Focusing on Reproductive Efficiency

This quarter's benchmarking article will concentrate on two areas. We will begin by comparing production records from the previous quarter and previous year with values from a year ago. This review will help us see what progress we have made in key performance indicators (KPIs).

Secondly, we will examine various factors that contribute to the inefficiencies of the breeding herd. We'll take a closer look at how the abilities of various artificial insemination (AI) technicians can impact reproductive levels. In addition, we'll study the impact that time of day, day of the week and multiple inseminations have on various reproductive performance indicators.

The database used for this analysis is again provided by Swine Management Services LLC (SMS), Fremont, NE. As in the past, we will use pigs weaned/mated female/year (PW/MF/Y) to evaluate whether progress has been made by comparing the 52-week average and 13-week average performance of the first quarter 2007 (Table 2) to the same time period in 2008 (Table 1).

Pigs Weaned Edges Upward

Some interesting points arise when comparing the 2007 and 2008 data. Using the 52-week average KPI values, it is apparent that the top 10% of herds has changed very little. This demonstrates how difficult it is to improve productivity at the upper echelon of the industry.

For example, PW/MF/Y, arguably one of the most important indicators of herd performance, shows a slight drop from 26.8 in 2007 to 26.5 in 2008 in the top-10% farm category. However, when PW/MF/Y is compared for all farms contributing to the database, the 2008 average (22.8) was 0.3 pigs higher than the 2007 average (22.5).

While the best herds failed to improve this important KPI for the 52-week period, the average producer made some gains.

When the 13-week PW/MF/Y averages were compared, again the top 10% of farms remained relatively static. However, the all-farm average jumped over half a pig (22.21 vs. 22.75) for the quarter.

A comparison of other KPI averages shows relatively minor changes when the top-10% and all-farm categories are reviewed. For example, farrowing rate for the top 10% of farms was 87.1% in 2007 and 87.4% in the 52-week comparison for 2008. Likewise, farrowing rate for all farms tallied in at 82.4% and 82.5%, respectively. The difference between the top producers and average producers was maintained at about 5%.

Still, farrowing rate is important when the focus is on improving herd efficiency. Obviously, sows must farrow in order to have a saleable product — whether producing weaned pigs or market hogs. While gains in this KPI do not make up for the adverse feed costs and low market hog prices, it is important to note that the productivity gains made in this area will likely prevail as the market hog price cycles recover to more profitable levels.

Additionally, farrowing rate for the average producer improved from 80.2% in 2007 to 81.8% in 2008 in the 13-week comparison.

Since the average values for the top 10% of operations actually declined slightly (87.9% vs. 87.5%) from 2007 to 2008, the improvement shown by all farms in the farrowing rate category would suggest the improvement was made by those who needed it most. It is a bit discouraging to see the farrowing rate of the bottom 25% slip, however.

Finally, it is worth noting that average parity of culled sows in the 52-week summary has dropped in the top 10% of herds (3.07 in 2007 vs. 2.45 in 2008). The reverse is true when all-farms parity averages are compared in the 52-week data. The same trend is reflected in the 13-week summary.

A Closer Look at the AI Data

The second objective of this quarter's benchmarking review is to examine the factors that contribute to breeding herd inefficiency. In an effort to drill down to find some deficiencies related to AI, SMS provided data focusing on the effects of the AI technician, day-of-the-week comparisons and time-of-day effects. Farrowing rate is the primary KPI used for these comparisons.

Clearly, there are differences between technicians when it comes to a successful AI program. Figure 1 shows the farrowing rate by AI technician, which shows whether the technician made both the first and second matings for a given sow or only the first, second or the third mating, respectively.

Figure 1 reinforces that it is difficult to have the same technician mate an individual sow when she is in estrus for consecutive days. The data serves as a good illustration of the team effort to get sows successfully mated. Additionally, the data can help identify those technicians who are part of a team where poor farrowing rates are consistently found. It is likely that the same technicians in that team would be even poorer if individual sows mated by individual technicians could be identified.

