Pilot project tracks effect of three feeding regimes on gilts' productivity through four parities.

How should developing gilts be fed to ensure they will be productive in the breeding herd?

Do gilts from one genetic line respond differently than another to changes in protein or energy levels?

Does the gilt development diet affect gilts' reproductive ability throughout their lives?

It is commonly believed that gilts of different genetic lines should be fed differently during their development. However, there is little scientific evidence to support this notion.

The National Pork Producers Council Genetics Programs Committee was anxious to find some definitive answers before initiating the Maternal Line Evaluation Project. They needed to know whether gilts representing each of the participating lines would require different developmental diets to fully express their genetic potential for reproductive and maternal genetic differences.

To find the answers, they designed a pilot study using gilts of five genetic lines representing a fairly broad cross section of growth, leanness and reproduction. Gilts from each line were fed one of three developmental diets. Reproduction was tracked from expression of first estrus through four parities.

If all gilts of the five lines responded the same to the diets tested, researchers felt they could identify a ration that would maximize reproductive performance across all the lines sampled.

But, if the lines responded differently to the three diets, then additional research could identify the best possible diet for each genetic line.

Pilot Project Protocol Groups of 140-145 gilts from each of five genetic lines were used (Duroc-Hampshire cross, Yorkshire-Landrace cross, plus three breeding company maternal lines from Genetic Improvement Services (GIS), National Genetic Technology (NGT) and Genetipork).

As the sampling methods did not permit reliable estimates of line differences, gilts were intentionally selected to represent a range of genetic variation for growth, backfat and reproduction. This allowed us to maximize the chance of detecting whether lines responded differently to the diets. The goal was not to evaluate genetic differences among lines.

All gilts were born the same week, weaned at 8-15 days of age, and transported to a SEW (segregated early weaning) nursery where they stayed until they weighed 50 lb. They were moved to an environmentally controlled building at the Minnesota Pork Producers Association (MPPA) Swine Testing Station. Typical SEW and nursery diets were fed. When they averaged 90 lb., they were placed in the MPPA modified open-front, test unit with about 19 gilts/pen.

All gilts had ad libitum access to a corn-soybean grower diet with 3% fat (.99% digestible lysine, 21% protein, 1548 Kcal Metabolizable Energy (ME) per lb.) from 50 to 150 lb.

When gilts averaged 150 lb., they were re-penned by genetic line, four or nine gilts per pen, and fed the developmental diets described below.

Diet 1: Ad libitum access to a high protein (18%, .95% lysine), moderate energy (1485 Kcal ME per lb.) corn-soybean diet without added fat; fed to an average weight of 250 lb. This is a typical diet commonly used in performance testing programs designed to maximize lean growth.

Diet 2: Ad libitum access to a low protein (13%, .6 % lysine), high energy (1595 Kcal ME per lb.) corn-soybean diet with 5% added fat; fed to an average weight of 250 lb. This ration is typically fed so gilts will build body fat reserves, which is thought to increase fertility and longevity.

Diet 3: Ad libitum access to the grower diet until gilts weighed 180 lb. Then, intake was restricted to 4 lb./day of a high protein (23%, 1.31% lysine), moderate energy (1459 Kcal ME/lb.), corn-soybean diet without added fat until the gilts averaged 180 days of age. This ration was fed on a restricted basis to slow growth and allow slower development. Protein, mineral and vitamin intake were expected to be equal to Diet 1.

As gilts reached the 250-lb. or 180-day treatment milestone, depending on the group, they were switched to a corn-soybean diet without added fat (1470 Kcal ME per lb, 16 % protein, .8% lysine). They received 4.5 lb./day of this diet until they averaged 200 days of age, when they were bumped to 6 lb./day from 200 days of age until they were mated. After mating, gilts were allotted 4 to 5 lb./day of this diet during gestation.

Estrus Detection, Breeding Gilts were checked twice daily for estrus. Cycling gilts that were at least 210 days old were artificially inseminated. One source of semen was used.

Gilts that did not cycle normally by 8.5 months of age were treated with PG600 (Intervet), then mated if they expressed estrus.

At approximately 100 days of gestation, pregnant gilts were moved to 10 cooperator farms. Cooperators farrowed the gilts, recorded litter size and rebreeding performance on all first parity sows. Some cooperators, including two university stations that received the largest numbers of gilts (University of Nebraska West Central Research and Extension Center and University of Tennessee Ames Plantation Experiment Station) recorded this information through four parities.

Results A total of 708 gilts (140-145/line) entered the SEW station at an average weight of 8.3 lb. (range of 7.4 to 8.6 lb. across lines. Their average weight out of the nursery was 35.3 lb., ranging from 33.5 to 37.4 lb. across lines. A total of 655 gilts (126-136/line) were placed on one of the three dietary treatments.

Performance data for gilts receiving each diet are in Table 1. Weight, 10th rib backfat thickness, and 10th rib longissimus muscle area were recorded at 120, 180 and 300 days of age.

