By Mikayla Spinler, Jordan Gebhardt, Joel DeRouchey, Mike Tokach, Robert Goodband, Katelyn Gaffield and Jason Woodworth; Kansas State University

Precision feeding can be used to decrease the environmental impact of production animal agriculture and improve animal welfare by preventing the underfeeding or overfeeding of nutrients. The effects of precision feeding sows were evaluated by Spinler et al. (2023) in two pilot studies.

The first study utilized NRC (2012) and INRA (2009) modeled lysine estimates for lactating sows. The NRC and INRA lysine estimates were used to create lysine intake curves to target a specific lysine intake for each day of lactation for an individual sow based on parity and litter size by blending a low and high lysine diet compared to a conventional feeding strategy of a single high lysine diet. The results of the study found that sows fed only the high lysine diet with no feed blending had higher litter performance compared to the NRC and INRA curves. Litters from sows fed the NRC curve had higher growth performance than litters from sows fed the INRA curve.

Because of this, a second study was conducted using the NRC model, but lysine intake targets for each day of lactation were increased by 20% to target an average of 60 g/d of standard ileal digestible lysine intake. No differences in litter growth performance were observed between litters from sows fed the blended diet curve and the control diet treatment. As a result of these experiments, the NRC model lysine curve was used as a base in the current trial but increased by 20% for each parity for sows with an average litter size of 17 piglets.

We hypothesized that precision feeding lysine to sows during lactation could be used to decrease feed cost and achieve similar sow and litter performance in a commercial production system if adequate g/d of lysine intake was achieved. Therefore, the objective of this study was to determine the effects of precision feeding lysine and other AA during sow lactation on sow and litter performance and feed cost compared to a conventional feeding strategy in a commercial setting.

Materials and methods

The Kansas State University Institutional Care and Use Committee approved the protocol used in this experiment. The study was conducted at a commercial sow farm in eastern Iowa (Brenneman Pork, Washington, Iowa). Sows were housed in individual farrowing stalls equipped with a self-automated feed system (Gestal Quattro Opti, Jyga Technologies, St-Lambert-de-Lauzon, Quebec, Canada) and a wet dry feed bowl with an additional nipple water. Creep feed was not offered throughout the trial.

Animals and diets

A total of 728 sows (average parity 1.5; Camborough, PIC, Hendersonville, Tennessee) and litters (Camborough × PIC 800) were used in the lactation study. Only gilts and parity 1 sows were farrowing on the farm at the time of the study. Sows were moved from gestation to farrowing at approximately d 114 of gestation. Upon entry into farrowing, sow weight, caliper, backfat depth and loin depth measurements were taken. Sow caliper score was taken at the last rib. Backfat and loin depth measurements were taken at the 10^th rib approximately 2.5 in from the midline using an ExaGo ultrasound machine.

Sows were allotted to one of two treatments at entry to the farrowing house, a control 1.07% SID lysine diet or a blended diet treatment. Dietary treatments were formed from two basal diets: a low lysine (0.60% SID lysine) and a high lysine (1.07% SID lysine) diet. Pre-farrow all sows were allowed 5 lb per day of the high lysine diet. After farrowing sows were given ad libitum access to feed and placed on their respective diets the day of farrowing. Sows fed the control diet were fed only the high lysine diet. Sows fed the blended diet treatment were fed a blend of the low and high Lysine diet based on parity, with one curve for gilts and one for parity 1 sows using the Gestal Quattro Opti feeders. Blends of the low and high lysine diets was created to target a specific lysine intake, and therefore other amino acids, because they were formulated on a ratio to lysine based on expected feed intake on the farm. Lysine targets were based on the shape of the NRC (2012) model estimate recommendations for each parity with a litter size of 17 piglets but increased by 20% to target an average lysine intake for gilts of 61 g/d and 65 g/d for parity 1 sows.

