The definition of a service is described as an estrus period that occurs every 18-25 days for open females. A second-service female is a female that was bred and returned to heat before farrowing and was rebred. We also list females with three-plus services that have been rebred after two returns to heat without farrowing.

Table 1 Farrowing Rate by Number of Services looks at 122,407 services from 22 sow farms which are different sizes, geographical locations and genetic companies. The data was broken out by parity for one service, two services and three-plus services. As you can see at these farms, the average farrowing rate was 85.6% for the 52 weeks of data. There was not a lot of variation in farrowing rate by parity except for Parity 0 females at 82.3%. In the data we review for producers who have the top farms for farrowing rates, we see a much higher farrowing rate for P0 females with several farms at 90%-plus farrowing rate. If your P0 farrowing rate is low, then you need to review development and breeding procedures on gilts. If you want to improve your overall farrowing rate, it starts with the gilts.

In the data set, about 9% of the sows that returned one time to heat were rebred and only about 9% of the females returning a second time were bred a third time. The farrowing rate for females on first service was recorded at 87.1%, dropping by 15.8% to 71.3% for females after two returns and down to 56.5% for the few females served three-plus times. That is a difference of 14.8% for the latter two groups and a 30.6% drop for females that conceive on first service vs. being rebred two times.

In looking at this data, we suggest that returns after two services be culled. There are a few farms starting to cull returns on the first repeat, especially in older parity females, because the feeling is that breeding a few more gilts will improve farrowing rate and performance at the farms.

Chart 1 shows again that there is not a lot of variation in farrowing rate by parity for first- and second-service females. However, there is some drop in farrowing rate for the few females with three-plus services with P7-plus females dropping to a low farrowing rate of 40.1%.

Table 2 Service Result Percent breaks down the returns by the reason for the returns to heat or being culled. Of the 122,407 sows bred, 104,725 farrowed with a farrowing rate of 85.6%. The returns were broken down into six areas with the percentage being of the total females bred: regular returns, 3.7%; irregular returns, 3.0%; negative pregnancy test, 2.8%; abortions, 1.1%; not in pig, 1.0% and removed at 2.9%. The removed groups were females that were still pregnant when they were either culled or died. We have not been able to find other data sets to validate our numbers especially for abortions at 1.1%. These numbers can be seasonal with most being affected by summer breeding.

In Chart 2, the data shows that P0 females have the highest percent regular returns at 5.4%, and more females were removed with the older parities.

In Table 3 Farrowing Rate by Returns to Estrus Days and Chart 3, the data for the 10,000-female second-service females are broken down by days the return females were found open and then by parity. Days to find returns are quite varied within this data base, with an average of 46.3 days. Several farms we do analysis on have average days to find returns as low as 23 days, with 80%-plus of the returns found by Day 25. The target of finding returns should be less than 30 days. This does a lot to lower nonproductive female days and reduce the cost of keeping open females on the farm.

The data shows that about 3% of the returns are found on Days 11-17 which are irregular returns. If we are seeing too many of these females, we question the timing of the first mating by the breeders. Regular returns on Days 18-24 are sows that did not conceive on the previous breeding and should be found with the heat check boars. The problem we sometimes see is with the regular returns on Days 39-45. We feel these open females were missed during the heat-checking process on Days 16-25 or during pregnancy checking. Top farms will find 60%-plus of the open females by Day 25 of gestation. In our data set, there were 30% of the returns found after Day 46 of gestation. This number should be less than 10%; if higher, review the whole process of heat and pregnancy checking.

Chart 3 shows the variation in farrowing rate by days the returns are found and the trend line for lower farrowing rate for females found later in gestation.

With increasing feed costs and the need to wean more pigs per female per year to reduce the cost to produce that weaned pig, more time and training need to go into finding returns and putting procedures in place to remove the returns to lower open sow days and improve farrowing rate.

Key Performance Indicators
Tables 4 and 5 provide 52-week and 13-week rolling averages for key performance indicators (KPI) of breeding herd performance. These tables reflect the most current quarterly data available and are presented with each column. The KPI’s can be used as general guidelines to measure the productivity of your herd compared to the top 10% and top 25% of farms, the average performance for all farms and the bottom 25% of farms in the SMS database.

If you have questions or comments about these columns, or if you have a specific performance measurement that you would like to see benchmarked in our database, please address them to: or

Click to view graphs.

Mark Rix and Ron Ketchem
Swine Management Services LLC