In this case, technicians #8, #2 and #9 have consistently been on teams that have an 85% farrowing rate and above. Conversely, technicians #7 and #5 have consistently been part of teams with a farrowing rate of 80% and below. Again, it is likely that these technicians' farrowing rates would have been even poorer if they mated sows exclusively.

These poor performers are candidates for additional AI technique training and, perhaps, for estrous detection training, which could be provided by their more successful co-workers. External trainers could also be called in.

The day of the week on which inseminations occur can have a significant impact on an AI program. Figure 2 features eight AI technicians and the farrowing rates they achieved, listed by the day of the week.

It is clear that some technicians have more success on different days of the week. For example, technician #1 has a farrowing rate that averages above 84% when inseminations occur on Sunday, Monday, Tuesday, Friday and Saturday. However, Wednesdays and Thursdays are not good days for this technician, whose farrowing rate average slips to about 75%.

Several factors could contribute to this situation. In this case, notice that virtually all technicians have their poorest farrowing rate when matings occur on Wednesday and Thursday. This slump could be caused by some factor or event that takes place on those days or it may just be that Wednesday and Thursday encompasses all of the problem breeding sows or those that have a higher-than-normal wean-to-estrus interval.

One possibility is that this farm may wean sows on Tuesdays and Thursdays, and this technician does a great job of identifying those that come into estrus during the normal time period after weaning. This same technician may have difficulty identifying sows that have an irregular return to estrus.

Another explanation for day-of-the-week differences in farrowing rate may reflect a technician's personal habits. For example, Tuesday night might be this technician's bowling night, where he/she stays out later than normal, and arrives tired on Wednesday morning.

There are a whole host of factors that can contribute to different performance levels. This type of data can help identify problem areas and serve as a basis to explore specific routines that contribute to varying results, whether they are abnormally high or low on specific days of the week.

The last area to examine in this data set is the hour of the day that inseminations occurred (Figure 3). Clearly, sows mated early in the morning or early in the afternoon have a higher success rate as measured by the farrowing rate.

Again, several factors may be in play here. One could be a technician effect. A technician charged with making many matings throughout the day may become fatigued. In that case, successful matings would likely be those made early in the day or at the beginning of a work shift. Others may do their best work after a noon lunch break. Figure 3 supports both scenarios.

Another contributing factor might be how strongly sows display estrus. It is well documented that when sows stand for a long period and get tired, they show less distinct signs of estrus. The results in Figure 3 could reflect this phenomenon where sows are well rested and show good signs of estrus early in the day.

Also worth considering, if the technician does not control the boar used for estrous detection, sows may be stimulated into a standing heat, but then tire and go into a refractory period. If the boar is removed from the breeding area during the worker's break times, for example, the sows may become rested and show good response when heat checked later.

Finally, some technicians have substantially better success with AI at certain times of the day (Figure 4). We all know people who can get more work done or have more energy in the morning vs. the afternoon. This is best illustrated by technician #6, whose farrowing rate with sows bred before noon is below 80% — and sometimes as low as 67%. However, when that person inseminates sows after lunch, farrowing rates jump to 85% or higher. Clearly, this individual is not a morning person.

Management of this operation might consider assigning this person to easier tasks in the morning, then put them back on the breeding team in the afternoon. Several factors could contribute to this situation. Regardless of the cause, the important point is that managers can use this type of information to identify these trends and take action to improve the efficiency of their operations.

This data clearly shows the importance of identifying each technician and the day and time that they are most effective. This information can help fine-tune procedures and identify employees who might benefit from additional training. Similar methods could help identify and track variations in semen handling or quality. Not all batches of semen are equal.

Targeting Greater Efficiency

Ever-increasing feed prices are a constant reminder that we must continue to improve efficiencies in pork production. Today, the focus is on feed costs and market prices. In the future, the emphasis could be totally different. Regardless, producers' abilities to identify areas that demand attention and afford the greatest payback potential will determine their success in the business and the U.S. pork industry's success in today's global marketplace.

If you have thoughts or questions about how to best utilize this benchmarking information, contact Stalder at stalder@iastate.edu; National Hog Farmer Editor Dale Miller at dpmiller@nationalhogfarmer.com; or the SMS staff at 402-727-6600.