Sow feed intake was analyzed on d 7 and 14 of lactation to identify sows that were eating below or above targeted feed intake based on the previous two days of feed intake. If a sow was eating below targeted feed intake by 25% or greater at either time point, they were fed a curve with a 10% increase in the blend of the high lysine diet for the remainder of the study. If a sow was eating above target feed intake by 10% or greater at either time point, they were fed a diet with a 10% decrease in the blend of the high lysine diet for the remainder of the study. This was done to more closely reach targeted lysine intake for low and high feed intake sows. Sow feed intake throughout lactation was tracked by the Gestal Quattro Opti feeders. Sow feed and lysine intake were analyzed assuming 5% feed wastage. Feeders were calibrated once a week for each diet by taking the average calibration value (weight of feed dropped in 5 turns of the feeder auger) from five random feeders.

On a subset of sows, blood and milk samples were taken on d 10 of lactation and the day before weaning. Blood samples were collected from 39 control-fed sows and 38 blend-fed sows. Sow blood was collected after a 6 h fasting period and were then analyzed for urea N concentration. Milk samples were collected from 20 control fed sows and 18 blend fed sows. Approximately 30 mL of milk was collected. To stimulate milk letdown, the ear was cleaned with isopropyl alcohol, and 0.3 mL of oxytocin was administered in the ear vein. Milk samples were analyzed for crude protein concentration.

Litters were equalized to approximately 16-17 piglets within 24 h of the end of farrowing. On d 2 of lactation after equalization, litter weight and size were recorded. Litter weight and size were also taken the day before weaning. Pre-weaning mortality for each sow was calculated by taking the litter size at weaning divided by the litter size at d 2 after equalization. The wean-to-service interval was analyzed for each sow that remained in the herd after weaning.

For the economic analysis, a low and high ingredient cost scenario was used. Feed cost in the low ingredient cost scenario was $0.08/lb for the low lysine diet and $0.10/lb for the high lysine diet. Feed cost in the high ingredient price scenario was $0.14/lb for the low lysine diet and $0.16/lb for the high lysine diet. Feed cost per lb of litter gain was calculated by taking feed cost per sow divided by litter gain for both the low and high ingredient price scenario. Feed cost per pig weaned was also calculated for both the low and high ingredient price scenario by taking feed cost per sow divided by pigs weaned. Revenue per weaned pig was calculated by taking the average pig weaning weight multiplied by a value of $0.50 per lb. Income over feed cost was calculated by subtracting sow feed cost per weaned pig from revenue per weaned pig.

Results

During the lactation period, the g/d of lysine intake was 87% of targeted lysine intake for sows fed the blended diet treatment because sows ate less feed during the study than expected. Expected feed intake was based on past feed intake records for the farm. The adjustments made to the feed curves during lactation were not great enough to bring average lysine intake closer to target.

There were no differences in sow BW at entry or weaning between treatments, as well as sow BW change from entry to wean (Table 2). No differences were observed for sow backfat, loin depth, or caliper score at entry or weaning, or changes from entry to wean. Lactation ADFI was similar between treatments. However, as expected, sows fed the control diet treatment had greater (P < 0.001) average daily lysine intake during lactation compared to sows fed the blended diet treatment because they were fed only the high Lysine diet. Sows fed the control diet treatment also had greater (P < 0.001) N intake compared to sows fed the blended diet treatment. No differences in wean to estrus interval were observed between treatments.

Differences were observed in litter growth performance during the study (Table 3). There were no differences in litter size at d 2 or at weaning as well as litter weight and average piglet weight at d 2. Litters from sows fed the control diet treatment tended to have greater litter weight (P = 0.082) at weaning as well as average piglet BW at weaning (P = 0.075) compared to litters from sows fed the blended diet treatment. A tendency for litters (P = 0.090) and piglets (P = 0.062) from sows fed the control diet treatment to have greater ADG was observed. Sows fed the control diet treatment had greater (P < 0.001) lysine intake per lb of litter gain compared to sows fed the blended diet treatment. No difference in pre-weaning mortality from d 2 to weaning was observed.

There was an interaction (P = 0.002; Table 4) between treatment and sampling time for serum urea N content. There was an increase (P < 0.05) over time in serum urea N content for sows fed the control diet but no difference for sows fed the blend diet treatment. Serum urea N concentration was lower (P < 0.05) in sows fed the blended diet treatment at d 10 and at weaning when compared to sows fed the control diet. Milk crude protein was greater (P = 0.050) for sows fed the control diet treatment at day 10 and weaning than sows fed the blended diet treatment.