As expected, gilts fed Diet 3 grew slower, had less backfat and smaller longissimus muscle area than those fed Diets 1 and 2. However, compensatory growth after 180 days of age occurred in gilts fed Diet 3. By the time they reached 300 days of age, the differences in weights, backfats and longissimus areas due to diets were small.

The results confirmed that we were successful in choosing lines with considerable variation in mean performance across lines for growth rate and composition of growth (lean:fat). The three diets affected growth of gilts in all lines similarly.

A total of 667 gilts completed the treatments (222, 222, and 223 for Diets 1, 2 and 3, respectively), although 12 were removed for umbilical hernias prior to breeding; 564 were mated (187 to 190 gilts/dietary treatment), and 422 had Parity 1 litter records (130, 142, and 150 for respective diets).

Diet-by-line interactions on proportion of gilts that cycled normally by 8.5 months of age and on proportion of gilts that had a litter were tested.

Gilts of all lines responded similarly to the diets. The proportion of gilts in each line that cycled did not differ among diets, but the number of gilts that farrowed was affected by the diet they received (Table 2). Gilts fed Diet 3 had a 15% higher farrowing rate than those fed Diet 1, and a 9% higher farrowing rate than those fed Diet 2.

But, we also wanted to know whether the diets fed affected sow productivity through second, third and fourth parities and whether the diets affected lines differently. We defined this as "stayability" (Table 2). Calculations were based on number of gilts completing their first parity.

Again, sows of all lines responded similarly to the diets. Stayability to second parity was similar for sows on all diets. But, significantly more gilts developed on Diets 2 and 3 had third (+16%) and fourth litters (+21%) when compared to gilts fed Diet 1. Complete data on decisions to cull sows were not available at all farms.

The next step was to determine whether diets affected litter size or litter weight differently, and, if any of the three diets affected those traits differently in any of the five lines sampled. This analysis was based on number of sows that actually produced a litter at each parity.

We found significant line-by-diet interaction for total number of pigs born and number born alive in Parity 1. Of all traits analyzed, these are the only ones we saw lines respond differently to the diet fed (Table 3). The lines listed are in no particular order and do not correspond to the order of participants listed above.

Diet differences within lines were not significant except for the line-by-diet interaction for litter size. Line B and C gilts had larger litters when fed Diet 3, whereas Line A, D and E gilts tended to have larger litters when fed Diet 2.

There was nothing about breed composition or rate of growth and composition of growth that was consistent among lines within each of these groupings. For example, Lines A, B and C had very similar weights at the end of the test period, while Lines D and E were similar to each other in final weight. On the other hand, backfat depths in Line A and C gilts were similar to each other, as were Line B and D gilts. But Line E was in a class by itself. Lines of very similar breed composition also landed in different groupings.

No biological explanation for interaction on litter size at birth was apparent. This may be one of those five times in a hundred when a statistical test of a treatment effects is significant, but no real treatment differences exist.

The diet fed did not affect litter size or litter weight at 21 days of any genetic line at any parity. However, total number of pigs produced per gilt (over four parities) was approximately six pigs less for gilts developed on Diet 1 than for those developed on Diets 2 and 3 (Table 4). This shows us that Diet 2 and 3 gilts bred back better and/or had lower culling rates and completed more parities than their counterparts on Diet 1.

The significant interaction for number born in first-parity litters is puzzling and has no clear biological explanation. If that interaction is real, some caution should be used in generalizing the results.

Which Diet Is Best? For lifetime production, it's a toss-up between Diets 2 and 3. A few more gilts farrowed at Parity 1 when they were developed on Diet 3; however, total output of the gilts through four parities was nearly identical for gilts on Diets 2 and 3. It does seem clear that Diet 1 is the poorest one to use for gilt development. It caused a lower proportion of gilts to have a Parity 1 litter, and of those that did, fewer continued on to third and fourth parities. As a result, total output per gilt was significantly less.

It is difficult to uniformly limit-feed developing gilts as required by Diet 3. Floor feeding can be practiced unless floors are totally slotted. But considerable variation in individual intake still occurs. Without floor-feeding or individual feeding stations, regulating individual feed intake is almost impossible. Therefore, the diet-management system described by Diet 2 treatment is the most practical for most producers. A similar diet was adopted for the NPPC Maternal Project.

Diet 2 treatment is easier to manage than Diet 3. The two treatments did not differ in total output per replacement gilt through four parities. The lower protein level of Diet 2 translates to lower costs of developing gilts on farm.

As many producers turn to purchasing early weaned replacement gilts, they are less concerned about maximizing growth of high-lean genetic lines, instead concentrating on capitalizing on their reproductive potential. The Diet 2 approach may be a less costly approach to enhancing the lifetime reproductive performance of replacement gilts.

Contributions of Ken Stalder, University of Tennessee, and Tom Long, University of Nebraska, for data collection and analysis are acknowledged and appreciated.