Sow studies evaluate the validity of National Research Council amino acid recommendations.

Sows are different enough in feed intake and milk production to make it unrealistic to expect the same diets to be appropriate for all of them.

So nutritionists have developed systems for estimating the amino acid requirements for sows, taking into account:

  • Body weight (maintenance requirement);

  • Litter growth rate (milk production); and

  • Body protein loss (contributes amino acids).

There is general agreement on the approach, but big differences in the relative needs (or supply) of the different amino acids.

Consider a situation that specifies a high level of milk production, a low level of protein intake, and a high rate of body protein loss. The National Research Council (NRC) estimates order of limitation of amino acids to be lysine, then valine, then threonine.

But numbers proposed by Sungwoo Kim (then at the University of Illinois, now at Texas Tech University) suggest the order of limitation to be first threonine, then lysine, then valine.

Other estimates based on uptake of amino acids by the mammary gland suggest even higher requirements for threonine and valine relative to lysine.

Research Farm Experiments

We conducted a series of experiments to guide nutritionists in the selection of the most appropriate set of numbers for estimating amino acid requirements in practice.

First, we conducted two intensive experiments with a few sows in our research farm. The first experiment used a basal diet with a strategically-selected ratio of threonine-to-lysine, using either or both of those amino acids added. The second experiment was similar, except we focused on threonine and valine.

Each experiment was divided into four periods of four days each, and the diets were rotated so each sow consumed all four diets. The critical measurement was plasma urea nitrogen (PUN) concentration, a measure that declines with improved amino acid balance.

The results clearly showed that under these circumstances, lysine was the first-limiting amino acid, and threonine was second.

Although these results were clear, we considered it important to determine whether feeding each individual diet throughout lactation would produce the same results.

Commercial Farm Experiments

Full-lactation studies require a lot of sows, so we collaborated with a team of swine nutrition researchers in Mexico, headed by Jose Cuaron, for an experiment in a 1,200-sow farm. This experiment was designed to determine whether threonine or valine is more limiting. We collected usable data on 351 sows.

The results (Table 1) showed that for first-litter sows, threonine was more limiting, in agreement with the short-term PUN experiment. Older sows (about 72 per treatment) did not respond to amino acid supplementation.

Collaborative Trial

A collaborative experiment with United Feeds was conducted with 378 sows on two farms in Indiana. In this trial, we tested the amount of crystalline lysine that could be used to replace soybean meal in a specially-formulated control diet. All three experimental diets contained the same amount of digestible lysine, but lysine HCl provided 0%, 0.1% or 0.3% lysine. As lysine HCl was added, the levels of threonine and valine declined; all other amino acids were added to ensure they would not be limiting.

As predicted by NRC and Kim, the results indicated that threonine and/or valine requirements, estimated by the mammary gland uptake studies, are too high.

In this experiment, litter growth rate was reduced by the higher level of lysine HCl and not by the lower level, but we urge caution in extending this observation to other diets.

The control diet used in this experiment was different than usually fed, because our purpose was to evaluate systems for calculating requirements. Our purpose was not to measure specifically the amount of lysine HCl that can be used, because that can vary.


These results provide confidence in the NRC estimates of amino acid requirements for lactating sows, except that the NRC estimate of the valine requirement is too high.

*K.T. Soltwedel is employed at SCA Nutrition as technical sales manager for Illinois and eastern Iowa.

Table 1. Effect of Amino Acid Supplementation on Growth Rate of Litters of First-Parity Sows
Control + Threonine + Valine + Both
Number of sows 16 16 14 15
Litter growth rate during first 8 days, lb./day 4.37 4.70 4.10 4.85
Threonine effect, P=0.09; threonine by parity interaction, P=0.07.