For the economic analysis, feed cost per sow for both the low and high ingredient price scenarios was lower (P < 0.05) for sows fed the blended diet treatment. No differences were observed for feed cost per lb of litter weight gain for either price scenario. When looking at feed cost per pig weaned, sows fed the blended diet treatment had a lower (P < 0.05) fed cost per pig weaned compared to sows fed the control curve treatment for both a low and high ingredient cost. There was no difference (P = 0.155) in revenue per weaned pig between treatments. In both the low and high ingredient price scenarios, no differences in income over feed cost per weaned pig were observed.

Because ADFI was lower than anticipated, litter growth performance was also analyzed for only sows who reached an average lactation feed intake of 16 lb or greater on both the control and blended dietary treatment (Table 5). One hundred and fourteen sows fed the control diet treatment and 100 sows fed the blended diet treatment reached an average feed intake of 16 lb or greater during lactation. Lysine and N intake were still greater (P < 0.001) for sows fed the control diet treatment compared to the blended as expected because they were fed only the high Lysine diet. However, when sows fed the blended diet treatment had an average feed intake of 16 lb or greater, there were no differences in litter growth performance compared to litters from sows fed the control diet treatment. Lysine intake per lb of liter weight gain was still greater (P < 0.001) for sows fed the control diet treatment compared to sows fed the blended diet treatment. This analysis would indicate that SID lysine intake of approximately 61 g/d is sufficient to maximize litter weight gain with a litter size of approximately 14.8 weaned pigs and that blend-fed sows were using dietary lysine more effectively toward milk production compared to control-fed sows.

1. A total of 214 gilts and parity 1 sows with an average lactation feed intake of 16 lbs or greater (Camborough, PIC, Hendersonville, TN) and their litters (Camborough × PIC 800) were used to compare litter growth performance between litters from control and blend fed sows. 2. Sows were allotted to 1 of 2 treatments upon entry to the farrowing house. A control high lysine diet (1.07% SID Lysine) or a blended diet of a low (0.60% SID Lysine) and high lysine diet to target a specific lysine requirement based on the NRC (2012) recommendations plus 20%. 3. Litter weight at d 2 was used as a covariate in the statistical analyses. 4. Feed cost of the low lysine diet was $0.08/lb for the low ingredient cost and $0.14/lb for the high ingredient cost. Feed cost of the high Lysine diet was $0.10/lb for the low ingredient cost and $0.16/lb for the high ingredient cost. 5. Feed cost, $ per lb of litter weight gain = feed cost ÷ lb of litter weight gain per sow 6. Feed cost, $ per pig weaned = feed cost ÷ pigs weaned per sow 7. Revenue, $/weaned pig = average pig weaning weight × $0.50 8. Income over feed cost, $/weaned pig = Revenue, $/weaned pig – Feed cost, $/weaned pig

When including only sows with 16 lb or greater average lactation feed intake, feed cost was greater (P < 0.05) for sows fed the control diet treatment in a low and high ingredient price scenario. Feed cost per lb of litter weight gain tended (P = 0.099) to be higher in the low ingredient price scenario for sows fed the control diet. Feed cost per weaned pig was lower (P < 0.05) for sows fed the blended diet in both the low and high ingredient price scenario, but no differences in revenue per weaned pig were observed. Income over feed cost per pig weaned was higher (P < 0.05) for sows fed the blended diet treatment when looking at only sows who met expected feed intake.

In conclusion, there were no differences in sow BW and composition throughout the study. Piglets and litters from sows fed the blended diet treatment tended to have decreased weaning weights and ADG. This is likely due to lower SID lysine intake than targeted, which was below the sow’s requirement to maximize litter growth. However, when sows fed the blended diet treatment achieved target Lysine intake, they had similar litter growth performance compared to litters from sows fed the control diet treatment. Feed cost per sow and feed cost per weaned pig were lower for sows fed the blended diet treatment. The next steps in this research would include using technology to automatically adjust diet blends to avoid the underfeeding of nutrients based on individual sow feed intake.

This and other data related to precision feeding of sows will be discussed more in-depth at the 2024 K-State Swine Industry Day on Nov. 21.